Chief Scientific Officer Optum Epidemiology Boston, United States
Background: Several vaccines have been developed to prevent severe COVID-19 disease caused by SARS-CoV-2. Post-market active surveillance for the occurrence of possible COVID-19 vaccine-related adverse events provides an ongoing assessment of vaccine safety in near real-time. Signals observed either through such surveillance or arising from external sources can be further evaluated through targeted methods, including validation.
Objectives: To calculate the positive predictive value (PPV) of a claims-based algorithm for myocarditis/pericarditis by age and sex using medical record review.
Methods: Patients aged 6 months to 17 years who received at least one dose of a COVID-19 vaccine from December 11, 2020 through December 16, 2022 were identified in the Optum claims database using pre-adjudicated claims. Potential cases of myocarditis or pericarditis were identified using ICD-10-CM diagnosis codes in the 42 days following a COVID-19 vaccination. Medical records corresponding to healthcare visits in any setting with claims for myocarditis/pericarditis for potential cases were sought to establish the diagnosis. Redacted medical records were reviewed independently by two cardiologists using the Brighton Collaboration criteria for confirmed or probable myocarditis/pericarditis. PPVs and 95% confidence intervals (CI) were calculated. Clinical features were described.
Results: Medical record review and adjudication was conducted for 14 potential myocarditis/pericarditis cases aged 6 months to 17 years. Of those, 11 were confirmed or probable cases corresponding to a PPV of 78.6% (95% CI 52.4-92.4). No reviewed cases were aged 6 months to 4 years. The PPV was high for all age categories (5-11 years, PPV 100.0% [95% CI 20.7-100.0]; 12-15 years, PPV 75.0% [95% CI 40.9-92.9]; 16-17 years, PPV 80.0% [95% CI 37.6-96.4]). The PPV was higher for males (PPV 81.8% [95% CI 37.6-96.4]) compared to females (PPV 66.7% [20.7-93.9]). Four of 11 (36%) confirmed or probable cases had a cardiac magnetic resonance imaging (MRI) and 11 of 11 (100%) had abnormal troponin levels.
Conclusions: The PPV of the claims-based algorithm for myocarditis/pericarditis was high among all pediatric age categories and males. The high PPV among pediatric patients emphasizes the importance of medical record review and validation of a claims-based definition for myocarditis/pericarditis as these codes are used in active surveillance.
