Association Between COVID-19 and Myocarditis Using Hospital-Based Administrative Data

United States, March 2020-January 2021

Tegan K. Boehmer, PhD; Lyudmyla Kompaniyets, PhD; Amy M. Lavery, PhD; Joy Hsu, MD; Jean Y. Ko, PhD; Hussain Yusuf, MD; Sebastian D. Romano, MPH; Adi V. Gundlapalli, MD, PhD; Matthew E. Oster, MD; Aaron M. Harris, MD

Disclosures

Morbidity and Mortality Weekly Report. 2021;70(35):1228-1232. 

In This Article

Abstract and Introduction

Introduction

Viral infections are a common cause of myocarditis, an inflammation of the heart muscle (myocardium) that can result in hospitalization, heart failure, and sudden death.[1] Emerging data suggest an association between COVID-19 and myocarditis.[2–5] CDC assessed this association using a large, U.S. hospital-based administrative database of health care encounters from >900 hospitals. Myocarditis inpatient encounters were 42.3% higher in 2020 than in 2019. During March 2020–January 2021, the period that coincided with the COVID-19 pandemic, the risk for myocarditis was 0.146% among patients diagnosed with COVID-19 during an inpatient or hospital-based outpatient encounter and 0.009% among patients who were not diagnosed with COVID-19. After adjusting for patient and hospital characteristics, patients with COVID-19 during March 2020–January 2021 had, on average, 15.7 times the risk for myocarditis compared with those without COVID-19 (95% confidence interval [CI] = 14.1–17.2); by age, risk ratios ranged from approximately 7.0 for patients aged 16–39 years to >30.0 for patients aged <16 years or ≥75 years. Overall, myocarditis was uncommon among persons with and without COVID-19; however, COVID-19 was significantly associated with an increased risk for myocarditis, with risk varying by age group. These findings underscore the importance of implementing evidence-based COVID-19 prevention strategies, including vaccination, to reduce the public health impact of COVID-19 and its associated complications.

Data for this study were obtained from the Premier Healthcare Database Special COVID-19 Release (PHD-SR), a large hospital-based administrative database. The monthly number of myocarditis§ and COVID-19 inpatient encounters was assessed before and during the COVID-19 pandemic, from January 2019 through May 2021.

A patient-level cohort was created to assess the association between COVID-19 and myocarditis. The cohort included all patients with at least one inpatient or hospital-based outpatient encounter with discharge during March 2020–January 2021. To minimize potential bias from vaccine-associated myocarditis,[6] 277,892 patients with a COVID-19 vaccination record in PHD-SR during December 2020–February 2021 were excluded. In addition, 37,896 patients for whom information on sex was missing were excluded. Patients with COVID-19 were defined as those who had their first inpatient or outpatient encounter with a COVID-19 International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) code during March 2020–January 2021. Patients with myocarditis were defined as those who had their first of at least one inpatient encounter, at least two outpatient encounters, or at least one outpatient encounter with a relevant specialist** with a myocarditis ICD-10-CM code during March 2020–February 2021.†† Among patients with COVID-19, the first myocarditis encounter could have occurred during or after the first COVID-19 health care encounter.

The risk for myocarditis was defined as the percentage of patients with myocarditis and was calculated among patients with and without COVID-19, overall and by sex (male or female) and age group (<16, 16–24, 25–39, 40–49, 50–64, 65–74, and ≥75 years). The percentage of myocarditis patients with a history of COVID-19 was calculated for each age group.

Associations between COVID-19 and myocarditis were estimated using a multiple logit model with the following covariates: three-way interaction between COVID-19, sex, and age group, including lower-order interactions and main effects; race/ethnicity; payer type; hospital U.S. Census region; and hospital urbanicity. Adjusted risk differences (aRDs, measure of absolute risk) were calculated as the difference between 1) the adjusted predicted risk for myocarditis (outcome) among patients with COVID-19 (exposed group) and 2) adjusted predicted risk for myocarditis among patients without COVID-19 (unexposed group); adjusted risk ratios (aRRs, measure of relative risk) were calculated as the ratio of the adjusted predicted risk among exposed to the adjusted predicted risk among unexposed§§.[7,8] All models used standard errors clustered on a unique hospital identifier. R (version 4.0.2; R Foundation) and Stata (version 15.1; StataCorp) were used to conduct all analyses. This activity was reviewed by CDC and was conducted consistent with applicable federal law and CDC policy.¶¶

During 2020, the number of myocarditis inpatient encounters (4,560) was 42.3% higher than that during 2019 (3,205). Peaks in myocarditis inpatient encounters during April–May 2020 and November 2020–January 2021 generally aligned with peaks in COVID-19 inpatient encounters (Figure 1).

Figure 1.

Number of myocarditis and COVID-19 inpatient encounters, by month* — Premier Healthcare Database Special COVID-19 Release, United States, January 2019–May 2021
*Data from recent months might be incomplete.

Within the cohort of 36,005,294 patients, 1,452,773 (4.0%) received a diagnosis of COVID-19 during March 2020–January 2021, and 5,069 (0.01%) received a diagnosis of myocarditis during March 2020–February 2021. Overall, 4,339 (85.6%) patients with myocarditis were identified by an inpatient encounter. Patients with myocarditis were slightly older than patients without myocarditis (median age = 54 years versus 50 years) and were more commonly male (59.3% versus 41.7%) (Supplementary Table, https://stacks.cdc.gov/view/cdc/109261).

Among patients with myocarditis, 2,116 (41.7%) had a history of COVID-19; this percentage was similar among males (42.4%) and females (40.9%) and differed by age group, with the lowest percentages among persons aged 16–24 years (23.7%) and 25–39 years (24.1%) and the highest among adults aged ≥75 years (64.6%) (Table). Among the 2,116 patients with COVID-19 and myocarditis, 1,895 (89.6%) received a diagnosis of COVID-19 and myocarditis during the same month; the remaining patients received a myocarditis diagnosis 1 month (139; 6.6%) or ≥2 months (82; 3.9%) after their COVID-19 diagnosis.

During March 2020–January 2021, the risk for myocarditis was 0.146% among patients with COVID-19 and 0.009% among patients without COVID-19. Among patients with COVID-19, the risk for myocarditis was higher among males (0.187%) than among females (0.109%) and was highest among adults aged ≥75 years (0.238%), 65–74 years (0.186%), and 50–64 years (0.155%) and among children aged <16 years (0.133%).

In adjusted analyses, patients with COVID-19 had, on average, 15.7 (95% CI = 14.1–17.2) times the risk for myocarditis compared with patients without COVID-19; however, because of the low risk for myocarditis in both groups, the aRD between patients with and without COVID-19 was small (aRD = 0.126%; 95% CI = 0.112%–0.140%) (Table) (Figure 2). The aRRs of myocarditis was higher among females (17.8; 95% CI = 15.6–20.0) than among males (13.8; 95% CI = 12.3–15.3), whereas the aRD was higher among males (0.165%; 95% CI = 0.146%–0.183%) than among females (0.100%; 95% CI = 0.087%–0.113%). The aRR and aRD were lowest for patients aged 25–39 years and were higher among younger and older age groups. The aRRs ranged from approximately 7.0 for patients aged 16–24 and 25–39 years to >30.0 for patients aged <16 years and ≥75 years.

Figure 2.

Adjusted risk ratio (A) and adjusted risk difference (B) of myocarditis comparing patients with and without COVID-19,* overall and by sex and age group — Premier Healthcare Database Special COVID-19 Release, United States, March 2020–January 2021
*The panels show adjusted risk ratios (A) and adjusted risk differences (B) of myocarditis comparing patients with COVID-19 to patients without COVID-19 (reference), obtained from a single logit model with the following covariates: a three-way interaction between presence of COVID-19, sex, and age group, including lower-order interactions and main effects; race/ethnicity; payer type; hospital U.S. Census region; and hospital urbanicity. 95% confidence intervals indicated by error bars.

*These authors contributed equally to this report.
PHD-SR, formerly known as the PHD COVID-19 Database, is a large U.S. hospital-based all-payer database that includes inpatient and hospital-based outpatient (e.g., emergency department or clinic) health care encounters from >900 geographically diverse, nonprofit, nongovernmental, community, and teaching hospitals and health systems from rural and urban areas. PHD-SR represents approximately 20% of inpatient admissions in the United States. Updated PHD-SR data are released every 2 weeks; release date August 4, 2021, access date August 4, 2021. http://offers.premierinc.com/rs/381-NBB-525/images/PHD_COVID-19_White_Paper.pdf
§Myocarditis was identified by the following International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) codes: B33.20, B33.22, B33.24, I40.0, I40.1, I40.8, I40.9, or I51.4.
COVID-19 was identified by ICD-10-CM code B97.29 (other coronavirus as the cause of diseases classified elsewhere) during March–April 2020 or ICD-10-CM code U07.1 (COVID-19, virus identified [laboratory-confirmed]) during or after April 2020. ICD-10-CM code B97.29 was recommended before the April 1, 2020, release of ICD-10-CM code U07.1. https://www.cdc.gov/nchs/data/icd/Announcement-New-ICD-code-for-coronavirus-3-18-2020.pdf
**Attending physician with one of the following specializations: cardiac electrophysiology, cardiovascular diseases, cardiovascular surgery, infectious diseases, or rheumatology.
††Myocarditis was assessed using all encounters for each patient from January 2019 to February 2021. Patients were defined as having new myocarditis during the study period if their first myocarditis encounter in PHD-SR occurred during or after March 2020. Myocarditis was assessed through February 2021 to allow for myocarditis diagnoses to occur after COVID-19 diagnoses in January 2021.
§§First, a multiple logit model was performed with the following covariates: three-way interaction between COVID-19, sex, and age group, including lower-order interactions and main effects; race/ethnicity; payer type; hospital U.S. Census region; and hospital urbanicity. Then the following average predicted probabilities (predictive margins) were estimated: 1) P1: the average predicted probability of myocarditis with COVID-19 set to be present and all other covariates set to their original values, and 2) P0: the average predicted probability of myocarditis with COVID-19 set to be absent and all other covariates set to their original values. aRD represents the difference in predicted probabilities (P1 minus P0); aRR represents the ratio of the predicted probabilities (P1 divided by P0). aRRs and aRDs were obtained for the full sample (where all covariates were set at their original values), for each age group (where age was set at the specific category of interest and other covariates were set at their original values), and for each sex (where sex was set at the specific category of interest and other covariates were set at their original values).
¶¶45 C.F.R. part 46, 21 C.F.R. part 56; 42 U.S.C. Sect. 241(d); 5 U.S.C. Sect. 552a; 44 U.S.C. Sect. Sect. 3501 et seq.

processing....