Skip to main content

Our take —

This is a case series of seven male patients (14-19 years) who developed myocarditis with or without pericardial inflammation within 4 days of receiving the second dose of Pfizer-BioNTech vaccine in the United States. All patients presented with chest pain, with or without myalgia and/or fever and had elevated cardiac specific enzymes and inflammatory markers and abnormal electrocardiograms. All patients recovered quickly with standard treatments for myo- and pericarditis. While this case series provides detailed descriptions of their clinical presentation and course of treatment, the relationship between the Pfizer-BioNTech vaccine and subsequent cardiac inflammation remains unclear given the following cautions: 1) the lack of adjustment for other factors that may have affected risk for acute cardiac inflammation; 2) the impact of potential selection biases; and, 3) the lack of an appropriate comparison group for evaluating potential risk differences.

Study design

Case Series

Study population and setting

This is a case series of seven adolescent males (14-19 years) hospitalized in the US with acute myocarditis with or without pericardial inflammation (myocarditis/myopericarditis) within four days of receiving their 2nd dose of Pfizer-BioNTech vaccine in April and May 2021. The authors document their presenting symptoms, diagnostic workup, treatment(s), and status at hospital discharge.

Summary of Main Findings

All patients presented to the hospital complaining of chest pain with or without arm pain, myalgia, or fever. Initial evaluation revealed elevated cardiac enzymes (Troponin T, from 1.09 to 22.1) and inflammatory markers (ESR range: 10-40; CRP range: 6.7-18.1) at presentation, All had negative SARS-CoV-2 PCR via nasopharyngeal swap and negative antibodies against SARS-CoV-2 nuclear capsid antigen, six patients had negative other respiratory viral panel and none met criteria for MIS-C.

All patients had evidence of ST segment elevation on their electrocardiogram (ECG); two had evidence of mild ventricular strain and 1 had mildly decreased systolic function on echocardiogram. Myocardial MRIs revealed evidence of myocardial with or without pericardial inflammation in all cases. No evidence of other viral infections were identified in six of the patients as a possible cause for their myocardial inflammation. One patient underwent cardiac catheterization, which revealed normal coronary arteries. Treatment included some combination of IVIG, corticosteroids, and non-steroidal anti-inflammatory drugs (NSAIDs). Three cases resolved with NSAIDs alone. Although 5 hemodynamically stable patients were initially admitted to the intensive care unit for close monitoring, all patients had significant improvement in their symptoms within few days of hospitalization.

Study Strengths

This case series provides detailed information about the presentation, diagnostic workup, treatment, and short-term outcomes among seven adolescents who developed myocardial with or without pericardial inflammation within four days of their second dose of the Pfizer-BioNTech vaccine.

Limitations

Despite testing for a variety of viral and bacterial infections that could possibly cause myocardial and pericardial inflammation, no mechanistic pathway was examined between Pfizer-BioNTech vaccine and these disorders. Cardiac biopsy was not performed, due to generally mild disease and rapid improvements in clinical status, precluding examining the potential underlying mechanism between Pfizer-BioNTech vaccine and this disorder. Furthermore, as a case study, this paper is not able to assess incidence of myocardial and/or pericardial inflammation after the second dose of the Pfizer-BioNTech vaccine in adolescents or compare incidence against background incidence in this age group. Finally, the cases were identified through personal communications, which may have resulted in significant selection biases if these individuals had other characteristics that affected their risk for acute cardiac inflammation.

Value added

The study provides detailed information on clinical presentation, treatment, and outcomes on male adolescents presenting with myocardial or pericardial inflammation following a second dose of the Pfizer-BioNTech vaccine. This will help pediatricians quickly identify this disorder, provide effective therapy, and encourage health care providers to report these events to the Vaccine Adverse Event Reporting System (VAERS).

Our take —

Between September and December 2020, there were 287 mucormycosis cases diagnosed across 16 healthcare centers in India, a two-fold increase compared to the same time period in 2019. Of these cases, 187 were COVID-19 associated mucormycosis (CAM). CAM was extremely rare (<0.3%) among hospitalized COVID-19 patients. Aside from CAM patients having more frequent hypoxia requiring ICU admission, CAM and non-CAM cases had similar clinical manifestations and outcomes. The majority of both CAM and non-CAM cases had uncontrolled diabetes. Most CAM cases were given glucocorticoids as part of their treatment for COVID-19, similar to previous smaller case series from India. It is possible that glucocorticoid treatment contributed to additional CAM cases in certain settings, but more research in larger sample sizes and with additional control for confounding is needed to further evaluate this finding.

Study design

Case Series, Retrospective Cohort

Study population and setting

This multicenter retrospective cohort study evaluated the prevalence, epidemiology, and outcomes of COVID-19 associated mucormycosis (CAM), a fungal disease, in India from September 1, 2020 to December 31, 2020, compared to mucormycosis not associated with COVID-19 (non-CAM) over the same time period. Mucormycosis was defined based on clinical and radiologic findings that showed fungi in tissue or sterile body fluids. COVID-19 diagnosis was based on detection of SARS-CoV-2 on RT-PCR or rapid antigen tests. The data from 16 health centers were used to evaluate predisposing factors and clinical manifestations of CAM and non-CAM patients, and data from 7 of these centers were used to estimate prevalence of CAM. Demographics, medical history, clinical presentation, and outcomes were extracted from patient medical records. Patients were treated for COVID-19 and mucormycosis in accordance with institutional protocols; inappropriate glucocorticoid use was defined as use of any steroid among non-hypoxic patients or when a dexamethasone-equivalent was used for 10 days or more with more than 6 mg/day.

Summary of Main Findings

Across the 16 participating centers, 295 cases of mucormycosis were diagnosed, among whom 287 had complete data and were included in the study population (mean age: 53.4 years, 25% women, 83% diagnosed on direct microscopy). Across the 7 centers with data on all hospitalized COVID-19 patients, the prevalence of CAM was 0.27% (28/10,517) in general wards and 1.6% (25/1579) in the ICU. The total number of mucormycosis cases between September 2020 and December 2020 was over twofold higher than the same time period in 2019, though the number of mucormycosis unrelated to COVID was relatively similar (92 in 2020 vs. 112 in 2019). Of the 187 CAM cases (65% of all mucormycosis cases), 61 (33%) did not have other underlying diseases, compared to only 19 of the 100 non-CAM cases without other underlying diseases. Among CAM and non-CAM cases, uncontrolled diabetes was common (63%) though newly diagnosed diabetes was more common among those with CAM (21% vs. 10%). Both groups experienced similar clinical manifestations and sites of involvement (~85% in each group with rhino-orbital or rhino-orbital-cerebral involvement), and mortality was similarly high (6-week mortality: 38%; 12-week mortality: 46%). However, patients with CAM were more likely to have hypoxia requiring ICU admission than non-CAM patients (31% vs. 9%). CAM patients were classified according to timing of onset (early: <8 days after COVID-19 diagnosis, n=29; late: ≥8 days after COVID-19 diagnosis, n=158), and although demographic, clinical characteristics, and outcomes were similar between early and late CAM groups, hypoxia and inappropriate or non-indicated glucocorticoid use was associated with development of late CAM as compared to early CAM (adjusting for age, sex, and underlying risk factors).

Study Strengths

This was a multisite study with detailed data on clinical characteristics, treatment history, and outcomes of COVID-19 associated mucormycosis, with a comparison group of adults with mucormycosis not associated with COVID-19.

Limitations

Despite its multi-site design, the sample size was still relatively small. Adjusted analyses, especially those among only CAM cases, were likely underpowered to detect associations or overfit. Consequently, many of the estimates had very wide confidence intervals and should be interpreted with caution. The study was retrospective and relied upon data available in the medical record, which may not have fully captured individual disease histories and comorbidities. The study did not report whether there was clustering of cases by healthcare center nor explore facility-level characteristics that might have enabled healthcare-associated sources of mucormycosis during the study period. Given the small sample size, analyses of mucormycosis and CAM outcomes were limited in their ability to fully adjust for treatment practices, which differed between sites.

Value added

This study is among the first and most detailed reports of COVID-19 associated mucormycosis, a rare fungal disease associated with high mortality.

Our take —

This study, conducted in Germany, aimed to estimate infectiousness through the SARS-CoV-2 infection time course. Findings demonstrated that younger individuals had a lower viral load and lower peak cell culture replication probability compared to adults. Those infected with B.1.1.7 variant had a higher viral load and higher estimated cell culture replication probability than those infected with other variants. Very high viral loads were detected in less than 10% of individuals, with one third of these higher viral loads among those who were pre-symptomatic, asymptomatic, or mildly symptomatic. This study allows for a comparison of infectiousness parameters by a number of different characteristics, however it should be noted that viral load and replicating virus are not a direct measure of infectiousness.

Study design

Case Series

Study population and setting

The goals of this study were to better quantify viral infection and shedding of SARS-CoV-2, by measuring and comparing viral load (viral RNA concentration) and replicating virus isolated in cell culture. Specifically, the study aimed to estimate: 1) differences in infectiousness by age and gender, along with differences by symptom severity (e.g. pre-symptomatic, asymptomatic, symptomatic); 2) differences in infectiousness by viral variants (specifically B.1.1.7),;and 3) timing and peak of infectiousness. Samples on 25381 German individuals with at least one positive RT-PCT test from February 24, 2020 to April 2, 2021 were examined. Samples were divided into three categories: hospitalized (n=9519, 37.5%), pre-symptomatic, asymptomatic, or mildly symptomatic (PAMS) individuals receiving testing at walk-in community testing centers (n=6110, 24.1%), and other (n=9752, 38.4%). A total of 1533 individuals had the B.1.1.7 variant. Viral loads were assessed using standard RT-PCR methodologies. The probability of viral culture was estimated using Baysesian regression by combining the estimated viral loads with cell culture isolation data from previously published work relating viral loads and culture probabilities. B.1.1.7 variants were identified using sequencing and comparisons were made using data from only testing centers that specifically reported B.1.1.7 cases.

Summary of Main Findings

Younger individuals had a lower mean log10 viral load than adults (those 20-65 years old) (0.5 lower for 0-5 year olds and 0.18 lower for older adolescents (those 15-20 years old)). Younger individuals also experienced 78% of the peak cell culture replication probability compared with older individuals, though the authors suggest this may have been due to a small swab size used for the youngest individuals in the sample. No differences were seen by gender. Very high viral loads (above 109 copies per swab) were detected in 8% of all individuals, with approximately one third of these high viral loads among presymptomatic, asymptomatic, or mildly symptomatic individuals. When comparing those with the B.1.1.7 variant with those without it, it was found that those with the B.1.1.7 variant had a higher mean log10 viral load (1.05 log10 higher), and a 2.6 times higher estimated cell culture replication probability. The estimated time from the start of viral shedding to peak viral load was 4.3 days.

Study Strengths

This study utilized multiple samples from individuals, and therefore it was possible to model viral dynamics over the time course of infection. Additionally, the large and generally varied population, by symptom severity, age, hospitalization, and variants allowed for the estimation of infectiousness parameters by a number of different characteristics.

Limitations

The authors note that the two parameters studied here: viral load and replicating virus isolate are not a direct measure of infectiousness, as behavior and context play an important role in infectiousness as well and these were not measured. Additionally, replicating virus was not measured in any of these samples, but instead relied on prior data to model culture positive probability. This analysis relies on the assumption that this prior work may be generalized to this study population.

Value added

This study furthers our understanding of SARS-CoV-2 infectiousness through the analysis of viral load, cell culture isolation, and genome sequencing data, especially to understand differences in infectiousness among those who are not symptomatic compared to those who are and those with the B.1.1.7 variant compared to others.

Our take —

A rare and serious adverse event involving abnormal blood clotting (vaccine-induced thrombosis and thrombocytopenia, or VITT) has been associated with the ChAdOx1 nCoV-19 (AstraZeneca) vaccine. This study reported on the clinical features and laboratory features of 23 patients who had received the AstraZeneca vaccine and who subsequently developed thrombosis and thrombocytopenia 6-24 days later, despite being mostly young and healthy. As seen in other studies, the clinical presentation of patients strongly resembled heparin-induced thrombocytopenia (HIT), even though patients had not been treated with heparin before symptom onset. All but two of the patients tested positive for anti-platelet factor 4 (anti-PF4) antibodies, which strongly supports the hypothesis that platelet activation induced by anti-PF4 antibodies leads to VITT. The results have implications for clinical guidance; the authors recommend avoidance of platelet transfusions, and suggest intravenous immunoglobulin and direct thrombin inhibitors as the first line of treatment. The study is limited by a relatively small sample size, which is a consequence of the rarity of this serious adverse event. It is still unknown how the vaccine might induce production of anti-PF4 antibodies, and what might predispose individuals to this reaction.

Study design

Case Series

Study population and setting

This study from the UK evaluated blood samples of 22 patients with venous thrombosis and thrombocytopenia and one patient with isolated thrombocytopenia and markedly elevated D-dimer after vaccination with the ChAdOx1 nCoV-19 vaccine (AstraZeneca). Because the clinical presentation in 22 patients highly resembled heparin induced thrombocytopenia (HIT), the investigators used enzyme-linked immunosorbent assays (ELISA) to test for antibodies against platelet factor 4 (PF4), which are the primary cause of HIT. A functional assay that tests platelet activation was used to confirm ELISA results for 7 of the 23 patients. The authors also described clinical characteristics and outcomes of the patients who provided samples.

Summary of Main Findings

All patients presented 6-24 days after receiving the first dose of ChAdOx1 nCoV-19, with a median time of 12 days. The median age was 46 years (range 21-77 years). Sixteen patients (70%) were younger than 50 years, and 14 (61%) were female. One patient had a history of deep vein thrombosis, and one patient was taking oral contraceptive pills at the time she sought medical attention. Clinically, 13 patients (56%) presented with cerebral venous thrombosis (one had deep vein thrombosis and pulmonary embolism along with cerebral venous thrombosis). One patient had deep vein thrombosis and bilateral adrenal hemorrhage, two patients had acute stroke due to thrombosis at the middle cerebral artery, and two patients had portal vein thrombosis: one with concurrent myocardial infarction and the other one had concurrent aortic thrombosis. Seven patients (30%) died. One patient had available data on post-mortem evaluation, which showed thrombosis at small arteries in the intestine, brain, lung, venous sinuses and intracerebral hemorrhage. Of note, additional thrombotic events occurred among patients who received platelet transfusion for thrombocytopenia and/or heparin for thrombosis. Only 10 patients (34%) had their sera available for antibody testing against coronavirus. All of these patients had negative antibodies against SARS-CoV-2 nucleocapsid protein, which ruled out recent COVID-19 infection. The levels of anti-SARS-CoV-2 receptor binding domain (RBD) and anti-spike antibodies were consistent with the level expected after ChAdOx1 nCoV-19 vaccination. The level of antibodies against non-SARS-CoV-2 coronavirus antibodies were similar to those in the general population. Thirteen patients (56%) had low fibrinogen levels, and all patients had markedly elevated D-dimer levels. Twenty-one patients (91%) had positive anti-PF4 antibodies via ELISA. One patient whose sample was collected five days after clinical presentation (and who had multiple platelet transfusions) tested  negative for anti-PF4 antibodies, and another had equivocal results. Of the 7 patients tested with the functional HIT assay, 5 had significant platelet activation in the absence of heparin. Adding heparin in excess of physiological doses fully suppressed platelet activation.

Study Strengths

In addition to ELISA, the authors used a functional platelet assay in a subset of patients to identify the presence of platelet activation. The clinical presentation of patients was reported in detail.

Limitations

The study had a small sample size (n=23). Only 10 patients were investigated for the possibility of previous SARS-CoV-2 infection, and only one-third of patients with detectable anti-PF4 antibodies were tested with the functional HIT assay. The course of treatments provided to the patients was not reported in detail, limiting inference about the benefits or harms associated with VITT treatment. Patient characteristics were also not reported in detail, and key demographic details were not available (e.g., ethnicity, country of origin, comorbidities); since it is not yet known what characteristics might predispose individuals to VITT, these data would have been particularly useful.

Value added

The study adds significant data pertaining to the mechanism of a rare adverse event known as vaccine-induced thrombosis and thrombocytopenia (VITT), which appears to be causally related to ChAdOx1 nCoV-19 (AstraZeneca) vaccine.

Our take —

A bar reopening event in rural Illinois led to a COVID-19 outbreak event with 46 cases identified, both among attendees and close contacts of attendees. At this indoor event, adherence to mask-wearing and physical distancing was poor. Though there may be limitations in measurement of cases (e.g. total number of attendees unknown, underreporting of close contacts, not all contacts were tested), this study provides further evidence that good public health prevention measures are needed to prevent the spread of SARS-CoV-2 in high-risk transmission settings like a bar.

Study design

Case Series

Study population and setting

In February 2021, the Illinois Department of Public Health and local health department conducted an investigation of an outbreak associated with an opening event at a rural Illinois bar. The event was held indoors in a 2800 square foot bar with limited ventilation. Cases associated with the outbreak were defined as COVID-19-like symptoms or a positive SARS-CoV-2 nucleic acid amplification test (NAAT) or antigen test within 14 days of the event. Cases could have been among attendees or employees who were at the event and reported no other known COVID-19 contacts. Secondary cases tied to the outbreak were defined as close contacts of the cases with a positive SARS-CoV-2 NAAT or antigen test. Interviews were conducted among all cases, both primary and secondary.

Summary of Main Findings

The total number of people who attended the event is unknown (the venue accommodated approximately 100 people), but there were six employees present. There was a total of 46 cases identified as linked to the event, including 26 attendees, three employees, and 17 secondary cases. At the event, there was one asymptomatic individual who had had a previous COVID-19 diagnosis from the day before, along with four people who had COVID-19 symptoms. Among attendees, it was reported that mask use was inconsistent and physical distancing of more than 6 feet apart was not adhered to during the event. There were 71 close contacts identified, of which 37 received testing (37/71, 52.1%). Among those who were tested, 46% received a positive result (17/37). Transmission associated with the event resulted in one school closure and one hospitalization of a long-term care facility resident.

Study Strengths

The use of contact tracing data allows us to understand the real-world implications of limited mask use and lack of consistent physical distancing.

Limitations

Because it is unknown how many people actually attended the event and because all close contacts were not tested, these numbers could be an underestimate of the total number of cases associated with the event. The cases identified are likely related to the bar opening event, but sequencing was not conducted and it is possible that the source of infection originated elsewhere or that there were multiple sources of infection.

Value added

This study builds on the evidence that when community transmission is relatively high, indoor events with poor mask use and limited physical distancing can lead to outbreaks.

Our take —

Reports of rare severe thrombotic events (blood clotting), associated with the Oxford-AstraZeneca vaccine have led to temporary suspensions of its use in several countries. This study reported on 11 patients (9 of whom were female, with a median age of 36 years) in Germany and Austria who had thrombotic events and low platelet counts. Antibodies against platelet factor 4 (PF4) were present in these patients, and platelet activation was enhanced by the presence of PF4. The clinical presentation of these cases resembled heparin-induced thrombocytopenia; however, none of the patients had been exposed to heparin before symptom onset. The following inhibited platelet activation in vitro: immune globulin, antibodies against platelet receptor factor IIA, and heparin. These results provide some detail on vaccine-induced thrombosis and thrombocytopenia, reports of which remain quite rare, though many questions about pathogenesis, risk factors, and treatment remain unanswered.

Study design

Case Series

Study population and setting

This study reported on clinical and laboratory characteristics of 11 patients who received the Oxford-AstraZeneca (ChAdOx1 nCov-19) vaccine in Germany and Austria and who developed thrombosis or thrombocytopenia. Laboratory testing was conducted on blood samples from 9 of the 11 patients, with ELISA used to test for the presence of platelet factor 4 (PF4)-heparin antibodies, and a modified (PF4-enhanced) platelet activation assay used for detection of platelet-activating antibodies under a range of conditions. Samples from 24 additional patients with suspected thrombocytopenia related to vaccination were also tested for validation purposes.

Summary of Main Findings

Nine of the 11 patients were women, with a median age of 36 years. Thrombotic complications began 5 to 16 days after vaccination, and all patients presented with moderate-to-severe thrombocytopenia (median platelet count nadir: 20,000 per cubic millimeter). Five patients had more than one thrombotic event; events included cerebral venous thrombosis (n=9), splanchnic-vein thrombosis (splenic, mesenteric, portal and hepatic veins; n=3), pulmonary embolism (n=3), and other types (n=4). One patient presented with fatal cerebral hemorrhage. Six patients died, and one had an unknown clinical outcome. None of the patients had received heparin before symptom onset, and only one patient was known to have prothrombotic blood disorder before symptom onset. There was evidence of disseminated intravascular coagulation in five patients, who all had d-dimer levels above 10mg/L and one or more of the following abnormalities: fibrinogen levels below 200ng/mL, elevation of the international normalized ratio (INR), and elevation of partial thromboplastin time(PTT). All samples that tested positive for PF4-heparin antibodies (n=24) showed strong reactivity; PF4-dependent platelet activation occurred in the absence of heparin. Platelet activation was enhanced by PF4 and inhibited by heparin, monoclonal antibody against platelet receptor FC gamma IIA, and immune globulin (10 mg/mL).

Study Strengths

Platelet activation patterns were assessed for additional samples from 24 patients suspected of thrombosis and/or thrombocytopenia, which allowed for comparison with the case series. The authors used two assays to measure antibodies against PF4.

Limitations

The sample of patients was small and detailed information on risk factors was not available. As this was a case series, data are needed on the prevalence of PF4-heparin antibodies among all individuals receiving the ChAdOx1 vaccine. More attention may have been paid to unusual thrombotic events, and so these may have been overrepresented (i.e., it is possible that the profile of thrombotic events associated with the ChAdOx1 vaccine is different from that observed in this study due to under-ascertainment). Detailed laboratory results are reported for only a subsample (n=4) of the 11 patients.

Value added

This study provides more detail about the mechanism of thrombotic adverse events that have led to pauses in the global rollout of the Oxford-AstraZeneca vaccine.

Our take —

At the start of the pandemic, 112 households in Bergen, Norway, with 112 index cases of SARS-CoV-2 and 291 total members were enrolled in a study to understand household transmission. The study used serological testing to understand attack rates, and found an overall attack rate of 45%. There were no significant differences by age or gender, and notably attack rates among young children were not lower than among adults. The results from serological testing proved to be better at capturing all those that had the virus (more sensitive) than PCR testing alone, suggesting a valuable place for serological testing to understand household transmission dynamics of SARS-CoV-2.

Study design

Case Series

Study population and setting

Between February 28 and April 4, 2020, 112 households in Begen, Norway, with a total of 291 participants, were enrolled into a study to understand household transmission of SARS-CoV-2. Households were selected based on index cases (positive RT-PCR test) identified at the Bergen Municipality Emergency Clinic. Household members included all those who lived in the same household as the index patient. Households where the index patient lived alone were excluded. SARS-CoV-2 antibodies were measured (SARS-CoV-2-spike- specific IgG) six to eight weeks after the index case tested positive in order to determine household attack rate. RT-PCR testing for household members was completed among those who reported COVID-19 related symptoms.

Summary of Main Findings

The secondary household attack rate was 45% (95% CI: 28-53%) based on serological testing (positive antibodies). The attack rate increased to 47% when adding those who tested negative for antibodies, but who were symptomatic and had positive RT-PCR tests. A total of 70 individuals received RT-PCR testing, and 32 tested positive. Using the standard RT-PCR testing algorithm which would have tested only symptomatic individuals, the attack rate would have been 17.9% (32/179). The risk of transmission was increased if the index patient had a fever (adjusted [a]OR: 3.31, 95%CI: 1.52-7.24), or difficulty breathing (aOR: 2.25, 95% CI: 1.80-4.62). There were no significant differences in attack rate by household member age, though those at risk for infection who were 60 years and older had an attack rate of 73% (95%CI 48-89%), and those 0-10 years old had an attack rate of 48% (95%CI: 29-67%).

Study Strengths

Testing of household members using antibody or serological tests provides a more sensitive, and likely more robust, estimate of household attack rate than RT-PCR alone.

Limitations

This study occurred at the beginning of the pandemic when fewer public health mitigation strategies had been put into place, which may make it difficult to generalize these results to other settings. It is possible that some selection bias occurred in which index cases and households agreed to participate in the study (228 initially eligible: 84/228 excluded, and 32/228 redefined as household members), though many were excluded because they resided in a single-person household (51/84).

Value added

This study shows the added value of utilizing serological tests for understanding transmission dynamics, particularly among household members.

Our take —

This large case series from more than 200 US hospitals provides additional evidence that obesity is associated with increased risk of COVID-19 severity and death, and that risk goes up with increasing obesity severity, with a 2-fold higher risk of death among adults with BMI ≥ 45 kg/m^2 relative to those with healthy weight. Notably, this study demonstrated that the risk persisted among adults aged 65 years and older, a finding in contrast to some previous studies. These findings reinforce the use of BMI as a factor in vaccine prioritization, and also suggest that clinicians may need to manage COVID-19 more intensively among patients with obesity.

Study design

Case Series

Study population and setting

This large case series included patients aged 18 and older who were diagnosed with COVID-19 in 238 hospitals across the United States from March 1 to December 31, 2020. The data were drawn from the Premier Healthcare Database Special COVID-19 Release, a hospital-based electronic database that includes patients from all health insurance/payer groups; data from the subset of hospitals (238/~800) that recorded height and weight were used. The primary analytic sample included adults with ICD-10 code confirmed COVID-19 diagnosis in either the emergency department (ED) or inpatient setting, and with non-missing height and weight. Multivariable logit regression analyses evaluated the association between body mass index (BMI; weight in kg/height in m^2) and four outcomes: 1) inpatient hospitalization (reference group: those seen in ED only), 2) intensive care unit (ICU) admission, 3) invasive mechanical ventilation, and 4) death. The analysis for inpatient hospitalization included all adults diagnosed in the ED or inpatient, while the remaining 3 outcomes were assessed among hospitalized patients. BMI was assessed as both a continuous and categorical variable; the latter grouped by underweight (<18.5 kg/m^2), healthy weight (18.5-24.9 kg/m^2), overweight (25-29.9 kg/m^2), and four classes of obesity (30-34.9; 35-39.9; 40-44.9; ≥45 kg/m^2).

Summary of Main Findings

Among the 148,494 patients with COVID-19 in the hospital/ED setting (median age 55 years; 46% male; 51% non-Hispanic white), 51% were obese, 28% were overweight, 9% were healthy weight, and 2% were underweight. Approximately half—71,491—of these COVID-19 patients were hospitalized (median age 65 years; 51% male; 56% non-Hispanic white), with similar BMI category distributions. Hospitalization was elevated for adults with obesity and underweight relative to those of normal weight, and risk increased with category of obesity, independent of age, sex, race/ethnicity, payer type, geospatial characteristics of the hospital (urbanicity and US Census region), and admission month. Similarly, when assessed as a non-linear continuous variable, BMI demonstrated a J-shaped association with hospitalization, ICU admission, and death. Among hospitalized patients, the risk of ICU admission was significantly elevated only in the two highest obesity classes relative to healthy weight, while the risk of mechanical ventilation increased in a graded relationship across all BMI categories. The risk of death among hospitalized patients increased with obesity severity; patients with the highest obesity class had a 2-fold greater risk of mortality relative to those with healthy weight. Most of the inferences were observed both among those younger than aged 65 years and those aged 65 years and older, although the effect sizes and uncertainty around them were larger in the younger age group. The magnitude of the risks attenuated after accounting for comorbidities (hypertension, diabetes, chronic kidney disease, asthma, coronary atherosclerosis or other heart disease, chronic obstructive pulmonary disease, and cancer), but the ordering of risk was retained.

Study Strengths

The large sample size provided sufficient power to assess BMI across seven categories (including four obesity classes) and to stratify the analyses above and below aged 65 years to assess potential differences in the risks of BMI by age group. The authors performed rigorous multivariable analysis, including sensitivity analyses to assess robustness to adjustment for comorbidities and to address missing BMI with a multiple imputation approach.

Limitations

The hospitalization outcome in this study, assessed only among patients in the hospital or ED, is not comparable to hospitalization in studies with population-based cases of COVID-19, because those seeking care in the ED are likely to differ in significant ways from the overall population with COVID-19 (i.e. more likely to have severe symptoms; less likely to have a primary healthcare provider; more likely to be underweight or obese). The authors justified not adjusting for known risk factors for COVID-19 severity–such as heart disease, lung disease, and diabetes–because these conditions may be partially caused by obesity, but it might be more balanced to consider the effect of BMI as falling between the primary results and the more modest results from the sensitivity analyses that adjusted for these comorbidities.

Value added

This study assessed the relationship between BMI and four COVID-19 severity outcomes in more granularity than has been possible in previous smaller studies, demonstrating the relative effects of seven mutually exclusive BMI categories, and showing graded increases in risk with increasing obesity severity for most outcomes. The study also evaluated BMI as a non-linear continuous variable, identifying specific BMI values in the healthy to overweight range as inflection points with the lowest risk of severe COVID-19 events using polynomial spline models. The large sample size supported age-stratified analyses, and demonstrated that obesity remained a strong risk factor for mortality and invasive mechanical ventilation among patients aged 65 and older as well as the younger population.

Our take —

This study investigated a cluster of COVID-19 cases linked to two instructors reported to the Hawaii Department of Health. They found that the attack rate in these classes varied based on proximity to symptom onset, and classes held <1 day from symptom onset had attack rates of 100% for Instructor A, and 95% for Instructor B, but classes held >=2 days from onset had no reported cases. The primary limitation of this study is that class participants may not seek out testing and underreport asymptomatic infections. However, this study showed that COVID-19 remains a major risk in exercise settings despite other infection control protocols being followed, including social distancing, sanitization, plexiglass installments, and intermittent mask use. Policies of six feet of physical distancing in indoor areas with limited ventilation are likely to be insufficient, particularly when individuals are breathing heavily due to physical exertion.

Study design

Case Series

Study population and setting

This study investigated a suspected COVID-19 outbreak first identified on July 2, 2020 by the Hawaii Department of Health. Fitness Instructor A tested positive by RT-PCR, and displayed signs and symptoms. Instructor A taught at two facilities (X and Y), and was linked to 21 cases. Another instructor, B, came into contact with Instructor A and later tested positive. Instructor A taught a class on June 27 at Facility X (60 hours before symptom onset), and another class on June 28, 38 hours before symptom onset at Facility Y. While Instructor A followed social distancing and sanitization protocols, they shouted instructions at students without a mask at a distance >6 ft. They taught another class on June 29 at Facility Y, 4 hours before symptom onset. A number of students had exposure to Instructor A during multiple classes. Instructor B worked at a third facility, Z, and was exposed via the June 28 class at Facility Y. Instructor B taught five personal training sessions and small-group classes at Facility Z 2 days after their first exposure, and more than 2 days before their symptom onset. They taught another person on July 1, and another 10 participants and 3 small classes on July 2, before developing symptoms 12 hours later.

Summary of Main Findings

The study calculated the attack rate for each class, ranging from 0% to 100%. The highest attack rate (10/10, 100%) was seen at Facility Y during a stationary cycling class with Instructor A where no one was masked on June 29. All participants tested positive by RT-PCR. For Instructor B at Facility Z, the highest attack rate was seen at the kickboxing small group class on July 2, 12 hours before symptom onset, where only 2 participants of 9 who attended were masked, and all (9/9, 100%) later tested positive by RT-PCR. There was a notable gradient of attack rates based on proximity to symptom onset for both instructors. Instructor A had an attack rate of 0% for classes taught >1 day from symptom onset, but 100% attack rate for <1 day. For Instructor B, more than 2 days from onset had an attack rate of 0% for 33 participants in the classes, and 13% attack rate for <2 and >1 day from symptom onset. Finally, <1 day from symptoms, Instructor B had a 95% attack rate in classes.

Study Strengths

The study strength was the depth of data collected for the type of classes, whether individuals were masked, and the timing of classes taught. This investigation allowed them to calculate attack rates for each class, as well as based on symptom onset. They also investigated the COVID-19 distancing and sanitation protocols in place, and whether they were being adhered to.

Limitations

The study notes that participants may have multiple dates of exposure, and so it may be challenging to uncover exactly when transmission could have occurred for some. Additionally, they did not require all contacts be tested, and so some individuals who did not experience symptoms or who refused testing were assumed to not have COVID-19 when calculating the attack rate. Thus, these estimates may underestimate the level of transmission among these classes.

Value added

This study shows the high risk of transmissibility of COVID-19 in exercise settings despite other COVID-19 mitigation protocols being followed and demonstrated that transmission in this setting was likely highest in the day of or before symptom onset of the infectious case.

Our take —

This large case-series found a low prevalence of myocardial injury through return to play cardiac screening after resolution of symptomatic or asymptomatic SARS-CoV-2 infection among professional league athletes in the US and Canada. Cardiac screening identified 30/789 (3.8%) athletes with abnormalities in one or more of the initial parameters. Subsequent evaluation among those with abnormalities confirmed just five athletes (0.6% of the total cohort) with evidence of myocardial (n=3) or pericardial inflammation (n=2) on cardiac MRI, all of whom had moderate symptoms while ill with COVID-19; these individuals were not cleared to return to play. There were no adverse cardiovascular outcomes in short term follow-up among those athletes who cleared cardiac screening and returned to play. The prevalence of myocardial injury after SARS-CoV-2 infection found in this population of healthy recovered young, predominantly male professional athletes, many who had asymptomatic infection, is likely to be lower than that among the general population, but highlighted the importance of disease severity/symptoms as a prognostic factor for myocardial injury. The low yield among athletes with prior asymptomatic infection supports the American College of Cardiology expert consensus panel recommendation that screening is limited to athletes who had symptomatic infection. Furthermore, the study findings substantiate the judicious use of advanced imaging such as cardiac MRI as a downstream test option rather than for broad, first line screening.

Study design

Case Series

Study population and setting

This study examined the prevalence of myocardial injury after the resolution of symptoms of laboratory-confirmed SARS-CoV-2 infection among 789 professional athletes in six major North American sports leagues (Major League Soccer, Major League Baseball, National Hockey League, the National Football League, and the men’s and women’s National Basketball Associations). Prior to resuming activity, athletes who tested positive for SARS-CoV-2, irrespective of symptoms, underwent downstream cardiac evaluation after the resolution of symptoms following the American College of Cardiology (ACC) return to play (RTP) screening algorithm. This study included data from all athletes who underwent RTP cardiac assessment between May and October 2020. Cardiac screening included cardiac enzymes (troponin), electrocardiogram (ECG) and transthoracic echocardiogram; those with abnormal results underwent further cardiac testing with cardiac magnetic resonance imaging (CMR) and/or stress echocardiography, at the discretion of the team physicians, and events were reported through December 2020. The case data were deidentified and pooled by Columbia University.

Summary of Main Findings

Of the 789 athletes with SARS-CoV-2 infection (98.5% male; mean age 25 years; 74% tested by PCR), 42% had been asymptomatic for the duration of their infection while 58% had mild to moderate or pauci-symptomatic COVID-19 illness. Only one athlete was monitored overnight in a healthcare facility with COVID-19; none were hospitalized for or had cardiopulmonary symptoms. The median time between SARS-CoV-2 test and cardiac screening was 17 days (3-156 days). There were 30/789 (3.8%) athletes with abnormalities on RTP screening tests; 6 (0.8%) had elevated troponin levels, 10 (1.3%) had ECG abnormalities concerning for myocardial injury and 20 (2.5%) had abnormalities in their echocardiograms. Those athletes underwent further downstream cardiac evaluation; 27 had CMR and 15 had stress echocardiograms (12 had both tests). There were no abnormalities in the stress echocardiograms that were performed. On CMR, three athletes met criteria for myocarditis (myocardial inflammation) and two had imaging evidence of pericarditis (pericardial inflammation) (0.6% of the total cohort). All 5 athletes with CMR abnormalities previously had mild to moderately symptomatic SARS-CoV-2 infection (loss of smell or taste, cough and/or fatigue). On the other hand, 15 athletes who had echocardiographic abnormalities and four out of six who had troponin elevation had normal CMR and returned to play. Resting ECGs had particularly low specificity as a screening test. All athletes with normal cardiac screening RTP testing and those with normal CMR returned to their normal sports activities without reporting any cardiac symptoms up to December 2020.

Study Strengths

The health care staff associated with each team followed the RTP cardiac screening algorithm recommended by an American College of Cardiology (ACC) expert consensus panel. However, they departed from the ACC guidelines by applying the same screening algorithm to all athletes who tested positive for SARS-CoV-2, including asymptomatic cases, which provides a broader and more systematic view of the prevalence of cardiovascular sequelae of SARS-CoV-2.

Limitations

Predominantly male athletes were included. The screening algorithm was performed after the resolution of symptoms and therefore represents a healthy recovered population, excluding athletes with unresolved symptoms who may have a higher likelihood of persisting myocardial abnormalities. There was no centralization of the cardiac testing protocol, which was individually directed by each team’s medical staff nor was cardiac testing interpretation centrally performed. Study athletes had wide variation in the time interval between their cardiac evaluation and diagnosis of COVID-19 infection (3-156 days). Some cardiac abnormalities due to COVID-19 may have been underestimated but missed due to delayed cardiac evaluation. There were no significant adverse outcomes observed after return to play, though follow-up duration was short.

Value added

The study adds valuable knowledge on the prevalence of myocardial injury among healthy professional athletes recovering from COVID-19 and supports the safe return to play with a judicious cardiac testing protocol.