(P-441) Risks of Cardiovascular and Kidney Events Associated with Proton Pump Inhibitor Use Among Medicare Beneficiaries: A Retrospective New User Cohort Study

Background: Proton Pump Inhibitors (PPIs) are widely used among the U.S. and worldwide populations. However, existing evidence provides mixed findings regarding the risks of myocardial infarction (MI), chronic kidney disease (CKD), and heart failure (HF) with PPI use.

Objectives: To examine the associations of MI, CKD, and HF events with PPI use among Medicare Beneficiaries.

Methods: A 5% random sample of the 2013-2016 Medicare administrative claims data was used to identify new users of PPI (n=106,884) and H2 blockers (n=25,447) (comparison) among fee-for-service Medicare beneficiaries with all ages. A new user cohort design with adequate washout (6-month) and follow up (15-month) periods was used. PPI use was evaluated for long term (>= 84days), any, and individual PPI (including both short and long term) use. Incident MI, CKD, and HF events were identified using diagnosis codes during the follow up period. Cox proportional hazards model was used to estimate the adjusted hazard ratios (AHR) and corresponding 95% confidence intervals (CI) as a measure of association between newly exposed PPIs use and incidence of MI, CKD, and HF. Propensity score matching and weighting techniques were used to balance confounding variables between PPIs vs. H2 blockers users.

Results: Compared with long-term H2 blocker users, long-term PPI users did not show increased risk of MI (AHR=0.98, 95%CI=0.79-1.22). Findings were similar when propensity score matching approach between long term PPI and long term H2 blocker users was used (AHR=0.88, 95%CI=0.71-1.21). Similarly, any PPI use (both long term and short term) did not show any difference in risk of MI compared with any H2 blocker users. Individual PPI users were compared with omeprazole users, and only pantoprazole users showed an increase in the risk of MI (AHR=1.15, 95%CI=1.05-1.26) while all other PPIs did not show any statistical difference. Furthermore, no association was observed between risk of CKD (AHR=1.03, 95%CI=0.98-1.08) and HF (AHR=1.03, 95%CI=0.96-1.05) among PPI users compared with H2 blocker users.

Conclusions: Our findings from this large, new user cohort study suggest PPI use does not contribute to the risk of MI, CKD, and HF among Medicare enrollees while pantoprazole showed an increased risk of MI. However, due to the nature of observational study designs, our findings should be interpreted as association rather than causality.