National mortality following upper gastrointestinal or cardiovascular events in older veterans with recent nonsteroidal anti-inflammatory drug use.

BACKGROUND

Upper gastrointestinal events (UGIE), myocardial infarction (MI) and cerebrovascular accident (CVA) are known morbidities among recent NSAID users.

AIM

To assess all-cause mortality following UGIE, MI or CVA among recent NSAID users.

METHODS

Veterans >65 prescribed an NSAID at any Veterans Affairs (VA) facility were identified using prescription fill data and their records linked to a merged VA-Medicare database. Each person-day was assessed for NSAID, coxib or proton pump inhibitor (PPI) exposure. Incidence density ratios and hazard rates of death were calculated following UGIE, MI and CVA adjusting for demographics, co-morbidity, prescription channeling, geographic location and pharmacological covariates.

RESULTS

Among 474 495 patients [97.8% male; 85.3% white; 73.9 years (s.d. 5.6)], death followed at a rate of 5.5 per 1000 person-years (95% CI: 5.4-5.6) post-UGIE, 17.7 per 1000 person-years (95% CI: 17.5-17.9) post-MI and 21.8 per 1000 person-years (95% CI: 21.6-22.0) post-CVA. CVA was associated with greatest risk of death [hazard ratio (HR) 12.4; 95% CI: 10.9-14.3] followed by MI (HR 10.7; 95% CI: 9.2-11.6) and UGIE (HR 3.3; 95% CI: 2.8-3.9). Predictors of mortality were advancing age and co-morbidity, increased use of coxibs and failure to ensure adequate gastroprotection.

CONCLUSION

Among elderly veterans with recent NSAID use, an UGIE, MI or CVA is a clinically relevant premorbid event.