Coronavirus Disease 2019 (COVID-19) Daily Research Briefs

May 13, 2021, Research Update

mRNA Vaccines Prevent Hospitalization in Adults 65 Years of Age and Older. A large epidemiological study from Israel found that the mRNA vaccines were 87% effective in preventing hospitalization from COVID-19 infection (Research Brief from NEJM posted March 15, 2021). In this case-control study from the CDC, investigators used data from a 24-hospital sentinel network to determine the effectiveness of partial or full vaccination with Pfizer-BioNTech or Moderna vaccines against COVID-19–associated hospitalization among older adults. Among the 417 adults 65 years of age and older hospitalized for COVID-19–like symptoms, 187 patients tested positive for SARS-CoV-2, and 230 tested negative. The median age was 73, 48% were female, 17% were non-Hispanic Black, and 6% were Hispanic. Only one of the patients in the group who tested positive for COVID-19 had been fully immunized, and 19 had received one dose. In the group that tested negative for acute COVID-19 infection, 18 patients were fully immunized; 44 had received one injection. The adjusted vaccine effectiveness against COVID-19–associated hospitalization was estimated to be 94% (95% CI, 49% to 99%) for full vaccination and 64% (95% CI, 28% to 82%) for partial vaccination. Although this is a small study and the confidence intervals are wide, it strongly suggests that the mRNA vaccines are effective in preventing illness severe enough to require hospitalization in older adults. These findings are consistent with the large cohort study from Israel.

Written by John Hickner, MD, MS, on May 1, 2021. (Source: Tenforde MW, Olson SM, Self WH, et al. Effectiveness of Pfizer-BioNTech and Moderna vaccines against COVID-19 among hospitalized adults aged ≥65 years — United States, January–March 2021. MMWR Morb Mortal Wkly Rep. 2021;70(18):674-679.)

May 12, 2021, Research Update

Case Series of Patients with Cerebral Venous Sinus Thrombosis with Thrombocytopenia after Janssen/Johnson & Johnson Vaccination in the United States. Soon after the introduction of the Oxford/AstraZeneca adenovirus-based COVID-19 vaccine, reports of patients with “clotting problems'' hit the media. These were patients with thromboses associated with thrombocytopenia, 72% of whom also had cerebral venous sinus thrombosis (CSVT). Similar reports occurred in the United States after the introduction of the Janssen/Johnson & Johnson adenovirus vaccine. These authors from the CDC mined the Vaccine Adverse Event Reporting System (VAERS) and identified reports of 12 patients with CVST with thrombocytopenia and described their clinical findings. The reporting period was for the six weeks between March 2 and April 21, 2021, during which about 7 million doses of the vaccine had been administered. After the CDC identified the people identified with CVST, they supplemented information provided to VAERS with medical records or discussions with treating clinicians. Eight of the patients were from 18 to 39 years of age, the remainder were from 40 to 60 years of age. Of the 12 patients, seven had at least one risk factor for CVST (obesity, hypothyroidism, or combined oral contraceptive use). The patients became symptomatic between six and 15 days of the vaccine and were hospitalized between two and 14 days from onset of symptoms. Eleven of the 12 presented with headache, and one developed a headache later. Eight of the 12 had additional venous thromboses (e.g., portal vein, internal jugular, etc.), and seven had intracerebral hemorrhages. All of the patients were hospitalized; 10 were in intensive care. At the time of publication, five patients were still in the hospital, three died, and four were discharged home. Because the VAERS is a passive reporting system, the true number of people with CSVT is higher, and people with mild or atypical findings have likely been missed. Nonetheless, this serious complication appears to be quite rare.

Written by Henry C. Barry, MD, MS, on May 1, 2021. (Source: See I, Su JR, Lale A, et al. US case reports of cerebral venous sinus thrombosis with thrombocytopenia after Ad26.COV2.S vaccination, March 2 to April 21, 2021 [published online April 30, 2021]. JAMA. 2021. DOI: //jamanetwork.com/journals/jama/fullarticle/2779731)

May 10, 2021, Research Update

Individuals with SARS-CoV-2 Persistence Are Not Infectious after 10 Days. The CDC has stated that 10 days after COVID-19 infection is diagnosed, a person should no longer be considered infectious. This is supported by prior epidemiological studies. What about people who persistently test positive for SARS-CoV-2? The National Basketball Association (NBA) “bubble” in Orlando, Florida, that isolated all NBA players, staff, and vendors during the second half of the season and playoffs provided a natural experiment for determining whether those who persistently test positive with a sensitive RT-PCR test are infectious. Of the 3,648 individuals who participated in the bubble, 36 individuals persistently tested positive after the 10-day isolation period, which began on the first day of symptoms of infection or the first positive test. The mean number of persistent positive days for these 36 individuals was 31 days. There were at least 1,480 person-days of close, direct contact among these individuals and others in the bubble, and there were no documented cases of transmission. This data is reassuring in supporting the recommendation that 10 days of isolation is sufficient for persons with COVID-19 infection. After 10 days, the likelihood of transmission appears negligible. 

Written by John Hickner, MD, MS, on May 1, 2021. (Source: Mack CD, DiFiori J, Tai CG, et al. SARS-CoV-2 transmission risk among National Basketball Association players, staff, and vendors exposed to individuals with positive test results after COVID-19 recovery during the 2020 regular and postseason [published online April 22, 2021]. JAMA Intern Med. //jamanetwork.com/journals/jamainternalmedicine/fullarticle/2779287)

May 4, 2021, Research Update

mRNA COVID-19 Vaccines Appear Safe in Pregnant Women. From December 14, 2020, to February 28, 2021, 35,691 pregnant women enrolled in the CDC v-safe voluntary reporting system for COVID-19 immunizations, which is a part of the Vaccine Adverse Event Reporting System. Of these women, 3,958 enrolled in the v-safe pregnancy registry, which follows pregnant women until delivery. There were 827 women who had a completed pregnancy at the time of this report. There were 724 live births (including 12 sets of multiple pregnancy; 86.1%), 104 spontaneous abortions (12.6%), and one stillbirth (0.1%). There were 96 of the 104 spontaneous abortions that occurred before 13-weeks’ gestation. Adverse events among the 724 live-born infants were preterm birth (9.4%), small size for gestational age (3.2%), and major congenital anomalies (2.2%); no neonatal deaths were reported. Although there is no parallel control group, these proportions of adverse outcomes are similar to pregnancy outcomes recorded in other studies prior to the COVID-19 pandemic. These preliminary data reassure us that COVID-19 mRNA vaccines are safe to use during pregnancy, although we need a much larger database and a comparison group to be absolutely certain.

Written by John Hickner, MD, MS, on April 29, 2021. (Source: Shimabukuro TT, Kim SY, Myers TR, et al. Preliminary findings of mRNA Covid-19 vaccine safety in pregnant persons [published online April 21, 2021]. NEJM. 2021. //www.nejm.org/doi/full/10.1056/NEJMoa2104983)

May 3, 2021, Research Update

Living Systematic Review Fails to Find Much for Prophylaxis against COVID-19. This is the first report of a systematic review that will be updated frequently from the same team that created the BMJ’s living systematic review of drugs to treat COVID-19 (Research Briefs posted on August 28, 2020, January 12, 2021, and April 16, 2021). In this paper, the authors searched multiple sources to identify randomized trials of medications, vitamins, or antibodies for people at risk of contracting COVID-19 and that reported data in a manner suitable to perform network meta-analyses. As part of their process, the team assessed the risk of bias for each study and then used standard statistical methods to perform the network meta-analyses—a technique that allows statistical estimates of the relative effectiveness of interventions even when head-to-head studies do not exist. At the time of the initial release, they identified 11 unique studies: six evaluated hydroxychloroquine, two ivermectin, one ivermectin and iota-carrageenan, one bromhexine, and one ramipril. Six studies evaluated preexposure strategies, four postexposure prophylaxis, and one looked at both. Overall, two of the studies (one of ramipril and one of bromhexine hydrochloride) were not included in the network meta-analysis because they were too small. Five of the hydroxychloroquine studies were at low risk of bias, and the rest of the studies were at high risk of bias. Hydroxychloroquine minimally, but statistically significantly, prevented hospitalization (one fewer per 1,000 participants) or death (one fewer per 1,000 participants). Additionally, hydroxychloroquine was ineffective in preventing lab-confirmed infection and caused more harms. The studies of ivermectin suggested that it was effective in preventing infection; however, the studies were poor enough to give little confidence about the data. As more studies are released, this systematic review will be updated.

Written by Henry C. Barry, MD, MS, on April 26, 2021. (Source: Bartoszko JJ, Siemieniuk RAC, Kum E, et al. Prophylaxis against covid-19: living systematic review and network meta-analysis. BMJ. 2021;373:n949.)

April 20, 2021, Research Update

Persistent Symptoms Common Even after Mild COVID-19 Eight Months Later. Although it has been well reported that persistent symptoms are common after hospitalization for moderate to severe COVID-19, there are fewer data regarding the prognosis for persons with a milder episode of illness. These Swedish researchers enrolled 2,149 health care workers between April 15 and May 8, 2020, and obtained baseline serology. They then had regular follow-up for eight months, drawing blood for serology and asking about the persistence of symptoms. The researchers excluded anyone who had an episode of severe COVID-19 illness and anyone who went from seronegative to seropositive during the study. This left 323 who were seropositive at the beginning of the study and had experienced no or mild symptoms and 1,072 who were seronegative at the beginning of the study and remained that way for the duration of the study. Groups were similar demographically and with regard to chronic disease. They identified patients with persistent anosmia, fatigue, ageusia, dyspnea, sleeping disorder, headache, palpitations, impaired concentration, myalgias, or memory impairment rated as moderate to severe. Any moderately severe or worse symptom was present more often in the seropositive patients at two months (26% vs. 8.9%), four months (21.4% vs. 7.2%), and eight months (14.9% vs. 3.4%). The most common persistent symptoms at eight months were anosmia (9.0%), fatigue (4.0%), ageusia (3.7%), and dyspnea (1.9%). They were also more likely to report that these symptoms at least moderately disrupted their work life, social life, and home life. It is encouraging that for each symptom, the percentage reporting it declined over time.

Written by Mark H. Ebell MD, MS, on April 11, 2021. (Source: Haverall S, Rosell A, Phillipson M, et al. Symptoms and functional impairment assessed 8 months after mild covid-19 among health care workers [published online April 7, 2021]. JAMA. 2021. //jamanetwork.com/journals/jama/fullarticle/2778528)

April 19, 2021, Research Update

Inhaled Budesonide Reduces Likelihood of Hospitalization in Patients with Mild COVID-19 within Seven Days of Symptom Onset (STOIC). Oral corticosteroids such as methylprednisolone are helpful only when used with more severely ill patients, and there was a trend toward outcomes in those with mild disease perhaps by suppressing immune function systemically (N Engl J Med. 2020; DOI:10.1056/NEJMoa2021436). But observational studies have shown an association between inhaled steroids and better outcomes in patients with mild disease. These researchers identified adults with fewer than seven days of cough and either fever or anosmia or both (n = 146). They randomized them to budesonide 400 mcg actuations, with two actuations twice daily or to usual care. A nurse completed a swab for SARS-CoV-2, which was positive for 94%. Groups were balanced at randomization. Four withdrew consent before receiving the allocated intervention, one in each group needed urgent care before they could be swabbed, and one withdrew because they found the inhaler too burdensome, leaving a per protocol population of 139 patients. Patients in the budesonide group were told to stop using the inhaler when they had recovered (median seven days), and all patients were followed for 28 days. It does not appear that outcomes were assessed in a blinded manner, and the study was stopped early due to the large benefit. In the entire population, the primary outcome of urgent emergency department visits or hospitalization occurred less often in the budesonide group (3% vs. 15%, p = 0.009, NNT = 8). The magnitude of benefit was similar in the per protocol population (1% vs. 14%, NNT = 7). Time to recovery was also about one day sooner, and symptom resolution was also more rapid. Adverse events (four with sore throat, one with dizziness) were minor and self-limited. The open-label design, early trial shutdown, and failure to mask outcome assessment are limitations, but the magnitude of benefit was larger enough that this should be considered for all patients with mild symptoms of COVID-19. Seven other trials are underway that are looking at inhaled budesonide or ciclesonide, and their results are urgently awaited.

Written by Mark H. Ebell MD, MS, on April 17, 2021. (Source: Ramakrishnan S, Nicolau DV, Langford B, et al. Inhaled budesonide in the treatment of early COVID-19 (STOIC): a phase 2, open-label, randomised controlled trial [published online April 9, 2021; correction on April 14, 2021]. Lancet Respir Med. 2021; //www.thelancet.com/journals/lanres/article/PIIS2213-2600(21)00160-0/fulltext)

April 16, 2021, Research Update

Neurologic and Psychiatric Consequences after COVID-19 Are More Common Than with Other Respiratory Illnesses and Influenza. These authors used a federated electronic health record database in the United States that house data on 81 million patients in 62 different health care organizations—pretty decent, but not quite a population-level database. They used the database to identify a primary cohort of persons diagnosed with COVID-19 (n = 236,379), another cohort of persons diagnosed with influenza during the same time period (n = 105,579), and a third cohort of persons diagnosed with any respiratory illness, including influenza (n = 236,038). They explicitly excluded any people also diagnosed with COVID-19 from the latter two cohorts. They used propensity matching on several covariates so that each cohort had similar baseline characteristics and then used additional statistical gymnastics (“greedy nearest neighbor matching,” caliper distances, etc.) to identify the occurrence of newly diagnosed neurologic or psychiatric conditions during the subsequent six months after the respiratory illnesses. For conditions that tend to recur or relapse (e.g., psychiatric illnesses, stroke, etc.), they estimated separately the incidence of first diagnosis as well as the incidence of any diagnosis. They also clustered 14 different neurologic or psychiatric conditions into a single composite outcome. Among the patients with COVID, 46,302 were hospitalized, 8,945 were in ICUs, and 6,229 had encephalopathy. In the six months after diagnosis, 33.6% of patients with COVID-19 had a subsequent diagnosis of any neurologic or psychiatric illness. These were more frequent among those hospitalized (38.7%), who were in ICUs (46.4%), and those with encephalopathy (62.3%). Compared with the people with influenza, the patients with COVID-19 were more likely to have subsequent care for any of the 14 neurologic or psychiatric conditions (HR 1.44; 95% CI, 1.40 to 1.47) or for a new diagnosis (HR 1.78; 95% CI, 1.68 to 1.89). These were also higher than in people with any respiratory illness (HR 1.16; 95% CI, 1.14 to 1.17 for any diagnosis; 1.32; 95% CI, 1.27 to 1.36 for a new diagnosis). Not surprisingly, hospitalized patients, patients receiving ICU care, and those with encephalopathy were incrementally worse off.

Written by Henry C. Barry, MD, MS, on April 8, 2021. (Source: Taquet M, Geddes JR, Husain M, et al. 6-month neurological and psychiatric outcomes in 236 379 survivors of COVID-19: a retrospective cohort study using electronic health records [published online April 6, 2021]. Lancet Psychiatry. 2021. //www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(21)00084-5/fulltext)

April 14, 2021, Research Update

Long-Term Consequences of Being Hospitalized in the United Kingdom with COVID-19 Are More Prevalent Than in the General Population. Using national health care databases, this research team identified nearly 54,000 people hospitalized in the United Kingdom with COVID-19 who were discharged alive by August 31, 2020, and another group from the general population who were exactly matched based on age group, sex, ethnicity, region of residence, living circumstances (level of deprivation), body mass index group, and several comorbid conditions: hypertension, cardiovascular disease, respiratory disease, stage 3 or worse chronic kidney disease (CKD), diabetes mellitus, and cancer. Because they had more than 56 million people in the general population, they were able to have matches for 47,780 of the patients with COVID-19, 4,745 of whom required intensive care. They “started the clock” for both sets of people at the same time based on the discharge day for the very first COVID-19 hospitalization. They then looked to see what happened subsequently. They had about five months of follow-up for both groups of patients. Compared with the general population before matching, hospitalized patients were more likely to be male 50 years of age or older, living in a low-income area, a former smoker, overweight or obese, and to have comorbid conditions. During the follow-up period, the hospitalized patients were more likely than the matched controls to die (12.3% vs. 1.7%) to have been hospitalized (29.4% vs. 9.2%) and to have newly diagnosed respiratory disease (21.5% vs. 0.8%). Although at much lower rates (3% or less), the rate of newly diagnosed cardiovascular disease, CKD, diabetes, and chronic liver disease were also much higher among the hospitalized patients than their matched pairs from the population. The authors also looked at differences across age categories, sex, race, etc., and found that generally the patterns were not uniform. If you haven’t figured it out yet, this COVID-19 stuff is serious business.

Written by Henry C. Barry, MD, MS, on April 5, 2021. (Source: Ayoubkhani D, Khunti K, Nafilyan V, et al. Post-covid syndrome in individuals admitted to hospital with covid-19: retrospective cohort study. BMJ. 2021;372:n693.)

April 13, 2021, Research Update

April 6, 2021, Update to “Living” Systematic Review of Drugs for Treating Patients with COVID-19. We have previously summarized (Research Brief posted on August 28, 2020, and again with Research Brief posted on January 12, 2021) a living systematic review from the BMJ. The authors use high-quality methods, including searching for unpublished studies. Their conclusions have evolved as more data have come along. In this update, the authors identified 196 trials with 76,767 patients. Their last update included 85 randomized trials with 41,669 patients! The researchers pooled data and performed network meta-analyses to drive their conclusions. The online version of the paper has interactive graphs that allow the reader to identify specific outcomes and to see the data on various therapies. They conclude that corticosteroids and possibly colchicine probably reduce mortality and that corticosteroids and interleukin-6 inhibitors probably reduce the need for mechanical ventilation. They conclude that janus-kinase inhibitors show promise—but stay tuned. They also conclude that azithromycin, hydroxychloroquine, lopinavir-ritonavir, and interferon-beta do not appear to have any important benefits. It remains uncertain whether remdesivir, ivermectin, and other drugs confer any patient-important benefit. They explicitly excluded vaccines and monoclonal antibody therapies from their review and plan separate analyses for those.

Written by Henry C. Barry, MD, MS, on April 7, 2021. (Source: Siemieniuk RA, Bartoszko JJ, Ge L, et al. Drug treatments for covid-19: living systematic review and network meta-analysis [published correction appears in BMJ. 2021;373:n967]. BMJ. 2020;370:m2980.)

April 12, 2021, Research Update

Three Studies Report That COVID-19 Vaccines Are Effective in Health Care Workers. These three studies (one from California, one from Israel, and one from Texas) report data on monitoring of outcomes of health care workers after receiving two-dose COVID vaccines (it is not clear which vaccines were used in each of the studies). The researchers from California actively tested their workers, whereas those in Israel tested health care workers opportunistically. It is unclear how the Texas researchers monitored their workers. In the California study, half of the 28,184 fully vaccinated health care workers were tested. Only 37 of them tested positive for COVID-19 (it is not clear whether they were symptomatic), and 22 of those occurred within one week of receiving the second dose. In the Israeli study, of the 5,297 workers who received the first dose, 98.9% received the second. A handful of cases occurred in the first weeks after the first dose, but the rates plummeted to near-zero within two weeks of the second dose. At the University of Texas Southwestern Medical Center, 30% of their 23,234 health care workers received both doses of vaccine. They report that COVID-19 occurred in 234 of 8,969 unvaccinated employees (2.6%), 112 of 6,144 partially vaccinated employees (1.8%), and in four of 8,121 fully vaccinated employees (0.05%). These data should provide reassurance for health care workers that the vaccines are effective.

Written by Henry C. Barry, MD, MS, on March 27, 2021. (Sources: Keehner J, Horton LE, Pfeffer MA, et al. SARS-CoV-2 infection after vaccination in health care workers in California [published online March 23, 2021]. N Engl J Med. 2021. //www.nejm.org/doi/10.1056/NEJMc2101927; Benenson S, Oster Y, Cohen MJ, et al. BNT162b2 mRNA Covid-19 vaccine effectiveness among health care workers [published online March 23, 2021]. N Engl J Med. 2021. //www.nejm.org/doi/10.1056/NEJMc2101951; and Daniel W, Nivet M, Warner J, et al. Early evidence of the effect of SARS-CoV-2 vaccine at one medical center [published online March 23, 2021]. N Engl J Med. 2021. //www.nejm.org/doi/10.1056/NEJMc2102153)

April 8, 2021, Research Update

AstraZeneca Vaccine Is 66.7% Effective in Preventing Symptomatic COVID and 100% Effective in Preventing Hospitalizations; Modeling Suggests Greater Effectiveness with Three-Month Interval between Doses. This study, funded by government agencies and foundations, including the Gates Foundation, pooled data from four randomized trials of the Oxford/AstraZeneca vaccine where 24,422 participants received the active vaccine or a control vaccine. There is a lot going on in this paper, including a primary analysis on full vaccine effectiveness against confirmed symptomatic COVID-19, secondary analysis on antibody responses, and exploratory analyses on first-dose vaccine effectiveness and the effect of vaccine timing. The main outcome it that lab-confirmed symptomatic COVID-19 infection more than 14 days after the last dose of vaccine is fairly standard across all the various COVID-19 vaccine trials. Among those receiving the active vaccine, 84 (1.0%) had confirmed COVID-19 compared with 248 (2.9%) control patients (NNT = 53; 95% CI, 43 to 67). The authors report that the overall vaccine effectiveness of the active vaccine was 66.7%. Only three participants were hospitalized within the first three weeks after vaccination. After 21 days, however, nobody receiving the active vaccine was hospitalized compared with 15 control patients (100% vaccine effectiveness; NNT = 390; 95% CI, 258 to 631). About 1% of participants in each group experienced serious adverse events. Using statistical modeling, they determined that the first dose of vaccine was 76% and that if the interval between doses was three months, the effectiveness was 81.3% compared with 55.1% for shorter intervals. Whereas others may get excited about antibody responses, the real deal is in preventing disease and its complications. The model suggests that a reasonable strategy would be to get the first dose into as many people as possible and the second dose three months later. This would allow time for vaccine supplies to ramp up.

Written by Henry C. Barry, MD, MS, on March 25, 2021. (Source: Voysey M, Costa Clemens SA, Madhi SA, et al.; Oxford COVID Vaccine Trial Group. Single-dose administration and the influence of the timing of the booster dose on immunogenicity and efficacy of ChAdOx1 nCoV-19 (AZD1222) vaccine: a pooled analysis of four randomised trials [published correction appears in Lancet. 2021;397(10277):880]. Lancet. 2021;397(10277):881-891.)

April 7, 2021, Research Update

Predictors of Severe COVID-19 or MIS-C in Hospitalized Children. This research consortium pooled medical record data on 281 children hospitalized with COVID-19 in 8 sites in New York, New Jersey, and Connecticut. They found 315 children who had positive tests for COVID-19, but 34 were hospitalized for reasons other than COVID-19. They used CDC criteria to classify the children with suspected multisystem inflammatory syndrome (MIS-C). They also classified severe disease as a child who was in the ICU for 48 hours or more. Just over half (50.9%) of the children had respiratory disease, and about 25% had MIS-C or had other syndromes (gastrointestinal manifestations, fever, etc.). The median age was 10 years, and 60% of the children were male. The majority of patients were Hispanic (51%) or non-Hispanic black (23%). Seven children (2%) died, and 114 children (41%) were admitted to the ICU. The authors used multivariate analyses to identify which factors had the strongest association with developing severe disease and MIS-C. Somewhat comparable with what we have seen in adults, obesity (OR 3.39; 95% CI ,1.26 to 9.10), hypoxia on admission (OR 4.01; 95% CI, 1.14 to 14.15), and bilateral pulmonary infiltrates (OR 3.69; 95% CI, 1.46 to 9.32) were significant predictors of severe disease. Although the majority of the children were Hispanic or non-Hispanic black, in the multivariate analyses that control for the aforementioned clinical factors, race, ethnicity, or sociodemographic factors were not predictive of disease severity. The authors also report that the following factors were most strongly associated with development of MIS-C: lower absolute lymphocyte count (OR 8.33 per unit decrease in 109 cells per L; 95% CI, 2.32 to 33.33) and higher levels of C-reactive protein (OR 1.06 per unit increase in mg per dL; 95% CI, 1.01 to 1.12).

Written by Henry C. Barry, MD, MS, on March 25, 2021. (Source: Fernandes DM, Oliveira CR, Guerguis S, et al.; Tri-State Pediatric COVID-19 Research Consortium. Severe acute respiratory syndrome coronavirus 2 clinical syndromes and predictors of disease severity in hospitalized children and youth. J Pediatr. 2021;230:23-31.e10.)

April 6, 2021, Research Update

Less Than 1% Chance of Reinfection with SARS-CoV-2. Prior studies, mostly limited to health care professionals, have shown a high degree of protection against reinfection with the SARS-CoV-2 virus for up to six months. This study is the largest, population-based study we have seen to estimate reinfection rates. The investigators used a large testing database in Denmark to estimate the protection against repeat infection. They determined reinfection rates from September 1 to December 31, 2020, of individuals who had positive and negative PCR tests during the first surge from March to May of 2020. They compared infection rates between those with and without a previous confirmed infection at least three months earlier. Of 533,381 people initially tested, 11,727 (2.2%) were PCR positive, and 11,068 (2.11%) had a positive PCR among the 525,339 who were tested again from September 1 to December 31. Seventy-two (0.65%) PCR-positive individuals from the first surge of the epidemic tested positive again during the second surge. Of the 514,271 who tested negative during the first surge, 16,819 (3.27%) tested positive during the second period. Infection during the second period, therefore, was 80.5% less likely in those who had a prior infection. Individuals older than 65, however, had only 50% protection against reinfection. Another way to look at this data is that one in 154 individuals with an initial infection were reinfected. A limitation of the study is that the investigators could not determine how many of these individuals had symptoms with reinfection.

Written by John Hickner, MD, MS, on March 19, 2020. (Source: Hansen CH, Michlmayr D, Gubbels SM, et al. Assessment of protection against reinfection with SARS-CoV-2 among 4 million PCR-tested individuals in Denmark in 2020: a population-level observational study. Lancet. 2021;397(10280):1204-1212.)

April 2, 2021, Research Update

AstraZeneca’s Vaccine Does Not protect against Mild to Moderate COVID-19 Infections Caused by the South African Strain B.1.351. These researchers in South Africa randomized 2,026 HIV-negative adults 18 to 65 years of age to receive either the AstraZeneca ChAdOx1 nCoV-19 vaccine (AZD1222) or placebo, each administered in two doses 21 to 35 days apart. The participants also had to have been COVID-free. The researchers prompted the study participants to report COVID-19 symptoms every two weeks and, if present, to be tested using a nucleic amplification test for COVID-19. This paper provides data on COVID-19 14 days or more after the second injection. The authors’ analysis plan included only those who received both doses of their shots and were grouped according to what they actually received as opposed to what they were randomized to receive. This approach is generally biased in favor of active interventions. Nonetheless, 3.2% placebo-injected participants developed lab-confirmed mild to moderate COVID-19 compared with 2.5% of active vaccine recipients. Furthermore, they report that 2.8% of placebo-injected persons developed mild to moderate infections with the B.1.351 (South African) strain of COVID-19 compared with 2.5% of those receiving the active vaccine. The authors also report various markers of immune response, but these are less relevant to clinicians and patients. Finally, the authors do not report on the development of more severe COVID-19 disease. Bottom line: In this study with biases that historically make interventions look better, the AstraZeneca ChAdOx1 nCoV-19 vaccine (AZD1222) was no better than placebo in preventing mild to moderate infections due to the B.1.351 COVID variant.

Written by Henry C. Barry, MD, MS, on March 17, 2021. (Source: Madhi SA, Baillie V, Cutland CL, et al.; NGS-SA Group Wits–VIDA COVID Group. Efficacy of the ChAdOx1 nCoV-19 Covid-19 vaccine against the B.1.351 variant [published online March 16, 2021]. N Engl J Med. 2021. //www.nejm.org/doi/full/10.1056/NEJMoa2102214)

April 1, 2021, Research Update

Rate of Anaphylaxis to COVID-19 mRNA Vaccines is one in 4,056 Doses at Mass General Brigham. Reports from the CDC Vaccine Adverse Events Reporting System (VAERS) have indicated that anaphylaxis to mRNA COVID vaccines have occurred at a rate of one in 90,000 doses for the Pfizer-BioNTech vaccine and one in 200,000 doses for the Moderna vaccine. These investigators report the rate of acute allergic reactions including anaphylaxis in 64,900 employees of Mass General Brigham who received one of these vaccines. Forty percent received the Pfizer-BioNTech vaccine and 60% received the Moderna vaccine. Employees were asked to complete a symptom survey for the first three days after the injection, and 81% completed it. Acute allergic reactions were reported by 1365 employees, a rate of 2.1%. Anaphylaxis was confirmed in 16 employees: 7 (0.027%) from the Pfizer-BioNTech vaccine and 9 (0.023%) from the Moderna vaccine. Fifteen of the 16 were female. Ten had a prior allergy history and 5 had a history of anaphylaxis. Mean time to anaphylaxis onset was 17 minutes and onset was 120 minutes for one recipient. One patient was admitted to the intensive care unit. The overall rate of anaphylaxis to the vaccines of 1 per 4,056 doses reported in this study is much greater than the CDC reports. The rate did not differ by vaccine type in contrast to the CDC data that shows a considerably higher rate of anaphylaxis to the Pfizer-BioNTech vaccine. It is unclear if this data is generalizable to the U.S. population as a whole, but this study suggests that the rate of anaphylaxis to COVID mRNA vaccines is overall low, but greater than the incidence reported through the CDC VAERS system.

Written by John Hickner, MD, MS, on March 15, 2021. (Source: Blumenthal KG, Robinson LB, Camargo CA, et al. Acute allergic reactions to mRNA COVID-19 vaccines [published online March 8, 2021]. JAMA. 2021. //jamanetwork.com/journals/jama/fullarticle/2777417)

March 23, 2021, Research Update

Risk of Death is Higher with the SARS-CoV-2 B.1.1.7 Variant Than with the Original Strain. We previously reported that the SARS-CoV-2 B.1.1.7 variant, first described in the United Kingdom, may have a higher mortality rate compared with the initially described SARS-CoV-2 strain. In the largest study to date, investigators matched 54,906 pairs of individuals who tested positive for the original strain or the B.1.1.7 variant, and they determined the relative mortality rate. Individuals were identified from community-based testing centers in the United Kingdom who tested positive between October 1, 2020, and January 29, 2021, and were followed up until February 12, 2021. Participants were matched on age, sex, ethnicity, index of multiple deprivation (including income, employment, education, skills and training, health and disability, crime, barriers to housing services, and living environment), location (lower-tier local authority region of about 190,000 people), and sample date of positive specimens; they differed only by detectability of the spike protein gene using the TaqPath assay. The mortality hazard ratio associated with infection with the B.1.1.7 variant compared with infection with previously circulating variants was 1.64 (95% CI, 1.32 to 2.04). In this comparatively low-risk group, this represents an increase in deaths from 2.5 to 4.1 per 1,000 detected cases. This study provides strong evidence that the B.1.1.7 SARS-CoV-2 variant is more lethal than the original strain.

Written by John Hickner, MD, MS, on March 14, 2021. (Source: Challen R, Brooks-Pollock E, Read JM, et al. Risk of mortality in patients infected with SARS-CoV-2 variant of concern 202012/1: matched cohort study. BMJ. 2021;372:n579.)

March 22, 2021, Research Update

Statewide Mask Mandates Reduced COVID-19 Hospitalizations. Prior studies have shown that wearing masks reduces the transmission of SARS-CoV-2. As of October 22, 2020, statewide mask mandates were in effect in 33 states and the District of Columbia. A statewide mask mandate was defined as the requirement to wear a mask anywhere outside the home or in retail businesses and in restaurants. This study examined change in hospitalization rates after a mask mandate was put in effect in 10 states participating in the COVID-19–Associated Hospitalization Surveillance Network. Weekly hospitalization growth rates declined by 2.9 percentage points (95% CI, 0.3 to 5.5) among adults 40 to 64 years of age during the first two weeks after implementing statewide mask mandates. After mask mandates had been implemented for 3 weeks or more, hospitalization growth rates declined by 5.6 percentage points among people 18 to 39 years of age (95% CI, 0.6 to 10.4) and those 40 to 64 years of age (95% CI, 0.8 to 10.2). This study suggests that mask mandates reduce hospitalizations for COVID-19 infection. The data would be stronger if there were a comparison group from states without mask mandates.

Written by John Hickner, MD, MS, on March 14, 2021. (Source: Joo H, Miller GF, Sunshine G, et al. Decline in COVID-19 hospitalization growth rates associated with statewide mask mandates—10 states, March–October 2020. MMWR Morb Mortal Wkly Rep. 2021;70(6):212-216.)

March 18, 2021, Research Update

Mask Mandates Associated with Fewer Cases and Deaths, Restaurant Openings Associated with More Cases and Deaths. This CDC study looked at county-level data, comparing counties that issued mask mandates (either anywhere outside the home or indoors in businesses and restaurants) and restaurant closures. They compared COVID-19 cases and deaths during the one to 20 days prior to implementation of the policies, and periods after the policies were or were not enacted. They used multivariate analysis to adjust for other kinds of bans, such as bans on gatherings and stay at home orders and various combinations of mask and restaurant policies. They found consistent and steady reductions in cases and deaths in counties that implemented mask mandates, steady increases in cases at one to 100 days after implementation, and deaths at 61 to 100 days after implementation in counties allowing on-premises dining. This should come as a surprise to no one.

Written by Mark H. Ebell MD, MS, on March 9, 2021. (Source: Guy GP, Lee FC, Sunshine G, et al. Association of state-issued mask mandates and allowing on-premises restaurant dining with county-level COVID-19 case and death growth rates – United States, March 1–December 31, 2020. MMWR. 2021;70(10):350-354.)

March 17, 2021, Research Update

Low Rate of Markers of Cardiac Pathology in Professional Athletes Who Contract COVID. We previously reported on a study of college athletes who contracted COVID and had cardiac assessments (Research Brief from Jama Cardiol. posted September 18, 2020). We now have another study of professional athletes in North America (Major League Baseball, National Hockey League, National Football League, Major League Soccer, and the men’s and women’s professional basketball associations). In these leagues, with differential approaches to monitoring for COVID, all athletes who have a positive test for COVID-19 are isolated and then screened after resolution of symptoms for cardiac involvement: serum troponin, electrocardiography (ECG), and resting transthoracic echocardiography. If the athlete had an abnormal screening test, they were referred for cardiac magnetic resonance imaging (MRI) or a stress echocardiography. The authors of this paper do not tell us how many athletes were tested or the number of COVID tests that were performed, but they report that 789 athletes had a positive test (PCR or antibody). Only 12 of the athletes were women, and 42% had no symptoms. Among these athletes, 30 (3.8%) had an abnormal cardiac test, six (0.8%) had elevated troponin levels, 10 (1.3%) had abnormal ECGs, and 20 (2.5%) had an abnormal echocardiography. Most (27 of the 30 or 90%) of these athletes underwent cardiac MRI, and 15 underwent both cardiac MRI and stress echocardiography. Among the three who didn’t undergo cardiac MRI, all stress tests were normal. Five of the 27 tested had abnormal cardiac MRIs; three had myocarditis, and two had pericarditis. All five were kept from sports participation. The remaining 25 were allowed to return to play, and, so far, none have experienced any cardiac events.

Written by Henry C. Barry, MD, MS, on March 5, 2021. (Source: Martinez MW, Tucker AM, Bloom OJ, et al. Prevalence of inflammatory heart disease among professional athletes with prior COVID-19 infection who received systematic return-to-play cardiac screening [published online March 4, 2021]. JAMA Cardiol. 2021. //jamanetwork.com/journals/jamacardiology/fullarticle/2777308)

March 16, 2021, Research Update

Underpowered RCT Finds Five Days of Ivermectin Does Not Significantly Decrease Duration of Symptoms in Outpatients with COVID-19. The latest guidelines from the World Health Organization (//www.bmj.com/content/370/bmj.m3379) and the Infectious Disease Society of America (//www.idsociety.org/practice-guideline/covid-19-guideline-treatment-and-management/) do not endorse the use of ivermectin in the treatment of patients with COVID-19. Their recommendations are based on systematic reviews that find the data are generally at high to serious risk of bias and that the data on mortality or symptom improvement is sparse. Since the release of their guideline updates in February 2021, this study has appeared in which the researchers randomized symptomatic outpatient adults with mild laboratory-confirmed COVID to receive either five days of ivermectin (300 μg per kg per day, n = 238) or placebo (n = 238). There is some noise in the implementation of this study. First, it was originally designed to detect a 2-point difference on a standard 8-point clinical severity scale. When they found that the expected rate of deterioration was much lower than anticipated, they altered the outcome of interest to symptom duration. For this, they recruited 400 patients, enough to detect a three-day difference. An additional hiccup is that among the 476 they randomized, they excluded 38 persons in each group due to an “error in labeling.” Post-randomization modifications are worrisome sources of bias. The time to resolution of symptoms among those they analyzed was 10 days for the ivermectin-treated patients and 12 days for the placebo-treated patients, but this two-day difference was not statistically significant. Many would say, however, that this is an important difference. Additionally, the authors report that only 2% of the ivermectin-treated patients and 3.5% of the placebo-treated patients experienced the aforementioned 2-point deterioration (they had anticipated 18%). Neither treatment was well tolerated: 77% of the ivermectin-treated patients reported adverse events compared with 81% of those treated with placebo. Only two people in each treatment group experienced serious adverse events, and 7.5% in the ivermectin-treated group and 2.5% in the placebo group discontinued treatment because of an adverse event. This study should not convince either the World Health Organization or the Infectious Disease Society of America to alter their recommendations.

Written by Henry C. Barry, MD, MS, on March 5, 2021. (Source: López-Medina E, López P, Hurtado IC, et al. Effect of ivermectin on time to resolution of symptoms among adults with mild COVID-19: a randomized clinical trial [published online March 4, 2021]. JAMA. 2021. //jamanetwork.com/journals/jama/fullarticle/2777389)

March 15, 2021, Research Update

Effectiveness of BNT162b2 mRNA Covid-19 Vaccine in Israel’s Nationwide Mass Vaccination Program Is Similar to Clinical Trial Data. A recurring tension in research is whether the ideal circumstances of a clinical trial generate data that can reflect the vagaries of real-world performance. These authors conducted an observational study from Israel’s national mass vaccination program to assess short-term vaccine effectiveness of the Pfizer-BioNTech clinical trial of its mRNA vaccine (BNT162b2) in people older than 16 years of age and with no prior history of a positive PCR for SARS-CoV-2. From the databases of one of the largest integrated healthcare systems in Israel, they identified persons who received the BNT162b2 vaccine between the start of the national vaccine campaign (December 20, 2020) until February 1, 2021. From this database, they matched unvaccinated persons on several factors: age, sex, religious sector (general Jewish, Arab, or ultra-Orthodox Jewish), neighborhood of residence, history of prior influenza vaccination, pregnancy, and number of comorbid conditions. Ultimately, from more than 1 million potential vaccinated people, they were able to create 596,618 matched pairs. Because of high vaccination rates among elderly persons, the pool of unvaccinated persons was smaller, resulting in controls that were slightly younger and had fewer comorbid conditions. Nonetheless, for the period beginning seven days after receiving the second dose, they estimate the vaccine effectiveness at 92% for preventing PCR-confirmed infections, 94% for preventing symptomatic infections, 87% effective for preventing hospitalizations, and 92% for preventing severe COVID. They were unable to provide mortality data for the fully immunized people. They also reported data for days 14 to 20 after the first dose. Those showed lower overall effectiveness compared with full dosing, but there was still clear protection, including 72% effectiveness for preventing death. These data show real-world effectiveness that rivals the data from clinical trials.

Written by Henry C. Barry, MD, MS, on February 25, 2021. (Source: Dagan N, Barda N, Kepten E, et al. BNT162b2 mRNA Covid-19 vaccine in a nationwide mass vaccination setting [published online February 24, 2021]. N Engl J Med. 2021. //www.nejm.org/doi/full/10.1056/NEJMoa2101765)

March 8, 2021, Research Update

Real-World Evaluation of Rapid Antigen Test Accuracy. Manufacturer reports of the accuracy of rapid antigen tests are inflated by using a diagnostic case-control design, highly trained personnel, and use of banked samples from before the pandemic as negative samples. They routinely cite sensitivities greater than 95% and specificity of 99% to 100%. But how do these perform in the real world, where sampling may not be perfect, and patients present with clinically suspected COVID-19 or who are asymptomatic? Two studies completed this evaluation, comparing the rapid antigen tests with reference laboratory PCR. One study evaluated the Abbott BinaxNOW rapid antigen in the Pima County, Arizona Health Department. In 827 symptomatic patients, 80% of whom were within seven days of the onset of symptoms, sensitivity was 64%, and specificity was 100%. In 2,592 asymptomatic persons, sensitivity was only 36% whereas specificity was 99.8%. A second study evaluated the Quidel Sofia SARS FIA test in patients at a Wisconsin urgent care center. In patients symptomatic five days or less, sensitivity was 82%, and specificity was 100%. However, if patients were symptomatic more than five days, sensitivity dropped to 55% and specificity to 97.3%. We previously reported a CDC study (Research Brief posted February 5, 2021) that found 80% sensitivity in symptomatic outpatients but only 41% sensitivity in asymptomatic outpatients at the University of Wisconsin. Taken together, the sensitivity of these rapid antigen tests in recently symptomatic patients is 80%, in less recently symptomatic patients it is closer to 60%, and in asymptomatic patients it is only about 40%. On the other hand, positive tests have a high specificity and positive likelihood ratio, so a positive result can be believed. Using rapid antigen tests to “clear” asymptomatic people prior to an event such as a vacation or birthday party will miss most infected persons.

Written by Mark H. Ebell, MD, MS, on February 21, 2021. (Sources: Beck ET, Paar W, Fojut L, et al. Comparison of the Quidel Sofia SARS FIA test to the Hologic Aptima SARS-CoV-2 TMA test for diagnosis of COVID-19 in symptomatic outpatients. J Clin Microbiol. 2021;59(2):e02727-20 and Prince-Guerra JL, Almendares O, Nolen LD, et al. Evaluation of Abbott BinaxNOW rapid antigen test for SARS-CoV-2 infection at two community-based testing sites—Pima County, Arizona, November 3-17, 2020. MMWR Morb Mortal Wkly Rep. 2021;70(3):100-105).

March 5, 2021, Research Update

Azithromycin Does Not Improve Outcomes in Patients Hospitalized with COVID (RECOVERY). The RECOVERY trial is an open-label randomized trial designed to compare multiple different medications in treating patients hospitalized with COVID-19. In this report, the researchers randomized 2,582 patients to receive azithromycin (500 mg once daily for 10 days or until discharge) and 5,181 patients to usual care alone. The 28-day mortality was the same (22%) for each treatment group. They also found no significant difference in length of stay or survival to discharge. Additionally, they found no improved outcomes in patients on mechanical ventilation. Most of the RECOVERY trial wings have been halted; however, the following are still underway: REGN-COV2, aspirin, and colchicine. One final point. I copied the official list of collaborators into a spreadsheet and counted 5,369 names! Not bad, at least one investigator for every control patient. This has to be a record.

Written by Henry C. Barry, MD, MS, on February 21, 2021. (Source: RECOVERY Collaborative Group. Azithromycin in patients admitted to hospital with COVID-19 (RECOVERY): a randomised, controlled, open-label, platform trial. Lancet. 2021;397(10274):605-612.)

March 4, 2021, Research Update

Update of WHO “Living” Guideline on Drugs for Treating Patients with COVID-19. The World Health Organization (WHO) created a standing guideline panel comprising content experts, clinicians, patients, and methodologists who use modern guideline development methods to evaluate the evidence related to managing patients with COVID-19 and to distill the data into a cohesive set of recommendations. This is their third update, which replaces the one issued in November 2020. The online version of the guideline has useful interactive graphics summarizing the supporting data and their final recommendations. The guideline makes strong recommendations against using hydroxychloroquine and against using lopinavir-ritonavir in patients with COVID-19. Additionally, due to low-quality evidence, the WHO makes a weak recommendation against using remdesivir. Finally, the WHO makes a strong recommendation for the use of corticosteroids only in patients hospitalized with severe or critical illness. Still, a fairly simple set of recommendations!

Written by Henry C. Barry, MD, MS, on February 16, 2021. (Source: Siemieniuk R, Rochwerg B, Agoritsas T, et al. A living WHO guideline on drugs for covid-19. Updated February 2020. BMJ. 2020;370;m3379.)

March 1, 2021, Research Update

Johnson & Johnson (Janssen Biotech) Vaccine Efficacy 66% in FDA Report. The FDA has released its report on the efficacy and safety of the Ad26.COV2.S vaccine developed by Janssen Biotech, a subsidiary of Johnson & Johnson. The vaccine was developed using a “replication incompetent” adenovirus that was encoded with a variant of the SARS-CoV-2 spike protein; this is a more traditional vaccine development approach than the novel mRNA approach used by Pfizer and Moderna. The vaccine was evaluated in 43,783 adults, and data for efficacy are reported for 39,058 adults. There was good racial and ethnic diversity; most of the patients were in the United States (44%) or Latin America (40.9%), with 15% in South Africa. Efficacy was 66.9% (95% CI, 59.0% to 73.4%) overall for preventing confirmed moderate to severe COVID-19 occurring at least 14 days after vaccination. Efficacy was 76.3% (95% CI, 61.6% to 86.0%) for people 60 years or older and 63.7% (95% CI, 53.9% to 71.6%) for those 18 to 59 years of age, although the confidence intervals overlap. Efficacy varied by region, with 74.4% efficacy seen in the United States, 64.7% in Latin America, and only 52.0% in South Africa, likely due to the South African variant. For prevention of severe/critical COVID-19, the vaccine had 76.7% overall efficacy. Results were similar for prevention of cases that required hospitalization, ICU admission, mechanical ventilation, or ECMO (75.0%), although numbers are small and confidence intervals are broad. All seven deaths attributed to COVID-19 occurred in the placebo group and were in South Africa; there were fewer all-cause deaths in the vaccine group in those vaccinated at least 14 days before (3 vs. 15). Similar to the Pfizer and Moderna vaccines, adverse effects included mild and transient injection site pain (49%), headache (39%), fatigue (38%), and myalgias (33%). Less than 2% had a reaction that was classified as grade 3 in severity, and no grade 4 reactions occurred. A single episode of a serious hypersensitivity reaction occurred, but it was not anaphylaxis. Important advantages of this vaccine include that it is a single dose and that it can be stored for up to three months at 2ºC to 8ºC. 

Written by Mark H. Ebell MD, MS, on February 24, 2021. (Source: FDA briefing document: Janssen Ad26.COV2.S vaccine for the prevention of COVID-19. Prepared for the Vaccines and Related Biological Products Advisory Committee Meeting, February 26, 2021. ,).

February 26, 2021, Research Update

Updated Infectious Disease Society of America Guideline on Managing Persons with COVID-19 Infections. The Infectious Disease Society of America (IDSA) used modern guideline development methods that included a multidisciplinary team of clinicians, public health workers, and methodologists who developed critical PICO-format questions and systematic searches of the research. In this most recent update, they made 19 recommendations for managing patients with confirmed COVID-19. The panel made strong recommendations against using anti-malarials, alone or in combination with macrolide antibiotics. Additionally, they recommend against using lopinavir/ritonavir. The panel made a strong recommendation to use dexamethasone (or equivalent dose of other glucocorticoids if dexamethasone is not available) in critically ill hospitalized patients and a conditional recommendation for use in patients hospitalized with severe COVID-19 infections and against its use in hospitalized patients with nonsevere illness and who do not need oxygen. Based on current data, the IDSA recommends against using tocilizumab in hospitalized patients unless they are severely or critically ill (conditional recommendation). It also makes a conditional recommendation in favor of using remdesivir in patients hospitalized with severe illness and those using supplemental oxygen but not on ventilators. The IDSA recommends against using remdesivir in hospitalized patients not needing oxygen. The panel recommends against using bamlanivimab in hospitalized patients and conditionally recommends against its routine use and against routinely using casirivimab/imdevimab in ambulatory patients (outside of high-risk patients who receive counseling for these agents). The IDSA also made a conditional recommendation when managing hospitalized patients with severe illness who have contraindications to corticosteroids that patients receive baricitinib with remdesivir rather than remdesivir alone. Among the additional recommendations, the IDSA recommends against using the following specific agents except in the context of a clinical trial: convalescent serum, famotidine, baricitinib plus remdesivir plus corticosteroids, and ivermectin.

Written by Henry C. Barry, MD, MS, on February 23, 2021. (Source: Infectious Diseases Society of America. ISDA guidelines on the treatment and management of patients with COVID-19. Updated February 22, 2021. //www.idsociety.org/practice-guideline/covid-19-guideline-treatment-and-management/)

February 25, 2021, Research Update

Initial Public Report of Janssen/Johnson & Johnson COVID-19 Vaccine Effectiveness. (CAUTION: This data is from a press release; there is no scientific publication as of February 14, 2021. EUA application was submitted to FDA February 4, 2021.) In a press release on January 29, 2021, Johnson & Johnson presented the initial results of the Janssen COVID Ad26.COV2-S vaccine effectiveness trials. The trials took place in the United States, Latin American, and South Africa. This phase III trial, called ENSEMBLE, randomized 43,783 participants to a single injection of vaccine or placebo. The primary endpoints were vaccine efficacy starting at 14 days and 28 days after the injection, although they report results only for 28-day efficacy in this initial press release. There were 468 symptomatic cases overall of COVID-19 at the time of this initial analysis. The vaccine was 66% effective overall in preventing moderate to severe COVID-19 28 days after vaccination. Vaccine efficacy in preventing moderate to severe COVID-19 infection ranged from 72% in the United States to 66% in Latin America and 57% in South Africa. The vaccine was 85% effective in preventing severe disease across all locations. Efficacy against severe disease increased over time, with no cases in vaccinated participants reported after day 49. There were no hospitalizations or deaths starting 28 days after vaccination. They do not report the effectiveness of prevention of any symptomatic COVID-19 infection, whereas other manufacturers have. Adverse events were not reported in detail, although the report says, based on a database of over 200,000 vaccine trial participants, that this vaccine adverse-event rate was similar to other vaccines. Serious events were no more common in the vaccination group compared with the placebo group, and no cases of anaphylaxis occurred. 

Written by John Hickner on February 14, 2021. (Source: Johnson & Johnson. Johnson & Johnson announces single-shot Janssen COVID-19 vaccine candidate met primary endpoints in interim analysis of its phase 3 ENSEMBLE trial. January 20, 2021. )

February 24, 2021, Research Update

Initial Results of the Oxford-AstraZeneca COVID-19 Vaccine. AstraZeneca published the initial results of the ChAdOx1 nCoV-19 vaccine, a chimpanzee adenoviral-vectored vaccine with full-length SAR-CoV-2 spike insert, on February 1, 2021. They presented data from phase III randomized trials of the ChAdOx1 nCoV-19 vaccine in the United Kingdom and Brazil, and phase I/II clinical trials in the United Kingdom and South Africa. The protocol called for participants 19 years and older to be randomized to receive two doses of the vaccine or two placebo injections 12 weeks apart, although there was variability in when the second dose was actually administered, allowing an analysis of effectiveness by timing of the second dose. In the UK trial, a subset of participants received a lower dose of the ChAdOx1 nCoV-19 for the first dose. The investigators randomized 17,177 participants, and 32 had a symptomatic infection greater than 14 days after the second dose. The overall vaccine efficacy was 66.7% (95% CI, 57.4% to 74.0%) in preventing symptomatic COVID-19 infection. No hospitalizations or deaths occurred in the vaccinated group, and there were 15 hospitalizations in the control group.  A subgroup analysis found the vaccine to be 63.1% effective (95% CI, 51.8% to 71.7%) in preventing symptomatic COVID-19 for those receiving two standard doses and 80.7% effective (95% CI, 62.1% to 90.2%) for those who received the low dose for the initial injection. Interestingly, efficacy was higher with a longer interval between the first and second vaccination: 82.4% when the second injection was given more than 12 weeks after the first injection compared with 54.9% when the second injection was given less than six weeks after the first injection. To determine whether the initial dose provided some protection, they analyzed the effectiveness of the first dose starting 22 days after the initial injection. The first dose was 76% effective from day 22 to day 90 in preventing symptomatic infections, although it was not effective in preventing asymptomatic infections during this time period. 

Written by John Hickner, MD, MS, on February 14, 2021. (Source: Voysey M, Costa Clemens SA, Madi SA, et al.; Oxford COVID Vaccine Trial Group. Single-dose administration and the influence of the timing of the booster dose on immunogenicity and efficacy of ChAdOx1 nCoV-19 (AZD1222) vaccine: a pooled analysis of four randomised trials [published online February 1, 2021]. Lancet. 2021. //www.thelancet.com/journals/lancet/article/PIIS0140-6736(21)00432-3/fulltext)

February 23, 2021, Research Update

Early Initiation of Anticoagulation Associated with Reduced Mortality in Hospitalized COVID-19. This was a retrospective cohort study of 4,297 patients hospitalized with COVID-19 between March 1, 2020, and July 31, 2020, in the U.S. Veterans Affairs health system. The researchers compared patients who had an initial dose of prophylactic anticoagulation in the first 24 hours of hospitalization with those who did not, with a primary outcome of 30-day mortality. The analysis used a propensity score-matched approach to adjust for comorbidities, tobacco use, medications, and laboratory results. This was a very sophisticated analysis that is as close as you can get to a randomized trial without doing one. Overall, 84.4% of patients received anticoagulation within 24 hours of admission; of those receiving it, most received subcutaneous heparin (30%) or enoxaparin (69%). At baseline, patients receiving early thromboprophylaxis were more likely to have oxygen saturation < 93%, tachycardia, and fever but had fewer comorbidities. In the propensity score-matched analysis, the 30-day mortality was significantly lower in the group receiving early thromboprophylaxis (14.3% vs. 18.7%; hazard ratio 0.73; 95% CI, 0.66 to 0.81; NNT = 23).

(Written by Mark H. Ebell MD, MS, on February 13, 2021. Source: Rentsch CT, Beckman JA, Tomlinson L, et al. Early initiation of prophylactic anticoagulation for prevention of coronavirus disease 2019 mortality in patients admitted to hospital in the United States: cohort study. BMJ. 2021;372:n311.)

February 19, 2021, Research Update

Moderna COVID-19 Vaccine Is 94% Effective. The mRNA-1273 SARS-CoV-2 vaccine developed by Moderna is the second of two messenger-RNA vaccines to be approved by the FDA for emergency use, approved in December 2020. This randomized, placebo-controlled trial, the basis for approval of this vaccine, enrolled 30,420 individuals. Participants were given two injections 28 days apart of either the vaccine or placebo. Starting 14 days after the second injection, symptomatic COVID-19 infection occurred in 185 participants in the placebo group and in 11 participants in the vaccinated group. Therefore, the vaccine effectiveness for preventing symptomatic COVID-19 infection was 94.1% (95% CI, 89.3 to 96.8%). The results were similar in subgroups such as those who had evidence of infection at baseline (2.2% of participants) and those 65 years of age and older. Results were similar 14 days after the first dose, indicating early protection even before the second dose was administered. There were 30 severe COVID infections, all in the placebo group, showing a high level of protection against severe COVID disease. Local reactions to the first injection occurred in 84% of participants, including swelling, redness, and pain, and in 89% after the second injection. Systemic adverse events occurred more often in the mRNA-1273 group than in the placebo group after the first dose (54.9% vs. 42.2%) and the second dose (79.4% vs. 36.5%). The frequency of serious adverse events was 0.6% in both the placebo and the vaccine group.

Written by John Hickner, MD, MS, on February 12, 2021. (Source: Baden LR, El Sahly HM, Essink B, et al.; COVE Study Group. Efficacy and safety of the mRNA-1273 SARS-CoV-2 vaccine. N Engl J Med. 2021;384(5):403-416.)

February 18, 2021, Research Update

Colchicine Did Not Significantly Decrease Deaths or Hospitalizations but Was Associated with More Pulmonary Emboli in Outpatients with COVID-19. This report of the COLCORONA study data comes from a preprint server (not yet peer-reviewed). They recruited patients with COVID who were not yet hospitalized and considered not likely to need hospitalization. The patients could be diagnosed clinically or with PCR testing. The patients also had to have at least one high-risk criterion: 70 years of age or older, obesity, diabetes, uncontrolled hypertension, known respiratory disease, known heart failure, known coronary disease, fever of at least 38.4°C (101°F) within the last 48 hours, dyspnea, and various laboratory parameters. The investigators terminated the study after 75% of the planned enrollment “due to logistical issues related to maintaining the central study call center, as well as the need to provide healthcare systems with study results in a timely fashion given the state of the COVID-19 pandemic.” The investigators randomized 4,488 patients to receive colchicine (0.5 mg twice daily for the first three days followed by once daily) or placebo. Using intention-to-treat, they found the rate of a composite 30-day outcome (death or COVID hospitalization) was similar in both groups of patients (4.7% vs. 5.8%, respectively; p = 0.08). There was also no difference in the rate of each of these events individually (0.2% vs. 0.4% deaths; 4.5% vs. 5.7% hospitalizations, respectively). When they restricted their analysis to the 4,159 patients with a positive PCR result, the differences became statistically significant for the composite outcome (4.6% vs. 6.0%, respectively; NNT = 71; 95% CI, 36 to 1,974), driven mainly by a barely statistically significant reduction in hospitalizations (4.5% vs. 5.9%). In the real world, clinicians seeing a high-risk patient with symptoms and a COVID exposure are more likely to apply data from the main analysis and not this last one. More colchicine-treated patients developed pulmonary emboli (0.5% vs. 0.1%; NNTH = 244; 95% CI, 124 to 1,163).

Written by Henry C. Barry, MD, MS, on February 12, 2021. (Source: Tardif JC, Bouabdallaoui N, L’Allier PL, et al.; COLCORONA Investigators. Efficacy of colchicine in non-hospitalized patients with COVID-19 [medRxiv preprint, not peer-reviewed January 27, 2021]. //www.medrxiv.org/content/10.1101/2021.01.26.21250494v1)

February 17, 2021, Research Update

Preliminary Studies Suggest an Increased Mortality Rate with SARS-CoV-2 B.1.1.7 Variant. Several United Kingdom government-sponsored analyses of COVID-19 mortality data in the United Kingdom suggest that infection with the B.1.1.7 SARS-CoV-2 strain results in a higher mortality rate compared with the original strain. In one analysis, 384 of 2,583 deaths were due to the B.1.1.7 strain, and the relative hazard of death within 28 days was 1.35 (95% CI, 1.08 to 1.68). Another case-control analysis of community testing data linked to mortality data found a mortality hazard ratio of 1.91 (95% CI, 1.35 to 2.71). In a matched cohort analysis of 14,939 people infected with the B.1.3.3.7 strain compared with 15,555 infected with the original strain, there were 104 deaths (0.2%) compared with 65 deaths (0.1%), a mortality ratio of 1.65 (95% CI, 1.21 to 2.25).  However, in a hospital-based study, the risk of death from the B.1.1.7 strain was not statistically increased (odds ratio 0.63; 95% CI, 0.20 to 1.69), although this small study could not exclude an excess mortality of up to 69%. More and better data are needed to confirm whether this new strain is more deadly.

Written by John Hickner, MD, MS, on February 11, 2021. (Source: Iacobucci, G. Covid-19: New UK variant may be linked to increased death rate, early data indicate. BMJ. 2021;372:n230.)

February 16, 2021, Research Update

SARS-CoV-2 B.1.1.7 Strain Emergence and Transmission in the United States. Since the onset of the COVID-19 pandemic, there has been concern about novel SARS-CoV-2 variants emerging that are more transmissible and possibly more deadly. This research paper tracks the emergence and accelerating transmission of the B.1.1.7 (otherwise known as United Kingdom) variant of SARS-CoV-2 in the United States. This strain emerged in the United Kingdom in September 2020 and is now the most common strain in the United Kingdom, accounting for 90% of COVID-19 infections during the last week of November 2020. The investigators sequenced 212 B.1.1.7 SARS-CoV-2 genomes collected during December 2020 and January 2021 from testing facilities in the United States, and they calculated the rate of growth and proportion of total COVID-19 infections due to this strain. They found that during December and January, the proportion of B.1.1.7 variants isolated in the United States was increasing at a rate similar to the rate in other countries, a doubling time of approximately 10 days. They determined the transmission rate was about 40% faster than the original strain in the United States. The proportion of B.1.1.7 cases reached an average of 2.1% in the United States by the last week of January 2021, although the rate varied by state. The investigators project that the B.1.1.7 variant is likely to become the dominant strain of SARS-CoV-2 in the United States. Preliminary reports from the United Kingdom suggest a slightly higher mortality rate for the B.1.1.7 variant compared to the original strain. 

Written by John Hickner, MD, MS, on February 11, 2021. (Source: Washington NL, Gangavarapu K, Zeller M, et al. Genomic epidemiology identifies emergence and rapid transmission of SARS-CoV-2 B.1.1.7 in the United States [medRxiv preprint, not peer-reviewed February 7, 2021]. //www.medrxiv.org/content/10.1101/2021.02.06.21251159v1)

February 9, 2021, Research Update

Anaphylaxis to the Moderna COVID-19 Vaccine Occurs in one in 400,000 Doses. The CDC Vaccine Adverse Event Reporting System (VAERS) detected 21 cases of anaphylaxis after the administration of 1,893,360 first doses of the Pfizer-BioNTech COVID-19 vaccine (11.1 cases per million doses, which is one per 90,000 doses). However, after publication of that report and more experience with the Pfizer BioNTech COVID-19 vaccine, the CDC revised this estimate downward on January 27, 2021, to five per 1,000,000 doses, which is one in 200,000 doses. This new report from the CDC VAERS summarizes allergic reactions to 4,041,396 first doses of the Moderna COVID-19 vaccine. Adverse events were reported to VAERS for 1,266 doses administered, 0.03%. Ten cases of anaphylaxis were reported, a rate of one per 400,000 doses, (This is approximately half the rate of anaphylaxis compared with the Pfizer-BioNTech vaccine.) Nine of the 10 individuals with anaphylaxis to the Moderna vaccine had a history of allergies or allergic reactions, seven due to medications. Five had a prior history of anaphylaxis. The onset was within 15 minutes of vaccine administration in nine individuals and after 30 minutes in one case. Six patients were hospitalized, and four required endotracheal intubation. No patients died. For the Moderna vaccine, all individuals with anaphylaxis were women; for the Pfizer vaccine, 19 of 21 were women. Limitations of these data are that not all adverse events are reported to VAERS and that this is a relatively small number of cases for making generalizations. Nonetheless, it seems clear that anaphylaxis to these two COVID vaccines occurs almost exclusively in those with a prior history of allergies to drugs or food and is much more common in women than men.

Written by John Hickner, MD, MS, on January 26, 2021. (Source: Allergic reactions including anaphylaxis after receipt of the first dose of Moderna COVID-19 vaccine — United States, December 21, 2020–January 10, 2021. MMWR. 2021;70(4):125-129.)

Addendum: “During December 14, 2020 through January 18, 2021, a total of 9 943 247 doses of the Pfizer-BioNTech vaccine and 7 581 429 doses of the Moderna vaccine were reported administered in the US.... Continued safety monitoring of mRNA COVID-19 vaccines in the US has confirmed that anaphylaxis following vaccination is a rare event, with rates of 4.7 cases/million [1 in 212,000 doses] Pfizer-BioNTech vaccine doses administered and 2.5 cases/million [1 in 400,000 doses] Moderna vaccine doses administered, based on information through January 18, 2021.”

(Source: Shimabukuro TT, Cole M, Su JR. Reports of anaphylaxis after receipt of mRNA COVID-19 vaccines in the US—December 14, 2020–January 18, 2021 [published online February 12, 2021]. JAMA. //jamanetwork.com/journals/jama/fullarticle/2776557)

February 8, 2021, Research Update

Anaphylaxis to the Pfizer-BioNTech COVID-19 Vaccine Occurs in One in 200,000 doses. During December 14 to 23, 2020, monitoring by the Vaccine Adverse Event Reporting System (VAERS) detected 21 cases of anaphylaxis after administration of a reported 1,893,360 first doses of the Pfizer-BioNTech COVID-19 vaccine (11.1 cases per million doses, which is one in 90,000 doses). A total of 4,393 (0.2%) adverse events after receipt of Pfizer BioNTech COVID-19 vaccine had been submitted to the VAERS. Among these, 175 case reports were identified for further review as possible cases of severe allergic reaction, and 21 were judged to be anaphylaxis. Although 71% of these 21 occurred within 15 minutes of vaccine administration, one occurred 150 minutes after vaccine administration. Seventeen of these people had a history of allergies, and seven had a prior history of anaphylaxis. Seventeen (81%) of 21 patients with anaphylaxis had a documented history of allergies or allergic reactions, including allergies to drugs or medical products, foods, and insect stings; seven (33%) patients had experienced an episode of anaphylaxis in the past. Nineteen of the 21 cases occurred in women. Seventeen were treated in the emergency department, and eight were hospitalized. All survived. After publication of this report and more experience with the Pfizer BioNTech COVID-19 vaccine, the CDC revised this estimate downward on January 27, 2021, to five per 1,000,000 doses, which is one in 200,000. 

Written by John Hickner, MD, MS, on January 26, 2021. (Source: Allergic reactions including anaphylaxis after receipt of the first dose of Pfizer-BioNTech COVID-19 Vaccine — United States, December 14–23, 2020. MMWR Morb Mortal Wkly Rep. 2021;70(2):46-51.)

February 5, 2021, Research Update

Rapid Antigen Tests Much Less Sensitive in Asymptomatic People (40% vs. 80%). This study is from the University of Wisconsin, an institution with a long history of excellent work in respiratory disease pathogen detection. They identified 1,098 paired midturbinate nasal swabs at two universities in Wisconsin between September 28, 2020, and October 9, 2020; 871 were from asymptomatic participants, and 227 were from symptomatic participants. The vast majority were students, although some older staff and faculty also participated (range 17 to 64 years of age). Each participant had one swab tested using rt-PCR, the reference standard, and one swab tested using the Sofia SARS Antigen Fluorescent Immunoassay (Quidel Corp.). A total of 2.0% of swabs from asymptomatic participants were PCR positive, compared with 18% from symptomatic participants. Specificity was almost identical for the two groups (98.4% for asymptomatic and 98.9% for symptomatic), but sensitivity was much lower for asymptomatic persons (41% vs. 80%). Of 21 positive rapid antigen tests in asymptomatic persons, 14 were false positives based on the PCR (false positive rate 67%), so it is important to confirm a positive rapid antigen test with PCR. In the symptomatic population, the positive and negative predictive values were 94% and 96% respectively.

Written by Mark H. Ebell MD, MS, on January 26, 2021. (Source: Pray IW, Ford L, Cole D, et al.; CDC COVID-19 Surge Laboratory Group. Performance of an antigen-based test for asymptomatic and symptomatic SARS-CoV-2 testing at two university campuses—Wisconsin, September–October 2020. MMWR Morb Mortal Wkly Rep. 2021;69(5152):1642-1647.)

February 4, 2021, Research Update

Bamlanivimab Appears to Decrease Hospitalizations or Emergency Department Visits in Outpatients with Mild to Moderate COVID-19. In an earlier research brief from NEJM posted on December 15, 2020, we presented an interim analysis of the ongoing BLAZE-I trial based on 452 outpatients and data submitted to the FDA based on 465 outpatients. The authors now present their final data on 577 ambulatory patients with confirmed mild to moderate COVID-19 infections. Over the first two months of the study, the researchers randomized patients to a single infusion of placebo or one of three doses of bamlanivimab (700 mg, n = 101; 2,800 mg, n = 107; 7,000 mg, n = 101), and over the last 10 days they added another wing where patients received a combination of bamlanivimab plus etesevimab (2,800 mg each, n = 112) or placebo. A total of 156 patients received placebo infusions. Their main outcome was viral load at 11 days, but the authors report having nine different primary and secondary outcomes and have an additional section for 75 more exploratory analyses. Having this many different outcomes increases the possibility that one outcome, by chance alone, might appear to be meaningful. The authors appropriately caution against using the exploratory analyses for decision-making. Having said that, the secondary outcome of hospitalizations or emergency department visits occurred in 5.8% of placebo-treated patients and in 1.4% of patients receiving bamlanivimab either as monotherapy or as combination therapy (NNT = 23; 95% CI, 11 to 85). The authors do not report whether these differences were statistically important; however, this reviewer’s Chi-square estimate results in a p-value of 0.004. This estimate is not very robust given the small number of outcomes and the relatively large number of outcomes assessed. They also report that viral loads are reduced, but this is less interesting. One final comment. Why was the final report published in JAMA when the interim analysis was published in NEJM?

Written by Henry C. Barry, MD, MS, on January 24, 2021. (Source: Gottlieb RL, Nirula A, Chen P, et al. Effect of bamlanivimab as monotherapy or in combination with etesevimab on viral load in patients with mild to moderate COVID-19: a randomized clinical trial [published online January 21, 2021]. JAMA. 2021. //jamanetwork.com/journals/jama/fullarticle/2775647)

February 3, 2021, Research Update

In Hospitalized Patients with COVID Who Require Supplemental Oxygen, Adding Tocilizumab Reduces Progression to Mechanical Ventilation. These investigators enrolled patients who were hospitalized with COVID-19 pneumonia with a blood oxygen saturation below 94% on ambient air but who were not yet receiving noninvasive positive airway pressure ventilation or mechanical ventilation. The trial was funded, conducted, analyzed, and authored with support from Genentech, the makers of tocilizumab. Patients were randomized in a 2:1 ratio to receive one or two doses of tocilizumab (n = 249) or placebo (n = 128) along with standard care. The study sites included those with large underserved and racial and ethnic minority populations in order to understand the effect of the drug on these patients. The two groups were balanced at baseline: 59% were men, mean age 56 years, 56% were Hispanic or Latino, 15% were Black, 13% were American Indian or Alaska Natives, and 13% were non-Hispanic White (race or ethnicity was self-reported). Although most patients in both groups received concomitant glucocorticoid or antiviral agents for the treatment of COVID-19 pneumonia, more patients in the placebo group received dexamethasone (55.4% vs. 67.2%). The authors do not state whether this was statistically significant, but it is possible it may have affected outcomes. For the primary efficacy outcome, significantly fewer patients in the tocilizumab group progressed to mechanical ventilation or death by day 28 (12% vs. 19%; hazard ratio 0.56; 95% CI, 0.33 to 0.97; P = .04), although this was driven entirely by less need for mechanical ventilation. This result was consistent when analyzed according to race or ethnic group. There was no effect on mortality alone. Median time to hospital discharge and median time to improvement in clinical status favored tocilizumab by 1.5 and 1.0 days, respectively, but these results were not statistically significant. Adverse events occurred in similar frequency in the two groups.

From the InfoPOEM written by Nita Kulkarni, MD, and reprinted with permission. For the full POEM see . (Source: Salama C, Han J, Yau L, et al. Tocilizumab in patients hospitalized with covid-19 pneumonia. N Engl J Med. 2021;384(1):20-30.)

February 2, 2021, Research Update

Baricitinib Plus Remdesivir Reduces Recovery Time Compared with Remdesivir Alone in Hospitalized COVID-19 Pneumonia Patients. Baricitinib, a disease-modifying antirheumatic drug, is a selective inhibitor of Janus kinase 1 and 2 and acts as an anti-inflammatory agent. In this international trial, investigators randomized hospitalized patients with COVID-19 and evidence of lower respiratory tract infection to remdesivir plus baricitinib (n = 515) or remdesivir plus placebo (n = 518). Remdesivir was administered as a 200-mg IV loading dose, followed by 100 mg IV daily on days 2 through 10 until hospital discharge or death. Baricitinib was given orally as a 4-mg daily dose for 14 days or until hospital discharge. The two groups had similar baseline characteristics (race and ethnic groups were self-reported): Mean age was 55 years, 63% were men, 48% were White, and 51% were Hispanic or Latinx. Disease severity was assessed on an ordinal scale from 1 to 8, with higher scores indicating more severe disease. Overall, two-thirds of the patients were in the moderate disease category at baseline, whereas the rest had severe disease. The primary outcome was time to recovery, defined as the first day after enrollment in which a patient attained scores of 1, 2, or 3 on the ordinal scale. Overall, patients who received remdesivir plus baricitinib recovered a median of one day faster than those who received remdesivir plus placebo (seven days vs. eight days; relative risk [RR] = 1.16; 95% CI, 1.01 to 1.32; P = .03). The outcome was most prominent in patients with a baseline ordinal score of 6 (those receiving noninvasive mechanical ventilation or high-flow oxygen) with a recovery time of 10 days compared with 18 days (RR = 1.51; 95% CI, 1.10 to 2.08). The odds of improvement in clinical status at day 15 were also greater in the remdesivir plus baricitinib group (odds ratio [OR] 1.3; 95% CI, 1.0 to 1.6), again more prominently in patients with a baseline ordinal score of 6 (OR 2.2; 95% CI, 1.4 to 3.6). Serious adverse events were less frequently noted in the remdesivir plus baricitinib group, and no significant mortality difference occurred between the two groups.

From the InfoPOEM written by Nita Kulkarni, MD, and reprinted with permission. For the full POEM see . (Source: Kalil AC, Patterson TF, Mehta AK, et al.; ACTT-2 Study Group Members. Baricitinib plus remdesivir for hospitalized adults with Covid-19 [published online December 11, 2020; updated January 5, 2021]. N Engl J Med. 2020. //www.nejm.org/doi/full/10.1056/NEJMoa2031994)

January 28, 2021, Research Update

Adding Tocilizumab to Standard Care Results in More Deaths at 15 Days Than Standard Care Alone in Patients with Severe or Critical COVID Infections. In a Research Brief posted October 29, 2020, we summarized several studies of tocilizumab, finding the data on its effectiveness unencouraging. In this study from Brazil, the researchers randomized adults with confirmed severe COVID infections to receive standard care (n = 64) or standard care plus tocilizumab (single dose of 8 mg per kg). To be included, the patients had to have pulmonary infiltrates on imaging and had to be receiving supplemental oxygen or mechanical ventilation. The patients also had abnormal levels of two or more biomarkers (D dimer > 2.74 nmol per L [1,000 ng per mL], C reactive protein > 50 mg per L [5 mg per dL], ferritin > 300 μg per L, or lactate dehydrogenase greater than the upper limit of normal). The main outcome was 15-day mortality. The study originally planned to enroll 150 patients, but the data-monitoring board stopped enrollment after 129 patients due to excess deaths in the patients randomized to receive tocilizumab. At day 15, 11 tocilizumab-treated patients had died (17%) compared with two (3%) in those receiving standard care alone (NNTH = 8; 95% CI, 4 to 30).

Written by Henry C. Barry, MD, MS, on January 21, 2021. (Source: Veiga VC, Prats JAGG, Farias DLC, et al. Effect of tocilizumab on clinical outcomes at 15 days in patients with severe or critical coronavirus disease 2019: randomised controlled trial. BMJ. 2021;372:n84.)

January 27, 2021, Research Update

Meta-Analysis: COVID Testing Using Saliva Is as Accurate as Nasopharyngeal Swabs. These authors systematically searched Medline and a preprint server (not yet peer-reviewed) to identify studies directly comparing saliva nucleic acid amplification testing with nasopharyngeal nucleic acid amplification testing. It is likely their analysis omitted studies because they searched only two databases, included only those with at least 20 participants, and excluded studies without random or consecutive samples. Two authors independently retrieved and evaluated studies for inclusion and independently assessed the risk of bias of the included studies. They resolved disagreements by consensus. Ultimately, they included 16 studies (5,922 unique patients), 10 of which came from peer-reviewed studies. Overall, the risk of bias in the included studies was high for patient selection, but the studies were generally at low risk of bias for other criteria. In spite of marked variation in study design, patient selection, reference standards, stage of illness, symptomatic persons, etc., the authors nonetheless chose to pool the data to estimate the sensitivity and specificity of saliva-based testing and for nasopharyngeal-based testing. Overall, the pooled sensitivity and specificity for saliva-based tests were 83.2% (95% credible interval [CrI], 74.7% to 91.4%) and 99.2% (95% CrI, 98.2% to 99.8%), respectively. This was comparable to those for nasopharyngeal-based tests: 84.8% (95% CrI, 76.8% to 92.4%) and 98.9% (95% CrI, 97.4% to 99.8%), respectively. Given the comparable diagnostic accuracy and simplicity, saliva-based testing is very attractive.

Written by Henry C. Barry, MD, MS, on January 16, 2021. (Source: Butler-Laporte G, Lawandi A, Schiller I, et al. Comparison of saliva and nasopharyngeal swab nucleic acid amplification testing for detection of SARS-CoV-2: a systematic review and meta-analysis [published online January 15, 2021). JAMA Intern Med. 2021. //jamanetwork.com/journals/jamainternalmedicine/fullarticle/2775397)

January 26, 2021, Research Update

High-Titer Convalescent Plasma Is Associated with Lower 30-Day Mortality in High-Risk Adults Hospitalized with COVID-19. We have previously reviewed several studies evaluating the use of convalescent plasma in treating patients with COVID-19, including a case series in our very first research brief! Randomized trials: PLACID – Research Brief from BMJ posted October 30, 2020; PlasmAr – Research Brief from NEJM posted December 10, 2020; and INFANT-COVID-19 Group – Research Brief from NEJM posted January 19, 2021. The outcomes have been mixed, but largely underwhelming. The limitations of the PLACID and PlasmAr studies include late administration to more severely ill people, and 29% of the donor samples in the PLACID study had no detectable neutralizing antibodies. In this next study, with many limitations because of its retrospective nature, the authors used a U.S.-based registry of high-risk adults hospitalized with lab-confirmed COVID-19. They were interested in evaluating the association between receipt of convalescent plasma and subsequent 30-day mortality. They also obtained 3,082 samples to assess antibody levels in the transfused plasma and classified the levels as low (561 samples), medium (2,006 samples), or high (515 samples). Among the patients who received high-titer plasma, 22.3% died compared with 27.4% and 29.6% in the medium- and low-titer recipients, respectively. These data alone can show only an association, but there is a clear dose-response that strengthens the association. The timing of administration was highly variable but consistent across each of the three dosing groups (average of four days; range from 0 days to more than a month). The authors do not report data on adverse events. Finally, these data are consistent with the randomized trial conducted by the INFANT-COVID-19 Group, illness severity notwithstanding.

Written by Henry C. Barry, MD, MS, on January 16, 2021. (Source: Joyner MJ, Carter RE, Senefeld JW, et al. Convalescent plasma antibody levels and the risk of death from Covid-19 [published online January 13, 2021]. N Engl J Med. 2021. //www.nejm.org/doi/10.1056/NEJMoa2031893)

January 25, 2021, Research Update

Some Symptoms Persist for at Least Six Months in Many Hospitalized COVID-19 Patients. Several studies have shown that symptoms persist for at least three months in about two-thirds of patients with COVID-19 infections. This follow-up cohort study of patients hospitalized with COVID-19 infection in Wuhan, China, describes patients’ symptoms and lung function abnormalities six months after the onset of the acute infection. Of 2,469 patients with COVID-19 who were discharged from Jin Yin-tan Hospital between January 7 and May 29, 2020, 1,733 were enrolled in the study; 736 patients were not included because they declined to participate (347), died (33), or were not available for a variety of logistical and medical reasons. Enrolled patients completed an in-person interview that included a checklist for physical symptoms and anxiety and depression, a physical exam, laboratory tests, and a six-minute walking test. A subgroup of 390 received follow-up imaging studies.  The median age of the participants was 57; 52% were men, and 48% were women. Of the 1,655 participants for whom survey data was available, 67% reported at least one symptom at follow-up. The most common symptoms were fatigue or muscle weakness (63%), sleep difficulties (26%), pain or discomfort (27%), and anxiety or depression (23%). Twenty-three percent performed under the lower limit of normal for the 6-minute walk test, and 53% had at least one abnormal lung CT pattern. A weakness of this study is the lack of a comparison group without COVID-19 infection. Nonetheless, this study illustrates that a high proportion of people hospitalized with COVID-19 infection have persistent symptoms and abnormal lung CT findings for at least a half year.

Written by John Hickner, MD, MS, on January 14, 2021. (Source: Huang C, Huang L, Wang Y, et al. 6-month consequences of COVID-19 in patients discharged from hospital: a cohort study. Lancet. 2021;397(10270):220-232.)

January 22, 2021, Research Update

At Least 50% of SARS-CoV-2 Transmission Is from Asymptomatic Individuals. The authors of this study used mathematical, probability modeling to estimate the proportion of Covid-19 infections that are due to contact with asymptomatic and presymptomatic individuals infected with the SARS-CoV-2 virus. They used standard probability computations with the following assumptions. (1) The median incubation period is five days, based on a meta-analysis of eight studies from China. (2) Peak infectiousness of the person infected occurs on the fifth day after contact with the infected individual, which is generally the day symptoms begin or one day prior to symptom onset. (3) Infectiousness varies over time and can last up to 10 days. (4) 30% of infected individuals do not develop any symptoms. (5) Asymptomatic individuals are 75% as infectious as symptomatic individuals. Under these baseline assumptions, about 59% of all infections come from asymptomatic individuals; 35% from presymptomatic individuals and 24% from those who remain asymptomatic. In a sensitivity analysis, the authors varied the baseline assumptions, and under any scenario they tested, at least 50% of SARS-CoV-2 transmission was from individuals with no symptoms. This mathematical exercise confirms prior epidemiological studies that suggest at least 40% of transmission of SARS-CoV-2 is from asymptomatic individuals. Therefore, social distancing measures and masks will continue to be important control measures to prevent spread of Covid-19 infection until herd immunity is developed through infections and immunizations.

Written by John Hickner, MD, MS, on January 11, 2021. (Source: Johansson MA, Quandelacy TM, Kada S, et al. SARS-CoV-2 transmission from people without COVID-19 symptoms. JAMA Network Open. 2021;4(1):e2035057.)

January 19, 2021, Research Update

Convalescent Plasma Reduced Progression of COVID-19 in Mildly Ill, Older Adults. Prior studies of convalescent plasma have had mixed results. It does not appear to be effective when given to seriously ill hospitalized patients. This randomized trial tested the effectiveness of convalescent plasma in preventing the progression of COVID-19 disease when given early in the course of illness in older adults. One hundred and sixty adults 65 to 74 years of age who had comorbidities or older than 75 years, regardless of comorbidities, with mild symptoms were randomized to convalescent plasma or placebo infusion within 72 hours of onset of illness. The primary endpoint was severe respiratory disease, defined as a respiratory rate of 30 breaths per minute or more, an O2 saturation of less than 93% on room air, or both. Thirteen of 80 patients (16%) in the treatment group and 25 of 80 (31%) in the placebo group developed severe respiratory distress (relative risk = 0.52; 95% CI, 0.29 to 0.94; p = 0.03) If one eliminates from the analysis the six patients who reached the primary end point prior to receiving the infusion, the relative risk was 0.40 (95% CI, 0.20 to 0.81). The sample size was too small to detect differences in the secondary endpoints of life-threatening respiratory disease and death, although the trend favored the convalescent plasma group because seven patients (9%) had any secondary end point in the convalescent plasma group and 12 patients (15%) in the placebo group did. The investigators conclude that convalescent plasma can prevent progression of COVID-19 in older adults with mild symptoms when administered within 72 hours. Whether it prevents very severe respiratory disease and death is uncertain.

Written by John Hickner, MD, MS, on January 10, 2021. (Source: Libster R, Pérez Marc G, Wappner D, et al.; Fundación INFANT–COVID-19 Group. Early high-titer plasma therapy to prevent severe Covid-19 in older adults [published online January 6, 2021]. N Engl J Med. 2021. //www.nejm.org/doi/full/10.1056/NEJMoa2033700)

January 15, 2021, Research Update

In-Person Instruction at Universities Associated with Increased Incidence of COVID-19 in the Community. Colleges and universities have responded in a range of ways to the COVID-19 pandemic, from going fully online to holding as much instruction in person as possible. This report from the CDC looked at the incidence of new cases in three kinds of counties: counties with a college or university holding in-person instruction, counties with a college or university holding online instruction only, and counties without a college or university. They compared several key metrics during the 21 days prior to the start of classes and the 21 days following the start of classes. Not surprisingly, incidence increased 56% in counties with in-person instruction and decreased in non-university and remote-instruction counties. Test positivity rates increased in in-person instruction counties and decreased in remote instruction and non-university counties. The percentage of counties in each group identified as a hotspot was similar prior to the start of classes, but after classes began it was much higher in the in-person instruction counties than in the remote-instruction or non-university counties (39.2% vs. 18.2% vs. 5.9%, respectively). These effects remained even after adjusting for the influx of students to these counties in the late summer. We do not know how many of the excess cases were in older or more vulnerable people in these communities, but certainly some were. My own university pushed hard for in-person instruction, brought students back to town, and as a result saw more than 4,000 confirmed cases and an estimated 13,000 total cases (largely asymptomatic) in the fall (//www.mdedge.com/familymedicine/article/231476/infectious-diseases/high-proportion-sars-cov-2-infected-university?sso=true). The authors from the CDC state that university administrators should work with local authorities to mitigate the impact of their decisions around instruction on the broader community.

Written by Mark H. Ebell MD, MS, on January 8, 2021. (Source: Leidner AJ, Barry V, Bowen VB, et al. Opening of large institutions of higher education and county-level COVID-19 Incidence—United States, July 6–September 17, 2020. MMWR. 2021;70(1):14-19.)

January 13, 2021, Research Update

Preliminary Data Show the Pfizer-BioNTech mRNA Vaccine Is Also Effective against Two Mutant Strains of SARS-CoV2. Two mutant strains of the SARS-CoV2 virus have emerged in the United Kingdom, South Africa, the United States, and probably elsewhere. These strains have slight alterations in the spike protein, which is part of the viral receptor binding site, and are the key targets for neutralizing antibodies. In this report from a preprint server (not peer-reviewed), researchers generated genome-specific antigens for these two mutant strains and tested them against the serum from 20 participants in the original randomized trial of the mRNA vaccine jointly developed by Pfizer and BioNTech. The outcome, neutralizing antibody generation as determined by a 50% plaque reduction neutralization assay, demonstrated the vaccine to be as effective against the mutant strains. Although reassuring, this is a fairly limited lab-based study, so we still need real-world performance data and ongoing monitoring.

Written by Henry C. Barry, MD, MS, on January 8, 2021. (Source: Xie X, Zou J, Fontes-Garfias CR, et al. Neutralization of N501Y mutant SARS-CoV-2 by BNT162b2 vaccine-elicited sera [published online January 7, 2021]. )

January 12, 2021, Research Update

Update to “Living” Systematic Review of Drugs for Treating Patients with COVID. We have previously summarized (Research Brief from BMJ posted on August 28, 2020) a living systematic review that concluded, based on all the available evidence on all the various drugs studied, that remdesivir decreased the duration of symptoms and did not cause any additional harm compared with standard care and that hydroxychloroquine increased the risk of adverse events. The authors use robust, high-quality meta-analytic methods. In this update, the authors identified 85 randomized trials with 41,669 patients (many more new studies since we last summarized the study). They conclude that, compared with standard care, corticosteroids probably reduce mortality and mechanical ventilation, whereas azithromycin, hydroxychloroquine, interferon-beta, and tocilizumab do not reduce either. In a reversal from their previous conclusions, the data on whether or not remdesivir confers any important patient-oriented outcomes is uncertain.

Written by Henry C. Barry, MD, MS, on January 5, 2021. (Source: Siemieniuk RA, Bartoszko JJ, Ge L, et al. Drug treatments for covid-19: living systematic review and network meta-analysis. BMJ. 2020;370:m2980.)

January 11, 2021, Research Update

Immunity from SARS-CoV-2 Lasts for at Least Six Months; Reinfection within Six Months Is Rare. Prior small studies have shown that antibodies to SARS-CoV-2 persist at least six months in most people who have been infected. Symptom recurrence within two months is thought to be due to continued infection and not to reinfection. Case reports of reinfection are rare. In this large study from the UK, the investigators followed up on health care workers who were initially positive or negative for anti-spike IgG antibodies to see whether those who tested positive were protected against subsequent COVID-19 infection compared with those who tested negative. There were 12,541 healthcare workers enrolled. At baseline, 11,364 individuals had a negative antibody test, and 1,265 individuals had a positive test. The health care workers were subsequently tested for SARS-CoV-2 infection every two months and were asked whether they developed COVID-19 symptoms. Eighty-eight seroconverted during the follow-up period, so a total of 1,353 individuals who had evidence of COVID-19 infection were followed up with repeat testing for an average of about six months. During the follow-up period, only two individuals who were positive for anti-spike IgG antibodies had a subsequent positive PCR test for SARS-Cov-2 compared with 223 who were antibody negative, an adjusted incidence rate ratio of 0.11 (95% CI, 0.03 to 0.44; P = 0.002). Both of these patients were asymptomatic. This study provides strong evidence that a second COVID-19 infection is highly unlikely within six months of the original infection. How long immunity lasts is not known at this time.

Written by John Hickner, MD, MS, on December 31, 2020. (Source: Lumley SF, O’Donnell D, Stoesser NE, et. al.; Oxford University Hospitals Staff Testing Group. Antibody status and incidence of SARS-CoV-2 infection in health care workers [published online December 23, 2020]. NEJM. 2020. //www.nejm.org/doi/full/10.1056/NEJMoa2034545)

January 7, 2021, Research Update

Bamlanivimab Does Not Improve Outcomes in Patients Hospitalized with COVID-19. We have previously reported a preliminary analysis of the BLAZE-I trial (Research Brief posted on December 15, 2020, on a study from NEJM) that suggested that an outpatient infusion of LyCov555 (bamlanivimab) reduced hospitalizations in patients with mild COVID. In this multicenter study, the researchers randomized patients hospitalized with COVID-19 to receive a single dose of 7,000 mg of bamlanivimab (n = 163) or placebo (n = 151). The active-treatment patients were slightly more likely to have comorbid conditions but were slightly less likely to be obese. All of the patients received remdesivir, and the protocol allowed for corticosteroid use “when indicated.” The authors report no difference in a composite outcome between the groups: death, serious adverse events, or clinical worsening.

Written by Henry C. Barry, MD, MS, on December 28, 2020. (Source: ACTIV-3/TICO LY-CoV555 Study Group. A neutralizing monoclonal antibody for hospitalized patients with Covid-19 [published online December 22, 2020]. NEJM. 2020. //www.nejm.org/doi/10.1056/NEJMoa2033130)

January 5, 2021, Research Update

COVID-19 Is More Severe Than Seasonal Influenza among U.S. Veterans. These researchers mined the databases of the U.S. Department of Veterans Affairs to compare features of veterans hospitalized for COVID-19 with those admitted for seasonal influenza. The 3,641 patients with COVID were admitted between February 1 and June 17, 2020, whereas the 12,676 with seasonal influenza were admitted between January 1, 2017, and December 31, 2019. Not surprising given the distribution in the military, well over 90% of the patients in each group were male. About half of the patients admitted with COVID were Black compared with 22% of those with influenza. The patients with COVID were slightly more likely to be obese and have dementia and diabetes mellitus, and those with seasonal influenza were slightly more likely to smoke, have cancer, cardiovascular disease, chronic lung disease, and use angiotensin-converting enzyme inhibitors or angiotensin receptor blockers and statins. After adjusting for many of these factors (which can also influence the severity of COVID or influenza), the authors report that those with COVID were more likely to experience complications such as acute kidney injury (37.2% vs. 29.0%; OR = 1.52; 95% CI, 1.37 to 1.69), severe septic shock (8.8% vs. 2.4%; OR 4.04; 95% CI, 3.38 to 4.83), pulmonary embolism (3.2% vs. 2.1%; OR 1.50; 95% CI, 1.18 to 1.90), and arrhythmias and sudden cardiac death (3.8% vs. 2.2%; OR 1.76; 95% CI, 1.40 to 2.20). Finally, more patients hospitalized with COVID died (18.6% vs. 5.3%; HR = 4.97; 95% CI 4.42 to 5.58), required mechanical ventilation (15.0% vs. 4.2%; HR 4.01; 95% CI, 3.53 to 4.54), and were in intensive care units (36.8% vs. 18.6%; HR 2.41; 95% CI, 2.25 to 2.59). On average, the patients with COVID were in the hospital three days longer than those with influenza. These data confirm what we have all suspected: COVID-19 is not just another flu-like illness. Keep in mind that these are hospitalized patients, representing the sickest of those afflicted.

Written by Henry C. Barry, MD, MS, on December 19, 2020. (Source: Xie Y, Bowe B, Maddukuri G, et al. Comparative evaluation of clinical manifestations and risk of death in patients admitted to hospital with covid-19 and seasonal influenza: cohort study [published online December 15, 2020]. BMJ. 2020;371:m4677.)

January 4, 2021, Research Update

WHO Solidarity Trial Finds No Benefit to Four Antiviral Drugs, Including Remdesivir. This was a large WHO sponsored trial set in 405 hospitals in 30 countries. A total of 11,330 adults hospitalized with COVID-19 who had no contraindications to study medications and who were expected to require at least three days of hospitalization were recruited. They were randomized to receive one of four study drugs (remdesivir, hydroxychloroquine, lopinavir, or interferon beta-1a) or usual care. Different drugs were available at each institution, and in some cases patients in the usual care group served as controls for patients in more than one active treatment group. Standard doses similar to those in other COVID-19 trials were used. This was an open-label trial, although groups were balanced at baseline and adherence to the assigned medication (or no medication) was approximately 95% in all groups. The primary outcome was in-hospital mortality, and analysis was by intention-to-treat. The hydroxychloroquine, lopinavir, and interferon arms were discontinued after it was confirmed that they had nonbenefit. Analysis of the Kaplan-Maier survival curves found that none of the drugs reduced mortality (p > 0.10), either for the entire group or for subgroups by age, ventilation status, geographic regions, or corticosteroid use. The authors also summarize the results of the published randomized trials of remdesivir, stratified by level of respiratory support. Patients not requiring mechanical ventilation had a trend toward reduced mortality (RR = 0.80; 95% CI, 0.63 to 1.01), whereas the groups requiring high levels of ventilatory support had a trend in the other direction (RR = 1.16; 95% CI, 0.85 to 1.60). This is consistent with the US ACTT-1 trial, where only patients having low-flow oxygen had a clear benefit. The absolute mortality difference in the low-risk group, even if real, was small (1.4%).

Written by Mark H. Ebell MD, MS, on December 9, 2020. (Source: WHO Solidarity Trial Consortium. Repurposed antiviral drugs for Covid-19—Interim WHO Solidarity Trial results [published online December 2, 2020]. N Engl J Med. 2020. //www.nejm.org/doi/full/10.1056/NEJMoa2023184)

December 24, 2020, Research Update

About One in Six Household and Family Contacts Becomes Infected with SARS-CoV-2. This is a meta-analysis of the proportion of secondary cases of SARS-CoV-2 infections in households and families. It includes 44 studies of transmission within households and 10 studies of transmission within families (where not all members of the family lived in the same household). The meta-analysis includes studies from many countries; 12 studies were from China and five from the United States. The investigators followed PRISMA meta-analysis guidelines, so the quality of the meta-analysis is good; however, the quality of the individual studies varied. Many were not rated as high quality, so the secondary infection rates may be under- or overestimated. The overall estimated secondary infection rate was 16.4% in households and 17.4% among families, meaning that roughly one in six household or family contacts become infected. However, the secondary infection rates varied greatly, from a low of less than 1% in a South Korean study to a high of 45% in an Italian study. Combining the five U.S. studies, 232 of 722 (32%) household contacts became infected. The infection rate was much lower when the index case was asymptomatic or presymptomatic (0.7%) vs. symptomatic (18%). Spouses were at greatest risk of secondary infection, at 38%. Adults were more likely to transmit infection than children, and the secondary infection rate was higher for adults (15.2%) than for children (7.9%). The percentage of individuals within a household who became infected varied greatly, even within studies. In one study, for example, in 25% of the households, 100% of the individuals exposed became infected. The authors conclude that household transmission remains an important venue for infection, especially when social mobility is reduced by social controls such as business and school closings. In a prior study from China, we reported that household transmission can be reduced about 75% by wearing masks and frequent handwashing if these are in effect prior to symptomatic infection in the household (Research Brief posted on June 6, 2020).

Written by John Hickner, MD, MS, on December 19, 2020. (Source: Madewell ZJ, Yang Y, Longini IM, et al. Household transmission of SARS-CoV-2: a systematic review and meta-analysis. JAMA Netw Open. 2020;3(12):e2031756. //jamanetwork.com/journals/jamanetworkopen/fullarticle/2774102)

December 23, 2020, Research Update

Some Consumer Face Masks Are as Good or Better Than Medical Procedure Masks. This study is not the first comparative test of the filtration effectiveness of various face masks, but it is a good addition to our understanding of the variability of protection, depending on construction and fit of the mask. The researchers tested seven “consumer-grade” masks and five medical procedure masks with modifications for better fit on a single adult male (for consistency of testing) in a laboratory setting. They describe the masks in detail, which included a variety of construction materials (cotton, nylon, polyester), styles (traditional mask, bandana, gaiter), use of ties, modifications, and number of layers. Medical mask modifications included tying the ear loops and tucking side pleats in, use of an ear guard to fasten loops behind the head, use of a hair clip to fasten ear loops behind the head, use of rubber bands over the mask to enhance the seal, and use of nylon hosiery over the fitted mask. They used the U.S. Occupational and Safety Health Administration quantitative fit testing protocol for testing the masks. The outcome is filtered filtration efficiency (FFE), which refers to the percentage of virus-sized particles that would be filtered out by the mask, thereby reducing exposure to the wearer. These results are taken directly from the study abstract: “The mean (SD) FFE of consumer grade masks tested ranged from 79.0% (4.3%) to 26.5% (10.5%), with the 2-layer woven nylon mask (with nose bridge, washed one time) having the highest FFE. Unmodified medical procedure masks with ear loops had a mean (SD) FFE of 38.5% (11.2%). All modifications evaluated in this study increased procedure mask FFE (range [SD], 60.3% [11.1%] to 80.2% [3.1%]), with a nylon hosiery sleeve placed over the procedure mask producing the greatest improvement.” One surprise was that the bandana (“bandit”) style mask was about 50% efficient, a bit better than the standard medical procedure mask. Read this article to see the detailed results and the great pictures of the masked man being tested.

Written by John Hickner, MD, MS, on December 14, 2020. (Source: Clapp PW, Sickbert-Bennett EE, Samet JM, et al. Evaluation of cloth masks and modified procedure masks as personal protective equipment for the public during the COVID-19 pandemic [published online December 10, 2020]. JAMA Intern Med. //jamanetwork.com/journals/jamainternalmedicine/fullarticle/2774266)

December 18, 2020, Research Update

Symptoms Associated with COVID-19 in Children in Alberta, Canada. This study from the province of Alberta, Canada, describes the symptoms of 2,463 children from birth through 17 years of age who had a first test for SARS-CoV-2 between April 13 and September 30, 2020. The investigators were interested in knowing whether any symptom or constellation of symptoms made COVID-19 infection more likely than other conditions. Children were tested either because they had symptoms compatible with COVID-19 or because they were close contacts of COVID-19 infected individuals; 1,987 children tested positive, and 476 children tested negative for SARS-CoV-2. Of those testing positive, 714 (36%) were asymptomatic at the time of testing. An extensive symptom checklist was available for every child. Cough (24.5%) and runny nose (19.3%) were the most common symptoms, but these symptoms did not distinguish between children with COVID-19 and children with other health conditions (positive likelihood ratio 0.96, 95% CI, 0.81 to 1.14 and 0.87, 95% CI, 0.72 to 1.06, respectively). Anosmia/ageusia was the symptom most highly associated with a positive SARS-CoV-2 test, (positive LR 7.33). Next was nausea/vomiting, (positive LR 5.51). Headache and fever were weakly associated with COVID-19 infection (positive LR 2.49 and 1.68, respectively). The positive LR for the combination of anosmia/ageusia, nausea/vomiting, and headache was highly associated with COVID-19 infection, positive LR 65.92.

Written by John Hickner, MD, MS, on December 5, 2020. (Source: King JA, Whitten TA, Bakal JA, et al. Symptoms associated with a positive result for a swab for SARS-CoV-2 infection among children in Alberta [published online November 24, 2020]. CMAJ. 2020. //www.cmaj.ca/content/early/2020/11/23/cmaj.202065.long)

December 17, 2020, Research Update

Summary of FDA Report on Pfizer/BioNTech COVID-19 Vaccine Efficacy and Safety. This is the FDA report summarizing the application for emergency-use authorization for the Pfizer/BioNTech vaccine. It provides the most detailed data as of yet of a large COVID-19 vaccine trial. A total of 40,277 persons were randomized to vaccine or placebo. Only 2.6% were found to have antibodies due to a prior infection, and results presented in the Table below are for all patients who received two doses of vaccine regardless of prior infection. Participants were 49.4% women, with a mean age of 50 years. By age group, 58% were 16 to 55 years of age, 42% were older than 55 years of age, and 21% were older than 65 years of age. The population was fairly representative by race and ethnicity of the United States as a whole, with 82% White, 10% Black, 4.4% Asian, and 26% Hispanic or Latino. Efficacy data are summarized in the Table below, both overall and by key subgroups, and show the number of cases of COVID-19 beginning seven days after the second dose of vaccine. Efficacy was consistent regardless of age, race, ethnicity, sex, or the presence of comorbidities. In regard to safety, injection-site pain, fatigue, and headache were common. Lymphadenopathy was uncommon but occurred more often in the vaccine group (64 vs. six cases). Bell palsy occurred in four vaccine participants and no placebo participants. Severe adverse events were rare, and only 0.2% of vaccine participants withdrew because of them. 

	Cases of COVID-19
	Vaccine group	Placebo group	Efficacy (95% CI)
All participants	9	169	94.6% (89.9% to 97.3%)
Age (years)
16 to 55	6	120	95.0% (88.7% to 98.2%)
> 55	3	49	93.8% (80.9% to 98.8%)
65 to 74	1	14	92.9% (53.2% to 99.8%)
≥ 75	0	5	100% (–12.1% to 100%)
Race
Black or African American (n = 3,516)	0	7	100% (30.4% to 100%)
White (n = 30,767)	7	153	95.4% (90.3% to 98.2%)
Ethnicity
Hispanic ethnicity (n = 10,164)	3	55	94.5% (83.2% to 98.9%)
Non-Hispanic (n = 26,889)	6	114	94.7% (88.1% to 98.1%)

 

Written by Mark H. Ebell, MD, MS, on December 8, 2020. Source: Food and Drug Administration. Vaccines and Related Biological Products Advisory Committee meeting; December 2020. )

December 15, 2020, Research Update

Interim Analysis of BLAZE-1 Trial Suggests That Intravenous Bamlanivimab Appears to Prevent Hospitalizations in Outpatients with Mild COVID. These researchers report a planned interim analysis of a Phase II trial that included 452 outpatients not using oxygen who had positive SARS-CoV-2 tests and at least one symptom attributable to COVID. These patients were randomized to receive a single intravenous infusion of a placebo (n = 143) or one of three doses of the neutralizing antibody, LY-CoV555 (now called bamlanivimab). The three doses were 700 mg (n = 101), 2,800 mg (n = 107), or 7,000 mg (n = 101). This analysis was planned after the last randomized patient had achieved 11 days of follow-up; in other words, the trial was still underway at the time of publication. The primary outcome, one that is not as clinically important, is that viral load after 11 days was significantly lowered only with the 2,800 mg dose; however, infectivity was not assessed. More importantly, after 29 days of follow-up, five of 309 (1.6%) bamlanivimab-treated patients were hospitalized compared with 9 of 143 (6.3%) of placebo-treated patients. This translates to a number needed to treat of 22 (95% CI, 10 to 96). The authors don’t report a p-value for this outcome, but it was 0.02 by my own crude analysis. The authors also report modest improvement in symptoms scores among the actively treated patients, generally 1 point or less on a 24-point scale. These differences are unlikely to be clinically meaningful, and, once again, they do not report whether the differences could have been due to chance. Finally, the authors report comparable rates of adverse effects in each group. The FDA has granted Emergency Use Authorization (EUA) for people 12 years of age and older weighing 88 or more pounds. In addition to the published study, the EUA includes longer follow-up with 465 patients, with an additional 13 who were all in the placebo group, none of whom were hospitalized. These updated data result in a slightly higher number needed to treat of 25 (95% CI, 11 to 140). The EUA includes analyses stratified by dose as well as a subgroup analyses of high-risk patients, but the numbers were small and not terribly robust. This is hopeful, but please keep in mind that this is a single report from a preliminary analysis of a small Phase II study and the usual cautions apply.

Written by Henry C. Barry, MD, MS, on December 2, 2020. (Source: Chen P, Nirula A, Heller B, et al.; BLAZE-1 Investigators. SARS-CoV-2 neutralizing antibody LY-CoV555 in outpatients with Covid-19 [published online October 28, 2020]. N Engl J Med. 2020. //www.nejm.org/doi/full/10.1056/NEJMoa2029849)

December 11, 2020, Research Update

Lopinavir–Ritonavir Equals Usual Care in Patients Hospitalized with COVID (RECOVERY Trial). The RECOVERY trial was a government-funded open-label trial that randomized more than 11,500 hospitalized patients with COVID-19 to one of six arms: azithromycin, lopinavir–ritonavir, tocilizumab, convalescent plasma, low-dose dexamethasone, or usual care. We have previously reported data from the steroid and hydroxychloroquine wings, and now we have a report on 1,616 patients who received daily lopinavir-ritonavir (400 mg and 100 mg, respectively) and 3,424 who received usual care. The authors report that 23% of the patients allocated to lopinavir-ritonavir died by 28 days compared with 22% of control patients. About 70% of patients in each group were discharged from the hospital, and about 10% in each group required mechanical ventilation. In general, open-label studies have biases that favor interventions, strengthening the conclusion that lopinavir-ritonavir does not improve outcomes in patients hospitalized with COVID-19.

Written by Henry C. Barry, MD, MS, on December 1, 2020. (Source: Horby PW, Mafham M, Bell JL, et al.; RECOVERY Collaborative Group. Lopinavir–ritonavir in patients admitted to hospital with COVID-19 (RECOVERY): a randomised, controlled, open-label, platform trial. Lancet. 2020;396(10259):1345-1352.)

December 10, 2020, Research Update

Convalescent Serum, in the Presence of Corticosteroid Co-Therapy, Does Not Improve Outcomes in Patients Hospitalized with Severe COVID-19 Pneumonia (PlasmAr). We have previously reviewed studies demonstrating the safety of administering convalescent serum (Research Brief, May 31, 2020) and their early inclusion into the IDSA management guidelines (Research Brief, May 6, 2020). We have also reported potential benefit based on early case series (Research Brief, April 1, 2020) and reports of convalescent plasma donors’ humoral immunity (Research Brief, May 14, 2020) and data from one trial (Research Brief, June 12, 2020). More recently, however, the PLACID trial in India (Research Brief, October 30, 2020) failed to demonstrate a benefit. In this study from Argentina, the research enrolled patients hospitalized with radiographically confirmed severe COVID-19 pneumonia based on oxygen saturation below 93% on room air, a ratio of the partial pressure of oxygen to the fraction of inspired oxygen below 300 mm Hg, or a sequential organ failure assessment (SOFA) or modified SOFA score of two or more points above their baseline status (higher scores indicate worse disease). All patients also had confirmed COVID by PCR. The researchers randomized the patients to receive convalescent serum (n = 228) or placebo (n = 105). The median age of the study participants was 62 years of age, and the median time between onset of symptoms and enrollment was eight days. Two-thirds of the patients were male, and 65% had a preexisting condition. Ninety percent of the patients were also receiving corticosteroids at the time of enrollment. After 30 days, 11% of patients receiving convalescent plasma died, compared with 11.4% of placebo-treated patients. Finally, no difference was found in the combined outcome of death, respiratory decline, or discharge status. The lack of improvement was also found in multiple planned subgroup analyses.

Written by Henry C. Barry, MD, MS, on December 1, 2020. (Source: Simonovich VA, Burgos Pratx LD, Scibona P, et al. A randomized trial of convalescent plasma in Covid-19 severe pneumonia [published online November 24, 2020]. N Engl J Med. 2020. //www.nejm.org/doi/full/10.1056/NEJMoa2031304)

December 9, 2020, Research Update

Face Mask Mandate Slowed Spread of COVID-19 Transmission in Kansas Counties. The value of face masks for preventing COVID-19 spread has been questioned by some individuals, despite epidemiological and laboratory simulation evidence suggesting a protective effect. In this study, the investigators documented trends in COVID-19 incidence before and after the Kansas governor’s July 2, 2020, nonmandatory executive order requiring face masks. Prior to the face mask mandate, the incidence of COVID-19 was increasing in both the mask and nonmask mandated counties, though much more rapidly in the counties that eventually mandated face masks: 467% vs. 50% comparing the seven-day rolling average during the first week of June to the first week of July 2020. After the governor’s executive order, as of August 11, 2020, 24 (23%) Kansas counties had a mask mandate in place, and 81 did not. By August 17 to 23, 2020, the seven-day rolling average COVID-19 incidence had decreased by 6% to 16 cases per 100,000 among mandated counties and increased by 100% to 12 per 100,000 among nonmandated counties. This study provides strong evidence that mandatory face mask orders are an effective method for preventing transmission of COVID-19. It is possible, however, that individuals in face-mask mandated counties were also more careful about social distancing and hand hygiene, which are other effective interventions for preventing spread. 

Written by John Hickner, MD, MS, on November 27, 2020. (Source: Van Dyke ME, Tia M, Rogers TM, et al. Trends in county-level COVID-19 incidence in counties with and without a mask mandate — Kansas, June 1–August 23, 2020. MMWR Morb Mortal Wkly Rep. 2020;69(47):1777-1781.)

December 8, 2020, Research Update

American College of Rheumatology Guideline on Managing Children with Multisystem Inflammatory Syndrome. To provide guidance on how to manage children with multisystem inflammatory syndrome, the American College of Rheumatology convened a task force that included pediatric specialists in rheumatology, cardiology, infectious disease, and critical care and adult rheumatologists. Four workgroups were assigned to address diagnostic evaluation, cardiac management, treatment, and management of hyperinflammation. In contrast with common guideline development processes, each workgroup developed an initial set of recommendations and then searched for the evidence to support the recommendations. The better practice would be to identify a set of questions to be addressed, search for the evidence, synthesize the evidence, and then create the recommendations. Their final guideline includes a summary of the criteria for case definitions from the World Health Organization, Centers for Disease Control and Prevention, and the Royal College of Paediatrics and Child Health, which differ, but include fever, a range of clinical symptoms and organ system involvement, various laboratory parameters suggesting inflammation, and testing for COVID-19. Additionally, it includes a suggested diagnostic algorithm (see figure 1 in the source material from Arthritis & Rheumatology). For cardiac management, the guideline recommends serial monitoring of abnormal cardiac biomarkers (e.g., BNP, troponin T levels), performing electrocardiography at least every 48 hours, performing follow-up echocardiography at least every one to two weeks and then four to six weeks after presentation, performing cardiac MRI after two to six months in those with signs of cardiac dysfunction and cardiac CT if there is a suspicion of distal coronary artery aneurysms if the arteries are not well seen on ultrasonography. The panel recommends completing the diagnostic evaluation before initiating immunomodulator therapy except in life-threatening illness. The panel also recommends intravenous immune globulin and/or glucocorticoids as first-line treatment followed by anakinra. Low-dose aspirin is recommended for at least four weeks unless the platelet count is less than 80,000 per microliter. They also provide guidance on when to use anticoagulants such as enoxaparin or warfarin. Finally, in addition to the use of glucocorticoids or anakinra, the guideline also recommends using tocilizumab in children with signs of hyperinflammation (acute respiratory distress syndrome; shock/cardiac dysfunction; substantially elevated LDH, d-dimer, IL-6, IL-2R, CRP, and/or ferritin levels; and depressed lymphocyte count, albumin levels, and/or platelet count). The panel’s strength of confidence in the recommendations ranged from moderate to high, but the underlying evidence is still in evolution.

Written by Henry C. Barry, MD, MS, on November 24, 2020. (Source: Henderson LA, Canna SW, Friedman KG, et al. American College of Rheumatology clinical guidance for multisystem inflammatory syndrome in children associated with SARS-CoV-2 and hyperinflammation in pediatric COVID-19: version 1. Arthritis Rheumatol. 2020;72(11):1791-1805.)

December 7, 2020, Research Update

World Health Organization’s “Living” Guideline on Drugs for Treating Patients with COVID-19. The WHO created a standing guideline panel comprising content experts, clinicians, patients, and methodologists who use modern guideline development methods to evaluate the evidence related to managing patients with COVID-19 and to distill the data into a cohesive set of recommendations. In September 2020, the WHO issued a guideline that is now replaced by this version, which is remarkably succinct. The online version of the guideline has useful interactive graphics summarizing the supporting data and their final recommendations. Based on a network meta-analysis of four randomized trials, the WHO guideline recommends remdesivir only for patients hospitalized with nonsevere COVID-19. Additionally, the WHO recommends corticosteroids for patients hospitalized with severe COVID-19.

Written by Henry C. Barry, MD, MS, on November 24, 2020. (Source: Lamontagne F, Agoritsas T, Macdonald H, et al. A living WHO guideline on drugs for covid-19. BMJ. 2020;370:m3379.)

December 1, 2020, Research Update

Face Masks, When Added to Social Distancing Measures, Did Not Prevent Wearers from COVID-19 Infection in Denmark. Epidemiological studies have demonstrated that mask wearing reduces the transmission of respiratory viruses. A prior randomized trial in a health care system demonstrated that surgical type, three-ply masks protect wearers from respiratory infections much better than simple cotton masks. Prior to this study, however, there had not been any randomized trials of mask wearing to protect wearers in the community from COVID-19 infection. This randomized trial in Denmark, appropriately named DANMASK-19, sought to determine whether wearing masks in a community setting, in addition to other public health measures, protected the wearer from COVID-19. There were 6,024 participants asked to follow social distancing, and half were randomly assigned to wear either a three-layer, disposable surgical face mask in public spaces or to not wear one for one month. Mask wearing in public was uncommon in Denmark at the time of the study. There were 4,862 (81%) participants who completed the study. Compliance was good in the intervention group; 46% reported wearing the mask as recommended, and 47% reported wearing the mask mainly as recommended. Infection with SARS-CoV-2 occurred in 42 (1.8%) participants randomized to wear masks and in 53 (2.1%) control participants (95% CI, −1.2 to 0.4 percentage points; P = 0.38, odds ratio 0.82 [CI, 0.54 to 1.23]; P = 0.33). Because the infection rate was low, the 95% confidence interval did not exclude a possible 46% reduction or 23% increase in infections in mask wearers. Limitations of the study include that other social distancing measures were in place at the time of the study and the low infection rate. This study did not seek to determine whether mask wearing prevents others from being infected.

Written by John Hickner, MD, MS, on November 20, 2020. (Source: Bundgaard H, Bundgaard JS, Raaschou-Pedersen DET, et al. Effectiveness of adding a mask recommendation to other public health measures to prevent SARS-CoV-2 infection in Danish mask wearers: a randomized controlled trial [published online November 18, 2020]. Ann Intern Med. 2020. //www.acpjournals.org/doi/10.7326/M20-6817)

November 30, 2020, Research Update

COVID-19 Symptoms Persist for at Least 60 Days for One-Third of Hospitalized Patients. Many patients with symptomatic COVID-19 infection have persistent symptoms after the acute illness. We previously summarized a study from France in which 60% of 150 patients with mild to moderate COVID-19 infection had persistent symptoms 30 days after illness onset and 30% had persistent symptoms 60 days after onset of the illness. The symptoms were mainly fatigue, shortness of breath, and loss of taste and smell (COVID Research Brief posted October 16, 2020). This larger observational study gives the outcomes of 1,648 patients with COVID-19 discharge from 38 Michigan hospitals between March 16 and July 1, 2020. Within 60 days of discharge, of the 1,250 (76%) patients who survived hospitalization, an additional 84 died, and 189 patients were rehospitalized. Of the patients alive 60 days after discharge, 488 (42%) were successfully contacted and completed a postdischarge telephone survey. Cardiopulmonary symptoms (such as cough and dyspnea) were reported by 159 (33%) patients, and 65 (13%) patients had persistent loss of taste or smell. This study from Michigan hospitals confirms that roughly 30% of patients hospitalized with COVID-19 infection will have persistent symptoms for at least two months. A limitation of this study is that less than 50% of discharge patients were included in the study, which could lead to over- or underestimation of the frequency of persistent symptoms.

Written by John Hickner, MD, MS, on November 18, 2020. (Source: Chopra V, Flanders SA, O'Malley M, et al. Sixty-day outcomes among patients hospitalized with COVID-19 [published online November 11, 2020]. Ann Intern Med. //www.acpjournals.org/doi/10.7326/M20-5661)

November 20, 2020, Research Update

Small Randomized Trial Finds Some Benefit from Fluvoxamine. Fluvoxamine is an SSRI primarily used to treat obsessive-compulsive disorder. It is also is a strong agonist for the sigma-1 receptor that regulates cytokine production, and a French cohort study on a preprint server found an association between SSRI use and a reduced likelihood of intubation or death (). In the current study, 181 outpatients with PCR-confirmed COVID-19 who had been symptomatic for fewer than seven days were randomized to fluvoxamine or placebo in a double-blind manner. Patients who were hospitalized at any time or who had an oxygen saturation at baseline < 92% were excluded. Fluvoxamine was given in a dose of 50 mg once daily on day 1, then 100 mg twice daily on days 2 and 3. If tolerated, the dosage was increased to 100 mg three times daily through day 15. The primary outcome was clinical deterioration, defined as the development of dyspnea or need for hospitalization plus an oxygen saturation < 92% and was available for 84% of participants. This outcome occurred in 0 of 80 patients in the fluvoxamine group and in six of 72 in the placebo group (0% vs. 8.3%, p = 0.009, NNT = 12). The secondary outcome was clinical deterioration on a seven-point ordinal scale ranging from none (stage 0) through stages of increasing oxygen requirements, to the need for mechanical ventilation, and ultimately death (stage 6). Of the six patients in the placebo group who deteriorated on this scale, four were hospitalized, and one required mechanical ventilation, but at 30 days of follow-up none had died. Based on this one small trial, fluvoxamine is a promising treatment for symptomatic outpatients with COVID-19, but this study requires confirmation.

Written by Mark H. Ebell, MD, MS, on November 15, 2020. (Source: Lenze EJ, Mattar C, Zorumski CF, et al. Fluvoxamine vs placebo and clinical deterioration in outpatients with symptomatic COVID-19: a randomized clinical trial [published online November 12, 2020]. JAMA. 2020. //jamanetwork.com/journals/jama/fullarticle/10.1001/jama.2020.22760)

November 19, 2020, Research Update

More than one-quarter of sailors aboard the USS Theodore Roosevelt contracted COVID early in the pandemic. The media covered the initial reports of an outbreak of COVID aboard the aircraft carrier USS Theodore Roosevelt but never reported the full data, probably because of all the other COVID stories they were covering, among other things. On Veterans Day, the New England Journal of Medicine published the full report of the outbreak. The ship had been at sea for 13 days when three sailors became ill and had positive PCR tests for SARS-CoV-2. Through contact tracing, they identified 400 close contacts. These contacts, as well as any symptomatic sailors, were tested. Sailors who tested positive were isolated, and the remaining 4,079 were quarantined in a hotel in single-occupancy rooms. Ultimately, between March 23 and May 18, 1,271 sailors (26.6% of the crew) tested positive, and an additional 60 crew members had suspected COVID but had negative tests. Of those who tested positive, 572 (43%) remained asymptomatic, 293 (22%) were already symptomatic when they tested positive, and 406 (30.5%) were presymptomatic at the time that they tested positive. Not unusual for a young, healthy, and presumably fit cohort (mean age was 27 years); only 23 of the infected sailors were hospitalized, four of whom needed intensive care. One sailor died.

Written by Henry C. Barry, MD, MS, November 14, 2020. (Source: Kasper MR, Geibe JR, Sears CL, et al. An outbreak of Covid-19 on an aircraft carrier [published online November 11, 2020]. N Engl J Med. 2020. //www.nejm.org/doi/full/10.1056/NEJMoa2019375)

November 16, 2020, Research Update

No Oxygen Desaturation Observed in Elderly Wearing Nonmedical Face Masks. Some people contend they cannot wear a face mask to reduce COVID transmission for “medical reasons.” It is unclear what medical conditions would qualify as legitimate medical reasons. One possible reason, explored in this study, is that the mask might reduce the amount of oxygen in the wearer’s blood, especially those with chronic pulmonary disease. To address this concern, the investigators recruited volunteers from a retirement condominium in Ontario, Canada, who were 65 years of age and older and who had cardiac or pulmonary conditions. Each participant was given an oximeter and told to record their oxygen saturation during three time intervals: for one hour before wearing the mask, for one hour while wearing the mask, and for one hour after taking off the mask, all during usual activities of daily living. The masks were three-layer disposable nonmedical face masks (Boomcare DH95), and participants were instructed in how to properly wear them. Twenty-eight people were approached, and 25 agreed to participate. Results: The mean oxygen saturation was 96.1% before wearing, 96.5% while wearing, and 96.3% after wearing the mask. None of the participants’ oxygen saturation fell below 92% while wearing the mask. The investigators conclude that, “these results do not support claims that wearing nonmedical face masks in community settings is unsafe.” It seems more likely that some people say they cannot wear masks because they hyperventilate, causing dizziness, rather than due to oxygen desaturation.

Written by John Hickner, MD, MS, on November 5, 2020. (Source: Chan NC, Li K, Hirsh J. Peripheral oxygen saturation in older persons wearing nonmedical face masks in community settings [published online October 30, 2020]. JAMA. 2020. //jamanetwork.com/journals/jama/fullarticle/2772655)

November 13, 2020, Research Update

Updated Age-Specific Mortality and Immunity Related to SARS-CoV-2 Infection. These authors used the results of 22 seroprevalence surveys to estimate the total number of cases in a population, both symptomatic and asymptomatic. They then used age-specific mortality data to estimate the infection fatality ratio (this is different from the case fatality ratio, which has only symptomatic cases in the denominator and will generally be higher). They fit models and show us lots of detailed and colorful graphs, reporting data by country, by continent, and by age groups. Most interestingly, the infection fatality rate is lowest in people five to nine years of age (0.001%), increasing to 0.1% among those 25 to 29 years of age, almost 1% among those 60 to 64 years of age, and is highest among those 80 years and older (8.4%). The IFR by country was very consistent for younger age groups, but it varied considerably for those older than 65 years, which the authors speculate may have to do with differences in nursing home outbreaks. The proportion of the population infected ranged from just more than 0.0% in Asia to approximately 13% in the United States (as of September 1, 2020) and more than 50% in Mexico and Peru.

Written by Mark H. Ebell MD, MS, on November 5, 2020. (Source: O’Driscoll M, Dos Santos GR, Wang L, et al. Age-specific mortality and immunity patterns of SARS-CoV-2 [published online November 2, 2020]. Nature. 2020. //www.nature.com/articles/s41586-020-2918-0).

November 12, 2020, Research Update

Case Fatality Ratio Due to COVID-19 Appears to Be Declining. Case fatality ratios appear to be declining compared with what they were in March and April 2020. However, it is not clear whether this is because of more testing, a demographic shift toward younger patients, or better treatment. These authors at New York University Langone Health gathered data on admissions of adults hospitalized with laboratory confirmed SARS-CoV-2 infection between March 1 and August 31, 2020, with mortality data available through October 8, 2020. They developed a multivariate model for mortality that included age, sex, race, tobacco use, the presence of each of nine comorobidities, and admission O2 saturation, d-dimer, serum ferritin, and c-reactive protein. They used this model to calculate the expected number of deaths given the patient mix for each month, compared that with the actual number of deaths to calculate standardized mortality ratios, and applied the standardized mortality ratio to the overall mortality to calculate adjusted mortality rates for each month. They used a second approach that calculates the average marginal effects in a model that included the month and determined the percentage difference in mortality between March and the months that followed. They found that the standardized mortality ratio decreased from 1.26 in March to 0.38 in August, and the adjusted mortality rate decreased from 25.6% in March to only 7.6% in August. The calculation using the average marginal effect had similar results. Thus, after adjusting for patient factors and severity of illness on admission, it appears that improvements in treatment may be related, probably because of a combination of factors such as use of corticosteroids and anticoagulants, less aggressive use of mechanical ventilation and proning during ventilation, a lower burden on the critical care system, and possibly the use of remdesivir and convalescent plasma. A UK study available on a preprint server had similar findings (//www.medrxiv.org/content/10.1101/2020.07.30.20165134v2).

Written by Mark H. Ebell, MD, MS, on October 27, 2020. (Source: Horwitz LI, Jones SA, Cerfolio RJ, et al. Trends in COVID-19 risk-adjusted mortality rates [published October 23, 2020]. J Hosp Med. 2020; //www.journalofhospitalmedicine.com/jhospmed/article/230561/hospital-medicine/trends-covid-19-risk-adjusted-mortality-rates)

November 11, 2020, Research Update

T-Cell Immunity after COVID Infection Lasts at Least Six Months. We have a report on a preprint server (not yet peer-reviewed) about the duration of post-COVID infection immunity. Researchers in the United Kingdom obtained blood samples from 100 adults between 22 and 65 years of age who had contracted SARS-CoV-2 six months previously. They don’t describe how the people were identified or how the initial diagnosis was made. None of the people required hospitalization, and 44 were asymptomatic; 77 of the people were women. Of the 95 samples that they were able to test for T-cell responses, 90 (95%) demonstrated a SARS-CoV-2-specific response to at least one protein. The researchers report significant heterogeneity in the magnitude of the response, but generally found that the participants who were initially symptomatic had 50% higher levels than those who were asymptomatic. They also conducted several other markers of immunity, but the T-cell response is the one most associated with longer-term cellular immunity. Because this is a lab-based study, it is unclear whether this translates to real-world immunity.

Written by Henry C. Barry, MD, MS, on November 4, 2020. (Source: Zuo J, Dowell A, Pearce H, et al. Robust SARS-CoV-2-specific T-cell immunity is maintained at 6 months following primary infection [published online November 2, 2020]. bioRxiv preprint online: )

November 5, 2020, Research Update

Frontline Health Care Workers in Scotland and Their Families Are at Increased Risk of Hospitalization from COVID. These researchers identified 158,445 Scottish health care workers between 18 and 65 years of age who were employed by the National Health Service. They used Scottish employment databases to create linkages to other health-related databases and to identify members of their households (n = 229,905). Additionally, the researchers used these same databases to identify people from the general population. They mined the databases to identify hospitalizations for COVID between March 1, 2020 (when cases of COVID were first identified in Scotland), and June 6, 2020. Among the health care workers, 90,733 (57.3%) had direct patient contact (they used the term “patient-facing”), and 79% were women. During the study period, they identified 6,346 hospitalizations for COVID, 2,097 of which occurred in people between 18 and 65 years of age. Of these, 1,737 (82.8%) occurred in the general population, and health care workers and their household members accounted for 243 (11.6%) and 117 (5.6%), respectively. After adjusting for age, sex, ethnicity, socioeconomic factors, and comorbidities, the overall risk of hospitalization for all health care workers and their households was similar to that of the general population. However, among patient-facing health care workers, while the numbers were small, the risk of hospitalization was much higher (hazard ratio 3.30, 95% CI, 2.13 to 5.13) as was the risk for their household members (hazard ratio 1.79, 95% CI, 1.10 to 2.91). Combining these data with the data in the Research Brief posted on October 23, 2020, summarizing the CDC’s October report on excess mortality attributed to COVID19 () illustrates the risk to health care workers, their families, and the public. It’s hard to imagine where any profit could be gained from naming COVID as a cause of death.

Written by Henry C. Barry, MD, MS, on November 1, 2020. (Source: Shah ASV, Wood R, Gribben C, et al. Risk of hospital admission with coronavirus disease 2019 in healthcare workers and their households: nationwide linkage cohort study. BMJ. 2020;371:m3582.)

October 30, 2020, Research Update

Convalescent Plasma Does Not Reduce Worsening nor All-Cause Mortality in Patients Hospitalized with Moderate COVID in India (PLACID Trial). In this open-label phase II trial from 39 hospitals in India, the researchers randomized patients to standard care (n = 229) or standard care plus convalescent plasma (n = 235). The patients had all been hospitalized with moderate illness (PaO2/FiO2 ratio between 200 mm Hg and 300 mm Hg or a respiratory rate of more than 24 per minute with oxygen saturation 93% or less on room air). To be eligible, there also had to be a matched plasma donor (adults with a previous PCR-confirmed symptomatic bout with COVID). The staff administered the convalescent plasma as two infusions administered 24 hours apart. Standard care, a kitchen sink of protocols that appeared to be unique to each hospital, could include hydroxychloroquine, oseltamivir, remdesivir, lopinavir/ritonavir, broad spectrum antibiotics, steroids, tocilizumab, and good supportive care. About two-thirds of the patients had preexisting comorbidities, and three-quarters were male. Most (91%) had dyspnea at baseline, and two-thirds had chest radiography with bilateral “patchy shadows.” After 28 days of follow-up, 41 (18%) of control patients died or had progression to severe disease compared with 44 (19%) of those receiving convalescent plasma. One potential limitation is that of the samples of convalescent plasma tested, 29% had no detectable neutralizing antibodies.

Written by Henry C. Barry, MD, MS, on October 27, 2020. (Source: Agarwal A, Mukherjee A, Kumar G, et al. Convalescent plasma in the management of moderate covid-19 in adults in India: open label phase II multicentre randomised controlled trial [PLACID Trial]. BMJ. 2020;371:m3939.)

October 29, 2020, Research Update

Evidence for Effectiveness for Tocilizumab for COVID-19 Treatment Inconclusive and Not Encouraging. We usually summarize single studies, but because of the inconclusive evidence for the efficacy of tocilizumab in the treatment of COVID-19 and the many clinical trials of tocilizumab recently published and underway, we refer to an editorial that summarizes the evidence as of October 22, 2020. Tocilizumab is a monoclonal antibody against interleukin 6 (IL-6) receptors that is used to treat inflammatory arthritis and giant cell arteritis. Because of its anti-inflammatory effects, it was believed to be a good candidate for severe COVID-19 infection, where much of the morbidity is due to severe inflammation. In this editorial, Parr summarizes the results of five studies: one observational study, two completed randomized trials, and two uncompleted randomized trials that have reported interim results. (These five studies are summarized in the accompanying table.) The observational study STOP-COVID found an improvement of 9.6% in estimated 30-day mortality, although it is subject to unmeasured or residual confounding. However, the two small, randomized trials published online in JAMA Internal Medicine on October 20, 2020 (RCT-TCZ-COVID-19 and CORIMUNO-TOCI-1), found little benefit. Differences in mortality at 28 and 30 days of 1.7% and 0.8%, respectively, were not statistically significant. Two larger randomized trials, COVACTA and EMPACTA, found nonsignificant differences in 28-day mortality of 0.3% and 1.8% favoring placebo, although these are preliminary results based on about 400 enrolled patients in each trial. There was a reduction in need for mechanical ventilation in the EMPACTA trial. Another small randomized trial published in NEJM in October and not included in the Parr editorial found that tocilizumab was not effective for preventing intubation or death in moderately ill hospitalized patients with COVID-19. These five randomized trials are considered inconclusive because of small sample sizes, but taken together, the results are not encouraging.

 

TABLE

Study	Design	Number of patients	28-day or 30-day mortality (tocilizumab vs. comparator)
STOP-COVID	Cohort	3,924	27.5% vs. 37.1% (p < 0.05)
RCT-TCZ-COVID-19	RCT	126	3.3% vs. 1.6% relative risk 2.1 (95% CI, 0.20 to 22.6)
CORIMUNO-TOCI-1	RCT	131	11.1% vs. 11.9% adjusted HR 0.92 (95% CI, 0.33 to 2.53)
COVACTA	RCT	450	19.7% vs. 19.4% adjusted risk difference 0.3% (95% CI, -7.6% to 8.2%)
EMPACTA	RCT	389	10.4% vs. 8.6% adjusted risk difference 2.0% (95% CI, -5.2% to 7.8%)

 

Written by John Hickner, MD, MS, on October 23, 2020. (Source: Parr JB. Time to reassess tocilizumab’s role in COVID-19 pneumonia [editorial; published online October 20, 2020]. JAMA Intern Med. 2020. //jamanetwork.com/journals/jamainternalmedicine/fullarticle/2772184, and Stone JH, Frigault MJ, Serling-Boyd, et al.; BACC Bay Tocilizumab Trial Investigators. Efficacy of tocilizumab in patients hospitalized with Covid-19 [published online October 21, 2020]. //www.nejm.org/doi/full/10.1056/NEJMoa2028836. References for the other studies can be found in the Parr editorial.)

October 27, 2020, Research Update

Systematic Review of Multisystem Inflammatory Syndrome in Children. We have previously presented data from individual studies (Research Brief on June 16, 2020, from JAMA and Research Brief on July 15, 2020, from NEJM) and have done our best to informally synthesize their findings. These authors systematically reviewed two databases, two preprint servers, and Google Scholar to identify studies that described at least five cases of the multisystem inflammatory syndrome in children with COVID-19 and “sufficiently described” (undefined) the inclusion criteria, demographic information, clinical signs or symptoms, outcomes, laboratory studies, cardiovascular function, and treatments administered. The authors do not describe assessing the methodologic quality of the included studies. They ultimately included eight studies with 440 children. Six studies came from Europe, and two came from the United States. Three of the studies required that the children have a diagnosis of Kawasaki disease to be included; several required cardiovascular involvement or suspected cardiovascular involvement. In other words, most of the studies included the most severely ill children. The median age of patients ranged from 7.3 to 10 years, and 59% of all patients (43% to 73% across studies) were male. In the six studies that reported race or ethnicity, the proportion of Black or Afro-Caribbean patients ranged from 31% to 62%. The two U.S. studies reported that 36% to 39% of patients were Hispanic. The proportion of children with positive PCR tests ranged from 13% to 69%, and the proportion with positive serology ranged from 75% to 100%. Seven (2%) of the 440 children died. Overall, the greatest proportion of children had gastrointestinal symptoms (87%), followed by skin or mucocutaneous symptoms (73%). Additionally, 71% reported cardiovascular symptoms. Keep in mind that the deck was stacked, so the spectrum of symptoms in children with milder and less obvious disease is likely to be quite different from what is reported in these studies. Finally, all the studies reported at least 75% of the children had elevated C-reactive protein, interleukin-6, and fibrinogen levels.

Written by Henry C. Barry, MD, MS, on October 22, 2020. (Source: Abrams JY, Godfred-Cato SE, Oster ME, et al. Multisystem inflammatory syndrome in children associated with severe acute respiratory syndrome coronavirus 2: a systematic review. J Pediatr. 2020;226:45-54.e41.)

October 26, 2020, Research Update

High Transmission Rate but Evidence of Protective Immunity on a Fishing Vessel Outbreak. This is a case report that investigated a “natural experiment” in disease transmission. Prior to departure, the entire crew of a commercial fishing vessel was tested using reference laboratory PCR and was also tested for IgG to SARS-CoV-2 and specifically the presence of neutralizing antibodies. Whereas everyone tested negative using PCR on departure, one person was missed, who went on to infect 103 of 122 persons on the ship (88.5%). Prior to departure, six crew members tested positive for antibodies to SARS-CoV-2, of whom three had evidence of neutralizing antibodies (the authors believe the other three were false positives, which are relatively common with antibody tests). Here is the interesting finding: Whereas 103 of 117 crew without neutralizing antibodies were infected, none of the three with neutralizing antibodies were infected (p = 0.002). This suggests good immunity for those with neutralizing antibodies due to previous infection, although how long lasting it is remains an open question.

Written by Mark H. Ebell, MD, MS, on October 21, 2020. (Source: Addetia A, Crawford KHD, Dingens A, et al. Neutralizing antibodies correlate with protection from SARS-CoV-2 in humans during a fishery vessel outbreak with a high attack rate. J Clin Microbiol. 2020;58(11):E02107-20).

October 23, 2020, Research Update

Excess Mortality in the United States, Two-Thirds of Which Are Directly Attributed to COVID-19. As of October 15, 2020, the Centers for Disease Control and Prevention (CDC) estimates that more than 216,000 Americans have died from COVID-19 and reports that this is likely an underestimate. Measures of excess mortality have been used in previous pandemics and health disasters to mitigate the potential underestimation of mortality. These authors looked at weekly all-cause mortality rates going back several years to establish stable temporal trends. They also used data from the National Vital Statistics System to identify deaths caused by COVID-19. Between January 26, 2020, and October 3, 2020, the United States experienced 299,028 more deaths than what would historically have been expected. Two-thirds (66.2%) were attributed to COVID-19, and the rest were attributed to other causes. Younger age groups have lower death rates, so the authors also reported percentage changes in excess mortality. Surprisingly, the greatest percentage increase occurred in those between 25 and 44 years of age, who had a 26.5% increase over expected deaths. In Americans between 45 to 64, 65 to 74, 75 to 84, and 85 years of age or older, the percentage increase in deaths was 14.4%, 24.1%, 21.5%, and 14.7%, respectively. The report does not adjust for socioeconomic status, but the authors report that the average percentage increase was largest for Hispanic persons (53.6%) and that deaths were 28.9% above average for non-Hispanic American Indian or Alaska Native people, 32.9% above average for Black people, 34.6% above average for those of other or unknown race or ethnicity, and 36.6% above average for Asian people. Frequently updated data are available at . Parenthetically, the blip seen in December 2017 to January 2018 (see the figure on the CDC website) is probably from the H1N1 influenza epidemic.

Written by Henry C. Barry, MD, MS, on October 20, 2020. (Source: Rossen LM, Branum AM, Ahmad FB, et al. Excess deaths associated with COVID-19, by age and race and ethnicity — United States, January 26–October 3, 2020 [published online October 20, 2020]. MMWR Morb Mortal Wkly Rep. 2020. )

October 22, 2020, Research Update

United States Ranks Poorly in COVID-19 Death Rates Compared with Other Developed Countries. In this report, the investigators compare COVID-19 specific mortality per 100,000 individuals and excess all-cause mortality per 100,000 individuals in the United States with that of 18 other developed countries (mostly European countries, though also including South Korea, Japan, Australia, Canada, and Israel). The measurement period was the 38 weeks of the COVID-19 pandemic through September 19, 2020. To determine overall excess mortality during these 38 weeks, they compared data from 2020 to average mortality for 2015–2019 in each country. Of the 19 countries, the United States had the fourth highest COVID-19 mortality rate: 60.3 deaths per 100,000 people. Belgium had the highest mortality rate: 86.8 per 100,000; South Korea had the lowest: 0.7 per 100,000. Regarding overall excess deaths during this time period, data were available for 15 of the 19 countries. The United States had the third highest excess all-cause mortality: 71.6 per 100,000. Spain had the highest excess all-cause mortality: 102.2 per 100,000; Norway had the lowest: –2.6 per 100,000. The investigators performed additional analyses by shorter time periods, since May 10, 2020, and since June 7, 2020. The United States had the highest COVID-19 and excess all-cause mortality of all countries during these time periods. Of course, these numbers will change because we have entered the second wave of infections in the United States and European countries. 

Written by John Hickner, MD, MS, on October 17, 2020. (Source: Bilinski A, Emanuel EJ. COVID-19 and excess all-cause mortality in the US and 18 comparison countries [published online October 12, 2020.] JAMA. 2020. //jamanetwork.com/journals/jama/fullarticle/2771841)

October 21, 2020, Research Update

Excess Deaths in the United States during the COVID-19 pandemic, March 1 to August 1, 2020. There have been several studies of the excess number of deaths in the United States during the COVID-19 pandemic. This report is the best summary of excess deaths in the United States during the COVID-19 pandemic to date. The investigators obtained population and mortality data for 48 of the 50 states for 2014–2020 from the U.S. Census Bureau and the National Center for Health Statistics. (Data from North Carolina and Connecticut were excluded because of incomplete data.) Expected deaths, calculated with data from 2014–2019, were compared with observed deaths in 2020. Between March 1 and August 1, 2020, there were 1,336,561 deaths, which included 225,530 (20%) more than expected deaths, of which 150,541 (67%) were attributed to COVID-19. The 10 states with the highest per capita rate of excess deaths were New York, New Jersey, Massachusetts, Louisiana, Arizona, Mississippi, Maryland, Delaware, Rhode Island, and Michigan. The 23% of excess deaths NOT attributed to COVID-19 were likely due to a combination of increased deaths from other causes, deaths from causes in which COVID-19 may have contributed but were not listed as the cause of death on the death certificate, and deaths from respiratory illness for which COVID-19 was not diagnosed. 

Written by John Hickner, MD, MS, on October 14, 2020. (Source: Woolf SH, Chapman DA, Sabo RT, et al. Excess deaths from COVID-19 and other causes, March–July 2020 [published online October 12, 2020]. JAMA. 2020;324(15):1562-1564.)

October 20, 2020, Research Update

Final Report of ACTT-1 Study: Remdesivir Shortens Time to Recovery in Patients Hospitalized with COVID. We have covered remdesivir rather extensively. First, we wrote about an uncontrolled case series that provided no useful data as to its effectiveness. Later we discussed preliminary reports from the Adaptive COVID-19 Treatment Trial (ACTT-1) that compared 10 days of remdesivir with usual care in patients hospitalized with COVID. We have also reported on two studies that demonstrated comparability in outcomes with five and 10 days of treatment. Finally, we have reviewed negative studies. One is a ”living” systematic review () that was last updated September 11, 2020, and found no overall benefit of remdesivir on mortality or length of stay. We now have the final data from ACTT-1 that compares 14-day and 28-day outcomes of patients hospitalized with COVID treated with 10 days of remdesivir (n = 541) or 10 days of placebo (n = 521). Time to recovery was significantly faster in the remdesivir-treated group than in those receiving placebo (10 vs. 15 days, p < 0.001). There was also a trend toward lower 14-day mortality in the remdesivir group (6.7% vs. 11.9%) and also at 28 days (11.4% vs. 15.2%, respectively), but these were not statistically significant. If the mortality difference were significant, this means one would need to treat 20 and 27, respectively, to prevent one death at 14 days and 28 days. We have previously commented that this represents a lack of power. The rate of serious adverse events in the remdesivir-treated patients was lower (24.6%) than in those receiving placebo (31.6%; p = 0.01).

Written by Henry C. Barry, MD, MS, on October 13, 2020. (Source: Beigel JH, Tomashek KM, Dodd LE, et al.; ACTT-1 Study Group Members. Remdesivir for the treatment of Covid-19— final report [published online October 8, 2020]. N Engl J Med. 2020. )

October 16, 2020, Research Update

Symptoms of COVID-19 Persist at Least 60 Days in Two-Thirds of Patients with Mild to Moderate Infection. This report from one hospital in France describes persistent symptoms in patients who had mild to moderate symptoms of COVID-19 during the initial illness. Persistent symptoms were defined as weight loss ≥ 5%, severe dyspnea or asthenia/fatigue, chest pain, palpitations, anosmia/ageusia, headache, cutaneous signs, arthralgia, myalgia, digestive disorders, fever, or sick leave. Of 293 patients presenting during the study period, 150 had mild to moderate disease (i.e., not admitted to ICU, not a nursing home resident, and not dead) and were available for 30-day follow-up via phone call and medical records review. One hundred and thirty were available for follow-up at 60 days. At day 30, 69% (103/150) had at least one persistent symptom, and at day 60, 66% (86/130) did. At day 30, 28% had anosmia/ageusia, 37% had shortness of breath, and 50% had asthenia/fatigue. At day 60, 23% had anosmia/ageusia, 30% had shortness of breath, and 40% had asthenia/fatigue. More than one-third still felt sick or felt worse at day 60 than during the initial phase of the illness.

Written by John Hickner, MD, MS, on October 9, 2020. (Source: Carvalho-Schneider C, Laurent E, Lemaignen A, et al. Follow-up of adults with non-critical COVID-19 two months after symptoms' onset [published online October 5, 2020]. Clinical Microbiology and Infection. 2020. //www.clinicalmicrobiologyandinfection.com/article/S1198-743X(20)30606-6/fulltext)

October 15, 2020, Research Update

Early Data Suggest That mRNA Vaccine Increases Immune Response and Is Well-Tolerated in Elders. This report from a Phase I trial of the mRNA-1273 vaccine co-developed by researchers at the NIAID Vaccine Research Center and Moderna in Cambridge, Massachusetts, focuses on 20 adults between 56 and 70 years of age and another 20 older than 71 years of age. They administered two doses of vaccine, 25 μg or 100 μg, in two intramuscular doses administered four weeks apart. The researchers evaluated each participant seven and 14 days after each dose and then again 57 days after the first dose and also plan to follow the participants for up to a year after the second dose. The researchers inquired about specific symptoms and found that the most common were similar to other vaccines: headache, fatigue, myalgia, chills, and injection-site pain. These were moderately severe and most prevalent after the second dose. One participant did not get the second dose and did not have blood collected. Each participant had increased binding-antibody titers, but those receiving the higher dose vaccine had about four times the titers as those receiving the lower dose. Additionally, all participants had evidence of neutralizing antibodies. Stay tuned for the definitive studies on the effectiveness and safety of this vaccine.

Written by Henry C. Barry, MD, MS, on October 7, 2020. (Source: Anderson EJ, Rouphael NG, Widge AT, et al. Safety and immunogenicity of SARS-CoV-2 mRNA-1273 vaccine in older adults [published online September 29, 2020). N Engl J Med. 2020. //www.nejm.org/doi/full/10.1056/NEJMoa2028436)

October 14, 2020, Research Update

Secondary Spread of COVID-19 Occurred from 29% of Those Infected in India. This report summarizes the findings of a large COVID-19 contact-tracing program in two states in India. Daily house-to-house surveillance was used to identify individuals with symptoms, and daily follow-up was used for all contacts of laboratory-confirmed or suspect COVID-19 cases, with the aim of testing these individuals 5 to 14 days after their contact with a primary case, irrespective of symptoms. As of August 2020, individual-level epidemiological data on cases and contacts were available for 575,071 tested contacts of 84,965 confirmed cases. COVID-19 positive contacts were identified for 29% of index cases for whom reliable contact-tracing data was available; 71% of index cases had no secondary cases. Assuming test-positive contacts were infected by the index case to whom they were traced, the overall secondary attack rate (or risk of transmission from an index case to an exposed contact) was 10.7% for high-risk contacts, those who had close social contact or direct physical contact with index cases without protective measures, and 4.7% for low-risk contacts, those who were in the proximity of index cases but did not meet these criteria for high-risk exposure. For the 78 individuals with high-risk travel exposure (close proximity to an infected individual in a shared vehicle for ≥6 hours, the secondary attack rate was 79.3%. These findings are consistent with the concept of “super-spreaders” and the high risk of being in small, enclosed spaces for prolonged periods of time.

Written by John Hickner, MD, MS, on October 7, 2020. (Source: Epidemiology and transmission dynamics of COVID-19 in two Indian states [published online September 30, 2020]. Laxminarayan R, Wahl B, Dudala SR, et al. Science. 2020. //science.sciencemag.org/content/early/2020/09/29/science.abd7672)

October 13, 2020, Research Update

Rapid Antigen Tests Lack Sensitivity for Diagnosing COVID-19. We mostly know the Cochrane Collaboration for writing systematic reviews of the effectiveness of interventions, but they have recently begun to explore diagnostic meta-analyses. This systematic review identified published and preprint studies that compared a rapid antigen test or a point-of-care molecular test to a reference laboratory rtPCR test as the reference standard. In a few studies, clinical criteria were used to define a positive case. Specificity was evaluated using prepandemic stored samples. The news is not good for rapid antigen tests, with a mean sensitivity of 56.2% (95% CI, 29.5% to 79.8%) and mean specificity of 99.5% (95% CI, 98.1% to 99.9%). Given a pretest probability of 5% (for example, 5% prevalence in a hypothetical sample of 1,000 patients), the corresponding positive predictive value (PV+) is 85%, and the negative predictive value (PV-) is 98%. If pretest probability is 10%, corresponding PV+ and PV- are 92% and 95%, respectively. Rapid molecular tests performed better, with a mean sensitivity of 95.2% (95% CI, 86.7% to 98.3%) and mean specificity of 98.9% (95% CI, 97.3% to 99.5%). The PV+ of these tests was very similar to the rapid antigen tests over a range of pretest probabilities from 5% to 15%, but the PV- was much better (99% to 100%). With regard to individual point-of-care molecular tests, the Abbott ID Now was less sensitive than the Cepheid Xpert Xpress (76.8% vs. 99.4%), but it was slightly more specific (99.6% vs. 96.8%). At a 10% pretest probability, that corresponds to PV+ of 96% for Abbott and 77% for Cepheid but PV- of 97% for Abbott and 100% for Cepheid. Importantly, the Abbott ID Now will miss about one-quarter of infected persons based on its sensitivity. This test was used in the White House; the correct strategy would be to first test using the highly sensitive Cepheid test and then confirm any positives with the highly specific Abbott test.

Table. Accuracy of rapid antigen and point-of-care tests with PV+ and PV- calculated based on 10% prevalence.  
 

Test (studies)	Sensitivity	Specificity	LR+	LR-	PV+	PV-
Rapid antigen tests (8)	56.2%	99.5%	112	0.44	92%	95%
Molecular tests (11)	95.2%	98.9%	87	0.05	90%	99%
Abbott ID Now (5)	76.8%	99.6%	192	0.23	96%	97%
Cepheid Xpert Xpress (6)	99.4%	96.8%	31	0.01	77%	100%

LR = likelihood ratio; PV = predictive value; + = positive; - = negative.

Written by Mark Ebell MD, MS, on October 6, 2020. (Source: Dinnes J, Deeks J, Adriano A, et al.; Cochrane COVID-19 Diagnostic Test Accuracy Group. Rapid, point‐of‐care antigen and molecular‐based tests for diagnosis of SARS‐CoV‐2 infection [published online August 26, 2020]. Cochrane Database Syst Rev. 2020. //www.cochranelibrary.com/cdsr/doi/10.1002/14651858.CD013705/full).

October 12, 2020, Research Update

PCR Sensitivity for Detection of COVID-19 Declines Rapidly over Time. These researchers identified seven published studies that evaluated the accuracy of reference laboratory rtPCR for the detection of SARS-CoV-2 at various points in time in relation to infection and symptom onset. They then fitted a Bayesian hierarchical logistic regression model using these data (the hierarchical part is necessary because there were seven different sets of data). They estimated a median duration from infection to symptom onset of five days and a pretest probability of infection of 11% based on studies of close household contacts in China. They then graphed the estimated rate of false negatives by day from infection. It begins at 100% (all infected cases are missed), then declines beginning on day 4 to 70%, then approximately 40% at symptom onset, bottoming out at 20% for the first five days of symptoms, then rising steadily to 50% by day 21. The Bayesian part of their model uses that pretest probability of 11% to estimate the posttest probability of infection given a negative rtPCR. That probability goes as low as about 3%, rising to 5% on day 21. The bottom line is that timing matters, and if a patient is clinically highly suspicious for COVID-19, repeat the test a day later, considering that the false negative rate was lowest eight days after exposure. On the other hand, PCR is highly specific, so if the patient is positive, do not repeat the test to confirm it; you will likely be misled by a false negative result if that second test is negative.

Written by Mark H. Ebell MD, MS, on October 6, 2020. (Source: Kucirka LM, Lauer SA, Laeyendecker O, et al. Variation in false-negative rate of reverse transcriptase polymerase chain reaction–based SARS-CoV-2 tests by time since exposure. Ann Intern Med. 2020;173(4):262-267.)

October 7, 2020, Research Update

Another Negative HCQ Study: Ineffective in Preexposure Prophylaxis in Health Care Workers (PATCH). This study took place in two hospitals in Philadelphia where the researchers recruited health care workers who worked at least 20 hours per week in emergency departments or dedicated COVID-19 units and who had no symptoms and no prior COVID. They randomized the staff to receive a placebo (n = 66) or hydroxychloroquine 600 mg daily (HCQ, n = 66). The researchers performed COVID PCR testing and antibody testing at baseline and again after four weeks and eight weeks. More of the HCQ-treated patients were women. At the end of the study, whereas 125 (95%) had complete follow-up, 22 of the 125 (18%) had discontinued their treatment (similar rate for each group). Four participants in each group tested positive for COVID. Using strict intention-to-treat, this would translate to 6.1% in each group. The authors report modified intention-to treat rates of 6.6% for the placebo-treated participants and 6.4% in the HCQ-treated ones. Regardless, no meaningful difference occurred. Of those who converted, six became symptomatic, but none required hospitalization. Adverse effects (diarrhea, anorexia, and nausea were most commonly reported) were generally mild but more prevalent in the HCQ-treated participants (45% vs. 26%; number needed to harm =6, 95% CI 3 to 50).

Written by Henry C. Barry, MD, MS, on October 4, 2020. (Source: Abella BS, Jolkovsky EL, Biney BT, et al. Efficacy and safety of hydroxychloroquine vs placebo for pre-exposure SARS-CoV-2 prophylaxis among health care workers: a randomized clinical trial [published online September 30, 2020]. JAMA Intern Med. 2020. //jamanetwork.com/journals/jamainternalmedicine/fullarticle/2771265)

October 6, 2020, Research Update

Outcomes in Hospitalized Young Adults with COVID-19 in the United States. With campuses reopening across the United States, a surge in cases has occurred among young adults in many communities this fall. This study used a commercial registry (Premier Healthcare Database) with information from more than 8 million admissions to 1,030 U.S. hospitals to identify nonpregnant adults 18 to 34 years of age with an initial hospitalization for COVID-19 between April 1 and June 30, 2020. They identified 3,222 nonpregnant young adults who were admitted to 419 hospitals for COVID-19, with a mean of 28 years of age. A majority (58%) were men, and 57% were Black or Hispanic patients. Common comorbidities included obesity (37%), morbid obesity (25%), diabetes mellitus (18%), asthma (17%), and hypertension (16%). Of the 3,222 patients, 21% required intensive care, 10% required mechanical ventilation, and 2.7% died during the hospitalization. Morbid obesity (adjusted odds ratio [aOR] 2.30), hypertension (aOR 2.36), and male sex (aOR 1.53) were independent risk factors for death or mechanical ventilation. Patients with the combination of morbid obesity, hypertension, and diabetes had a mortality risk similar to that of adults 35 to 64 years of age without these conditions.

Written by Mark H. Ebell MD, MS, on October 1, 2020. (Source: Cunningham JW, Vaduganathan M, Claggett BL. Clinical outcomes in young US adults hospitalized with COVID-19 [published online September 9, 2020]. JAMA Intern Med. 2020. //jamanetwork.com/journals/jamainternalmedicine/fullarticle/2770542)

September 30, 2020, Research Update

Systematic Review Finds Children and Adolescents Have Lower Susceptibility to COVID-19 but Has Limited Data on Infectivity. These authors did a limited systematic review (they searched PubMed, a preprint server, and a public health website) to identify studies reporting contact tracing or prevalence of proven COVID-19 in children and adolescents. The studies also had to include secondary infection rates. Adding this inclusion criterion may have limited the number of potential studies compared with looking at prevalence and secondary infection alone. In any case, they included 32 studies with 41,640 children and adolescents plus 268,945 adults. Eighteen studies were contact tracing or school-based studies, and 14 were population-based. Only two of the studies, both population-based, were judged to be high quality. Overall, compared with adults, the pooled odds ratio of a child or adolescent being infected was 0.56 (95% CI, 0.37 to 0.85). The authors stratified their analysis by study quality and by age (children vs. adolescents) and still found marked residual heterogeneity. Not surprising, the studies came from all over the world, raising the background and unmeasured issue of how testing strategies vary and likely contributed to the marked heterogeneity in the data. Three medium-quality school-based contact-tracing studies with only 36 index cases found no evidence of child-to-adult transmission. Bottom line, whereas it appears that children and adolescents are less susceptible to contract SARS-CoV-2 than adults, the data on infectivity is inadequate to draw conclusions. Once again, our mantra is to proceed with an abundance of caution.

Written by Henry C. Barry, MD, MS, on September 27, 2020. (Source: Viner RM, Mytton OT, Bonell C, et al. Susceptibility to SARS-CoV-2 infection among children and adolescents compared with adults: a systematic review and meta-analysis [published online September 25, 2020]. JAMA Pediatr. 2020. //jamanetwork.com/journals/jamapediatrics/fullarticle/2771181)

September 29, 2020, Research Update

SHERLOCK One-Pot Testing Is 93% Sensitive and 98% Specific in Detecting SARS-CoV-2. If you are a nerdy type who is fascinated by lab methods and application of new technology, this paper is for you! I won’t go into all the details, but the terminology is truly fascinating and colorful. These researchers used nasopharyngeal swabs from 402 patients who were tested with the CDC’s standard antigen assay, 202 of whom had tested positive. They applied their new test that is simpler to perform and can be performed at a single temperature in less than an hour: specific high-sensitivity enzymatic reporter unlocking (SHERLOCK) testing in one pot. The authors report that in blind testing at an external laboratory, the new test was 93.1% sensitive and 98.5% specific. That translates to a positive likelihood ratio of 62 and negative likelihood ratio of 0.07. In other words, this test is pretty good at ruling in and ruling out SARS-CoV-2 infection. Now we just need confirmatory studies so that it can become available!

Written by Henry C. Barry, MD, MS, on September 24, 2020. (Source: Joung J, Ladha A, Saito M, et al. Detection of SARS-CoV-2 with SHERLOCK One-Pot Testing [published online September 16, 2020]. N Engl J Med. 2020. //www.nejm.org/doi/full/10.1056/NEJMc2026172)

September 24, 2020, Research Update

Stopping the COVID-19 Pandemic with a Vaccine Would Require an Effective Vaccine and a High Vaccination Rate. We do not know how effective a vaccine will be in stopping the COVID-19 pandemic. These researchers used a computational modeling technique to determine how effective a vaccine must be and what proportion of residents of the United States would need to take the vaccine to reduce transmission sufficiently to prevent a pandemic or to stop an existing pandemic. To prevent a pandemic, if 100% of U.S. residents received the vaccine, vaccine efficacy would need to be 60%; if 75% of U.S. residents received the vaccine, it would need to be 70% effective in preventing individual infection; and if 60% of U.S. residents received the vaccine, it would need to be 80% effective to prevent a pandemic. To stop the current COVID-19 pandemic, the vaccine efficacy would need to be at least 60% if received by 100% of people and at least 80% if received by 75% of people to reduce the peak incidence by 85% to 86%, 61% to 62%, and 32% when vaccination occurs after 5%, 15%, and 30% of the population, respectively, have already been exposed to the COVID-19 coronavirus. The bottom line, according to the investigators, is that to extinguish the ongoing COVID-19 pandemic, the vaccine must have an efficacy of at least 80%, and 75% of people must receive it. Because this high vaccination coverage seems unlikely to happen soon, other measures such as social distancing and wearing masks will probably be important preventive measures for the foreseeable future.

Written by John Hickner on September 22, 2020. (Source: Bartsch SM, O’Shea KJ, Ferguson MC, et al. Vaccine efficacy needed for a COVID-19 coronavirus vaccine to prevent or stop an epidemic as the sole intervention. Am J Prev Med. 2020;59(4):493−503.)

September 23, 2020, Research Update

Antibodies Persist for Four Months after Infection; Infection Fatality Ratio 0.3%, Much Higher in Older Patients. This Icelandic study measured a variety of IgG and IgM antibodies to SARS-CoV-2 in six groups of Icelanders who had either never been tested or had tested negative and in two groups of Icelanders who had tested positive and had recovered. They weighted the sample by age and sex to estimate the number of Icelanders who had been infected and used that to estimate the infection fatality ratio (deaths/[symptomatic + asymptomatic positives]) and case fatality ratio (deaths/symptomatic positives). Among patients who had experienced a symptomatic infection, IgG antibody levels remained stable over a four-month period, which is encouraging in terms of at least medium-term immunity. Those who smoked and those taking anti-inflammatory medications had lower antibody levels, although it is not clear whether this difference is clinically significant. This was not the case for IgM and IgA antibodies, which is what you would expect. They also estimate that the infection fatality ratio was 0.3% (95% CI, 0.2% to 0.6%), and the case fatality ratio was 0.6% (95% CI, 0.3% to 1.0%). The infection fatality ratio was 0.1% (95% CI, 0.0% to 0.3%) in people 70 years or younger and 4.4% in those older than 70 years (95% CI, 1.9% to 8.4%).

Written by Mark H. Ebell MD, MS, on September 20, 2020. (Source: Gudbjartsson DF, Norddahl GL, Melsted P, et al. Humoral immune response to SARS-CoV-2 in Iceland [published online September 1, 2020]. N Engl J Med. 2020. DOI: 10.1056/NEJMoa2026116).

September 22, 2020, Research Update

Clinical Data and Labs Can Predict Mortality in Patients Hospitalized with COVID-19. This consortium of researchers used data from 35,463 patients hospitalized with COVID-19 to develop a clinical score, the Coronavirus Clinical Characterisation Consortium Mortality Score (4C Mortality Score) to predict mortality and then validated it on another sample of 22,361 patients. All the patients were adults admitted to one of 260 hospitals in England, Scotland, and Wales. They used robust approaches to guide this development and validation process. From 41 candidate variables, the final model (0 to 21 points, see Table 1 in the cited source) included age, sex, number of comorbidities, respiratory rate, oxygen saturation, level of consciousness, urea level, and C reactive protein. The overall mortality rate in the development cohort was 32% and 30% in the validation cohort. In the validation cohort, the model was 77% accurate, which they report is better than other scores. The model was generally best at ruling out mortality than ruling it in, as summarized in Table 2 in the source cited here. Finally, the score has limited applicability and would not be useful in the community setting with patients at lower risk of death because well over half of the patients were in the high-risk group.

Written by Henry C. Barry, MD, MS, on September 15, 2020. (Source: Knight SR, Ho A, Pius R, et al. Risk stratification of patients admitted to hospital with covid-19 using the ISARIC WHO Clinical Characterisation Protocol: development and validation of the 4C Mortality Score. BMJ. 2020;370:m3339.)

September 21, 2020, Research Update

People Dining at a Restaurant or Going to a Bar/Coffee Shop More Likely to Have COVID-19. In this case-control study, 615 potential case-patients and 1,212 control-participants were randomly selected from those tested for SARS-CoV-2 at 11 sites from July 1 to 29 and were contacted by phone 14 to 23 days after the date they received SARS-CoV-2 testing to complete a survey. Among 802 adults contacted, 314 individuals were included in the final analyses; 154 who tested positive, and 160 who tested negative. All participants were symptomatic at the time of their SARS-CoV-2 test. Participants were asked about close contact (within six feet for ≥ 15 minutes) with a person with known COVID-19, workplace exposures, mask-wearing behavior, and community activities ≤ 14 days before symptom onset. Close contact with a person with known COVID-19 was more commonly reported among cases (42%) than among controls (14%). No significant differences were observed in the bivariate analysis between case-patients and control-participants in shopping, gatherings with ≤ 10 persons in a home, going to an office setting, going to a salon, gatherings with > 10 persons in a home, going to a gym, using public transportation, going to a bar/coffee shop, or attending a church/religious gathering. Case-patients were more likely to have reported dining at a restaurant (adjusted odds ratio [aOR] = 2.4, 95% CI = 1.5 to 3.8) in the two weeks before illness onset than controls. When the analysis was restricted to the 225 participants who did not report recent close contact with a person with known COVID-19, cases were more likely than were controls to have dined at a restaurant (aOR = 2.8, 95% CI = 1.9 to 4.3) or going to a bar/coffee shop (aOR = 3.9, 95% CI = 1.5 to 10.1). One limitation the authors report is the inability to evaluate bar and coffee shop exposure separately, which may represent different risks. 

Written by John Hickner, MD, MS, on September 14, 2020. (Source: Fisher KA, Tenforde MW, Feldstein LR, et al. Community and close contact exposures associated with COVID-19 among symptomatic adults ≥18 years in 11 outpatient health care facilities—United States, July 2020. MMWR Morb Mortal Wkly Rep. 2020;69(36):1258–1264.)

September 18, 2020, Research Update

MRI Evidence of Myocarditis in Ohio State University Athletes with COVID-19. Due to reports of cardiac involvement in people infected with COVID-19, all collegiate athletes at Ohio State University with COVID were referred for cardiac magnetic resonance imaging (MRI) after quarantine (11 to 53 days). On the day of the MRI, the athletes also had electrocardiograms, echocardiograms, and serum troponin I performed. Twenty-six athletes (15 males, 11 females) agreed to participate. They do not report how many declined to participate nor what proportion of all athletes this represented. None of the athletes required hospitalization, 12 (46%) had mild symptoms and the remainder were asymptomatic. None of the athletes had electrocardiographic evidence of myocarditis or elevated troponin I levels. The authors also report that none of the athletes had echocardiographic or MRI evidence of abnormal ventricular enlargement or dysfunction. Four athletes (15%), all male, however, had MRI evidence of myocarditis. Additionally, 12 athletes (46%) had late gadolinium enhancement, indicative of past myocardial injury or perhaps part of the athlete's heart syndrome. Finally, it is unclear as to the clinical meaning and the prognosis of finding subclinical myocarditis. As with many things we have learned from COVID, we ought to proceed “with an abundance of caution.” 

Written by Henry C. Barry, MD, MS, on September 12, 2020. (Source: Rajpal S, Tong MS, Borchers J, et al. Cardiovascular magnetic resonance findings in competitive athletes recovering from COVID-19 infection [published online September 11, 2020]. JAMA Cardiol. 2020. //jamanetwork.com/journals/jamacardiology/fullarticle/2770645)

September 17, 2020, Research Update

Systematic Review Summarizes Myriad of Cardiovascular Manifestations among People with COVID-19. In a clumsily written article on a preprint server (not peer-reviewed) where the paper is long on narrative and mechanistic hypotheses and short on synthesis, the authors systematically searched PubMed and EMBASE as well as the reference lists of retrieved publications and review articles to identify papers of any variety that reported on cardiovascular outcomes among people with COVID-19. They do not describe assessing the quality of the included studies. Ultimately, they reviewed 86 papers and reported multiple cardiovascular manifestations with little summary data. They report that about 20% (range 19% to 28%) of patients with COVID-19 sustained cardiac injury or manifestations of myocarditis and 50% of those persons died. They also report that about 20% of patients experience thromboembolic manifestations and 30% of those persons died. Additionally, they report that arrhythmias are more common among patients admitted to intensive care units (44%) compared with 6% of those not requiring intensive care. There were 3% of patients experienced a stroke, and they estimate about that 40% of those persons died. Finally, they report data among patients who present with myocardial infarction but found no data on the rate of myocardial infarction or acute coronary syndrome among patients with COVID.

Written by Henry C. Barry, MD, MS, on September 10, 2020. (Source: Thakkar S, Arora S, Kumar A, et al. A systematic review of the cardiovascular manifestations and outcomes in the setting of coronavirus-19 disease [pre-print, not peer reviewed, published online August 12, 2020]. )

September 16, 2020, Research Update

Low Vitamin D Level Associated with Higher Risk of COVID-19. This retrospective cohort study determined the incidence of COVID-19 in those with and without vitamin D deficiency, as measured within the prior year. Of 489 people, vitamin D status was categorized as likely deficient for 124 participants (25%), likely sufficient for 287 (59%), and uncertain for 78 (16%). Overall, 71 participants (15%) tested positive for COVID-19. In the multivariate analysis, testing positive for COVID-19 was associated with likely deficient vitamin D status (relative risk, 1.77; 95% CI, 1.12 to 2.81; P = .02) compared with likely sufficient vitamin D status. Whether supplementation will reduce the risk of contracting COVID-19 or the severity of infection is not yet known. Low vitamin D levels are associated with many adverse outcomes, but nearly all randomized trials have NOT shown any benefit of vitamin D supplementation. More than 40 randomized trials of vitamin D supplementation to treat or prevent COVID-19 are listed on clinicaltrials.gov, so an answer about vitamin D and COVID-19 should be available within the next year.

Written by John Hickner, MD, MS, on September 6, 2020. (Source: Meltzer DO, Best TJ, Zhang H, et al. Association of vitamin D status and other clinical characteristics with COVID-19 test results. JAMA Netw Open. 2020;3(9):e2019722.)

September 11, 2020, Research Update

Meta-Analysis Shows Pregnant Women Less Likely to Manifest COVID-19 Symptoms but More Likely to Go to Intensive Care Than Nonpregnant Women. These authors used modern meta-analytic methods, including searching multiple databases, to find studies that reported data on pregnant women with suspected or confirmed COVID-19. This is a “living systematic review” with frequent updates, including monthly reporting through a dedicated website (). For this publication, they included 77 studies with 13,118 pregnant and recently pregnant women with COVID-19 and 83,486 nonpregnant women of reproductive age with COVID-19. Roughly one-third of the studies came from the United States and one-third from China. All of the studies used PCR to confirm the presence of COVID-19. Overall, for study trustworthiness (internal validity), 75 (96%) of the studies were at low risk of bias. However, only 10 of the studies were thought to be low risk of bias when it comes to representativeness (external validity). Overall, the rate of COVID-19 in pregnant and recently pregnant women was 10% (95% CI 7% to 14%), although the rate varied by the testing strategy. For example, in the studies that used universal testing, 7% of the pregnant women (range 4% to 10%) had COVID compared with 18% (range 10% to 28%) in the studies that tested based on symptoms. In the 26 studies that reported mortality data, 73/11,580 (0.1% but with high heterogeneity) of pregnant women with COVID-19 died. They fail to report the same rates for nonpregnant women or for uninfected pregnant women. Overall, pregnant women were less likely to have fever (OR 0.43, 95% CI 0.22 to 0.85, but high heterogeneity) and myalgias (OR 0.48, 95% CI 0.45 to 0.51; no heterogeneity). Compared with nonpregnant women of reproductive age with COVID-19, pregnant women with COVID-19 were more likely to require intensive care (OR 1.62, 95% CI 1.33 to 1.96; no heterogeneity) or mechanical ventilation (OR 1.88, 95% CI 1.36 to 2.60; no heterogeneity). Among the women who died, higher mortality was associated with increased maternal age, high body mass index, chronic hypertension, and preexisting diabetes. The presence of comorbidity was also associated with a higher risk of intensive care admission and mechanical ventilation. Overall, 17% (95% CI 13 to 21) of pregnant women with COVID-19 had preterm births and 6% (95% CI 3 to 9) had a spontaneous preterm delivery. Additionally, 18 of 2,837 (0.6%) pregnancies among women with COVID resulted in a stillbirth, and six of 1,728 neonates (0.3%) died. Overall, these data suggest that pregnant women who have COVID-19 are less likely to have typical symptoms and are more likely to require intensive care or mechanical ventilation. Although this paper is data intense, the limited reporting of comparison data prevents making too many conclusions.

Written by Henry C. Barry, MD, MS, on September 4, 2020. (Source: Allotey J, Stallings E, Bonet M, et al. Clinical manifestations, risk factors, and maternal and perinatal outcomes of coronavirus disease 2019 in pregnancy: living systematic review and meta-analysis. BMJ. 2020;370:m3320.)

September 10, 2020, Research Update

Meta-analysis of Seven RCTs Confirms That Systemic Corticosteroids Reduce 28-Day COVID-19 Mortality. The World Health Organization has created a Rapid Evidence Appraisal for COVID-19 Therapies (REACT) working group to perform an ongoing series of meta-analyses. This meta-analysis did a careful search of clinical trial registries to identify all ongoing or completed studies comparing a systemic corticosteroid with placebo or usual care in patients who were critically ill with COVID-19. Trials that included patients with mild or moderate disease or that had not recruited patients were excluded. They identified nine clinical trials, and seven agreed to share data with the REACT team for meta-analysis. Trials were done in six European countries, the United States, Canada, Brazil, New Zealand, Australia, and China. The median age of participants was 60 years, 29% were women, and all but one study required ICU admission. Data from the large UK RECOVERY trial was limited to the subgroup of patients who were mechanically ventilated. Risk of bias was assessed as low using the Cochrane Risk of Bias Tool for six of seven trials. In total, 678 patients were randomized to a corticosteroid and 1,025 to usual care or placebo. There was good consistency across trials in their results, with the I2 of 16% (that means that only 16% of the identified variance was between studies and that 84% was within studies). Corticosteroids significantly reduced mortality (odds ratio 0.66, 95% CI 0.53 to 0.82; risk ratio 0.80, 95% CI 0.70 to 0.91). I think risk ratios are more appropriate because odds ratios are only an approximation of relative risk and are most reliable when outcomes are rare (not the case here). Given a 41.4% mortality rate in the usual care/placebo group, applying the risk ratio of 0.80 yields a 20% mortality reduction to 33.2%, which corresponds to a number needed to treat to prevent one death of 12. They found no difference in outcomes by age, dose of steroid, or duration of illness prior to enrollment. Benefit was smaller for patients not receiving mechanical ventilation at randomization and possibly for those receiving vasoactive agents at randomization. This provides support for use of low-dose corticosteroids in critically ill patients but should not be applied to outpatients or patients not requiring supplemental oxygen.

Written by Mark H. Ebell MD, MS, on September 3, 2020. (Source: The WHO Rapid Evidence Appraisal for COVID-19 Therapies (REACT) Working Group. Association between administration of systemic corticosteroids and mortality among critically ill patients with COVID-19: a meta-analysis [published online September 2, 2020]. JAMA. 2020. //jamanetwork.com/journals/jama/fullarticle/2770279)

September 9, 2020, Research Update

Hydrocortisone Likely to Reduce Need for Organ Support Interventions in Critically Ill COVID-19 Patients. The REMAP-CAP randomized clinical trials are adaptive trials designed to test several COVID-19 interventions in 121 sites in eight countries. For the hydrocortisone study, 384 of 403 enrolled critically ill patients with COVID-19 completed the open-label randomized trial in these three groups: fixed-dose hydrocortisone (137), shock-dependent (146), and no cortisone (101). The primary outcome was organ support-free days, defined as days alive and free of ICU-based respiratory or cardiovascular support within 21 days. The mean for organ support-free days was 11.5 days for the fixed-dose hydrocortisone group, 9.5 days for the shock-dependent hydrocortisone group, and 6 days for the no-hydrocortisone group. In the Bayesian analysis, there was a 93% probability of superiority for the fixed-dose hydrocortisone group over the no-hydrocortisone group and an 80% probability of superiority of shock-dependent dosing of hydrocortisone over no hydrocortisone. The trial was stopped early because of publication of the positive results from the RECOVERY trial on June 16, 2020, because the investigators believed that there was no longer clinical equipoise and that hydrocortisone should not be withheld from critically ill COVID-19 patients. A meta-analysis published simultaneously with this study confirms the benefit of corticosteroids for severe COVID-19 infection.

Written by John Hickner, MD, MS, on September 4, 2020. (Source: The Writing Committee for the REMAP-CAP Investigators. Effect of hydrocortisone on mortality and organ support in patients with severe COVID-19: The REMAP-CAP COVID-19 corticosteroid domain randomized clinical trial [published online September 2, 2020]. JAMA. 2020. //jamanetwork.com/journals/jama/fullarticle/2770278)

September 8, 2020, Research Update

Clinical Course and Viral Shedding of Children with COVID-19 in Korea. This case series of children younger than 19 years with laboratory-confirmed COVID-19 infections includes 91 of the 119 cases, symptomatic and asymptomatic, diagnosed in Korea from February 18, 2020, to March 31, 2020. These children were identified by the extensive contact-tracing system in Korea, and they were monitored closely for symptoms and viral shedding. The median age was 11 (range, 27 days to 18 years). Twenty children (22%) remained asymptomatic. Sixty percent had respiratory symptoms, 55% had nonspecific systemic symptoms, and 18% had gastrointestinal symptoms. Only two required supplemental oxygen, and none died. Mean viral shedding was 14.1 days in asymptomatic patients, 18.7 days for those with upper respiratory symptoms, and 19.9 days for those with lower respiratory track symptoms. Despite close surveillance, 52% of children had symptoms a median of three days prior to diagnosis, indicating that their symptoms were initially mild and nonspecific. The authors conclude that asymptomatic children and children with minimal symptoms may be a significant source of exposure to SARS-CoV-2.  

Written by John Hickner, MD, MS, on August 31, 2020. (Source: Han MS, Choi EH, Chang SH, et al. Clinical characteristics and viral RNA detection in children with coronavirus disease 2019 in the Republic of Korea [published online ahead of print August 28, 2020]. JAMA Pediatr. 2020. //jamanetwork.com/journals/jamapediatrics/fullarticle/2770150)

September 4, 2020, Research Update

Incidence of SARS-CoV-2 among Asymptomatic Children at 28 U.S. Hospitals Ranges from 0% to 2.2%. These authors compiled data from the charts of pediatric otorhinolaryngologists at 28 hospitals around the United States. As part of a quality-improvement initiative, 33,041 asymptomatic children had PCR testing before elective surgery, clinic visits, or hospital admissions. Overall, 250 (0.8%) children had positive tests, but the rates varied among the hospitals from 0% to 2.2%. They also found that the weekly incidence among the children generally followed the incidence in the general population. It is unclear whether children seeking care from otorhinolaryngologists are a reflection of the general population. Although the overall incidence among asymptomatic children is low, the amount of variability will make it difficult to make general recommendations about return to school.

Written by Henry C. Barry, MD, MS, on August 30, 2020. (Source: Sola AM, David AP, Rosbe KW, et al. Prevalence of SARS-CoV-2 infection in children without symptoms of coronavirus disease 2019 [published online August 25, 2020]. JAMA Pediatr. 2020. //jamanetwork.com/journals/jamapediatrics/fullarticle/2769878)

September 3, 2020, Research Update

Racial and Ethnic Disparities in COVID-19 Infection and Hospitalization Rates in the United States. It is well known that because of health and social inequities, COVID-19 disproportionately affects Black and Hispanic people in the United States. The following two epidemiological studies provide quantitative estimates of this increased risk of COVID-19. In the first study, from the CDC, the investigators summarize data from 205 hotspot counties in 33 states where COVID-19 cases were rising rapidly between June 5 to 18. Seventy-nine of these counties in 22 states reported race and ethnicity for at least 50% of cases, and in 96.2% of these counties, the proportion of cases in minorities was significantly greater than the population proportion in the county (see table).

Racial/Ethnic group	Mean of estimated differences, % (range)	Mean of estimated ratios of proportion of cases to proportion of population (range)
Hispanic/Latino	30.2 (8.0 to 68.2)	4.4 (1.2 to 14.6)
Black/African American	14.5 (2.3 to 31.7)	2.3 (1.2 to 7.0)
American Indian/Alaska Native	39.3 (16.4 to 57.9)	4.2 (1.9 to 6.4)
Asian	4.7 (2.7 to 6.8)	2.9 (2.0 to 4.7)
Native Hawaiian/Other Pacific Islander	4.5 (0.1 to 31.5)	8.5 (2.7 to 18.4)

The other study analyzed hospitalizations in 12 U.S. states during a two-month period. There were 48,788 hospitalizations among a population of 66,796,666 individuals. The proportion of White individuals hospitalized for COVID-19 was considerably less than the population proportion in all 12 states, and the proportion of Black patients was greater than their population proportion in all 12 states. The greatest differences for Black patients were in Ohio (31.8% hospitalized vs. 13.0% of the state population), Minnesota (24.9% vs. 6.8%), Indiana (28.1% vs. 9.8%), and Kansas (22.0% vs. 6.1%). Eleven states report Hispanic ethnicity, and in 10 of these the hospitalized proportion was greater than the population proportion (Virginia [36.2% vs. 9.6%], Utah [35.3% vs. 14.2%], and Rhode Island [33.0% vs 15.9%]). Similar findings were reported for American Indian and Alaskan Native populations in eight states (Arizona, 15.7% of hospitalizations vs. 4.0% of the population). Asian Americans were less likely to be hospitalized.  

Written by John Hickner, MD, MS, on August 29, 2020. (Source: Moore JT, Ricaldi JN, Rose CE, et al. Disparities in incidence of COVID-19 among underrepresented racial/ethnic groups in counties identified as hotspots during June 5–18, 2020 — 22 States, February–June 2020. MMWR Morb Mortal Wkly Rep. 2020;69(33):1122–1126. And Karaca-Mandic P, Georgiou A, Sen S. Assessment of COVID-19 hospitalizations by race/ethnicity in 12 states [published online, August 17, 2020]. JAMA Intern Med. 2020. //jamanetwork.com/journals/jamainternalmedicine/fullarticle/2769369)

September 2, 2020, Research Update

Similar Outcomes of Five Days of Remdesivir Vs. 10 Days of Remdesivir Vs. Usual Care in Patients with Moderate COVID-19 Pneumonia. We have previously summarized several studies of remdesivir published in NEJM. A Research Brief posted on April 17, 2020, described a case series with 53 patients with severe COVID-19. The ACTT study reported in a Research Brief posted on April 30, 2020 (updated in a Research Brief on June 2, 2020), was a randomized trial of 1,063 patients with COVID-19 with lung involvement that found speedier recovery and a trend toward lower mortality. In contrast, a Chinese randomized trial with 237 patients with COVID-19 pneumonia and hypoxia found no improvement in mortality. We also summarized a flawed study (Research Brief from June 3, 2020) that found comparable outcomes of five days and 10 days of remdesivir in patients with severe COVID-19. Finally, a recent systematic review (Research Brief from August 28, 2020, based on a study in BMJ) found that remdesivir had no overall benefit on mortality, use of mechanical ventilation, duration of mechanical ventilation, viral clearance, or hospital length of stay but that it decreased symptom duration and was unlikely to cause adverse events. We now have another randomized trial from the same team that compared five days and 10 days in patients with severe COVID-19 that was not included in the aforementioned systematic review. The researchers randomized patients hospitalized with moderate COVID-19 pneumonia (not requiring mechanical ventilation, no ARDS) to receive, unblinded, 10 days of remdesivir (n = 197), five days of remdesivir (n = 199), or usual care (n = 200). This study used a 7-point scale (ranging from death to hospitalized with various measures to not hospitalized) and had significant differences in baseline characteristics of the patients. For example, the patients randomized to 10 days of remdesivir were slightly more likely to have diabetes (44% vs. 37% and 38%), and about 10% of patients randomized to either remdesivir arms also received hydroxychloroquine compared with nearly half of the patients receiving usual care. Regardless, they report that after 11 days, compared with those receiving usual care, the distribution of clinical status based on the 7-point scale was significantly better statistically for those receiving five days of remdesivir but no different for those receiving 10 days of remdesivir. This oddity might be explained by serious remdesivir toxicity that accrues after five days or by inadequate accounting for baseline differences. It may also be purely a random finding. There may be other explanations, but this finding suggests that better studies are needed. The authors reasonably question whether the difference is clinically important.

Written by Henry C. Barry, MD, MS, on August 26, 2020. (Source: Spinner CD, Gottlieb RL, Criner GJ, et al. Effect of remdesivir vs standard care on clinical status at 11 days in patients with moderate COVID-19: a randomized clinical trial [published online August 21, 2020]. JAMA. 2020. //jamanetwork.com/journals/jama/fullarticle/2769871)

September 1, 2020, Research Update

High Incidence of Thrombotic Events in Hospitalized COVID-19 Patients. This study reports a consecutive series of 3,334 patients with COVID-19 hospitalized between March 1 and April 17, 2020, at NYU Langone Health Center, of whom 829 were in the ICU and 2,505 were not in the ICU. During this period most patients received thromboprophylaxis. The authors used medical record review, including natural language processing, to identify episodes of thrombosis. Rates of pulmonary embolism were 6.2% in the ICU and 2.2% on the ward; for deep vein thrombosis, rates were 9.4% in the ICU and 2.0% on the ward. Rates of stroke were 3.7% in the ICU and 0.9% on the ward, and for myocardial infarction were 13.9% in the ICU and 7.3% on the ward. All-cause mortality was significantly higher in those with thrombosis (43.2% vs. 21.0%, NNT = 5). Factors associated with thrombosis included increasing age, self-identified Hispanic ethnicity, underlying coronary artery disease, and increasing d-dimer on admission. Although there is no comparison with non-COVID-19 patients, the authors argue that these rates are higher than with other respiratory infections such as influenza. These findings are consistent with other studies and reinforce the value of measuring d-dimer on admission (it is an independent risk factor for mortality) to strongly consider providing a low-dose anticoagulant to all patients without a contraindication and to be vigilant for thrombotic complications.

Written by Mark H. Ebell MD, MS, on August 25, 2020. (Source: Bilaloglu S, Aphinyanaphongs Y, Jones S, et al. Thrombosis in hospitalized patients with COVID-19 in a New York City health system. JAMA. 2020;324(8):799-801.)

August 31, 2020, Research Update

China Contains Beijing Outbreak in One Month. This research letter describes the response to an outbreak of COVID-19 in Beijing, China, that began in early June 2020. After 56 days of no new cases in Beijing, three cases exposed to SARS-CoV-2 at a seafood market on June 3 and 4 were diagnosed as confirmed cases on June 11 and 12. The first case triggered an immediate response that included aggressive contact tracing, testing, movement restrictions, and quarantine of cases in a hospital. The market was shut down on June 12, and all workers and residents of the surrounding community were tested. A total of 335 confirmed cases were identified, plus 33 asymptomatic infections; 93 people were presymptomatic at the time of diagnosis. The peak of diagnosed cases was on June 13 to 14, and the last case was reported on July 5 after a period of relatively steady decline. The combination of a rapid response, extensive testing, quick results reporting, and comprehensive contact resulted in an impressive and rapid end to the outbreak. It is impossible to imagine this happening in the United States with our fragmented health system and underfunded public health infrastructure.

Written by Mark H. Ebell MD, MS, on August 25, 2020. (Source: Wu Z, Wang Q, Zhao J, et al. Time course of a second outbreak of COVID-19 in Beijing, China, June-July 2020 [published online August 24, 2020]. JAMA. 2020. //jamanetwork.com/journals/jama/fullarticle/2769930)

August 28, 2020, Research Update

BMJ “Live” Systematic Review of Drugs for Treating SARS-CoV-2. This team performed a high-quality systematic review of randomized trials of treatments for people with suspected, probable, or confirmed COVID-19. Their review process included searching for studies from more than 31 databases and preprint servers every day and assessing the risk of bias of the included studies. The team extracted several kinds of data, including study characteristics (e.g., publication status, study status, design features, etc.), patient demographics, smoking habits, comorbidities, setting and type of care, COVID severity, and outcomes. The outcomes included a spectrum of less patient-relevant events such as time to viral clearing to more clinically important ones such as mechanical ventilation and death. In addition to the systematic review, the authors pooled data and performed network meta-analyses for the various outcomes. Their process identified 13 published trials (2,568 persons), 10 preprints (7,167 persons), and nine more (6,156 persons) upcoming. All except two studies were at “probably high” or “high” risk of bias. In other words, the data that exist are of questionable reliability, and there are lots more data that have yet to be peer-reviewed. This is a true work in progress. The BMJ online version of the paper includes a really nice interactive feature that allows you to quickly see summary data and conclusions for each of the eight individual outcomes, a few of which are summarized. In 15 trials (8,654 persons), only glucocorticoids were likely to reduce mortality compared with standard care (37 fewer deaths per 1,000 persons). In eight trials (6,953 participants), only glucocorticoids were likely to reduce the need for mechanical ventilation (30 fewer per 1,000 persons). In the 11 trials (1,875 participants) that reported adverse events, the authors found that remdesivir decreased the duration of symptoms and did not cause any additional harm compared with standard care and that hydroxychloroquine increased the risk of adverse events. It appears that more data are out there, and fortunately, the authors plan to revisit the data frequently.

Written by Henry C. Barry, MD, MS, on August 25, 2020. (Source: Siemieniuk RAC, Bartoszko JJ, Ge L, et al. Drug treatments for covid-19: living systematic review and network meta-analysis. BMJ. 2020;370:m2980.)

August 27, 2020, Research Update

Hospital Mortality Similar for Black and White Patients in the United States. A prior single health system study in New Orleans found no difference in mortality between White and Black hospitalized COVID-19 patients after adjusting for age, chronic illness, and sociodemographic factors. This is a larger, similar study that included 92 hospitals in 12 states in the Ascension hospital system. In the analysis of mortality, the investigators adjusted for age, sex, insurance status, comorbidity, neighborhood deprivation, and site of care. This study included 7,139 adult patients hospitalized with COVID-19 infection between February 19, 2020, and May 31, 2020. Race was determined by self-report: 2,812 patients were Black, 3,127 were White, and the remainder were another race or not identified. Compared with White patients, the Black patients were on average younger, more likely to be women, and a greater burden of chronic illness and higher neighborhood deprivation scores. Overall, all-cause, in-hospital mortality was 20.3%, with 34.7% among those with an ICU stay and 38.1% for those receiving mechanical ventilation. There were 23.1% of White patients and 19.2% of Black patients who died during hospitalization. After adjusting for the factors noted previously, race was not significantly associated with increased risk of death (HR for Black vs. White, 0.93: 95% CI, 0.83 to 1.09). A limitation of this study is that the investigators were not able to identify ethnicity, so they were not able to study mortality for Hispanic White patients compared with Black or non-Hispanic White patients.

Written by John Hickner, MD, MS, on August 24, 2020. (Source: Yehia BR, Winegar A, Fogel R, et al. Association of race with mortality among patients hospitalized with coronavirus disease 2019 (COVID-19) at 92 US Hospitals. JAMA Netw Open. 2020;3(8):e2018039.)

August 26, 2020, Research Update

Face Shields Reduced COVID-19 Infections in Community Health Workers. In this observational before-and-after study in India, the investigators assessed the effectiveness of adding face shields to the personal protective equipment of community health workers who provided in-home counseling to family members of COVID-19 patients. At the home visits, the workers met with the family in the front room of the house to explain important preventive measures and isolation procedures, and they wore three-layered surgical masks, gloves, and shoe covers. Family members were asked, but not required, to wear face masks. During the study, the community health workers lived in separate rooms in hostels and did not visit their homes or any public places, so their only possible exposure was work related. In the two weeks before the workers started using face shields in addition to the other protection, four of 62 workers developed symptomatic COVID-19 infection, and eight developed asymptomatic infections, a 19% attack rate. During these two weeks, they had visited 5,880 homes with 31,164 persons (222 subsequently tested positive for COVID-19). After face shields were introduced, 50 workers visited 18,228 homes and counseled 118,428 persons (of whom 2,682 developed COVID-19 infections), and none of the 50 workers developed symptomatic or asymptomatic COVID-19 infection. Face shields provided added protection from COVID-19 to these community health workers, and they have also been shown to reduce infections in hospital health workers.

Written by John Hickner, MD, MS, on August 23, 2020. (Source: Bhaskar ME, Arun S. SARS-CoV-2 infection among community health workers in India before and after use of face shields [published online ahead of print August 17, 2020]. JAMA. 2020. )

August 25, 2020, Research Update

Peak COVID-19 Mortality rate in NYC Higher Than Peak 1918 Flu Mortality Rate. These investigators compared the estimated excess deaths in New York City from the COVID-19 pandemic with the estimated excess deaths due to the 1918 flu pandemic. They calculated the all-cause mortality rate per person-months for 61 days during the peak months of the 1918 flu epidemic (October and November 1918) and for the same months in 1914 to 1917 before the pandemic. They made the same calculation for COVID-19, using the peak 61 days of March 11 through May 11, 2020, and for the three prior years, 2017 to 2019. They divided this number by two to give a person-month unit. Using this method, they were able to determine the excess number of deaths due to flu or COVID-19 compared with the prepandemic periods.  The mortality incident rate ratio for the 1918 flu epidemic compared to the prepandemic years was 2.80, and the mortality incident rate ratio for COVID-19 in 2020 compared to the prepandemic years was 4.15. The mortality incident rate ratio, comparing the 1918 flu pandemic with the 2020 COVID-19 pandemic was 0.70, indicating that the proportional mortality increase for COVID-19 in March through May in New York City was 30% greater than the 1918 flu. In conclusion, in New York City, COVID-19 was more deadly in March through May of 2020 than the 1918 flu epidemic was during its peak months.

Written by John Hickner, MD, MS, on August 17, 2020. (Source: Faust JS, Lin Z, del Rio C. Comparison of estimated excess deaths in New York City during the COVID-19 and 1918 influenza pandemics. JAMA Network Open. 2020;3(8):e2017527.)

August 24, 2020, Research Update

COVID-19 Hospitalization Rates Higher for Hispanic and Black Children Than for White Children. The COVID-19–Associated Hospitalization Surveillance Network (COVID-NET) is a population-based surveillance system that collects and reports to the CDC data on laboratory-confirmed COVID-19–associated hospitalizations in 14 states. From March 1 to July 25, 2020, the hospitalization rate for children younger than 18 years of age was 8.0 per 100,000, with a total of 576 children hospitalized in these 14 states. The highest hospitalization rate by age group was for children younger than 2 years of age (24.8 per 100,000). The hospitalization rate was much higher for Hispanic and Black children (16.4 and 10.5 per 100,000, respectively) compared with non-Hispanic White children (2.1 per 100,000). One in three hospitalized children required intensive care. The higher hospitalization rate for Hispanic and Black children is not fully understood but is possibly due to higher exposure rates and underlying conditions such as obesity. An improved understanding of social determinants of health is critical in reducing these disparities.

Written by John Hickner, MD, MS, on August 11, 2020. Source: Kim L, Whitaker M, O’Halloran A, et al. Hospitalization rates and characteristics of children aged <18 years hospitalized with laboratory-confirmed COVID-19 — COVID-NET, 14 States, March 1–July 25, 2020. MMWR Morb Mortal Wkly Rep. 2020;69(32):1081-1088.)

August 21, 2020, Research Update

Symptomatic and Asymptomatic COVID-19–Infected Individuals in Korea Test Positive for SARS-CoV-2 for 19.5 and 17 Days, Respectively. The investigators evaluated the clinical course and viral loads of 303 individuals with symptomatic (193) and asymptomatic (110) COVID-19 infections who were isolated in a community treatment center in Korea. These were mostly healthy young adults with a median of 25 years of age. PCR testing for SARS-CoV-2 was performed on days 8, 9, 15, and 16 of isolation, and additional tests were completed at physician discretion. Twenty-one of the 110 individuals (19%) who were initially asymptomatic developed symptoms, and the median time to develop symptoms was 15 days. By day 21, 75% of asymptomatic patients and 70% of symptomatic patients had converted to a negative COVID-19 test. The median time it took to convert to negative was 17 days in asymptomatic patients and 19.5 days in symptomatic patients. The conclusion is that asymptomatic COVID-19–infected patients may be infectious for nearly as long as symptomatic patients. A prior community contact-tracing study in Taiwan, however, found that the risk of transmitting COVID-19 after day 5 of infection was very low for symptomatic individuals (Cheng H-Y, Jian S-W, Liu D-P, et al.; Taiwan COVID-19 Outbreak Investigation Team. Contact tracing assessment of COVID-19 transmission dynamics in Taiwan and risk at different exposure periods before and after symptom onset [published online May 1, 2020]. JAMA Intern Med. 2020. ).

Written by John Hickner, MD, MS, on August 11, 2020. (Source: Lee S, Kim T, Lee E, et al. Clinical course and molecular viral shedding among asymptomatic and symptomatic patients with SARS-CoV-2 infection in a community treatment center in the Republic of Korea [published online August 6, 2020]. JAMA Intern Med. 2020. )

August 20, 2020, Research Update

Not All Face Coverings Offer Similar Degrees of Protection. Face masks have become a foundation of public health measures to mitigate the spread of respiratory illnesses such as COVID-19, but they have also become lightning rods for various political agendas. Some have used face coverings (e.g., bandanas, gaiters, etc.) as a more comfortable or fashionable alternative to face masks. Analogous to other studies that we have reported on the spread of respiratory droplets (April 22, 2020, Research Brief from a study in NEJM and May 19, 2020, Research Brief from the National Academy of Sciences), these researchers used fancy laser technology to evaluate the effectiveness of 14 different face coverings. They included commercial and homemade masks, various materials, number of layers, bandanas, gaiters, and a simple swath of mask material. They also included an unmasked control. One operator wearing each type of facial garb spoke into the darkened laser chamber, and the flow of respiratory droplets were recorded and analyzed. An additional four operators wore a selection of the face coverings. The authors are proud to point out that the main equipment costs about $200; using a cellphone camera means you can inexpensively replicate their work! The clear winner was the commercially made, fitted N95 masks that, compared with the unmasked controls, transmitted less than 0.1% of the droplets. The clear loser was the gaiter-type neck fleece that actually transmitted more droplets than the unmasked controls (110%)! Bandanas were only slightly better than controls. Although the study does not assess real-world performance, we have other empiric data showing various degrees of protection from commercial and homemade masks. 

Written by Henry C. Barry, MD, MS, on August 11, 2020. (Source: Fischer EP, Fischer MC, Grass D, et al. Low-cost measurement of facemask efficacy for filtering expelled droplets during speech. Science Advances. 2020:eabd3083. )

August 19, 2020, Research Update

Hydroxychloroquine Not Effective in Large British Randomized Controlled Trial. This is another report from the UK RECOVERY trial, which previously concluded that low-dose dexamethasone significantly reduces mortality in more severely ill patients with COVID-19 (preprint server, not peer reviewed). In this arm of the study, 1,561 hospitalized patients with COVID-19 were randomized to hydroxychloroquine (HCQ) 800 mg at 0 and 6 hours, followed by 400 mg every 12 hours for up to nine days, whereas 3,155 were allocated to usual care. They excluded anyone known to have a prolonged QTc interval. This was an open label trial, and allocation was appropriately concealed. The mean age of participants was 65 years of age, 38% were female, 57% had at least one major comorbidity, and 76% were receiving supplemental oxygen or assisted ventilation. Very little crossover occurred, with 92% in the HCQ group getting at least one dose (mean duration six days) and only 0.4% in the usual care group receiving the drug. No difference occurred between groups with regard to 28-day mortality (26.8% HCQ vs. 25.0% usual care, rate ratio [RR] 1.09, 95% CI 0.96 to 1.23). Patients in the HCQ group had a longer time to hospital discharge (16 vs. 13 days) and were less likely to be discharged alive within 28 days (60.3% vs. 62.8%, RR 0.92, 95% CI 0.85 to 0.99, number needed to treat [NNT] = 40). Patients in the HCQ group were also more likely to experience the combined outcome of mechanical ventilation or death (29.8% vs. 26.5%, RR 1.12, 95% CI 1.01 to 1.25, NNT = 30). This represents the strongest evidence to date of the lack of benefit and potential harm of HCQ as a treatment for COVID-19.

Written by Mark Ebell MD, MS, on August 9, 2020. (Source: Horby P, Mafham M, Linsell L, et al. Effect of hydroxychloroquine in hospitalized patients with COVID-19: preliminary results from a multi-centre, randomized, controlled trial [pre-print, not peer reviewed, published online July 15, 2020]. )

August 12, 2020, Research Update

Hydroxychloroquine Alone or Combined with Azithromycin Does Not Improve 15-Day Outcomes in Patients Hospitalized with Mild to Moderate COVID in Brazil. In this open-label study, funded by research institutes and industry, adult patients hospitalized at one of 55 Brazilian hospitals with mild to moderate COVID-19 (not requiring supplemental oxygen) were randomized to receive standard care (n = 227), standard care plus hydroxychloroquine (HCQ; 400 mg twice daily; n = 221), or standard care plus hydroxychloroquine plus azithromycin (HCQ+AZ; 500 mg daily for seven days; n = 217). Standard care was pretty much wild west: corticosteroids, antibiotics, antivirals, immunomodulators, or whatever else the treating team could imagine. The researchers assessed each patient’s status after 15 days. For patients who were no longer hospitalized, a research assistant unaware of treatment allocation contacted the patient to assess vital status—this appears to be the only time blinding occurred. The researchers evaluated the main outcome, a seven-point scale, using modified intention-to-treat based on those who had confirmed COVID-19. Of the 665 randomized patients, 504 (76%) had a positive PCR for COVID-10. Regardless of all of the design elements that tend to make interventions look better, the researchers were unable to find any improvements in the main outcome nor in any single category of outcome. For example, about two-thirds of patients in each group were no longer hospitalized and had no limitations on activities at 15 days. Only two patients died, both of whom received HCQ+AZ. About 15% of patients receiving HCQ or HCQ+AZ experienced prolonged QT intervals compared with 2% of those who received neither HCQ nor AZ.

Written by Henry C. Barry, MD, MS, on August 3, 2020. (Source: Cavalcanti AB, Zampieri FG, Rosa RG, et al. Hydroxychloroquine with or without azithromycin in mild-to-moderate Covid-19 [published online July 23, 2020]. N Engl J Med. 2020. )

August 11, 2020, Research Update

Accuracy of Antibody Tests for COVID-19 Highly Variable. This meta-analysis identified 40 studies of serologic tests for IgG or IgM antibodies to SARS-CoV-2. The researchers searched standard databases and preprint servers. Studies were evaluated in parallel by at least two researchers, and study quality was assessed using the QUADAS-2 tool. Most of the studies were from China, used a case-control design, and studied inpatients. To estimate specificity, half of the samples were collected before the pandemic (good), whereas half of the samples were collected from patients not suspected to have COVID-19 (bad, as many cases are asymptomatic). The risk of bias for patient selection was high for all but one study because of case-control design and were high or unclear for the index test because of lack of blinding and uncertainty about how the cutoff for an abnormal test was determined. A mix of commercial and home-grown tests were used. The results are shown in the table below. The chemiluminescent had the best sensitivity by far and had good specificity. For all of the tests, sensitivity was unacceptably low for IgG and IgM until at least three weeks after symptom onset. Given the high risk of bias and lack of evaluation in outpatients, these tests should be used with caution and are not useful for acute diagnosis. An interactive calculator can be found at .

Written by Mark H. Ebell, MD, MS, on July 31, 2020. (Source: Bastos ML, Tavaziva G, Abidi SK, et al. Diagnostic accuracy of serological tests for covid-19: systematic review and meta-analysis.  BMJ. 2020;370:m2516.)

August 10, 2020, Research Update

Outbreak in Israeli High School with High Rates of Transmission to Staff. As schools struggle with the decision to reopen in the fall, this report of an outbreak in an Israeli high school is informative. Because of COVID-19, all schools in Israel were closed on March 13, 2020, and reopened on May 17, 2020, with mandatory daily health reports, facemasks, sanitizers, social distancing, and reduced interaction between classes. During a heatwave from May 19 to 21, the mask requirement was lifted. The school in question had 162 staff and 1,190 students between 12 to 18 years of age. On May 26 and 27, two students from different grades were diagnosed with COVID-19. Especially because the cases were not epidemiologically linked, the school was closed, and almost all staff and students were tested. The attack rate was 17.3% to 32.6% in grades 7 to 9 and was 1.6% to 4.5% in grades 10 to 12. Overall, 13.2% of students and 16.6% of staff tested positive for COVID-19. Cases were symptomatic in 43% of students and 75% of staff. There were also 87 secondary cases among family and close contacts of the school cases, including people in their 50s, 60s, and older. As of June 30, about one-third of students and staff had still not recovered. Nationally, the proportion of cases in school-aged children was 19.8% before reopening and 40.9% after reopening. A review of conditions at the school found that social distancing was not possible and that recirculating air conditioners (implicated in other indoor outbreaks) ran continuously during the heatwave.

Written by Mark H. Ebell, MD, MS, on July 27, 2020. (Source: Stein-Zamir C, Abramson N, Shoob H, et al. A large COVID-19 outbreak in a high school 10 days after schools’ reopening, Israel, May 2020. Euro Surveill. 2020;25(29):2001352.)

August 7, 2020, Research Update

Prevalence of Antibodies to SARS-CoV-2 about 10 Times Higher Than Reported Case Rate in United States. We previously summarized a Centers for Disease Control and Prevalence study of COVID-19 sero-prevalence in six regions of the United States on July 2, 2020. This report expands the scope of the same study to include 10 areas in the United States: San Francisco Bay area, California; Connecticut; south Florida; Louisiana; Minneapolis-St Paul-St Cloud metro area, Minnesota; Missouri; New York City metro area, New York; Philadelphia metro area, Pennsylvania; Utah; and western Washington state. A convenience sample of serum specimens drawn by a commercial lab for other reasons was tested for antibodies to SARS-CoV-2 antibodies, and the proportion testing positive was standardized to the site populations for age and sex and was then adjusted for test sensitivity and specificity. The prevalence of antibodies to SARS-CoV-2 ranged from 1.0% in the San Francisco Bay area to 6.9% in New York City. These prevalences were six to 24 times greater than the prevalence of reported cases reported during the same time period. For seven of the 10 sites, the estimated infections were 10 times greater than the reported number of cases. Therefore, it is likely that the true COVID-19 infection rate is roughly 10 times that of diagnosed cases. 

Written by John Hickner, MD, MS, July 25, 2020. (Source: Havers FP, Reed C, Lim T, et al. Seroprevalence of antibodies to SARS-CoV-2 in 10 sites in the United States, March 23–May 12, 2020 [published online July 21, 2020]. JAMA Intern Med. 2020. )

August 6, 2020, Research Update

Successful Chinese Phase 2 Trial of a COVID-19 Vaccine. These researchers identified 508 healthy volunteers and randomized them to low-dose vaccine (5 x 1010 viral particles), high-dose vaccine (1 x 1011 viral particles), or placebo vaccine in a 2:1:1 ratio (this was an adenovirus type-5 vectored vaccine). Allocation was concealed, and patients and outcome assessors were masked. The mean was 40 years of age, with a range of 18 to 83 years, but only 13% of recipients were 55 years or older. Seroconversion of the binding antibody was high at 96% to 97% in the two vaccine groups at 28 days. Seroconversion of the neutralizing antibodies to SARS-CoV-2 occurred in 59% receiving the high-dose vaccine and 49% receiving the low-dose vaccine at 28 days. Note that binding antibodies bind to the viral antigen but do not prevent infection of the host cell, whereas neutralizing antibodies prevent infection. Evidence of humoral (T-cell mediated) immunity was detected in 90% of patients receiving the high-dose vaccine and 88% with the low-dose vaccine at 28 days. Patients with high preexisting titers to adenovirus type 5 and older patients had less vigorous immune responses and would likely require a second immunization. Severe adverse events (most commonly fever but also induration, headache, fatigue, dyspnea, and myalgia/arthralgia) occurred in 9% given the high dose and only 1% given the low dose. No serious or life-threatening adverse effects occurred.

Written by Mark H. Ebell, MD, MS, on July 21, 2020. (Source: Zhu F-C, Guan X-H, Li Y-H, et al. Immunogenicity and safety of a recombinant adenovirus type-5-vectored COVID-19 vaccine in healthy adults aged 18 years or older: a randomised, double-blind, placebo-controlled, phase 2 trial [published online July 20, 2020]. Lancet. 2020. )

August 5, 2020, Research Update

Oxford’s COVID-19 Vaccine Also Appears Promising. We have previously reported promising data from two small vaccine trials (a Research Brief on June 25 from a Lancet study of 108 participants and a Research Brief on July 24 from a New England Journal preliminary report of 45 participants). By far the largest study to date, this team randomized more than 1,000 healthy adults in the United Kingdom to receive a single intramuscular dose of Oxford University’s chimpanzee adenovirus-vectored vaccine (at a dose of 5 x 1010  viral particles) that targets the spike protein (n = 543) or a meningococcal vaccine (n = 534). They used an active comparator vaccine because of known local and systemic reactions that are likely to be less apparent with a simple saline placebo vaccine. The participants had no lab evidence of COVID-19 infection, no COVID symptoms, and could not be at high risk of exposure. The staff administering the vaccines, the investigators, and the laboratory teams all were blinded as to which vaccine the participants received. To illustrate some of the complexities involved with vaccine trials, the researchers divided the participants into four distinct study groups. They evaluated Group 1 participants for safety and immunogenicity at 3, 7, 14, 28, and 56 days after vaccination. The researchers took greater volumes of blood from Group 2 participants to allow for assessing humoral and cellular immunogenicity. Group 3 included 10 participants who received a booster dose 28 days after the first dose. Group 4 participants were evaluated only for humoral immune response. About 10% of participants in each vaccine wing took prophylactic paracetamol (acetaminophen). The rate of fatigue was the same whether the participant received paracetamol or not (70% vs. 71%), but headaches were slightly less common in those taking paracetamol (68% vs. 61%). Other common reactions included myalgias, malaise, chills, and feverishness, most of which were also slightly less in those taking paracetamol. Two weeks later, those receiving the active vaccine already had T-cell responses and by four weeks also had an antibody response: 91% had neutralizing antibodies. All participants receiving the booster had developed neutralizing antibodies. The report does not say what proportion of patients had developed cellular immunity. This study moves us a little closer to having a safe, effective vaccine.

Written by Henry C. Barry, MD, MS, on July 20, 2020. (Source: Folegatti PM, Ewer KJ, Aley PK, et al. Safety and immunogenicity of the ChAdOx1 nCoV-19 vaccine against SARS-CoV-2: a preliminary report of a phase 1/2, single-blind, randomised controlled trial [published online July 20, 2020]. Lancet. 2020. )

August 4, 2020, Research Update

Household Transmission of COVID-19 Highest in People 10 to 19 Years of Age in South Korea. South Korea instituted a comprehensive contact-tracing program to control COVID-19 transmission. These investigators from the South Korean Centers for Disease Control and Protection studied COVID-19 transmission to 59,073 contacts of 5,706 COVID-19 infected index patients identified from January 20–March 27, 2020. They monitored the contacts for an average of 9.9 days after COVID-19 infection was detected in the index case. The transmission rate was much higher for the 10,592 household contacts (11.8%) than for the 48,481 nonhousehold contacts (1.9%). They also analyzed transmission rates by age group of the index case. The household transmission rate was highest for the 10 to 19 age group (18.6% [95% CI, 14.0% to 24.0%]) and lowest for the 0 to 9 age group (5.3% [95% CI, 1.3% to 13.7%]). For adults, transmission rates ranged from 7% for the 20 to 29 age group to 18% for the 70 to 79 age group. The finding of a higher transmission rate for children/teens 10 to 19 years of age has significant implications for school re-openings.

Written by John Hickner, MD, MS, on July 20, 2020. (Source: Park YJ, Choe YJ, Park O, et al. Contact tracing during Coronavirus disease outbreak, South Korea, 2020 [published online July 16, 2020]. Emerg Infect Dis. 2020;26(10). )

August 3, 2020, Research Update

Proton Pump Inhibitors and Associated Risk of Contracting COVID-19. In this prepublication release from the American Journal of Gastroenterology, the authors report data from an online survey. To obtain the participants, the researchers recruited a nationwide, representative sample based on the U.S. census data on age, sex, and region. The participants had to be at least 18 years of age. The survey was administered solely in English, and the participants had to have internet access and language fluency, limiting the generalizability of their findings. The survey included questions about gastrointestinal symptoms and the frequency and duration of use of acid suppressing medications. Of the 264,058 people invited, 128,847 (48.8%) accessed the survey, and 86,602 (32.7%) eligible respondents completed it. The authors report on only 53,130 (20.1%), the subset of those who reported having gastrointestinal symptoms. Of these, 3,386 (6.4%) reported a positive COVID-19 test. After adjusting for a slew of other factors that might be associated with contracting COVID-19, they report that proton pump inhibitor use was independently associated with increased odds for reporting a positive test. Once daily use of proton pump inhibitors was associated with a lower risk (OR 2.15; 95% CI, 1.90 to 2.44) compared with twice daily use (OR 3.67; 95% CI, 2.93 to 4.60). In contrast, once daily use of histamine receptor antagonists was associated with a lower risk (OR 0.85, 95% CI, 0.74 to 0.99), and twice daily use was not significantly associated with reporting a positive COVID test. Much of the important data in this study is based on self-report; therefore, these findings should be considered exploratory.

Written by Henry C. Barry, MD, MS, on July 19, 2020. (Source: Almario CV, Chey WD, Spiegel BMR. Increased risk of COVID-19 among users of proton pump inhibitors [published online July 7, 2020]. Am J Gastroenterology. 2020. 

July 31, 2020, Research Update

Physical Distancing Measures Are Effective in Reducing COVID-19 Transmission. Most information about the effectiveness of physical distancing (e.g., school closures, workplace closures, restrictions on mass gatherings, public transportation closure, lock-down orders) in reducing COVID-19 transmission comes from modeling studies. This study uses empirical data regarding the effects of physical distancing on COVID-19 transmission in an interrupted time series analysis and random effects meta-analysis of data from interventions in 149 countries, the most comprehensive data to date. Investigators calculated the incidence rate ratios (IRR) of COVID-19 cases before and after physical distancing interventions as measured 30 days after the intervention or on May 30, wherever came first. Implementation of any physical distancing intervention reduced COVID-19 cases an average of 13% (IRR 0.87; 95% CI, 0.85 to 0.89; n = 149 countries). The results were essentially the same whether four or five interventions were in place; most countries implemented four or five of these interventions. The findings did not change in a detailed sensitivity analysis.

Written by John Hickner, MD, MS, on July 18, 2020. (Source: Nazrul Islam N, Sharp SJ, Gerardo Chowell G, et. al. Physical distancing interventions and incidence of coronavirus disease 2019: natural experiment in 149 countries. BMJ. 2020;370:m2743.)

July 30, 2020, Research Update

Spanish RCT of Hydroxychloroquine for Patients with Early, Nonsevere COVID-19 Finds No Benefit. This Spanish study sought to determine whether early treatment with hydroxychloroquine is beneficial for patients with COVID-19. The researchers identified adults in Catalonia with a positive test for SARS-CoV-2, were symptomatic but did not require hospitalization, and were within five days of symptom onset. The mean age of the 293 included patients was 42 years, 69% were women, 87% were health care workers, and the median time from symptom onset to enrollment was three days. This was an open label study, but allocation to groups was concealed from patients and investigators; lab workers were masked to treatment assignment. Patients were followed for up to a month. Patients were randomized to hydroxychloroquine 800 mg on day 1 and 400 mg once daily on days 2 through 7 or usual care alone. The primary outcome was viral load at three and seven days, but they also tracked symptom improvement. Absolutely no difference occurred between groups at days 3 or 7 in viral load. No difference occurred in hospitalization rates (7.1% vs. 5.9% in the intervention group), and no deaths occurred during the study period. Also, no difference occurred in the mean days to resolution of symptoms (12 in the control arm vs. 10 in the treatment arm, p = 0.21). Adverse events (mostly gastrointestinal and nervous system related) were much more common in the treatment group (72% vs. 9%). No serious adverse events attributed to hydroxychloroquine occurred. 

Written by Mark H. Ebell, MD, MS, on July 17, 2020. (Source: Mitjà O, Corbacho-Monné M, Ubals M, et al.; BCN PEP-CoV-2 Research Group. Hydroxychloroquine for early treatment of adults with mild Covid-19: a randomized controlled trial [published online July 16, 2020]. Clin Infect Dis. 2020. )

July 29, 2020, Research Update

United States/Canadian RCT of Hydroxychloroquine for Patients with Early, Nonsevere COVID-19 Finds No Benefit. It has been hypothesized that whereas hydroxychloroquine does not appear to work in severely ill patients, perhaps it could be helpful earlier in the course of illness or in patients with less severe illness. This study enrolled 491 nonhospitalized patients with confirmed or probable COVID-19 who were enrolled online. More than half were enrolled within one day of symptom onset. A limitation of this study is that only 149 had a positive polymerase chain reaction (PCR) test, whereas 280 had a PCR positive contact to enter the study and 37 had neither but had typical symptoms. Patients were randomized to hydroxychloroquine 800 mg, followed by 600 mg six to eight hours later, and then 600 mg once daily for four additional days. They initially assumed a 10% hospitalization rate, but this did not happen, so they changed the primary outcome to the difference between groups in symptom severity over 14 days. The median age was 40 years, 56% were women, 32% had a comorbidity, and 3% were Black. They found no difference in overall symptom severity or the percentage with symptoms between groups. Adverse effects were twice as common in the treatment group (43% vs. 22%, p < 0.001), mostly gastrointestinal. The number of hospitalizations (12 total) and deaths (two) was small and did not differ between groups. The lack of diagnostic confirmation in most patients is concerning; however, a second study in Spain with 293 patients with confirmed infection published July 16, 2020, in Clinical Infectious Diseases has similar results.

Written by Mark H. Ebell, MD, MS on July 17, 2020. (Source: Skipper CP, Pastick KA, Engen NW, et al. Hydroxychloroquine in nonhospitalized adults with early COVID-19: a randomized trial [published online July 16, 2020]. Ann Intern Med. 2020. )

July 28, 2020, Research Update

Children Less Likely Than Adults To Be the Index Case in Household COVID-19 Transmission. From March 10 to April 10, 2020, the Geneva University Hospital’s surveillance network in Switzerland identified 4,310 patients with COVID-19 infection, including 40 patients (0.9%) 16 years of age or younger. In addition to describing the symptoms and clinical outcomes of these children (all recovered), the study describes household transmission for these 40 children. Among the 111 household contacts of these 40 children, there were 88 adults and 23 children.  Adult household contacts were suspected or confirmed to have COVID-19 infection before the study child in 79% (31/39) of cases. The study child developed symptoms before any other household contact in only 8% (3/39) of households. In a similar international study of COVID-19 transmission in 31 household clusters from China, Singapore, South Korea, Japan, and Iran, the investigators found that in only three of the 31 (9.7%) household clusters was a child the first (index) case. There are several plausible explanations for these findings: Children may be less likely than adults to be infective; asymptomatic children may have spread the virus more frequently and not been identified; or because of quarantine measures, children were less likely than adults to be exposed to COVID-19.

Written by John Hickner, MD, MS, on July 15, 2020. (Sources: Posfay-Barbe KM, Wagner N, Gauthey M, et al. COVID-19 in children and the dynamics of infection in families. Pediatrics. 2020;146(1):e20201576. Zhu Y, Bloxham CJ, Hulme KD, et al. Children are unlikely to have been the primary source of household SARS- CoV-2 infections [published online March 30, 2020]. medRxiv preprint. // www.medrxiv.org/content/10.1101/2020. 03.26.20044826v1)

July 24, 2020, Research Update

Promising Early Data from the United States on Another New Vaccine. Around the world, more than 100 candidate vaccines against SARS-CoV-2 are in various stages of development and testing. In a Research Brief published on June 25, 2020, we reviewed early data from a Chinese study in Lancet that included 108 people. The current study was an open-label Phase I trial on 45 healthy adults. The participants received two doses four weeks apart of Moderna’s messenger RNA vaccine (mRNA-1273) at doses of 25 μg, 100 μg, or 250 μg. There was no comparison group, and, a significant limitation, the researchers did not screen the participants for SARS-CoV-2 infection by serology or polymerase chain reaction before enrollment. The study protocol calls for assessments at multiple points after each vaccination: 7, 14, 57, 119, 209, and 394 days. On days 1, 15, 29, 36, 43, and 57, the researchers tested each participant for binding antibodies and for neutralizing activity. Three of the 45 participants did not receive the second dose, one of whom developed urticaria and one who was in isolation for suspected COVID. On day 57, the researchers were able to detect binding and neutralizing antibodies with all three doses by 14 days after immunization with a peak at 28 days. The high-dose vaccine induced the greatest immunogenic responses but also the highest rate of adverse effects. Additionally, as we have observed with the new recombinant zoster vaccine, the second dose triggers more frequent adverse events: 54%, 100%, and 100% for each of the three doses, respectively. Fortunately, most of these were considered to be mild, and only three participants receiving the highest dose reported one or more serious adverse effects (21%). The adverse events included fever, arthralgias, injection site pain, headache, and chills. These are early and incomplete data from a small study with no control group, but the data provide hope that an effective vaccination is feasible.

Written by Henry C. Barry, MD, MS, on July 15, 2020. (Source: Jackson LA, Anderson EJ, Rouphael NG, et al. An mRNA vaccine against SARS-CoV-2—preliminary report [published online July 14, 2020]. N Engl J Med. 2020. )

July 23, 2020, Research Update

Higher Incidence of Stillbirth during COVID-19 Pandemic. To determine whether more adverse obstetric/neonatal outcomes occurred during the COVID-19 pandemic, investigators from St George’s University Hospital in London compared stillbirth, preterm birth, cesarean delivery, and neonatal unit admissions from October 1, 2019, to January 31, 2020 (prepandemic), to these outcomes from February 1, 2020, to June 14, 2020 (during the pandemic). There were 1,681 births during the first period and 1,718 births during the pandemic period. No differences were found in preterm birth, cesarean delivery, or neonatal unit admissions. Stillbirth incidence, however, was higher during the pandemic (n = 16 [9.31 per 1,000 births]) than prior to the pandemic (n = 4 [2.38 per 1,000 births]), a difference of 6.93 per 1,000 births (95% CI, 1.83 to 12.0; P = .01) The significant difference remained after excluding late terminations for fetal abnormalities (6.98 vs. 1.19 per 1,000 births, a difference of 5.79 [95% CI, 1.54 to 10.1]). None of the women who experienced stillbirth had symptoms of COVID-19, nor did the postmortem or placental examinations show evidence of COVID-19 infection. But these women were not tested for COVID-19, so they may have had asymptomatic infections. Nineteen pregnant women were hospitalized for COVID-19, and none of these women had stillbirths. Many possible explanations can account for these findings, including asymptomatic COVID-19 infection, inferior prenatal care during the pandemic, or more high-risk referrals to St George’s Hospital during the pandemic. The study is small, retrospective, and from one hospital. It will take a larger, multicenter study to see whether this association is real. 

Written by John Hickner, MD, MS, on July 12, 2020. (Source: Khalil A, von Dadelszen P, Draycott T, et al. Change in the incidence of stillbirth and preterm delivery during the COVID-19 pandemic [published online July 10, 2020]. JAMA. 2020. )

July 22, 2020, Research Update

Results of Study That Tested an Entire Town in Italy during the Peak of the Epidemic. This is one of the only whole population studies of COVID-19 to have been performed, and it is likely more representative of the real world than similar studies on cruise ships, aircraft carriers, prisons, and maternity wards. The Italian town of Vo’ was one of the first affected, and it was entirely locked down earlier in the epidemic. Researchers tested everyone they could, which included 86% of the population between February 21 and 29, 2020, and 72% on March 7, 2020. This is a very rich dataset; highlights are as follows. In late February, 2.6% of all samples were positive. Importantly, they found that 42.5% of those infected were asymptomatic at the time of testing and never became symptomatic. The time to clear the infection for those positive on the first survey and negative on the second ranged from eight to 13 days, with a mean of 9.3 days. The viral load was similar for symptomatic and asymptomatic individuals, and through contact tracing they determined that asymptomatic individuals can transmit the infection. The lockdown reduced the effective reproductive number from 2.49 to 0.41. Finally, they estimated an infectious period between 3.6 to 6.5 days, with infectiousness peaking on the day of symptom onset.

Written by Mark H. Ebell MD, MS, on July 12, 2020. (Source: Lavezzo E, Franchin E, Ciavarella C, et al. Suppression of a SARS-CoV-2 outbreak in the Italian municipality of Vo’ [published online June 30, 2020]. Nature. 2020. )

July 21, 2020, Research Update

Super-Spreading Study of COVID-19 in Georgia. These investigators used sophisticated modeling to determine the temporal-spatial spread of COVID-19 infection in five counties in Georgia, focusing on super-spreaders—defined as “individuals who disproportionately infect a large number of secondary cases relative to an ‘average’ infectious individual.” Data sources included the Georgia Department of Public Health surveillance and Facebook’s Data for Good program, which is designed to track movement of cell phone users over time and space. The authors state, “Our model mechanistically integrates detailed individual-level surveillance data, geo-spatial location data and highly-resolved population density (grid-) data, and aggregate mobility data.” There were 9,559 consecutive cases reported between March 1, 2020, and May 3, 2020, available for analysis. They found that the top 2% of infected individuals (those that generate the highest number of infections per individual) are responsible directly for about 20% of total infections. These super-spreaders were more likely to be children and adults younger than 60 years of age. These data have significant implications for controlling the COVID-19 pandemic, especially with the recent resurgence of infections in the United States.

Written by John Hickner, MD, MS, on July 10, 2020. (Source: Lau MSY, Grenfell B, Nelson K, et al. Characterizing super-spreading events and age-specific infectivity of COVID-19 transmission in Georgia, USA [published preprint online June 22, 2020]. medRxiv preprint. )

July 20, 2020, Research Update

Persistent Fatigue, Dyspnea, and Other Symptoms Common Following Hospitalization for COVID-19. In this study set in Rome, Italy, patients who had been hospitalized but now met criteria for discontinuing quarantine (afebrile for three days, symptoms improved, and two negative polymerase chain reaction [PCR] tests 24 hours apart) were invited to attend a COVID-19 follow-up clinic. Of 179 eligible patients, 165 agreed to participate, but 22 had a positive PCR test and were excluded. The mean age was 57 years, and 63% were men. Patients were a mean of 60 days post symptom onset, and 87% still reported symptoms, with 32% reporting one or two symptoms and 55% reporting three or more persistent symptoms. Fully 44% reported a clinically significant 10-point decline in a 100 points quality of life scale compared with their pre–COVID-19 state. The most common post–COVID-19 symptoms at follow-up were fatigue (53%), dyspnea (43%), joint pain (27%), chest pain (22%), cough (17%), and anosmia (16%). Strengths of this study include a standard protocol for data collection and a good response rate among invited patients. This is important prognostic information for patients who were sick enough to be hospitalized; longer term follow-up in these patients as well as in patients who were less symptomatic will be important. Comparison with patients who have survived other serious pulmonary infections such as community acquired pneumonia and acute respiratory distress syndrome are also needed to provide context.

Written by Mark H. Ebell MD, MS, on July 10, 2020. (Source: Carfi A, Bernabei R, Landi F, et al. Persistent symptoms in patients after acute COVID-19 [published online July 9, 2020]. JAMA. 2020. )

July 16, 2020, Research Update

Healthcare Workers in Wuhan, China, Wearing Personal Protective Equipment and Who Are Exposed to SARS-CoV-2 Did Not Become Infected. These authors followed 420 healthcare workers who were “on loan” to Wuhan from two hospitals in Guangzhou. For six weeks, they worked four to six hours per day for an average of 5.4 days per week. On average, they worked about 16 hours per week in intensive care units. All had direct contact with patients and reported at least one aerosol-inducing encounter. All of them wore personal protection equipment—the full body armor of masks, goggles, face shields, gowns, gloves, etc. Upon returning home, they were quarantined for two weeks and tested. None developed symptoms of COVID-19, and none had a positive polymerase chain reaction nor IgG or IgM antibodies. 

Written by Henry C. Barry, MD, MS, on July 9, 2020. (Source: Liu M, Cheng S-Z, Xu K-W, et al. Use of personal protective equipment against coronavirus disease 2019 by healthcare professionals in Wuhan, China: cross sectional study. BMJ. 2020;369:m2195.)

July 15, 2020, Research Update

Clinical Manifestations of Children with Multisystem Inflammatory Syndrome Associated with COVID-19 in Two New Studies from the United States. We previously reviewed three case series of children with multisystem inflammatory syndrome in a Research Brief published on June 16. Two more recent papers report data from medical record audits of children hospitalized in New York and in 26 states in the United States. As in the other studies, there were differences in reporting. Overall, 285 children were included in these studies, 77 (27%) were Black children, 195 (68%) had lab-confirmed COVID-19, and 110 (39%) also met criteria for Kawasaki Disease. Gastrointestinal symptoms were seen in 250 (88%) of the children, much higher than what was previously reported. One of the studies (99 children) reported that 60 children (61%) had abdominal pain and that 57 (58%) had nausea or vomiting. Rash occurred in 169 (59%) of the children, but neither study reported on desquamation. Conjunctivitis occurred in 158 (55%) of the children. One study reported that 27 (27%) children had redness or swelling of the lips or mucous membranes. Coronary artery dilation and shock, reported in one study, occurred in 9 (9%) and 10 (10%), respectively. Myocarditis occurred in 158 (55%) children. One of the studies reported chest radiograph data; 79 of the 186 (42%) were abnormal. 

Written by Henry C. Barry, MD, MS, on July 9, 2020. (Sources: Feldstein LR, Rose EB, Horwitz SM, et al. Multisystem inflammatory syndrome in U.S. children and adolescents [published online June 29, 2020]. Updated July 2, 2020. N Engl J Med. 2020. ; Dufort EM, Koumans EH, Chow EJ, et al. Multisystem inflammatory syndrome in children in New York state [published June 20, 2020]. N Engl J Med. 2020. )

July 14, 2020, Research Update

Guillain–Barré Syndrome Associated with SARS-CoV-2 in Italy. These authors report on five patients from three hospitals in northern Italy. Four of the patients developed lower limb weakness and paresthesias, and one developed facial diplegia followed by ataxia and then paresthesias. The patients were ultimately diagnosed with Guillain–Barré Syndrome. At the onset of neurologic symptoms, four of the patients had positive swabs for SARS-CoV-2, and the other eventually developed antibodies. None of the patients had a positive polymerase chain reaction in their cerebrospinal fluid. The development of Guillain–Barré Syndrome should not be a surprise because it has been frequently associated with viral infections and with active immunizations.

Written by Henry C. Barry, MD, MS, on July 9, 2020. (Source: Toscano G, Palmerini F, Ravaglia S, et al. Guillain-Barre syndrome associated with SARS-CoV-2. N Engl J Med. 2020;382(26):2574-2576.)

July 13, 2020, Research Update

Hydroxychloroquine Associated with Lower Mortality in Patients Hospitalized with COVID-19: Observational Study in Detroit. In this observational study from the Henry Ford Health System in Detroit, investigators examined the in-hospital mortality of 2,541 consecutive hospitalized COVID-19 patients in four treatment categories: hydroxychloroquine (HCQ) plus azithromycin (AZM), either drug alone, or neither drug. They excluded patients who died in the first 24 hours after admission and another 10% of patients for whom final outcome data were unavailable (e.g., still hospitalized, left against medical advice, or transferred to another facility). The median time to follow-up was 28.5 days. Overall in-hospital mortality was 18.1% (95% CI, 16.6% to 19.7%). Mortality for patients taking HCQ+AZM was 20.1% (95% CI, 17.3% to 23.0%); for HCQ alone, 13.5% (95% CI, 11.6% to 15.5%); for AZM alone, 22.4% (95% CI, 16.0% to 30.1%); and neither drug, 26.4% (95% CI, 22.2% to 31.0%). In the Cox regression analysis that adjusted for age, gender, comorbid conditions, and disease severity, the hazard ratio for mortality was reduced 66% (p < 0.001) compared with the group receiving neither HCQ nor AZM. The authors report wide variations in the use of corticosteroids among the different treatment groups: 36% in those treated with neither medication, 39% of those receiving AZM alone, 74% of those treated with both, and 79% of those receiving HCQ alone. This is an observational study. These kinds of studies can find associations but are fairly weak in determining a causal link between an exposure and an outcome. In this study, the findings are inconsistent with other observational studies and with data we have from the few randomized trials that exist. Additionally, they found no effect of steroids in the outcome, which is at odds with other studies (e.g., see Research Brief from June 30 on the effects of dexamethasone from the RECOVERY Collaborative Group), at least in those with severe COVID. Observational studies are subject to all kinds of bias and are subject to alternative explanations for the findings. For example, about 25% of the patients had missing measures of disease severity and were excluded from the regression model. The co-treatment with steroids is likely to reflect differences in disease severity. Observational studies are also unable to account for the “hidden” factors involved in how physicians decide the treatments they choose based on other patient characteristics. Finally, the regression analysis took into account only factors the authors chose and could not address residual confounding. In a randomized trial, the known and unknown factors associated with the outcome of interest are evenly distributed between the intervention groups. 

Written by John Hickner, MD, MS, and Henry Barry, MD, MS, on July 7, 2020. (Source: Arshad S, Kilgore P, Chaudhry ZS, et al.; Henry Ford COVID-19 Task Force. Treatment with hydroxychloroquine, azithromycin, and combination in patients hospitalized with COVID-19 [published online July 1, 2020]. Int J Infect Dis. )

July 10, 2020, Research Update

Gastrointestinal Symptoms Associated with Delayed Diagnosis and Prolonged Viral Shedding in Patients in Wuhan, China. These authors reviewed the medical records of 850 consecutive patients hospitalized for COVID-19 over a two-week period in February 2020 in Wuhan, China. From this group, they identified 206 patients with mild disease who had demonstrated viral clearing and were subsequently discharged. They defined mild disease when patients had no dyspnea and had no evidence of respiratory distress and were able to maintain oxygen saturation above 93% while at rest. It may seem curious that these patients were admitted to the hospital, but this was in the early days of the outbreak, and hospitals were used to facilitate quarantine measures. They identified patients with gastrointestinal (GI) symptoms (“cases”) and matched them to the next admitted patient who had no GI symptoms (“controls”). They classified the case patients by whether they experienced GI symptoms alone (n = 48) or both GI and respiratory symptoms (n = 69). The authors matched them to 69 control patients. Roughly 25% of the patients had hypertension and 10% had diabetes; only 10% could confirm an exposure to a person known to be infected with SARS-CoV-2. Of the 117 patients with GI symptoms, 67 (57%) reported diarrhea that lasted an average of 5.4 days (range 1 to 14 days). Additionally, only 73 (62.4%) patients with GI symptoms also had fever compared with 65 (73%) of patients without GI symptoms. Patients with GI symptoms had a longer interval from onset of symptoms to hospitalization (16.0 ± 7.7 days) than patients with respiratory symptoms alone (11.6 ± 5.1 days) and also had longer durations of viral shedding: 40.9 days for patients with GI symptoms alone, 42.0 days for patients with GI and respiratory symptoms, and 33.5 days for patients with respiratory symptoms alone. Unfortunately, they tested the stool for viral RNA in only 22 of the patients and found it in 3/5 (60%) patients with only GI symptoms, in 8/10 (80%) of those with GI plus respiratory symptoms (net for all patients with GI symptoms 11/15 or 73%), and in only 1/7 (14.3%) of patients with respiratory symptoms alone.

Written by Henry C. Barry, MD, MS, on July 7, 2020. (Source: Han C, Duan C, Zhang S, et al. Digestive symptoms in COVID-19 patients with mild disease severity: clinical presentation, stool viral RNA testing, and outcomes. Am J Gastroenterol. 2020;115(6):916-923.)

July 8, 2020, Research Update

Ischemic Stroke in COVID-19 Patients Hospitalized in New York. This was a retrospective cohort study of 1,916 patients with COVID-19 who were hospitalized or had an emergency department visit at two New York hospitals between March 4, 2020, and May 2, 2020. They were compared with 1,486 patients who presented to the same hospitals with influenza A or B between January 1, 2016, and May 31, 2018. A stroke registry was used to determine the incidence of acute ischemic stroke in the influenza patients, and chart review by two board certified neurologists was used for the patients with COVID-19. Estimates of the incidence of stroke were adjusted for age, sex, and race. Among patients with COVID-19, 31 of 1,916 (1.6%, 95% CI, 1.3% to 2.6%) were diagnosed with acute ischemic stroke, compared with only three of 1,486 of the influenza patients (0.2%, 95% CI, 0.0% to 0.6%). After adjusting for age, sex, and race, the likelihood of acute ischemic stroke was far more likely in the patients with COVID-19 (OR 7.6, 95% CI, 2.3 to 25.2). Sensitivity analyses adding adjustment for vascular risk factors, symptoms, inpatient status, and location of treatment found similar results (OR 4.6 to 9.3).

Written by Mark H. Ebell MD, MS, on July 6, 2020. (Source: Merkler AE, Parikh NS, Mir S, et al. Risk of ischemic stroke in patients with coronavirus disease 2019 (COVID-19) vs patients with influenza [published online July 2, 2020]. JAMA Neurol. 2020. )

July 7, 2020, Research Update

Counting Deaths in the United States due to the COVID-19 Pandemic. There has been controversy about the number of deaths in the United States due to the COVID-19 pandemic. Causes of death can be difficult to determine, and a significant reporting lag also occurs, making real-time estimates inaccurate. A good way to measure the impact of COVID-19 is to compare all-cause mortality before and during the pandemic. Two recent studies that used National Center for Health Statistics data and sophisticated modeling techniques reached similar conclusions about the under-reporting of COVID-19–related deaths. Woolf’s group compared weekly all-cause mortality counts for 2014-2019 with all-cause mortality during the COVID-19 epidemic for eight weeks in March and April 2020 to determine the excess deaths in 2020 over the expected number based on 2014-2019 statistics. The excess number of deaths was estimated to be 87,001, averaging 10,875 deaths per week. Weinberger used a similar methodology, comparing total deaths in the United States from March 1 through May 30, 2020, with average total deaths during the same period from 2015 through 2019. Comparing expected deaths to observed deaths, there were 122,300 excess deaths (95% prediction interval, 116,800 to 127,000), or 9,408 excess deaths per week. There were 95,235 reported deaths directly attributed to COVID-19 during the same time period, a figure that underestimates deaths due to the pandemic by about 28%. The excess mortality is likely because of a combination of undercounting of COVID-19 deaths, people for which COVID-19 was a significant contributing factor but not the “official” cause of death, and an excess number of deaths due to other causes for which people did not seek care because of the pandemic.

Written by John Hickner, MD, MS, on July 3, 2020. (Sources: Woolf SH, Chapman DA, Sabo RT, et al. Excess deaths from COVID-19 and other causes, March-April 2020 [published online July 1, 2020]. JAMA. 2020. ; and Weinberger DM, Chen J, Cohen T, et al. Estimation of excess deaths associated with the COVID-19 pandemic in the United States, March to May 2020 [published online July 1, 2020]. JAMA Intern Med. 2020. )

July 6, 2020, Research Update

Colchicine Yields Small Clinical Improvement in a Preliminary Randomized Trial. In this preliminary clinical trial in Greece, 105 hospitalized patients with COVID-19 were randomized to receive colchicine or standard medical therapy for up to three weeks. The outcomes were cardiac and inflammatory marker levels and clinical outcomes. The primary clinical outcome was time from baseline to clinical deterioration, defined as a two-grade deterioration on a seven-point ordinal scale that ranges from 1 = ambulatory, normal activities to 7 = death. There were no significant differences in the levels of high-sensitivity troponin and time to maximum c-reactive protein level. Mean event-free survival time was 18.6 days in the control group vs. 20.7 in the colchicine group (log rank P = .03). Seven patients (14%) in the control group and one patient (1.8%) in the colchicine group had a two-point deterioration in the primary clinical outcome (odds ratio, 0.11; 95% CI, 0.01 to 0.96; P = .02). The sample size was too small to test for differences in mortality. A larger trial is needed to see whether colchicine leads to meaningful clinical benefit for patients with COVID-19.

Written by John Hickner, MD, MS, on July 2, 2020. (Source: Deftereos SG, Giannopoulos G, Vrachatis DA, et al. Effect of colchicine vs standard care on cardiac and inflammatory biomarkers and clinical outcomes in patients hospitalized with coronavirus disease 2019: the GRECCO-19 randomized clinical trial. JAMA Netw Open. 2020;3(6):e2013136. )

July 3, 2020, Research Update

Classroom Spread of SARS-CoV-2. The Centers for Disease Control and Prevention reports a fascinating case report with contact tracing in a classroom setting. A schoolteacher had been in Europe in February and became ill with headache, sore throat, myalgia, and fever. The teacher taught 16 classes, each with no more than 30 students, while still symptomatic. Six of the classes were non-interactive, but the remainder were interactive where the teacher walked about the classroom. The teacher tested positive for SARS-CoV-2 by polymerase chain reaction. The exposed students were asked to self-quarantine. Of the 120 students contacted, 21 volunteered to be tested serologically, which took place about two weeks after the teacher’s return. The students were between five and 18 years of age. Of the 21, five students were in the interactive classes. One of these students had a positive antibody test, one had an indeterminate result, and three tested negative. Whereas the symptoms may be “suggestable,” the student who tested positive reported one day of myalgias and rhinorrhea and three days of cough. The indeterminate student and one other student reported no symptoms. Two students who tested negative reported one day of mild symptoms: fever, headache, and rhinorrhea. None of these symptomatic children had polymerase chain reaction testing. The authors provide no information about the remaining 16 students. It is difficult to draw conclusions about modes of teaching and risk of transmission given the selective testing and reporting.

Written by Henry C. Barry, MD, MS, on July 1, 2020. (Source: Brown NE, Bryant-Genevier J, Bandy U, et al. Antibody responses after classroom exposure to teacher with coronavirus disease, March 2020 [published online June 29, 2020]. Emerg Infect Dis. 2020;26(9). )

July 2, 2020, Research Update

Centers for Disease Control and Prevention Seroprevalence Study Reports Rate of 1.9% to 6.9% in Six States and Approximately 11 Total Cases Per Confirmed Case. This is the initial report of a seroprevalence survey conducted by the Centers for Disease Control and Prevention (CDC) in six states (Connecticut, south Florida, Missouri, New York metro area, Utah, and Washington State Puget Sound region). They identified a convenience sample of 11,933 blood specimens submitted to commercial laboratories for reasons other than COVID-19. The specimens were evaluated for antibodies to SARS-CoV-2 using a test that was 96% sensitive and 99.3% specific, with the results adjusted for the estimated false negative and false positive rates. The age-standardized positive rates were 1.1% in Washington, 1.9% in south Florida, 2.2% in Utah, 2.7% in Missouri, 4.9% in Connecticut, and 6.9% in the metro New York City region. Rates were higher in men than women in five of six jurisdictions, and no clear pattern by age group occurred, which varied in different jurisdictions. Based on the number of confirmed cases, they estimate a range of six to 24 total cases per confirmed case, with four of six jurisdictions having 10.8 to 11.9 total cases per confirmed case. They plan to perform ongoing surveillance every three to four weeks.

Written by Mark Ebell MD, MS, on June 27, 2020. Centers for Disease Control and Prevention. Commercial laboratory seroprevalence survey data. Updated June 26, 2020.  //www.cdc.gov/coronavirus/2019-ncov/cases-updates/commercial-lab-surveys.html)

July 1, 2020, Research Update

ACEIs and ARBs Not Associated with COVID-19 Infection and Mortality. This Danish study is one of several investigations to find no association between use of angiotensin-converting enzyme inhibitors (ACEIs)/angiotensin receptor blockers (ARBs) and the incidence and adverse outcomes of COVID-19 infection. Using Danish national registries and adjusting for age, sex, and medical history, in a cohort study of 4,480 patients with COVID-19 that included 895 ACEI/ARB users and 3,585 nonusers, there was no significant increased risk of death (adjusted hazard ratio [HR] 0.83 [95% CI, 0.67 to 1.03]) or death or severe disease (adjusted HR 1.04 [95% CI, 0.89 to 1.23]) within 30 days of diagnosis of COVID-19. In the companion case-control study of patients with hypertension, there was no significant difference in ACEI/ARB use between those who contracted COVID-19 and those who did not (adjusted HR 1.05 [95% CI, -0.80 to 1.36]). In a similar case-control Spanish study of 1,139 patients taking ACEIs/ARBs and 11,390 population controls, adjusted for age, sex, cardiovascular comorbidities, and risk factors, there was no difference in hospital admission between ACEI/ARB users and nonusers (adjusted odds ratio 0.94 [95% CI, 0.77 to 1.15]).  Along with prior studies, we have substantial data to conclude that use of ACEI/ARBs does not increase risk of or worse outcomes from COVID-19 infection.

Written by John Hickner, MD, MS, on June 26, 2020. (Sources: Fosbøl EL, Butt JH, Østergaard L, et al. Association of angiotensin-converting enzyme inhibitor or angiotensin receptor blocker use with COVID-19 diagnosis and mortality [published online June 19, 2020]. JAMA.; and de Abajo FJ, Rodríguez-Martín S, Lerma V; MED-ACE2-COVID19 Study Group. Use of renin–angiotensin–aldosterone system inhibitors and risk of COVID-19 requiring admission to hospital: a case-population study. Lancet.2020;395(10238):1705-1714.)

June 30, 2020, Research Update

Corticosteroids Reduce Mortality in Hospitalized Patients with COVID-19 Requiring Oxygen or Ventilatory Support. The RECOVERY trial (Randomised Evaluation of COVid-19 thERapY) randomized 11,303 hospitalized patients with suspected or confirmed COVID-19 to one of six arms: azithromycin, lopinavir-ritonavir, tocilizumab, convalescent plasma, low-dose dexamethasone (6 mg once daily for 10 days), or usual care. This was an open-label trial, and allocation to groups was appropriately concealed. This study reports the results from the comparison of low-dose dexamethasone in 2,104 patients with 4,321 patients randomized to usual care. The primary outcome was 28-day mortality. The mean age of participants was 66 years of age, 36% were women, and 56% had at least one major comorbidity; groups were balanced at baseline. At the time the manuscript was written, 89% of patients had a positive test for COVID-19, 0.4% had a test result pending, and 10% apparently had a negative result. Follow-up was complete, and few crossovers occurred (8% in the usual care group received dexamethasone, whereas 5% in the dexamethasone group did not receive the drug). Mortality was significantly reduced overall (22.9% vs. 25.7%; rate ratio [RR], 0.83; 95% CI, 0.75 to 0.93; number needed to treat [NNT] = 36). The degree of benefit was strongly associated with the severity of illness. For patients requiring mechanical ventilation, mortality reduction was greatest (29.3% vs. 41.4%; RR, 0.64; 95% CI, 0.51 to 0.81; NNT = 8), whereas patients requiring oxygen but who were not mechanically ventilated benefitted somewhat less (23.3% vs. 26.2%; RR, 0.82; 95% CI, 0.72 to 0.94; NNT = 35). Patients randomized to dexamethasone were also less likely to require mechanical ventilation (5.7% vs. 7.8%; RR, 0.77; 95% CI, 0.62 to 0.95). No benefit, and in fact a trend toward harm, was observed for hospitalized patients not requiring oxygen or mechanical ventilation (mortality 17.8% vs. 14.0%; RR, 1.19; 95% CI, 0.91 to 1.55). The latter finding has important implications for the care of outpatients not receiving oxygen—they should not be started on a glucocorticoid in the absence of another compelling indication. 

Updated by Mark H. Ebell, MD, MS, on July 27, 2020. (Source: The RECOVERY Collaborative Group. Dexamethasone in hospitalized patients with COVID-19—preliminary report [published online July 17, 2020]. N Engl J Med. 2020. )

June 29, 2020, Research Update

About 12 Undetected Cases Per Confirmed Case in Swiss Prevalence Study. In the Swiss canton around Geneva, the first confirmed case of COVID-19 was detected on February 26, with 5,160 confirmed cases by May 9, 2020. To determine the true prevalence of previous infection in this population, they planned a series of 12 weekly seroprevalence surveys; these are the results from the first five weeks. Each week, 1,300 participants in an ongoing health study were randomly selected and invited to participate, and if they agreed, they underwent antibody testing with a test that the research team determined was 93% sensitive and 100% specific. In their calculations, they adjusted for differences in the age and sex distribution of their sample and the overall population. At five weeks, 35% of invitees had become participants, 3% were ineligible, and the remainder were waiting to book an appointment or waiting to be recontacted. The seroprevalence ranged from 3.5% in week 1 to 6.0% in weeks 2 and 4, and 10.6% in weeks 3 and 5. Seropositivity was lower for children younger than 10 years than in those older (0.8% vs. 4.1% to 9.9%, depending on age stratum). Based on the seroprevalence and the number of confirmed cases, they estimate that there are 11.6 infections in the community for every confirmed case.

Written by Mark Ebell, MD, MS, on June 25, 2020. (Source: Stringhini S, Wisniak A, Piumatti G, et al. Seroprevalence of anti-SARS-CoV-2 IgG antibodies in Geneva, Switzerland (SEROCoV-POP): a population-based study [published online June 11, 2020]. Lancet. 2020. )

June 26, 2020, Research Update

High Mortality Rate for Cancer Patients with COVID-19 Infection. This is a cohort study from the United States, Canada, and Spain of 928 cancer patients with COVID-19 infection. It is the largest outcome study of COVID-19 infection in cancer patients to date. The primary outcome was 30-day mortality from the time of diagnosis of COVID-19. The mean age was 66 years. Thirty-nine percent of patients were on active anticancer treatment. Forty-three percent had active cancer. At the time of analysis on May 7, 2020, 13% of patients had died. Risk factors for death were increased age, male sex, smoking, number of comorbidities, and active cancer. This mortality rate is similar to the 11% mortality reported in a cohort of 334 patients with cancer and COVID-19 infection from the Mount Sinai Health System. (Mehta V, Goel S, Kabarriti R, et al. Case fatality rate of cancer patients with COVID-19 in a New York hospital system [published online May 1, 2020]. Cancer Discov. 2020. ). The investigators do not provide a comparison group of patients with cancer and no COVID-19 infection. Nonetheless, the mortality rate is high.

Written by John Hickner, MD, MS, on June 25, 2020. (Source: Kuderer NM, Choueiri TK, Shah DP, et. al. Clinical impact of COVID-19 on patients with cancer (CCC19): a cohort study. Lancet. 2020;395(102410):1907-1918. )

June 25, 2020, Research Update

Promising Early Data from China on a Vaccine. Around the world, over 100 candidate vaccines against SARS-CoV-2 are in various stages of development and testing. These authors from China present data from a small open-label, nonrandomized study of CanSino’s nonreplicating adenovirus type-5 vectored COVID-19 vaccine. The study included 108 persons who had no evidence of SARS-CoV-2 infection, as indicated by negative serum specific IgG or IgM antibodies, negative SARS-CoV-2 nucleic acid on pharyngeal swabs or sputum and anal swabs, and a clear chest CT. The study participants were sequentially enrolled to receive an intramuscular infection of this recombinant vaccine in one of three doses: 5 x 1010, 1 x 10¹¹, and 1.5 x 10¹¹ viral particles). Note that there was no control group. The first group (n = 36) received the low-dose injection and were followed up for a minimum of three days before proceeding to recruit further participants (n = 36) who received the middle dose. After safety observation for three days, the last group of participants (n = 36) received the high dose. The researchers collected blood 14 and 28 days after the vaccinations to assess antibody response, including neutralizing antibodies. During the first 14 days after the vaccination, study investigators confirmed patient-reported adverse events; afterward, they used only patient-reported events. Most (81%) of the participants reported at least one adverse event within seven days of vaccination, and the overall rate was similar across all three doses. Just over half (54%) experienced injection site pain with little difference among the three doses. The most commonly reported systemic adverse reactions overall were fever (46%), fatigue (44%), headache (39%), and muscle pain (17%). More than half (56%) of the persons receiving the high dose developed fever. Most of the adverse effects were mild. The researchers were able to detect specific antibodies and neutralizing antibodies with all three doses by 14 days after immunization with a peak at 28 days. The high-dose vaccine induced the greatest immunogenic responses. This is early data from a small study with no control group but provides hope that an effective vaccination is feasible.

Written by Henry C. Barry, MD, MS, on June 24, 2020. Source: Zhu F-C, Li Y-H, Guan X-H, et al. Safety, tolerability, and immunogenicity of a recombinant adenovirus type-5 vectored COVID-19 vaccine: a dose-escalation, open-label, non-randomised, first-in-human trial. Lancet. 2020;395(10240):1845-1854.)

June 24, 2020, Research Update

Fecal Viral Shedding of SARS-CoV-2  and Reports of Gastrointestinal Symptoms. These authors searched two bibliographic databases and two preprint servers as well as the reference lists of all retrieved studies to identify studies reporting clinical symptoms of patients infected with SARS-CoV-2 and for studies that reported testing of stool samples in patients with COVID-19. They excluded studies that were only reported as abstracts. They ultimately included 23 peer-reviewed papers and 6 pre-print studies with 4,805 patients. Most of the studies were observational and most came from mainland China. Overall, the studies were of moderate quality, and 12% of patients presented with gastrointestinal symptoms. Most of the studies reported on diarrhea, roughly occurring in 7.8% of patients, but there was marked variability in the reports of frequency. Twelve studies measured the presence of vomiting or nausea and found these occurred in 3.9% of the patients (with significant variability). Eight studies reported on viral shedding, finding that 40.5% (significant heterogeneity) of patients with COVID-19 shed the virus in their stools.

Written by Henry C. Barry, MD, MS, on June 17, 2020. (Source: Parasa S, Desai M, Thoguluva Chandrasekar V, et al. Prevalence of gastrointestinal symptoms and fecal viral shedding in patients with coronavirus disease 2019: a systematic review and meta-analysis [published online June 11, 2020]. JAMA Netw Open. 2020;3(6):e2011335. 

June 23, 2020, Research Update

Natural History of Asymptomatic COVID-19 Infection in Cruise Ship Passengers. These researchers identified 712 of 3,711 passengers and crew members of the Diamond Princess cruise ship who had tested positive polymerase chain reaction (PCR) for SARS-CoV-2, 410 of whom (57.6%) had no symptoms. It is not clear how they identified them, but 96 of these asymptomatic persons and 32 of their cabin mates who tested negative were transferred to a hospital in Japan for seven days of observation. During the subsequent week, 11 (11.5%) developed symptoms, typically three to five days after the first positive test. Eight of the 32 (25%) cabin mates had a positive PCR within three days of quarantine but were asymptomatic. That means that 85 infected passengers and their eight infected cabin mates in this subset of the asymptomatic cruise passengers remained asymptomatic, for a total of 93 persons (73%) with asymptomatic COVID-19. Resolution, defined as two consecutive negative PCR tests, occurred in 90 (97%) of these asymptomatic persons. A longer time to resolution occurred more frequently in older persons and in those with comorbid conditions such as hypertension or diabetes.

Written by Henry C. Barry, MD, MS, on June 17, 2020. (Source: Sakurai A, Sasaki T, Kato S, et al. Natural history of asymptomatic SARS-CoV-2 infection [published online June 12, 2020]. N Engl J Med. 2020. )

June 22, 2020, Research Update

COVID-19 Patients with Chronic Diseases Are Much More Likely To Be Hospitalized and To Die. This Center for Disease Control and Prevention (CDC) report summarizes COVID-19 cases reported to the CDC from January 22 through May 30, 2020. There were 1,761,503 cases reported and 103,700 deaths. The seven-day average number of cases peaked in the United States on April 12 (31,994 cases), and daily deaths peaked on April 21 (2,856). The CDC had sufficient information on 287,320 individuals with COVID-19 infection to report outcomes of patients with and without chronic diseases. Cardiovascular disease was present in 32% of cases, diabetes in 30%, and chronic lung disease in 18%. Of all patients with COVID-19 reported to the CDC during this time period, 184,673 (14%) patients were hospitalized, 29,837 (2%) were admitted to an ICU, and 71,116 (5%) died. The hospitalization rate was six times higher among patients with a chronic disease (45.4% vs. 7.6%), and the death rate was twelve times higher (19.5% vs. 1.6%).

Written by John Hickner, MD, MS, on June 17, 2020. Source: Stokes EK, Zambrano LD, Anderson KN, et al. Coronavirus disease 2019 case surveillance—United States, January 22–May 30, 2020. MMWR Morb Mortal Wkly Rep. 2020;69(24):759-765.)

June 19, 2020, Research Update

Corticosteroids Reduce Mortality in Hospitalized Patients with COVID-19 Requiring Oxygen or Ventilatory Support. Editor’s Note: This research brief is based on a press release. We will update it when the article is published. We are covering it because a highly respected team of researchers from Oxford University is reporting the results in considerable detail, and it is funded by the UK government.

The RECOVERY trial (Randomised Evaluation of COVid-19 thERapY) randomized more than 11,500 hospitalized patients with COVID-19 to one of six arms: azithromycin, lopinavir-ritonavir, tocilizumab, convalescent plasma, low-dose dexamethasone (6 mg once daily for 10 days), or usual care. This announcement reports the results from the comparison of low-dose dexamethasone in 2,104 patients, with 4,321 patients randomized to usual care. The primary outcome was 28-day mortality. The risk ratio for mortality in patients requiring mechanical ventilation was 0.65 (95% CI, 0.48 to 0.88); for those receiving only oxygen, it was 0.80 (95% CI, 0.67 to 0.96). No benefit was observed for patients not requiring oxygen or mechanical ventilation. I applied those risk ratios to the mortality rates in the usual care groups to calculate absolute risk reductions and numbers needed to treat (NNT). For patients on mechanical ventilation, the mortality rates are 26.6% vs. 41%, for an NNT of 7. For patients receiving only oxygen, the mortality rates are 20% vs. 25%, for an NNT of 20 (the authors report slightly different NNTs of eight for patients on a ventilator and 25 for those receiving oxygen). They provide no data on the patient demographics, any harms or complications, or the effect on time to recovery. We await the full report.

Written by Mark H. Ebell MD, MS, on June 16, 2020. (Source: press release from Oxford University [released June 16, 2020]. )

June 18, 2020, Research Update

Econometric Study: Large Number of Cases Prevented by Nonpharmaceutical Interventions. This report is from a team of economists and public health experts primarily at the University of California, Berkeley. They take econometric methods that are generally used to study the impact of nonpharmaceutical interventions on economic growth (e.g., does raising a minimum wage increase unemployment?) and apply them to the question of whether policy changes affect a different kind of growth (e.g., that of the COVID-19 pandemic). The researchers gathered data from 1,717 administrative units in six countries (United States, Iran, Italy, South Korea, China, and France). They estimated a daily growth rate of 43% overall in new cases in the absence of any measures, ranging from 34% in the United States to 68% in Iran. Their econometric modeling concludes that implementation of all policy measures (such as school closures, social isolation guidelines, and travel bans) resulted in reductions in the rate of growth of cases, with larger reductions in Iran, China, South Korea, and Italy and smaller reductions in the United States and France. They estimate that these measures reduced the number of confirmed cases in those six countries by more than 60 million, with a reduction of confirmed cases in the United States by 4.8 million and total cases by 60 million (assuming that many cases were asymptomatic or never confirmed). This study provides an estimate of the magnitude of benefit of these nonpharmaceutical interventions and is consistent with that of a study applying epidemiologic methods to the same question by Imperial College researchers. These estimates are broadly consistent with those of the University of California, Berkeley, researchers using completely different econometric modeling approaches.

Written by Mark H. Ebell MD, MS, on June 10, 2020. Source: Hsiang S, Allen D, Annan-Phan S, et al. The effect of large-scale anti-contagion policies on the COVID-19 pandemic [published June 8, 2020]. Nature. 2020.; and Flaxman S, Mishra S, Gandy A, et al. Estimating the effects of non-pharmaceutical interventions on COVID-19 in Europe [published online June 8, 2020]. Nature. 2020. )

June 17, 2020, Research Update

A Large Proportion of Individuals with COVID-19 Infection Are Asymptomatic. This narrative review quantifies the proportion of people with COVID-19 infections who are asymptomatic at the time of testing. It includes 16 cohorts of COVID-19–positive individuals from Iceland, Italy, Greece, France, Japan, Argentina, and the United States, including four ship outbreaks, ranging in size from 76 to 13,080 individuals. The percentage of people who tested positive who had no symptoms at the time of testing ranged from a low of 6.3% in nursing home residents at a facility in King County, Washington, to highs of 87.8% of occupants in a Boston homeless shelter, 87.9% in an obstetric service in New York City, and 96% of 3,146 inmates in state prison systems in Arkansas, North Carolina, Ohio, and Virginia. The three cohorts that came from representative samples of the population suggest that the rate of asymptomatic infection is in the range of 40% to 45%. As we noted in a study of infector–infectee pairs in China, transmission from asymptomatic individuals may account for approximately 40% of cases. Therefore, quarantine alone is insufficient to prevent spread; social distancing and use of face masks are important preventive measures.

Written by John Hickner, MD, MS, on June 10, 2020. (Source: Oran DP, Topol EJ. Prevalence of asymptomatic SARS-CoV-2 infection: a narrative review [published online June 3, 2020]. Ann Intern Med. 2020. )

June 16, 2020, Research Update

Clinical Manifestations of Children with Multisystem Inflammatory Syndrome Associated with COVID-19 in Three Studies from New York, France, and England. The pediatric multisystem inflammatory syndrome shares many features with Kawasaki disease and causes great concern because it carries a worse prognosis than what is otherwise expected in children infected with SARS-CoV-2. Three recent papers report data from medical record audits of children hospitalized in New York, France, and England. These case series are challenging to summarize because of differences in reporting, but we can still glean some common elements. Overall, 96 children were included in these studies, 38 (40%) were black, 81 (84%) had lab-confirmed COVID-19, and 33 (34%) also met criteria for Kawasaki disease. Two studies reported that 36 (38%) had gastrointestinal symptoms. The other study reported that 31 children (32%) had abdominal pain and that 26 (27%) had vomiting. Rash occurred in 58 (60%) of the children, and desquamation, only reported in two studies, occurred in 7 of 28 (25%). Fifty-four (5%) had conjunctivitis, and 42 (44%) had redness or swelling of the lips or mucous membranes. Coronary artery dilation occurred in 13 (14%) of the children, myocardial dysfunction in 24 (25%), and shock in 57 (59%). Only two of the studies reported chest radiograph data; 22 of the 38 (58%) were abnormal. 

Written by Henry C. Barry, MD, MS, on June 10, 2020. (Sources: Toubiana J, Poirault C, Corsia A, et al. Kawasaki-like multisystem inflammatory syndrome in children during the covid-19 pandemic in Paris, France: prospective observational study. BMJ. 2020;369:m2094. Cheung EW, Zachariah P, Gorelik M, et al. Multisystem inflammatory syndrome related to COVID-19 in previously healthy children and adolescents in New York City [published online June 8, 2020]. JAMA. 2020. . Whittaker E, Bamford A, Kenny J, et al. Clinical characteristics of 58 children with a pediatric inflammatory multisystem syndrome temporally associated with SARS-CoV-2 [published online June 8, 2020]. JAMA. 2020. )

June 15, 2020, Research Update

A Third of Patients with COVID-19 Early in the China Outbreak Had Neurologic Symptoms. These authors reviewed the records of 214 consecutive patients hospitalized for COVID-19 between January 16 and February 19, 2020. The study took place in three hospitals in Wuhan, China. In addition to collecting demographic information, comorbidities, and common respiratory symptoms, the authors also looked for the presence of neurologic symptoms. Because they do not tell us whether the hospitals had an explicit approach to prospectively collect this information, we can assume only that the frequency of each element potentially is underestimated. Additionally, because they were trying to minimize direct contact with patients, much of the information was based on subjective symptoms and initial exam findings. They classified neurologic manifestations into three categories: central nervous system (CNS; dizziness, headache, impaired consciousness, stroke, ataxia, seizure), peripheral nervous system (PNS; altered taste or smell, vision impairment, nerve pain), and skeletal muscle injuries (pain and elevated serum creatine kinase). Overall, 78 patients (36%) had neurologic manifestations. Patients with severe COVID-19 infections were more likely to have neurologic symptoms (46%) than those with mild disease (30%). About 25% of all patients had CNS manifestations; 31% of those with severe COVID and 21% of those with mild infection. About 9% had PNS manifestations, with no difference based on COVID severity. Skeletal muscle injuries occurred once in patients with severe COVID and once in patients with mild COVID.

Written by Henry C. Barry, MD, MS, on June 10, 2020. (Source: Mao L, Jin H, Wang M, et al. Neurologic manifestations of hospitalized patients with coronavirus disease 2019 in Wuhan, China. JAMA Neurol. 2020;77(6):683-690.)

June 12, 2020, Research Update

Convalescent Plasma Marginally Improved COVID-19 Outcomes in Severely Ill Patients in China. In this open-label, multicenter, randomized clinical trial (RCT), 103 patients (median 70 years of age) with severe or life-threatening COVID-19 infection in Wuhan, China, were randomized to receive convalescent plasma or to not receive it. The primary outcome was time to clinical improvement within 28 days, defined as the patient was discharged alive or a reduction of 2 points on a 6-point disease severity scale (1 = discharged, 6 = death). Clinical improvement occurred in 52% (27/52) of the convalescent plasma group and 43% (22/51) in the control group, not statistically significant (hazard ratio [HR], 1.40 [95% CI, 0.79 to 2.49]; P = .26). Among those with severe disease, the primary outcome occurred in 91% (21/23) of the convalescent plasma group vs. 68% (15/22) of the control group (HR, 2.15 [95% CI, 1.07 to 4.32]; P = .03); among those with life-threatening disease (the most severely ill patients) the primary outcome occurred in 21% (6/29) of the convalescent plasma group vs. 24% (7/29) of the control group (HR, 0.88 [95% CI, 0.30 to 2.63]; P = .83). There was no significant difference in 28-day mortality (15.7% vs. 24.0%; OR, 0.65 [95% CI, 0.29 to 1.46]; P = .30) or time from randomization to discharge (51.0% vs. 36.0% discharged by day 28; HR, 1.61 [95% CI, 0.88 to 2.93]; P = .12). The trial was terminated early after 103 of the planned 200 patients were enrolled because of the rapid decline in COVID-19 cases in Wuhan. A larger study is needed to confirm or refute the results in this underpowered RCT.

Written by John Hickner, MD, MS, on June 5, 2020. (Source: Li L, Zhang W, Hu Y, et al. Effect of convalescent plasma therapy on time to clinical improvement in patients with severe and life-threatening COVID-19: a randomized clinical trial [published online June 3, 2020]. JAMA. 2020.  )

June 11, 2020, Research Update

Self-Collected Swabs Are Accurate in Detecting SARS-CoV-2 in Ambulatory Settings. These researchers in the United States identified 530 symptomatic patients seen in ambulatory clinics in the Puget Sound area of Washington. After brief training, the patients were also asked to collect tongue, nasal (inserted vertically into nasal passage), and midturbinate (inserted horizontally) samples, in that order. Afterward, trained staff collected swab samples from the nasopharynx. The authors estimated they would need to have 48 patients test positive for SARS-CoV-2 to have adequate power. Not surprisingly, most of the time patients tested positive at more than one site, regardless of who gathered the sample. Using the staff-collected sample as the first-choice standard, the authors determined that self-collected samples were 89.8% sensitive (one-sided 97.5% CI, 78.2 to 100.0) for tongue samples, 94.0% sensitive (97.5% CI, 83.8 to 100.0) for nasal samples, and 96.2% sensitive (97.5% CI, 87.0 to 100.0) for midturbinate samples. This study suggests that patient-collected samples are reasonably accurate and have the potential to decrease the frequency of exposing staff to potentially infectious material.

Written by Henry C. Barry, MD, MS, on June 4, 2020. (Source: Tu Y-P, Jennings R, Hart B, et al. Swabs collected by patients or health care workers for SARS-CoV-2 testing [published online June 3, 2020]. N Engl J Med. 2020. )

June 10, 2020, Research Update

Effects of Physical Distance, Face Masks, and Eye Protection on Virus Transmission in Health Care and Community Settings. This meta-analysis, sponsored by the World Health Organization, identified 172 observational studies of the effects of physical distancing, face masks, and eye protection on virus transmission in health care and community settings. The investigators found no randomized trials, but they found 44 observational comparative studies that included 25,697 people. Transmission of viruses was lower with physical distancing of 1 meter (roughly 3 feet) or more compared with a distance of less than 1 meter, an absolute risk difference (RD) of -10.2%, a finding with moderate certainty. Protection improved as distance increased (relative risk [RR] of 2.0 per meter). Face mask use was associated with RD of -14.3%, but with low certainty of evidence. There was stronger evidence for risk reduction with N95 or similar respirators compared with disposable surgical masks or reusable cotton masks. Eye protection also was associated with lower infection risk (RD -10.6%). Randomized trials are needed, but this meta-analysis summarizes the best available evidence for effectiveness of these interventions to date.

Written by John Hickner, MD, MS, on June 5, 2020. (Source: Chu DK, Akl EA, Duda S, et al. Physical distancing, face masks, and eye protection to prevent person-to-person transmission of SARS-CoV-2 and COVID-19: a systematic review and meta-analysis [published online June 1, 2020]. Lancet. 2020. )

June 9, 2020, Research Update

Postexposure Prophylaxis with Hydroxychloroquine Is Not Effective. This study is different from all other hydroxychloroquine (HCQ) studies in that it studied postexposure prophylaxis in healthy patients. Patients were identified who had a high-risk exposure to someone with confirmed COVID-19—defined as at least 10 minutes, less than six feet away, and without facial covering—or a moderate-risk exposure that allowed a face mask but no eye shield. They randomized 821 initially asymptomatic persons within four days of exposure to HCQ (800 mg once, then 600 mg in six to eight hours, and then 600 mg once daily for four days) or matching placebo. The primary outcome was laboratory-confirmed or clinically suspected COVID-19 (testing was not yet widely available) in the next 14 days after enrollment. Median age was 40 years, and healthcare workers accounted for two-thirds of the participants. No difference occurred between groups in the primary outcome, with 49/414 (11.8%) reporting infection in the HCQ group and 58/407 (14.3%) in the placebo group (risk difference -2.4%; 95% CI, -7.0 to 2.2). The findings were the same at 5, 10, and 14 days.

Written by Mark H. Ebell, MD, MS, on June 3, 2020. Boulware DR, Pullen MF, Bangdiwala AS, et al. A randomized trial of hydroxychloroquine as postexposure prophylaxis for Covid-19 [published online June 3, 2020]. N Engl J Med. 2020. ) 

June 8, 2020, Research Update

Systematic Review: Fever, Cough, Elevated C-Reactive Protein, and Ground Glass Opacities Common in Patients Diagnosed with COVID-19 in China. These authors identified 43 studies with 3,600 patients in China who were diagnosed with COVID-19. They excluded studies that took place outside of mainland China. They extracted and summarized the frequency of symptoms and laboratory and radiographic findings. They found substantial variability in the data as well as significant publication bias. Only nine of the studies were at low risk of bias, and 30 were of moderate risk of bias. Among the symptoms they report, the most common include fever (83%) and cough (60%), fatigue (38%), myalgias (29%), and dyspnea (25%). In addition, 8% had diarrhea; other studies have shown significantly higher risk of transmission among patients with diarrhea. The following laboratory findings were also commonly elevated: C-reactive protein (68.6%), lactate dehydrogenase (51.6%), and d-dimer (29.3%). Finally, 80% of patients had radiographic findings of ground glass opacities, and 73.2% had bilateral pneumonia. 

Written by Henry C. Barry, MD, MS, on June 2, 2020. (Source: Fu L, Wang B, Yuan T, et al. Clinical characteristics of coronavirus disease 2019 (COVID-19) in China: a systematic review and meta-analysis. J Infect. 2020;80(6):656-665.)

June 7, 2020, Research Update

Increased COVID-19 Mortality among Blacks, Asians, and Other Minority Groups in the UK. These researchers used data from the United Kingdom’s National Health System and identified patients who died while hospitalized for COVID-19 and applied estimated standardized mortality rates (SMR) using the entire English population as a reference. The SMRs, adjusted for age and geographic region, were lowest among white Irish (SMR 0.52; 95% CI, 0.45 to 0.60) and white British ethnic groups (0.88; 95% CI, 0.86 to 0.0.89) and were higher among black African (3.24; 95% CI, 2.90 to 3.62), black Caribbean (2.21; 95% CI, 2.02 to 2.41), Pakistani (3.29; 95% CI, 2.96 to 3.64), Bangladeshi (2.41; 95% CI, 1.98 to 2.91), and Indian (1.70; 95% CI, 1.56 to 1.85) minority ethnic groups. So, in a nation with universal healthcare coverage, socioeconomic factors are still important predictors of adverse health outcomes.

Written by Henry C. Barry, MD, MS, on June 1, 2020. (Source: Aldridge RW, Lewer D, Katikireddi SV, et al. Black, Asian and minority ethnic groups in England are at increased risk of death from COVID-19: indirect standardisation of NHS mortality data. Wellcome Open Research. 2020;5(88). )

June 6, 2020, Research Update

Face Masks and Disinfecting Reduced Household COVID-19 Transmission in China. Most transmission of COVID-19 occurs among household and close social contacts. The investigators studied transmission in Beijing, China, among household members in 124 families that had at least one COVID-19 infected individual. Secondary transmission occurred in 41/124 families (77 cases). The other 83 families had no secondary transmission cases. The secondary attack rate in families was 23% (77/335). Face masks, if worn by the index case prior to symptoms and by other family members, were 79% effective in reducing secondary infections. Daily use of chlorine or ethanol-based disinfectant was 77% effective in preventing spread. Household transmission was 18 times more likely with frequent daily close contact with the primary case and four times higher if the primary case had diarrhea.

Written by John Hickner, MD, MS, on June 1, 2020. (Source: Wang Y, Tian H, Zhang L, et. al. Reduction of secondary transmission of SARS-CoV-2 in households by face mask use, disinfection and social distancing: a cohort study in Beijing, China [published May 28, 2020]. BMJ Global Health. 2020. )

June 5, 2020, Research Update

Patients Actively Being Treated for Cancer Not at Increased Risk for Death from COVID-19 Compared with Other Cancer Patients. This UK study prospectively identified 800 patients with a broad range of active cancers (defined as treated in the past 12 months) who were newly diagnosed with symptomatic COVID-19. Of 800 patients, 34% were not currently being treated, 35% were receiving cytotoxic chemotherapy, 10% radiotherapy, 10% surgery, 8% hormone therapy, 6% immunotherapy, and 4% surgery within four weeks of their diagnosis. The mean age was 69 years, and 28.3% died of COVID-19. Comparing the 66% of patients who were recently treated with the 34% not currently being treated, multivariate analysis found no association between current treatment and increased risk of death. The primary independent predictors of death were increasing age, male sex, and comorbidities such as hypertension and cardiovascular disease.

Written by Mark Ebell MD, MS, on May 29, 2020. (Source: Lee LYW, Cazier JB, Starkey T, et al. COVID-19 mortality in patients with cancer on chemotherapy or other anticancer treatments: a prospective cohort study [published online May 28, 2020]. Lancet. 2020. )

June 4, 2020, Research Update

Adjusted In-Hospital Mortality from COVID-19 in Louisiana No Different for Black and White Patients. In this retrospective study from a single Louisiana health care system, the association between race and hospital mortality was determined for 1,382 hospitalized patients from a cohort of 3,481 black/non-Hispanic and white/non-Hispanic COVID-19 positive patients. Of these patients, 70.4% were black/non-Hispanic and 29.6% were white/non-Hispanic. The mortality results were adjusted for age, burden of chronic disease using the Charlson Comorbidity Index, public insurance status, residence in a low-income area, and obesity, factors all associated with increased odds of hospital admission. During the study period, 326 patients died. Among the patients who died, 70.6% were black/non-Hispanic, and 29.4% were white/non-Hispanic. The unadjusted case-fatality rate was 30.1% for white patients and 21.6% for black patients. Race was not associated with higher mortality (HR 0.89 for black/non-Hispanic vs. white/non-Hispanic, 95% CI, 0.68 to 1.17) when adjusted for the variables noted above and for severity of clinical illness at presentation to the hospital.

Written by John Hickner, MD, MS, on May 29, 2020. (Source: Price-Haywood EG, Burton J, Fort D, et al. Hospitalization and mortality among black patients and white patients with Covid-19 [published online May 27, 2020]. N Engl J Med. 2020. )

June 3, 2020, Research Update

Outcomes Similar for 5 vs. 10 Days Remdesivir but Trial Had Flaws. The results of the ACTT-1 trial suggest that remdesivir for 10 days reduces the time to recovery and possibly mortality in patients with COVID-19. In this study sponsored by Gilead, 397 hospitalized patients were randomized to 5 vs. 10 days of remdesivir (200 mg loading dose on day 1 followed by 100 mg daily). The study had two major flaws that limit our ability to interpret its conclusion that results were similar for the two groups. First, this was an open-label trial, so patients and physicians knew who was getting the medication and for how long. Second, clinically important differences occurred between groups at baseline, with more patients requiring high flow oxygen, mechanical ventilation, or ECMO in the 10-day group at baseline (69 vs. 53). In the ACTT-1 trial, those patients appeared to benefit less from remdesivir. There is also no description of how randomization was performed or whether allocation to groups was concealed, so it seems likely that sicker patients were nonrandomly allocated to the 10-day group. The primary outcome was improvement of at least two grades on a 7-point scale ranging from death (grade 1) to not hospitalized (grade 7); all patients started at grades 2 to 5. Improvement occurred more often in those in the five-day group (65% vs. 54%), but this difference was no longer statistically significant after adjustment for baseline differences. The final assessment of the authors is that no difference occurred in outcomes between 5 and 10 days, so 5 days should be considered, especially if supplies are limited.

Written by Mark Ebell MD, MS, on May 28, 2020. (Source: Goldman JD, Lye DCB, Hui DS, et al. Remdesivir for 5 or 10 Days in patients with severe Covid-19 [published online May 27, 2020]. N Engl J Med. 2020; )

June 2, 2020, Research Update

Remdesivir Shortens Recovery Time in Hospitalized Patients with COVID-19 and Lower Respiratory Tract Infections (ACTT-1). This preliminary report of a multicenter, multinational randomized trial caught a lot of news attention (see COVID-19 Research Brief from April 30, 2020), and it has now been published! The study, the Adaptive Covid-19 Treatment Trial (ACTT-1), took place in 60 different sites in the United States, Germany, Denmark, the United Kingdom, Greece, Korea, Mexico, Spain, and Japan. The researchers randomized 1,063 patients hospitalized with COVID-19 and evidence of lower respiratory tract involvement to receive remdesivir (n = 538; 200 mg IV on day 1 and then 100 mg IV daily for up to 10 days or hospital discharge) or matching placebo infusion (n = 521). There was a little controversy about this study because in mid-April, the primary endpoint was switched from mortality to time to recover. Time to recovery was significantly faster in the remdesivir-treated group than in those receiving placebo (11 vs. 15 days, p < 0.001). There was also a trend toward lower 14-day mortality in the remdesivir group (7.1% vs. 11.9%), but this was not statistically significant. If the mortality difference were significant, this would translate to a number needed to treat of 21. Projecting statistically, it is likely the study would have needed to continue for one to two weeks to accrue enough patients and events for the mortality data to become statistically significant. The rate of serious adverse events in the remdesivir-treated patients was slightly lower (21.1%) than in those receiving placebo (27.0%).

Written by Henry C. Barry, MD, MS, on May 27, 2020. (Source: Beigel JH, Tomashek KM, Dodd LE, et al. Remdesivir for the treatment of Covid-19—preliminary report [published online May 22, 2020]. N Engl J Med. 2020. )

June 1, 2020, Research Update

Large Multinational Cohort Study Finds the Use of Macrolide Antibiotics, Hydroxychloroquine, or Chloroquine Is Not Associated with Better Outcomes but Is Associated with Increased Harms.

(Source: Mehra MR, Desai SS, Ruschitzka F, et al. Hydroxychloroquine or chloroquine with or without a macrolide for treatment of COVID-19: a multinational registry analysis [published online May 29, 2020]. )

May 31, 2020, Research Update

Patients Receiving Convalescent Plasma at Mayo Clinic Had Few Transfusion-Related Serious Adverse Events. Many new treatments are being proposed for COVID-19, but safety assessment is often lacking. This report from the Mayo Clinic on a preprint server describes the safety outcomes for 5,000 consecutive patients from around the United States with COVID-19 treated with convalescent plasma as part of an FDA-approved Expanded Access Program that began in early April. The median was 62 years of age, and 81% had current severe or life-threatening disease while the remainder were at risk for severe disease. There were 36 serious adverse events (SAEs) within four hours of transfusion, of which only two were felt to be definitely related to the transfusion by the treating physician. The SAEs included 15 deaths (0.3), of which 4 (0.08%) were thought to be related to the transfusion. Other SAEs included transfusion-associated circulatory overload in seven patients, transfusion-related acute lung injury in 11 patients, and severe allergic transfusion reaction in three. Mortality at seven days was 14.9%, but without a randomized trial of convalescent plasma compared to usual care we do not know its effectiveness. At least it does appear to be safe. Primum non nocere.

Written by Mark Ebell MD, MS, on May 22, 2020. (Source: Joyner M, Wright RS, Fairweather D, et al. Early safety indicators of COVID-19 convalescent plasma in 5,000 patients [published online May 14, 2020]. MedRxIV.  )

May 30, 2020, Research Update

Patients Taking Biologic Agents Did Not Appear To Be at Increased Risk with COVID-19 in New York City. An unanswered concern is whether patients taking immune-modulating drugs for the treatment of inflammatory arthritis or inflammatory bowel disease are at greater risk of poor outcomes than the general population. This is a case series from a New York City health system of 86 such patients with COVID-19, of whom 14 were hospitalized, one died, and the remaining were managed as an outpatient. Most (87%) were taking an immune-modulating drug for their condition, with 72% taking any biologic or Janus kinase inhibitor, 44% a tumor necrosis factor inhibitor, and 20% methotrexate. The rate of hospitalization was slightly lower than that for New York City as a whole (16% vs. 26%); patients who were hospitalized tended to be older and have other comorbidities such as chronic lung disease. Use of nonbiologic agents such as methotrexate (43% vs. 15%), glucocorticoids (29% vs. 6%), and hydroxychloroquine (21% vs. 7%) was more common in the hospitalized patients. Overall this supports that chronic use of biologic agents does not significantly increase risk of adverse outcomes with COVID-19.

Written by Mark Ebell MD, MS, on May 22, 2020. (Source: Haberman R, Axelrad J, Chen A, et al. Covid-19 in immune-mediated inflammatory diseases—case series from New York [published online April 29, 2020]. N Engl J Med. 2020. )

May 29, 2020, Research Update

How Did Social Distancing Policies Affect Local Spread of COVID-19 in Georgia? Clarke County, Georgia, implemented mandatory social distancing policies on March 20, 2020, 14 days prior to statewide implementation on April 3. Counties surrounding Clarke followed the statewide guidelines. Using statewide case reports by county, the investigators estimated doubling time, an accepted measure of spread of infection, for Clarke compared with surrounding counties. By April 26, the five-day rolling average doubling time was twice as long in Clarke compared with surrounding counties (roughly 40 days vs. 20 days). A social distancing policy that prohibited events and gatherings of more than 10 people and that required individuals to remain at home 24 hours a day with very limited exceptions for essential travel was effective in slowing the spread of COVID-19.

Written by John Hickner, MD, MS, on May 20, 2020. (Source: Ebell MH, Bagwell-Adams G.  Mandatory social distancing associated with increased doubling time: an example using hyperlocal data [published online May 7, 2020]. Am J Prevent Med. 2020. )

May 28, 2020, Research Update

Stay-at-Home Orders Reduced Incidence of COVID-19 in Illinois Compared with Iowa Border Counties. Do stay-at-home orders reduce COVID-19 spread? This study compared daily changes in COVID-19 cases in eight Iowa counties bordering Illinois (where no stay-at-home order was in place) with seven Illinois counties bordering Iowa (where stay-at-home orders were in place). Sensitivity analysis was used to account for differences in other factors such as closing of schools and nonessential business, county population density, and poverty rates. Trends in cumulative COVID-19 incidence were compared before and after March 21, the day the stay-at-home order went into effect in Illinois. At 10, 20, and 30 days after that date, cases increased more quickly in Iowa than Illinois counties, with a difference of 4.7 additional cases per 10,000 residents (95% CI, -8.64 to -0.78; p = 0.02) at 30 days. The estimated excess cases in Iowa was as high as 217 cases during that 30 days, which represents 30.4% of the 716 cases diagnosed in those Iowa counties by that date. The sensitivity analysis supported these findings.

Written by John Hickner, MD, MS, on May 20, 2020. (Source: Lyu W, Wehby GL. Comparison of estimated rates of coronavirus disease 2019 (COVID-19) in border counties in Iowa without a stay-at-home order and border counties in Illinois with a stay-at-home order. JAMA Netw Open. 2020;3(5):e2011102.)

May 27, 2020, Research Update

Seroprevalence of COVID-19 Infection in Los Angeles County. Counting only confirmed cases significantly underestimates the true incidence of COVID-19 infection because mild and asymptomatic cases are generally not diagnosed. Seroprevalence studies of antibodies to COVID-19 provide better estimates of population infection. A random sample of 1,952 Los Angeles (LA) county residents were invited to participate in a seroprevalence study, and 863 adults were tested. Thirty-five (4%) individuals tested positive; after adjusting for the sensitivity and specificity of the test and population weighting, the estimate was adjusted slightly upward to 4.6%. Assuming a similar infection prevalence across LA county, the estimated number of infected individuals as of April 11, 2020, is about 400,000 rather than 8,430, which was the cumulative number of confirmed COVID-19 cases reported by April 11. This is roughly a 50-fold difference, suggesting that about 98% of COVID-19 cases in LA county were not diagnosed. These results are similar to a prior COVID-19 seroprevalence study in Santa Clara, California, that we reported in a Daily Research Brief on April 27. (Source: Bendavid E, Mulaney B, Sood N, et al. COVID-19 antibody seroprevalence in Santa Clara County, CA [published online April 17, 2020]. )

Written by John Hickner, MD, MS, on May 20, 2020. (Source: Sood N, Simon P, Ebner P, et al.  Seroprevalence of SARS-CoV-2-specific antibodies among adults in Los Angeles county, California, on April 10-11, 2020 [published online May 18, 2020]. JAMA. 2020. )

May 26, 2020, Research Update

Neither Hydroxychloroquine nor Azithromycin Are Associated with Decreased In-Hospital Mortality in New York. These authors reviewed the records of 1,438 patients hospitalized with COVID-19. These patients represented nearly 90% of all patients with COVID-19 hospitalized in the New York metropolitan region. The overall in-hospital mortality was 20.3%. The mortality rate for patients taking hydroxychloroquine (HCQ) alone was 19.9%, whereas the rate for those taking HCQ plus azithromycin was 25.7%. The mortality rate for patients taking only azithromycin was 10%, whereas it was 12.7% for those taking neither. After taking into account other factors associated with in-hospital mortality, they found no association with either drug alone or in combination and decreased hospital mortality. They also found, however, that patients taking the combination of HCQ plus azithromycin were twice as likely to suffer a cardiac arrest (adjusted OR, 2.13; 95% CI, 1.12 to 4.05), whereas no increase in risk occurred for patients taking either drug alone.

Written by Henry C. Barry, MD, MS, on May 20, 2020. (Source: Rosenberg ES, Dufort EM, Udo T, et al. Association of treatment with hydroxychloroquine or azithromycin with in-hospital mortality in patients with COVID-19 in New York state [published online May 11, 2020]. ) 

May 25, 2020, Research Update

Hydroxychloroquine Ineffective in Oxygen-Requiring Patients in France. These authors analyzed care data from 181 oxygen-requiring patients with COVID-19 hospitalized at one of four tertiary care centers in France but who were not requiring intensive care. They compared outcomes of patients who received standard care with those who also received hydroxychloroquine (600 mg per day). They also tried to account for potential confounding factors. The survival rate without transfer to the intensive care unit at day 21 was 76% in the treatment group and 75% in the control group. Additionally, there was no difference in the overall survival at day 21 (89% and 91%, respectively). They also found no differences in the rates of developing acute respiratory distress syndrome or in the rate of successful weaning from supplemental oxygen. Eight patients (10%) taking hydroxychloroquine developed electrocardiographic abnormalities requiring discontinuation of treatment.

Written by Henry C. Barry, MD, MS, on May 20, 2020. (Source: Mahévas M, Tran V-T, Roumier M, et al. Clinical efficacy of hydroxychloroquine in patients with covid-19 pneumonia who require oxygen: observational comparative study using routine care data. BMJ. 2020;369:m1844.)

May 24, 2020, Research Update

BCG Vaccine Not Associated with COVID-19 Infection in Israel. Some ecologic studies have pointed to lower rates of COVID-19 and less morbidity and mortality in countries where higher rates of BCG immunization occurred. These investigators took advantage of a natural experiment in their country. In Israel, BCG was no longer routinely administered after 1982. A study compared COVID-19 infection rates in 3,064 patients born three years before this cutoff (39 to 41 years of age) and 2,869 patients born three years after this cutoff (35 to 37 years of age). They found no difference in infection rates (11.7% vaccinated vs. 10.4% unvaccinated). No difference in rates of COVID-19 positive per 100,000 population occurred, either.

Written by Mark Ebell, MD, MS, on May 20, 2020. (Source: Hamiel U, Kozer E, Youngster I. SARS-CoV-2 rates in BCG-vaccinated and unvaccinated young adults [published online May 13, 2020]. )

May 23, 2020, Research Update

Hydroxychloroquine Ineffective in Patients with Mild or Moderate COVID-19 in China. Researchers conducted an open-label randomized trial in one of 16 designated COVID-19 treatment centers in China. They randomized 75 patients to standard care and 75 patients to standard care plus hydroxychloroquine (HCQ; 1,200 mg daily for three days followed by 800 mg daily for two weeks in those with mild or moderate disease or three weeks for those with severe disease). All but two of the patients had mild or moderate COVID. Overall, the rate of resolution of viral shedding was similar (81% in controls and 85% in those receiving HCQ). Only 9% of the control patients experienced adverse events compared with 30% of the HCQ-treated patients. This was a research design with biases that generally favor the intervention group, strengthening the findings of no differences in outcomes.

Written by Henry C. Barry, MD, MS, on May 20, 2020. (Source: Tang W, Cao Z, Han M, et al. Hydroxychloroquine in patients with mainly mild to moderate coronavirus disease 2019: open label, randomised controlled trial. BMJ. 2020;369:m1849.) 

May 22, 2020, Research Update

Outdoor Transmission of COVID-19 Rare in China. In a case series, investigators studied the likely site of transmission for 7,324 of the 10,980 confirmed COVID-19 cases that occurred outside of Hubei province in China by February 11, 2020. They collected descriptions of each confirmed case from the local Municipal Health Commission website of 320 cities in mainland China, excluding Hubei province. They identified clusters of three or more cases linked temporally to the same infection venue. Among the identified clusters, 54% involved three cases, 26% involved four cases, and only 2% involved 10 or more cases. Exposure at home was the most common (254 of 318 clusters; 80%), followed by transportation (108; 34%). Many clusters had more than one exposure venue. Only one outbreak occurred involving two cases for which only outdoor exposure was identified. They conclude that COVID-19 transmission during outdoor exposure is rare in China. It is not known whether this is true in other countries where use of masks outdoors is less common. 

Written by John Hickner, MD, MS, on May 15, 2020. (Source: Qian H, Miao T, Liu L, et al. Indoor transmission of SARS-CoV-2 [published online April 7, 2020]. )

May 21, 2020, Research Update

Anticoagulation May Benefit COVID-19 Patients Who Have Significant Coagulopathy. Severe COVID-19 infection has a high mortality rate that may be improved with anticoagulation. In an observational study, 449 consecutive patients with severe COVID-19 out of 1,786 total COVID patients admitted to Tongji Hospital in Wuhan, China, were enrolled in the study, and 99 of them received heparin for seven days or longer. Overall no improvement in mortality occurred with heparin, except in patients with a sepsis-induced coagulopathy (SIC) score ≥ 4 (40.0% vs. 64.2%, P = .029) or D-dimer greater than sixfold upper limit of normal (32.8% vs. 52.4%, P = .017). These results suggest that anticoagulation of severely ill COVID-19 patients who have coagulopathy may benefit from heparin anticoagulation. Given that there were multiple post hoc comparisons, we should continue to be watching for better studies. 

Written by John Hickner, MD, MS, on May 15, 2020. (Source: Tang N, Bai H, Chen X, et al. Anticoagulant treatment is associated with decreased mortality in severe coronavirus disease 2019 patients with coagulopathy. J Thromb Haemost. 2020;18(5):1094-1099.)

May 20, 2020, Research Update

Outcomes with Anticoagulation in Hospitalized COVID-19 Patients in New York City. This observational study assessed the relationship between anticoagulation and the outcomes of mortality and length of stay among 2,773 hospitalized patients with COVID in the Mount Sinai Health System in New York City after adjusting for important confounders. A total of 786 patients received systemic anticoagulation (SAC), including intravenous, oral, and subcutaneous forms. There was no overall survival benefit of SAC (22.5% vs. 22.8%), although length of survival was greater with SAC (median of 21 days vs. 14 days without SAC). Longer duration of SAC treatment was associated with lower mortality (adjusted HR of 0.86 per day, 95% CI, 0.82 to 0.89, p < 0.001). Patients treated with SAC who required mechanical ventilation (N = 395) had a much lower mortality rate (29.1% vs. 62.7%). Significant bleeding was slightly more common with anticoagulation (3% vs. 1.9%). These results suggest that ventilator-requiring patients with COVID-19 may benefit substantially from anticoagulation.

Written by John Hickner, MD, MS, on May 15, 2020. (Source: Paranjpe I, Fuster V, Lala A, et. al. Association of treatment dose anticoagulation with in-hospital survival among hospitalized patients with COVID-19 [published online May 5, 2020]. J Am Coll Cardiol. 2020. )

May 19, 2020, Research Update

Respiratory Droplets Can Linger for Several Minutes after Normal Speech in Closed Spaces. We reported in a Daily Research Brief on April 22, 2020 (), that respiratory droplets can travel up to 7.5 cm during normal speech and that droplets travel farther with louder speech. This same research team, using the same laser-based technology, now reports that louder speech also generates thousands of oral fluid droplets per second and that in a closed, stagnant air environment the droplets linger in the air for 8 to 14 minutes. This in vitro experiment demonstrates the significant potential for COVID-19 transmission in closed spaces. Imagine how much spray is generated in a loud bar.

Written by Henry C. Barry, MD, MS, on May 14, 2020. (Source: Stadnytskyi V, Bax CE, Bax A, et al. The airborne lifetime of small speech droplets and their potential importance in SARS-CoV-2 transmission [published online May 13, 2020]. Proceedings of the National Academies of Sciences; 2020. )

May 18, 2020, Research Update

Decreased Hospitalization Rates for ACS during COVID-19 Outbreak in Italy, These researchers compared the rate of hospitalization for acute coronary syndrome (ACS) during three time periods: the first six weeks after the first COVID-19 case was reported in northern Italy (February 20 to March 31, 2020); the six weeks preceding this (January 1 to February 19, 2020); and the same time period the previous year (February 20 to March 31, 2019). Each hospital was a hub for performing primary coronary interventions. During the index period, an average of 13.3 patients were admitted daily for ACS compared with 18.0 during the preceding six weeks and 18.9 for the same time period the previous year. Combining these findings with recent data suggesting increased deaths from ACS during the COVID-19 pandemic suggests that many patients are not seeking care for important health issues.

Written on May 14, 2020, by Henry C. Barry, MD, MS. (Source: De Filippo O, D'Ascenzo F, Angelini F, et al. Reduced rate of hospital admissions for ACS during Covid-19 outbreak in northern Italy [published online April 28, 2020]. )

May 17, 2020, Research Update

Reduced Symptoms and Viral Shedding in Low-Risk Patients with COVID-19 on Four-Drug Regimen in China. A previous study of 199 patients found no benefit to lopinavir 400 mg and ritonavir 100 mg for COVID-19 (N Engl J Med. 2020; 382(19):1787-1799). In this study set in Hong Kong, 127 patients hospitalized with COVID-19 were randomized in a 2-to-1 ratio to a combination of ribavirin 400 mg twice daily, interferon beta-1b 8 million IU every other day, lopinavir 400 mg and ritonavir 100 mg twice daily, or to lopinavir 400 mg and ritonavir 100 mg alone twice daily as a control group. They initially recruited patients within seven days of symptoms, and patients recruited later in that window received fewer doses of interferon. The average age was 51, and the mean national early warning score 2 (NEWS2 - ) was 2 at baseline for both groups, indicating low clinical risk. The mean time to a score of 0 was four days for the combination therapy group and eight days for the control group (p < 0.001). The time to a negative nasopharyngeal swab was seven days for the combination therapy group and 12 days for the control group (p = 0.001). There were no deaths in either group.

Written by Mark H. Ebell MD, MS, on May 14, 2020. (Source: Hung IF, Lung KC, Tso EY, et al. Triple combination of interferon beta-1b, lopinavir–ritonavir, and ribavirin in the treatment of patients admitted to hospital with COVID-19: an open-label, randomised, phase 2 trial [published online May 8, 2020]. Lancet. 2020. )

May 16, 2020, Research Update

Hydroxychloroquine Lacks Benefit for COVID-19 in a Large New York City Hospital. To date, several small, relatively poor quality RCTs have failed to find benefits of hydroxychloroquine (HCQ) for the treatment of COVID-19. This study from Columbia University and New York Presbyterian Hospital studied 1,376 consecutive patients who were admitted with COVID-19  and not on mechanical ventilation after 24 hours. They compared those treated with HCQ (58.9%) with those who were not for the outcome of mechanical ventilation or death. There were significant differences between those groups, with patients receiving HCQ somewhat older and also sicker based on the PaO2/FIO2 ratio. They used multivariate Cox regression and propensity score matching to match patients who were similar on all measured parameters other than receipt of HCQ. They did this analysis several ways, and each time the result was the same: no benefit of HCQ, with nonsignificant hazard ratios ranging from 0.97 to 1.04 depending on the method used.

Written by Mark H. Ebell MD, MS, on May 12, 2020. (Source: Geleris J, Sun Y, Platt J, et al. Observational study of hydroxychloroquine in hospitalized patients with Covid-19 [published online May 7, 2020; updated May 14, 2020]. N Engl J Med. 2020. )

May 15, 2020, Research Update

Descriptions of Children with COVID Infections in Italy and China. Several studies have described the characteristics of children infected with COVID-19. A research letter in the NEJM summarized an Italian case series of 100 childhood cases and compared their results to four other studies. Of the Italian cases, only 54% presented with fever, 44% with cough, and 11% with shortness of breath. Mortality ranged from 0% to 0.6% in the five studies summarized. The most recent case series comes from China. They identified 2,135 cases, of which 728 (34.1%) were laboratory confirmation cases and 1,407 (65.9%) were suspected cases. Cases were initially identified based on clinical signs and symptoms and exposure history. The median age was seven years of age. More cases were boys (56.6%) than girls (43.4%). Of the confirmed cases, 12.9% were asymptomatic, 43.1% mild, 41% moderate, 2.5% severe, and 0.4% critical. Severe and critical cases were more prevalent in those under one year of age. On average, children were less severely ill compared to adult cases. Better data is needed to fully understand COVID-19 infections in children.

Written by John Hickner, MD, MS, on May 8, 2020. (Sources: Parri N, Lenge M, Buonsenso D. Children with Covid-19 in pediatric emergency departments in Italy [published online May 1, 2020]. N Engl J Med. 2020. . Dong Y, Mo X, Hu Y, et al. Epidemiology of COVID-19 among children in China. Pediatrics. 2020;145(5):e20200702.)

May 14, 2020, Research Update

SARS-CoV-2 Infection Confers Humoral Immunity in at Least 99% of Patients. Researchers at the Icahn School of Medicine (preprint, nonpeer-reviewed) in New York identified 624 patients with previous PCR-confirmed SARS-CoV-2 infection who were volunteering to provide convalescent sera. In week 1, when most patients had just recovered, they tested both PCR and serum for IgG. As time went on, they dropped the PCR and tested only IgG. The IgG test was 92% sensitive and more than 99% specific. Strongly positive patients (IgG titer > 1:320) were referred for donation of sera, whereas those who were weakly positive or who were initially negative for IgG returned at least a week later for retesting. At the initial testing, 82% were strongly positive, 7% weakly positive, and 11% were negative. Of the 113 who were initially weakly positive or negative, 64 returned a median of 13 days later, and at that time 57 were strongly positive, four were weakly positive, and three were negative at follow-up (the PCR for these three was self-reported, a possible source of error). The median duration from onset of symptoms to development of IgG antibodies is 24 days, with a range of 7 to 50 days. This suggests that the vast majority of patients develop a vigorous antibody response, similar to influenza, which should provide some degree of protection against re-infection. Testing for antibodies should be delayed until at least three to four weeks after the onset of symptoms and repeated in one to two weeks if initially negative.

Written by Mark H. Ebell MD, MS, on May 7, 2020. Wajnberg A, Mansour M, Leven E, et al. Humoral immune response and prolonged PCR positivity in a cohort of 1343 SARS-CoV 2 patients in the New York City region [published online May 5, 2020]. )

May 13, 2020, Research Update

No Antihypertensive Classes Associated with Contracting COVID-19 or Its Severity in New York City. These researchers, using their electronic medical record system, identified 12,594 patients tested for COVID-19 from March 1 to April 15, 2020. Of these, 5,894 (46.8%) were positive for COVID, 1,002 of whom (17.0%) had severe illness (ICU admission, mechanical ventilation, or death). The high rate of positive testing is reflective of highly selective test ordering, which makes comparisons with other settings challenging. A total of 4,357 (34.6%) of the tested patients had a history of hypertension, and 2,573 (59.1%) tested positive for COVID. Not surprisingly, antihypertensive medication (angiotensin-converting enzyme inhibitors, angiotensin-receptor blockers, beta blockers, calcium-channel blockers, or thiazide diuretics) use was higher in the COVID-infected patients than in those who tested negative. However, using a propensity analysis to account for known confounders, the authors found no association between any antihypertensive class and contracting COVID-19 or in the risk of having severe disease among those who tested positive. 

Written by Henry C. Barry, MD, MS, on May 7, 2020. (Source: Reynolds HR, Adhikari S, Pulgarin C, et al. Renin-angiotensin-aldosterone system inhibitors and risk of Covid-19 [published online May 1, 2020; updated online May 6, 2020]. N Engl J Med. 2020. )

May 12, 2020, Research Update

Saliva May Be Better Source for SARS-CoV-2 Samples Than Nasopharyngeal Swab. We usually use nasopharyngeal (NP) swabs to sample for respiratory viruses, but recent studies have reported similar sensitivity for detection of SARS-CoV-2 in saliva (Clin Infect Dis. February 2020; doi:10.1093/cid/ciaa149). The current study from Yale University (preprint server, nonpeer-reviewed) took paired saliva samples with NP swabs from 29 inpatients every three days for a total of 121 samples. They found that viral titers were higher in saliva, with fewer false negatives. A total of eight patients were negative using NP swab but positive using saliva, compared with only three positive using NP but negative using saliva. In five patients, NP swabs went from negative to positive over time, but this did not occur in the saliva samples of any patients. They also tested 98 asymptomatic healthcare workers and detected SARS-CoV-2 in saliva of two who were repeatedly negative by NP swab. Saliva samples are attractive because they can be self-collected and do not require personal protective equipment for the person obtaining the sample. 

Written by Mark H. Ebell MD, MS, on May 6, 2020. (Source: Wyllie AL, Fournier J, Casanovas-Massana A, et al. Saliva is more sensitive for SARS-CoV-2 detection in COVID-19 patients than nasopharyngeal swabs [published online April 22, 2020].  ) 

May 11, 2020, Research Update

ACEI/ARB Not Associated with Increased Risk of Contracting COVID-19 at Cleveland Clinic. Researchers at the Cleveland Clinic reviewed the records of patients with suspected COVID-19 seen in their healthcare system in Ohio and Florida between March 8, 2020, and April 12, 2020. They tested based on clinical suspicion of COVID based on symptoms, travel, and exposure history and being in a high-risk group (based on age and chronic diseases). They also tested healthcare workers purely based on symptoms. Because patients prescribed ACEIs or ARBs are more likely to also have serious comorbid conditions, the researchers performed a propensity-weighted analysis to attempt to adjust for known confounders. Of 18,472 patients tested, 1,735 (9.4%) had a positive PCR for SARS-CoV-2. Whereas patients taking an ACEI had a slightly lower risk and patients taking an ARB had a slightly higher risk of having been diagnosed with COVID-19, the odds were tiny and not statistically significant. However, patients on ACEIs were more likely to be hospitalized (OR 1.8, 95% CI, 1.2 to 2.8) or go to the ICU (OR 1.8, 95% CI, 1.07 to 2.9). Whereas patients taking ARBs were also at higher risk of hospitalization (OR 1.6, 95% CI, 1.04 to 2.5), there was no association with going to the ICU. The use of ACEIs or ARBs was not associated with the use of mechanical ventilation.

Written by Henry C. Barry, MD, MS, on May 6, 2020. (Source: Mehta N, Kalra A, Nowacki AS, et al. Association of use of angiotensin-converting enzyme inhibitors and angiotensin II receptor blockers with testing positive for coronavirus disease 2019 (COVID-19) [published online May 5, 2020]. JAMA Cardiol. )

May 10, 2020, Research Update

About 1% of Close Contacts Get Infected with COVID-19 in Taiwan. Close contacts of 100 confirmed COVID-19 patients in Taiwan were identified and followed to ascertain what proportion of them became infected with the coronavirus. Among 2,761 contacts, there were 22 secondary infections, an attack rate of 0.7% (95% CI, 0.4% to 1.0%) The attack rate was highest among the 1,818 individuals exposed within five days of symptom onset (1.0%) compared with those with later exposure (zero cases from 852 contacts). The rate was also highest among household contacts (4.6%) and nonhousehold family contacts (5.3%). Contacts with only presymptomatic exposure had a 0.7% incidence of infection. This study suggests that isolation of infected individuals is insufficient to halt transmission of COVID-19. 

Written by John Hickner, MD, MS, on May 4, 2020. (Source: Cheng HY, Jian SW, Liu DP, Ng TC, Huang WT, Lin HH. Contact tracing assessment of COVID-19 transmission dynamics in Taiwan and risk at different exposure periods before and after symptom onset. JAMA Intern Med. 2020 [published online May 1, 2020]. )

May 9, 2020, Research Update

High Incidence of Thrombotic Events in Dutch ICU COVID-19 Patients. Investigators studied the incidence of thrombotic complications of all 184 ICU patients with COVID-19 pneumonia hospitalized at three Dutch academic medical centers. 31% of the patients had thrombotic complications, including 27% with venous thromboembolism and 3.7% with arterial thrombotic events. Pulmonary embolism was the most frequent complication (25% of all patients). The investigators recommend that all patients with COVID-19 admitted to the ICU receive thrombosis prophylaxis. This is consistent with the observation that high levels of d-dimer are associated with a worse prognosis.

Written by John Hickner, MD, MS, May 1, 2020. (Source: Klok F, Kruip M, van der Meer NJM, et. al. Incidence of thrombotic complications in critically ill ICU patients with COVID-19. Thromb Res. 2020 [published online ahead of print April 10, 2020]. )

May 8, 2020, Research Update

Viral Shedding in Feces in Patients Hospitalized in China with COVID-19. Researchers in Zhejiang province evaluated 96 consecutive patients hospitalized with COVID-19, 22 of whom had mild disease and 74 had severe disease. To determine the amount of SARS-CoV-2 RNA, the patients had daily polymerase chain reaction assays of sputum, saliva, blood, urine, and stool. The researchers evaluated 3,497 samples and found SARS-CoV-2 RNA in 59% of the patients but only in one of the urine samples. The median duration of fecal shedding of SARS-CoV-2 RNA was 22 days compared with 18 days in the respiratory samples and 16 days in the serum samples. Patients with severe disease shed virus for one week longer than those with mild disease (21 days vs. 14 days, respectively). At this time, it is unclear how much virus is needed to infect another individual, but it appears to be low.

Written by Henry C. Barry, MD, MS, on April 30, 2020. (Source: Zheng S, Fan J, Yu F, et al. Viral load dynamics and disease severity in patients infected with SARS-CoV-2 in Zhejiang province, China, January-March 2020: retrospective cohort study. BMJ. 2020;369:m1443.)

May 7, 2020, Research Update

Case Series of Large-Vessel Stroke in Young Persons with COVID-19 in New York City. A case series reported findings regarding five persons with SARS-CoV-2 infection who presented during a two-week period with a large vessel stroke (seven times the usual rate of stroke in persons under 50 years). Their ages ranged from 33 to 49 years; two were healthy, one had hypertension and hyperlipidemia, one had undiagnosed diabetes, and one had diabetes and a history of mild stroke. The vessels involved include the middle cerebral artery in three, posterior cerebral artery in one, and internal carotid artery in one. All were treated with antiplatelet therapy, one with tissue plasminogen activator, and four with clot retrieval. All but one experienced at least some improvement in their symptoms, and three had been discharged at the time the article was written. D-dimer levels were significantly elevated in three patients; two patients delayed seeking care because of fears about COVID-19.

Written by Mark H. Ebell MD, MS, on April 29, 2020. (Source: Oxley TJ, Mocco J, Majidi S, et al.  Large-vessel stroke as a presenting feature of Covid-19 in the young. N Engl J Med. 2020 [published online April 28, 2020]. )

May 6, 2020, Research Update

IDSA Guidelines on Treatment of COVID-19. The Infectious Diseases Society of America (IDSA) used modern guideline development methods that included a multidisciplinary team of clinicians, public health workers, and methodologists to develop critical PICO-format questions and to perform systematic searches of the research. They made seven recommendations for managing patients hospitalized with confirmed COVID-19, all contingent upon using these specific agents in the context of a clinical trial: administer hydroxychloroquine or chloroquine; give hydroxychloroquine/chloroquine plus azithromycin only (emphasis added) in the context of a clinical trial; give lopinavir/ritonavir; do not give corticosteroids to patients with COVID-19 pneumonia (conditional recommendation based on low certainty evidence); give corticosteroids to patients with respiratory distress syndrome; give tocilizumab only (emphasis added) in the context of a clinical trial; administer convalescent serum. The most important recommendation of the IDSA was the need to recruit patients into randomized trials to overcome the limitations of the existing research.

Written by Henry C. Barry, MD, MS, on April 23, 2020. (Source: Bhimraj A, Morgan R, Shumaker AH, et al. Infectious Diseases Society of America guidelines on the treatment and management of patients with COVID-19 [published online April 11, 2020; updated April 13, 2020]. )

May 5, 2020, Research Update

Mortality 21% for Hospitalized Patients with COVID-19 in New York City. This is a case series of the characteristics of 5,700 consecutive individuals infected with COVID-19 who were hospitalized at one of 12 hospitals in the New York City area between March 1, 2020, and April 4, 2020. The median age was 63 years of age, and 60% were men. Of the individuals, 57% of patients had hypertension, 34% had diabetes, and 42% were obese. Only 31% of patients were febrile, and 28% were receiving supplemental oxygen at triage. Outcomes are presented for the 2,634 patients who had been discharged or died by the end of the study period. A total of 553 (21%) died, 373 (14%) were treated in the ICU, and 320 (12%) received mechanical ventilation. Of those who were on a ventilator, 3% had been discharged alive, 25% had died, and 72% were still hospitalized at the end of the study period. In this case series, COVID-19 infection had a high mortality rate in hospitalized patients, especially those requiring mechanical ventilation.

Written by John Hickner, MD, MS, on April 23, 2020. (Source: Richardson S, Hirsch JS, Narasimhan M, et. al. Presenting characteristics, comorbidities, and outcomes among 5700 patients hospitalized with COVID-19 in the New York City area. JAMA. [published online April 22, 2020]. )  

May 4, 2020, Research Update

Asymptomatic Carriage of COVID-19 in Pregnancy is Common in New York Sample. Two obstetric services in New York City instituted universal COVID-19 PCR testing in March 2020. This report summarizes screening of the first 215 pregnant women presenting in labor. Four women had symptoms consistent with COVID-19, and all tested positive. Of the remaining 211 asymptomatic patients, 29 tested positive. Twenty-nine of the 33 patients (88%) testing positive, therefore, were asymptomatic at the time of testing on admission to the hospital. The high rate of asymptomatic carriage of COVID-19 may have been due to the high prevalence of COVID-19 infections in New York City at the time of testing, though other surveillance reports also have found a high proportion of asymptomatic COVID-19 infections. This and other studies suggest that COVID-19 spread will be difficult to contain with medical isolation, quarantine, and contact tracing only.

Written by John Hickner, MD, MS, on April 23, 2020. (Source: Sutton D, Fuchs K, D'Alton M, et al. Universal screening for SARS-CoV-2 in women admitted for delivery. N Engl J Med. 2020 [published online April 13, 2020]. )

May 3, 2020, Research Update

Characteristics of Adult Inpatients with COVID-19 in Wuhan. Researchers from Wuhan conducted a retrospective cohort study of 191 adults hospitalized with confirmed COVID-19 who had been discharged or had died by January 31, 2020. Most of the patients (72%) were discharged, and 28% died in the hospital. About half had a comorbidity: 30% had hypertension, 19% had diabetes, and 8% had coronary heart disease. The likelihood of in-hospital death was associated with older age, higher Sequential Organ Failure Assessment (SOFA) score, and d-dimer greater than 1 μg per mL on admission. The duration of viral shedding was a median of 20 days in survivors, and SARS-CoV-2 was detectable until death in nonsurvivors. The longest observed duration of viral shedding in survivors was 37 days.

Written by Henry C. Barry, MD, MS, on April 23, 2020. (Source: Zhou F, Yu T, Du R, et al. Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study [published correction appears in Lancet. 2020;395(10229):1038]. Lancet. 2020;395(10229):1054-1062.)

May 2, 2020, Research Update

Anosmia Common in SARS-CoV-2 Infection. Alteration in taste and smell has been reported informally in patients with COVID-19. The virus attaches selectively to angiotensin-converting enzyme 2 receptors, which are numerous in the nasal epithelium. This Italian study identified 374 consecutive adults who tested positive for SARS-CoV-2 during a four-day period at the regional hospital in Treviso, of whom 283 had contact information and 202 completed a telephone survey. Fully two-thirds (64.4%) reported any alteration in taste or smell, which was described as moderate, severe, or as bad as it can be in 102 out of 130 (78%). In 101 of 130 patients, the onset was concomitant with or after the onset of other symptoms. Although there is no comparison with patients who have noncoronavirus respiratory infections, and anosmia is a feature of sinus infections, this seems much higher than would ordinarily be the case with other viral respiratory infections.

Written by Mark H. Ebell MD, MS, on April 23, 2020. (Source: Spinato G, Fabbris C, Polesel J, et al. Alterations in smell or taste in mildly symptomatic outpatients with SARS-CoV-2 infection. JAMA. 2020 [published online April 22, 2020].  )

May 1, 2020, Research Update

Chinese Study Reports Lack of Efficacy of Remdesivir in Patients Hospitalized with COVID-19 Pneumonia. In 10 hospitals in Hubei, China, researchers randomly assigned 237 patients hospitalized with COVID-19 pneumonia and hypoxia who were within 12 days of symptom onset to receive intravenous remdesivir (200 mg on day 1 followed by 100 mg on days 2 to 10 in single daily infusions) or the same volume of placebo infusions for 10 days. There were 158 patients in the remdesivir group, and 79 were in the placebo group. The treating teams were allowed to administer other therapies. In this study, there was no statistically significant improvement in time to recovery. About two-thirds of patients in each group reported adverse events, and more remdesivir-treated patients stopped treatment early due to adverse events than did placebo-treated patients (12% vs. 5%). They found no difference in mortality (12% vs. 13%). This contrasts with the NIH-sponsored ACTT trial that was terminated early because of clear improvement in the remdesivir-treated patients.

Written by Henry C. Barry, MD, MS, on April 29, 2020. (Source: Wang Y, Zhang D, Du G, et al. Remdesivir in adults with severe COVID-19: a randomised, double-blind, placebo-controlled, multicentre trial. Lancet. 2020 [published online April 29, 2020]. )

April 30, 2020, Research Update

Initial Report of Efficacy of Remdesivir for COVID-19 (ACTT). The National Institute of Allergy and Infectious Diseases (NIAID) released initial results (not peer reviewed, full data not yet available) for the Adaptive COVID-19 Treatment Trial. The study recruited 1063 hospitalized patients with SARS-CoV-2 infection and evidence of lung involvement. They were randomized to remdesivir (200 mg IV on day 1 and then 100 mg IV daily for up to 10 days or hospital discharge) or matching placebo injection. The primary outcome was time to recovery, which was significantly faster in the remdesivir group (11 vs. 15 days, p < 0.001). There was also a trend toward lower mortality in the remdesivir group (8.0% vs. 11.6%, p = 0.059, NNT = 28). Harms or other outcomes such as need for mechanical ventilation are not reported.

Written by Mark H. Ebell MD, MS, on April 29, 2020. (Source:  and )

April 29, 2020, Research Update

Transmission of COVID-19 in a Restaurant. The investigators report a cluster of 10 COVID-19 cases that were most likely spread from one infected person at a restaurant in Guangzhou, China. The index case and three family members traveled from Wuhan on January 23, 2020, and had lunch at a small restaurant. The index case reportedly did not feel ill at lunch but developed cough and fever later that day. Within two weeks, nine other people who were in the restaurant became ill; four family members of the index case, three members of a family sitting at a neighboring table, and two family members sitting at the table on the other side. The distance between tables was three feet. The exposure time to the other two families was about one hour in the restaurant. The investigators could find no other exposures to COVID-19 for the other two infected families. The investigators hypothesize that the infection may have been spread by the direction of airflow from the air-conditioning unit. If so, this report has significant implications as the United States opens restaurants and other business venues requiring close physical proximity and ventilation. Maintaining social distancing and possibly the use of well-constructed face masks seem essential for reducing spread of COVID in close public spaces.

Written by John Hickner, MD, MS, on April 23, 2020. (Source: Lu J, Gu J, Li K, et al. COVID-19 outbreak associated with air conditioning in restaurant, Guangzhou, China, 2020. Emerg Infect Dis. 2020;26[7] [published online April 2, 2020]. )

April 28, 2020, Research Update

Hydroxychloroquine Not Effective for COVID-19 in U.S. Veterans. This is the first published evaluation of hydroxychloroquine (HCQ) with or without azithromycin (AZ) in U.S. patients. The authors identified patients hospitalized with COVID-19 in all U.S. Veteran’s Administration hospitals. They identified 97 patients who had been given HCQ, 113 who had been given HCQ plus AZ, and 158 who had not received HCQ. They used propensity score matching to identify patients who looked similar other than the fact that one had received HCQ (or HCQ + AZ) and one had not received HCQ. In the unadjusted analysis, the risk of death was 27.8% in the HCQ group, 22.1% in the HCQ + AZ group, and 11.4% in the no HCQ group. After propensity score matching, the adjusted hazard ratio for death was 2.61 (95% CI, 1.10 to 6.17) for HCQ compared with no HCQ. However, the risk of death was not significantly increased for patients given the combination HCQ and AZ compared with those receiving no HCQ (adjusted hazard ratio 1.14, 95% CI, 0.56 to 2.32). There was no significant difference between groups with regard to the need for mechanical ventilation in the adjusted analysis.

Written by Mark H. Ebell, MD, MS, on April 22, 2020. (Source: Magagnoli J, Narendran S, Pereira F, et al. Outcomes of hydroxychloroquine usage in United States veterans hospitalized with Covid-19 [updated online April 23, 2020]. )

April 27, 2020, Research Update

The initial report on a preprint server of a community serosurvey is from Santa Clara County, California (population 1.9 million). At the time of the study, 956 cases had been confirmed, and the county had some of the earliest reported cases (January 31, 2020). On April 3 and 4, they obtained serum from 3,330 adults and children. After adjusting for age distribution in the population and trying to account for false positives and negatives, they estimate a prevalence of 2.5% to 4.2% of antibodies to SARS-CoV2. That provides a range of 48,000 to 81,000 infections, far higher than the number reported on April 1 of 956 cases.

Written by Mark H. Ebell MD, MS, on April 18, 2020. (Source: Bendavid E, Mulaney B, Sood N, et al. COVID-19 antibody seroprevalence in Santa Clara County, CA [published online April 17, 2020]. )

April 26, 2020, Research Update

A Chinese study published on a preprint server (not peer reviewed) randomized 150 patients to hydroxychloroquine (HCQ) plus usual care vs. usual care alone. All of them but two patients had mild or moderate disease, and the mean time from symptom onset to randomization was 16 days. They found no difference in the rate of conversion to nondetectable SARS-CoV-2 between groups within 28 days (85% for HCQ plus usual care vs. 81% for usual care alone). There was also no difference in symptom alleviation at 28 days or the number of days until symptom alleviation. They did many comparisons, including post hoc analyses, that report a benefit during the second week of therapy, but this is speculative at best. Mortality is not reported, and adverse events were more common in the HCQ group.

Written by Mark H. Ebell MD, MS, on April 18, 2020. (Source: Tang W, Cao Z, Han M, et al. Hydroxychloroquine in patients with COVID-19: an open-label, randomized, controlled trial. MedRxiv. 2020 [published online April 14, 2020]. )

April 25, 2020, Research Update

COVID-19 is highly contagious. Spread during the presymptomatic and early symptom phase, such as with the influenza virus, could make it difficult to institute effective quarantine procedures. Researchers from China studied the temporal pattern of viral shedding in 94 COVID-19 positive patients and modeled viral shedding in another 77 infector-infectee transmission pairs. The 94 COVID-19 infected individuals had a total of 414 throat cultures for COVID-19 from symptom onset to 32 days after onset. The greatest viral load shedding was at the time of symptom onset, and they surmised that infectiousness peaks at or before symptom onset. In the separate study of 77 infector-infectee pairs, based on epidemiological modeling and a mean incubation period of 5.2 days, they inferred that infectiousness starts 2.3 days before symptom onset with a peak infectiousness at 0.7 days prior to symptom onset. They estimated that 44% of secondary cases they studied were infected during the presymptomatic phase of the person who infected them. They conclude that there is substantial presymptomatic transmission. 

Written by John Hickner, MD, MS, on April 19, 2020. (Source: He X, Lau EHY, Wu P, et al. Temporal dynamics in viral shedding and transmissibility of COVID-19. Nat Med. 2020 [published online April 15, 2020]. )

April 24, 2020, Research Update

The National Academies of Sciences reviewed the evidence for effectiveness of homemade fabric masks in prevention of the spread of influenza and SAR-CoV-2 (COVID-19). These viruses can be spread by visible and invisible droplets as small as 5 microns and even smaller bioaerosols. Which size is most dangerous is unknown. Effectiveness depends on how the mask is made and how well it is made. Leakage around the mask is a problem, so fit must be as tight as possible. One must consider filtration efficiency and how much the mask impedes breathing (wearability). They found seven studies that evaluated the ability of the mask to protect the wearer or to prevent spread of infection from a wearer. Performance ranged from very poor to reducing exposure to the wearer by about 60% depending on the material used. Jayaraman found a filtration of only 0.7% of 0.3 micron-sized particles with a four-layer woven handkerchief fabric, 35.3% for five-layered woven brushed fabric, and 50% for four layers of polyester-knitted cut-pile fabric. A recent study of COVID-19 infected patients found that surgical and cotton masks were not effective at blocking the virus from disseminating during coughing. On the other hand, two studies of cotton mask wearers suggest moderate protection against inhalation of infectious-sized particles. In the only randomized trial, however (performed in healthcare workers), cotton two-layer masks were much worse than medical masks (three layers of nonwoven material) in protecting from respiratory infection (RR = 13).

Written by John Hickner, MD, MS, on April 16, 2020. (Source: National Academies of Sciences, Engineering, and Medicine, 2020. Rapid expert consultation on the effectiveness of fabric masks for the COVID-19 pandemic; April 8, 2020. Washington, DC: The National Academies Press. )

April 23, 2020, Research Update

On April 7, 2020, using existing research, the National Academies Press published a rapid report on whether heat or humidity have any impact on SARS-CoV-2 survival. Basically, they found conflicting data, poor quality studies, multiple confounders, and insufficient time since the start of the COVID-19 pandemic to make much judgment. Because we are seeing COVID-19 on both sides of the equator, in tropical and temperate climates, we really should not assume any seasonality to COVID-19. 

Written by Henry C. Barry, MD, MS, April 16, 2020. (Source: National Academies of Sciences, Engineering, and Medicine 2020. Rapid expert consultation on SARS-CoV-2 survival in relation to temperature and humidity and potential for seasonality for the COVID-19 pandemic; April 7, 2020. Washington, DC: The National Academies Press. )

April 22, 2020, Research Update

Aerosolized droplets generated during speech have been implicated as a possible source of transmitting COVID-19. In an in vitro experiment published online April 15, 2020, researchers used fancy laser gadgets to determine how far respiratory droplets travel during speech, the sizes of droplets, and the effect of speaking through a slightly damp washcloth (not sure why they did not try a homemade mask as a better test in the real world). They repeated the phrase "stay healthy" at different volumes. The droplets traveled between 5 to 7.5 cm, and they ranged in size from 20 to 500 micrometers; louder speech resulted in droplets that traveled farther. Using the damp washcloth resulted in dramatic reductions in number and size of droplets as well as distance traveled. The authors are clear to report that they did not assess viral transmission.

Written by Henry C. Barry, MD, MS, April 16, 2020. (Source: Anfinrud P, Stadnytskyi V, Bax CE, et al. Visualizing speech-generated oral fluid droplets with laser light scattering. N Engl J Med. 2020 [published online April 15, 2020]. )

April 21, 2020, Research Update

The widespread use of quarantine measures has created a need for telehealth in a setting where few are trained or have planned infrastructures. Although these are not research, the BMJ and the CDC have published two useful tools. On March 25, 2020, Greenhalgh, Koh, and Car published a nice overview of how telehealth sessions ought to occur as well as the necessary infrastructures that need to be in place. The article has a one-page summary (). The CDC has created an explicit script with branch points and specific care messages (www.cdc.gov/coronavirus/2019-ncov/hcp/phone-guide/index.html). For example, patients reporting symptoms of hypoxia or hypotension are directed to the emergency department, and the script also cautions the patient to inform the medical providers of possible COVID-19 exposure.

Written by Henry C. Barry, MD, MS, April 16, 2020.

April 20, 2020, Research Update

On April 8, 2020, the National Academies Press published a rapid report based on existing research on the duration of viral shedding and antibody response in patients infected with SARS-CoV-2. By consensus, they suggest that two sequential negative PCR tests be used to indicate that the virus is no longer being shed. They report that viral shedding occurs two to three days before the onset of symptoms and that the amount of shedding is greatest earlier in the illness. They found no correlation between COVID-19 severity and amount of shedding, but patients with more severe disease shed for longer durations. The virus can be detected for up to a week after resolution of symptoms, although they report a single case of persistent shedding for 49 days. As for antibody response, the report describes issues related to accuracy of various tests and which viral antigen is used. Their best guesses are that IgM can be detected a median of five days after onset of symptoms and IgG 10 to 18 days (median 14) after symptom onset. They found no correlation between antibody titers and disease severity, and they found no correlation between timing of seroconversion and clinical outcomes.

Written by Henry C. Barry, MD, MS, April 16, 2020. (Source: National Academies of Sciences, Engineering, and Medicine 2020. Rapid expert consultation on SARS-CoV-2 viral shedding and antibody response for the COVID-19 pandemic; April 8, 2020. Washington, DC: The National Academies Press. )

April 17, 2020, Research Update

An uncontrolled case series published in the New England Journal of Medicine reported the outcomes for 53 patients treated with remdesivir, with 40 receiving the full course of therapy (200 mg on day 1, 100 mg once daily on days 2 through 10). The median age was 64 years of age, and median follow-up duration was 18 days. Of the 34 receiving invasive ventilation, six died, nine were still on a ventilator, three had improved to noninvasive ventilation, and 16 were on room air or had been discharged. Of the seven on noninvasive ventilation, one died, one was on invasive ventilation, and five had been discharged. Of the 12 on low flow oxygen or room air, all had been discharged. Although the outcomes compare favorably with other published series, randomized trials are ongoing to provide suitable evidence that this drug is effective.

Written by Mark H. Ebell, MD, MS, on April 11, 2020. (Source: Grein J, Ohmagari N, Shin D, et al. Compassionate use of remdesivir for patients with severe COVID-19. N Engl J Med. 2020 [published online April 10, 2020]. )

April 16, 2020, Research Update

The Cochrane Collaboration recently published a Rapid Review (April 8, 2020) evaluating 29 studies (10 modeling studies of COVID-19; four observational studies; and 15 modeling studies of SARS or MERS) that demonstrates that quarantine measures are consistently found to be effective in damping the spread of an epidemic. Quarantine is most effective when implemented early and in conjunction with other public health measures such as closing schools, restricting travel, and social distancing. The Ohio Department of Health just released a clever public service announcement illustrating why social distancing can work: . 

Written by Henry C. Barry, MD, MS, on April 10, 2020. (Source: Nussbaumer-Streit B, Mayr V, Dobrescu AI, et al. Quarantine alone or in combination with other public health measures to control COVID-19: a rapid review. Cochrane Database Syst Rev. 2020;[4]:CD013574.)

April 15, 2020, Research Update

In this open-label trial, researchers in Wuhan enrolled patients hospitalized with COVID-19 and oxygen saturation 94% or less on room air. They randomized the patients to receive standard care (n = 100) or standard care plus an HIV medication, lopinavir/ritonavir (Kaletra; 400 mg/100 mg; n = 99), twice a day for 14 days. They found no difference between groups in time to clinical improvement (median 16 days for both). Although the patients treated with lopinavir/ritonavir had fewer deaths at 28 days (19.2% vs. 25.0%), this was not statistically significant, but the study was underpowered to detect this difference. A study would need just more than 800 patients in each group, so hopefully this will be studied further.

Written by Henry C. Barry, MD, MS, on April 8, 2020. (Source: Cao B, Wang Y, Wen D, et al. A trial of lopinavir-ritonavir in adults hospitalized with severe Covid-19. N Engl J Med. 2020 [published online March 18, 2020; updated online April 14, 2020]. ) 

April 14, 2020, Research Update

The Chinese Center for Disease Control and Prevention reported the characteristics of 72,314 cases of COVID-19 infected patients in China. The first case occurred December 8, 2019, and the COVID-19 virus wasa identified January 7, 2020. Most cases (62%) had a laboratory confirmed diagnosis, 37% were diagnosed clinically, and 1% were asymptomatic with a positive test; 87% were between 30 to 70 years of age, 3% were older than 80 years, and only 2% were younger than 20 years. The overall case fatality rate of laboratory confirmed cases was 2.3% and was 14.8% for those older than 80 years. Most cases were mild (81%), and only 5% were considered critical. Mortality of critical cases was 49%. COVID-19 spread very rapidly, from one city to the entire country in 30 days. 

Written by John Hickner, MD, MS, on April 9, 2020. (Source: Wu Z, McGoogan JM. Characteristics of and important lessons from the coronavirus disease 2019 [COVID-19] outbreak in China: summary of a report of 72 314 cases from the Chinese Center for Disease Control and Prevention. JAMA. 2020;323[13]:1239-1242.)

April 13, 2020, Research Update

In a systematic review of studies reporting diagnostic and prognostic prediction models, Wynants and colleagues identified 27 studies that reported 31 prediction models. Only one study took place outside of China, and most were at high risk of bias. Nonetheless, the most frequently reported predictors of the presence of COVID-19 included age, body temperature, and signs and symptoms. The most reported predictors of severe prognosis (deterioration to severe or critical disease, prolonged hospitalization, or death) in patients with COVID-19 included age, sex, computed tomography findings, C-reactive protein, lactic dehydrogenase, and lymphopenia. Given the high risk of bias, estimates of model accuracy are unreliable. 

Written by Henry C. Barry, MD, MS, on April 8, 2020. (Source: Wynants L, Van Calster B, Bonten MMJ, et al. Prediction models for diagnosis and prognosis of COVID-19 infection: systematic review and critical appraisal. BMJ. 2020;369:m1328 [published online April 7, 2020].)

April 10, 2020, Research Update

In a New England Journal of Medicine editorial, Harvey Fineberg, MD, PhD, former dean of the Harvard School of Public Health, provost of Harvard University, and president of the Institute of Medicine, strongly recommends six steps to "crush" the coronavirus outbreak in the United States. (1) Establish unified command. (2) Make millions of diagnostic tests available. (3) Supply health workers with personal protective equipment and equip hospitals to care for a surge in severely ill patients. (4) Differentiate the population into five groups: infected, presumed to be infected, exposed, not known to be exposed or infected, and adequately immune. (5) Inspire and mobilize the public. (6) Learn while doing through real-time, fundamental research.

Written by John Hickner, MD, MS, on April 5, 2020. (Source: Fineberg H. Ten Weeks to Crush the Curve. N Engl J Med. [published online April 1, 2020]. )

April 9, 2020, Research Update

In a prospective study, French investigators assessed virologic and clinical outcomes of 11 consecutive hospitalized patients (seven men and four women with mean age of 59 years) who received hydroxychloroquine ([HCQ] 660 mg/d for 10 days) and azithromycin (500 mg on day 1 and 250 mg on days 2 to 5); eight patients had significant comorbidities. At the time of treatment initiation, 10 out of 11 had fever and received nasal oxygen therapy. Within five days, one patient died and two were transferred to the ICU. In one patient, HCQ and azithromycin were discontinued after four days because of a prolongation of the QT interval. Repeated nasopharyngeal swabs in 10 patients (not completed in the patient who died) using a qualitative PCR assay were still positive for SARS-CoV2 RNA in eight of 10 patients at days 5 to 6 after treatment initiation. These virologic results stand in contrast with those reported by Gautret, et al., and cast doubts about the strong antiviral effectiveness of this combination.

Written by John Hickner, MD, MS, on April 5, 2020. (Source: Molina JM, Delaugerre C, Goff JL, et al. No evidence of rapid antiviral clearance or clinical benefit with the combination of hydroxychloroquine and azithromycin in patients with severe COVID-19 infection. Médecine et Maladies Infectieuses. [published online March 30, 2020]. )

April 8, 2020, Research Update

A UK trial randomized 20,066 persons living in households with others to receive either access to a brief online handwashing intervention or no access. The primary outcome was the likelihood that a participant had a respiratory tract infection during a four-month period, which was decreased in the intervention group (51% vs. 59%, p < 0.001, number needed to treat [NNT] = 12). The intervention also decreased the number of gastrointestinal infections (21% vs. 25%, p < 0.001, NNT = 25) and the number of respiratory infections in a household member (44% vs. 49%, p < 0.001, NNT = 20). The site has been updated for COVID-19 and is freely available at germdefence.org.

Written by Mark H. Ebell, MD, MS, on April 4, 2020. (Source: Little P, Stuart B, Hobbs FDR, et al. An internet-delivered handwashing intervention to modify influenza-like illness and respiratory infection transmission (PRIMIT): a primary care randomised trial. Lancet. 386(10004):1641-1639.)

April 7, 2020, Research Update

C-reactive protein (CRP) is associated with more severe illness and higher mortality in COVID-19 cases. These authors retrospectively evaluated the impact of age and CRP to identify three risk groups for mortality in 577 patients in Wuhan, China. Low-risk group (0% 12-day mortality): age < 60 years and CRP < 34 mg/L. Moderate risk group (6% 12-day mortality): age 60+ years and CRP < 34 mg/L OR < 60 years and CRP ≥ 34 mg/L. High-risk group (33% 12-day mortality): ago 60+ years and CRP ≥ 34 mg/L. 

Written by Mark H. Ebell, MD, MS, on March 27, 2020. (Source: Lu J, Hu S, Fan R, et al. ACP risk grade: a simple mortality index for patients with confirmed or suspected severe acute respiratory syndrome coronavirus 2 disease (COVID-19) during the early stage of outbreak in Wuhan, China. Preprint only, not peer reviewed. )

April 6, 2020, Research Update

A trial randomized 62 patients with mild COVID-19 infection to hydroxychloroquine (HCQ) 200 mg twice daily for five days or usual care. Although the authors describe the study as double-blinded, there is no mention that patients in the control group received placebo. All had oxygen saturation greater than 93% or PaO2/FIO2 ratio > 300. Patients in the HCQ group experienced relief from fever (2.2 vs. 3.2 days, p < 0.001) and cough (2.0 vs. 3.1 days, p = 0.002) about one day sooner than controls. Of 31 patients in the control group, nine worsened, five were unchanged, and 17 improved at five days. This compares with two worsened, four unchanged, and 25 improved in the HCQ group. However, concerns about the study design and the fact that it is on a preprint server awaiting peer review means we should interpret these results cautiously.

Written by Mark H. Ebell, MD, MS, on March 27, 2020. (Source: Chen Z, Hu J, Zhang A, et al. Efficacy of hydroxychloroquine in patients with COVID-19: results of a randomized clinical trial. Preprint only, not peer reviewed. )  

April 3, 2020, Research Update

This report from Imperial College in London and others provides updated estimates of the infection fatality ratio (IFR, deaths/all infections, including mild and asymptomatic) and the case fatality ratio (CFR, death/symptomatic or confirmed infections). They estimate a mean duration from symptom onset to death of 18 days and for survivors the time from symptom onset to hospital discharge of 25 days. They do a good job of trying to adjust for biases in the data attributable to oversampling of severe cases early in a pandemic, failure to adjust for age, and the lag between case identification and death. The overall IFR is estimated to be 0.66%, and the overall CFR is 1.38%. The CFR increases from 0.06% for those in their 20s to 0.15% in their 30s, 0.30% in their 40s, 1.3% in their 50s, 4.0% in their 60s, 8.6% in their 70s, and 13.4% for those 80 and older. The proportion hospitalized increases from 1% in their 20s to 4% in their 40s to 12% in their 60s.

Written by Mark H. Ebell, MD, MS, on March 31, 2020. (Source: Verity R, Okell LC, Dorigatti I, et al. Estimates of the severity of coronavirus disease 2019: a model-based analysis. Lancet Infect Dis. [published online March 30, 2020]. )

April 2, 2020, Research Update

In this initial report of a small randomized trial in Shanghai, China, 30 patients with COVID-19 infection were randomized to hydroxychloroquine (HCQ) 400 mg per day for five days plus usual care versus usual care alone. At seven days, there was no difference in the rates of negative viral swabs by PCR (87% in the HCQ group versus 93% in the control group). There was also no difference regarding incidence of severe disease (one patient in the HCQ group and none in the usual care group), time to discharge, and time to being afebrile. While larger trials are needed and are ongoing, these results are not encouraging.

Written by Mark H. Ebell, MD, MS, on March 27, 2020. (Source: Chen J, Liu D, Liu L, et al. A pilot study of hydroxychloroquine in treatment of patients with common coronavirus disease-19 (COVID-19) [published online March 2020]. J XheJiang Univ. )

April 1, 2020, Research Update

A case series identified five patients with severe COVID-19 pneumonia and acute respiratory distress syndrome (ARDS) who were mechanically ventilated and had a PaO2/FIO2 < 300. They were all given convalescent sera from donors who had recovered from SARS-CoV-2 infection; fever resolved in four out of five patients within three days, and ARDS resolved in four out of five patients by 12 days after the transfusion of convalescent sera. At the time of writing, three had been discharged from the hospital, and two were in stable condition. While promising, clinical trials are needed given the small number of patients and uncontrolled design.

Written by Mark H. Ebell, MD, MS, on March 27, 2020. (Source: Shen C, Wang Z, Zhao F, et al. Treatment of 5 critically ill patients with COVID-19 with convalescent plasma [published online March 27, 2020]. JAMA. )

March 31, 2020, Research Update

A report published March 19 reports on a case series of 21 critically ill COVID-19 patients at a single ICU in Washington state. Patients were admitted between February 20 and March 5, 2020. The mean age of patients was 70 years (range 42 to 90), and 11 of 21 were male. Patients had symptoms an average of 3.5 days prior to hospital admission. Similar to published series in China, lymphopenia was seen in 67% of patients. Comorbidities including chronic kidney disease (48%), heart failure (43%), diabetes (33%), COPD (33%) and obstructive sleep apnea (29%) were common, with 86% having at least one comorbidity. Fifteen patients required mechanical ventilation, all due to ARDS, with one-third developing cardiomyopathy. As of March 17, 14 had died, five remained in the ICU, and two had been discharged.

Written by Mark H. Ebell MD, MS, on March 25, 2020. (Source: Arentz M, Yim E, Klaff L, et al. Characteristics and outcomes of 21 critically ill patients with COVID-19 in Washington state [published online March 19, 2020]. JAMA. )

March 25, 2020, Research Update

While the case fatality ratio (CFR) is higher in Italy than in China (7.2 % vs 2.3%), once stratified by age the CFRs are very similar within each age group up to age 70. For example, 0.4% for both countries for ages 40 to 49, 1.0% for Italy versus 1.3% for China for ages 50 to 59, and 3.5% vs 3.6% for patients aged 60 to 69. For patients in their 70s (12.8% vs 8.0%) and 80’s (20.2% vs 14.8%), CFRs were higher in Italy. This may be in part due to more very old in Italy, and differences between countries in determining cause of death in patients with multiple comorbidities.

Written by Mark Ebell MD, MS, on March 25, 2020. (Source: Onder G, Rezza G, Brusaferro S. Case-fatality rate and characteristics of patients dying in relatin to COVID-19 in Italy [published online March 23, 2020]. JAMA. 

Source: //www.aafp.org/journals/afp/content/covid-briefs.html