Evaluating adverse events after vaccination in the Medicare population.

PURPOSE: Post-licensure observational studies using large linked databases can provide important data about whether adverse events are associated with vaccines, but databases that have been used may not have sufficient statistical power to examine rare events, and may underrepresent the elderly. We assessed the utility of Medicare data for evaluating adverse events after influenza and pneumococcal vaccines, by using an example involving selected clinical conditions, and evaluating aspects of data quality relevant to vaccine safety analyses.
METHODS: We used 2001 data from the National Claims History File and Enrollment Database to determine if hospitalization for urinary tract infection (not likely associated with vaccination) or for cellulitis and abscess of the upper arm and forearm is associated with vaccination.
RESULTS: For influenza vaccine, the 7-day period after vaccination did not demonstrate an elevation in hospitalization with cellulitis and abscess of the upper arm and forearm; for pneumococcal vaccine, a clear peak was evident. No increase in urinary tract infection was found after either influenza or pneumococcal vaccine. Having a prior Medicare claim for pneumococcal vaccine within 5 years was a risk factor for hospitalization with cellulitis and abscess of the upper arm and forearm (relative risk, 2.6; 95% confidence limits (CL), 1.3, 5.0).
CONCLUSIONS: Medicare data are a useful source for evaluating adverse events after vaccination. Screening analyses can be performed using administrative data, but medical record review to validate diagnoses will often be needed for rigorous study of vaccine-adverse event associations.