BACKGROUND: More United States (US) data are needed that quantify the risk of serious cardiovascular adverse events (AEs) associated with COVID-19 diagnoses to provide context for COVID-19 vaccine safety surveillance.
OBJECTIVES: To estimate the association of 9 cardiovascular AEs with a COVID-19 diagnosis in adults in the US.
METHODS: This study used cohort and self-controlled risk interval (SCRI) designs in 2 US claims data sources—Merative™ MarketScan® Commercial Database (ages 18-64 years) and Medicare fee-for-service data (ages ≥ 65 years). AEs included nonhemorrhagic stroke, hemorrhagic stroke, acute myocardial infarction, myocarditis/pericarditis, deep vein thrombosis, pulmonary embolism (PE) (evaluated both in any setting and inpatient PE only), disseminated intravascular coagulation, unusual site thrombosis with thrombocytopenia syndrome (TTS), and common site TTS. AEs were analyzed separately with AE-specific exclusion criteria. The cohort (study period: 1 April 2020 through 10 December 2020, before the introduction of COVID-19 vaccines) matched adults with a COVID-19 diagnosis and a comparator group on calendar date and patient characteristics. Follow-up began the day after the COVID-19 diagnosis or matched index date. Stabilized inverse probability of treatment-weighted hazard ratios (HRs) and 95% confidence intervals (CIs) were estimated. The SCRI (study period: 1 June 2020 through 10 December 2020) used post-COVID-19 risk windows and reference windows within individuals with a COVID-19 diagnosis and the AE. Relative incidences (RI) and 95% CIs were estimated with seasonality-adjusted conditional Poisson regression models accounting for outcome-dependent observation windows. SCRI analyses were not performed for some AEs due to high case fatality rates.
RESULTS: HR and RI estimates were above 1 for all cardiovascular AEs in both data sources and study designs. Estimated effect sizes were generally larger in MarketScan than in Medicare, and SCRI RIs were generally larger than the cohort HRs. The strongest associations were observed for PE (inpatient only): MarketScan, HR = 8.65 (95% CI, 6.06-12.35), RI = 27.4 (95% CI, 23.0-32.5); Medicare HR = 3.06 (95% CI, 2.88-3.26), Medicare SCRI not performed. The weakest associations were observed for unusual site TTS: MarketScan, HR = 1.98 (95% CI, 0.35-11.05); Medicare HR = 1.26 (95% CI, 0.85-1.88); SCRI not performed.
CONCLUSIONS: COVID-19 diagnosis was associated with cardiovascular AEs, though the strengths of the associations varied by data source and study design. These results may apply only to the time period before the introduction of COVID-19 vaccines. However, these results suggest that COVID-19 contributed to potentially large numbers of cardiovascular AEs in the US.
