Safety of Proton Pump Inhibitor Use among Community-Dwelling Adults with Gastroesophageal Reflux Disease



Journal Title

Journal ISSN

Volume Title



Objectives The primary objectives of this study were: 1) to examine the factors associated with persistent use of Proton pump inhibitors (PPIs) among adults with gastroesophageal Reflux Disease (GERD); 2) to examine the risk of pneumonia associated with users of PPIs versus non-users in community dwelling adults with GERD; and 3) to examine the risk of Clostridium difficile infection associated with the use of PPIs among community dwelling adults with GERD.

Methods The study utilized retrospective cohort study design and nested case control study design using 2003-2004 Truven MarketScan database. The study population included adults aged more than 18 years of age, diagnosed with GERD anytime between July 1, 2003, and December 31 2003 for the first two aims of the study. The study cohort was limited to patients who had not filled a prescription for PPIs for 6 months preceding the index PPI prescription. The study population was followed for one year after first PPI prescription to examine the persistence and factors associated with persistence for the first aim of the study; patients were followed for one year after first GERD identification until pneumonia diagnosis for second aim of the study. For the third aim, a nested case control analysis was conducted and cases were identified from July 1, 2003 to December 31 2004. Presence of PPI medication was evaluated 30 days before the diagnosis of C. difficile infection. Descriptive statistics were used to evaluate the characteristics of patients in the three aims. Accelerated Failure Time model was used for the first aim, time- varying Cox proportional model and Fractional Polynomial model were used for the second aim and conditional logistic regression was utilized for the third aim of the study.

Results In the first objective, the study cohort consisted of 33,492 patients with a GERD indication and those using PPIs. The mean duration of PPI treatment was 136.8 days (95% CI: 135.2-138.5). After adjusting for other factors, several predisposing, enabling and need factors were associated with PPI persistence. Among predisposing factors, increased PPI persistence was observed among males (Survival Time Ratio (STR) 1.05); summer season (STR: 1.03), northeast region (STR: 1.10), north-central region (1.12) and west region (STR: 1.09) compared to south region; visit to gastroenterology specialty care (STR: 1.17). Increased PPI persistence was observed with enabling factors, namely, comprehensive plan coverage (STR: 1.13) and decreased PPI persistence observed among HMO insured patients (STR: 0.95) compared to PPO/EPO insurance type. Among the need factors, increased PPI persistence was observed among patients with duodenal ulcer (STR: 1.26), corticosteroids (STR: 1.09) and antidepressants (STR: 1.10) and decreased PPI persistence was observed among patients with H. Pylori infection (STR: 0.47), gastritis (STR: 0.92), gastro-intestinal hemorrhage (STR: 0.92) and abdominal pain (STR: 0.93) For the second objective, there were 89,378 patients with GERD indication in the s tudy cohort and among thes e 60,207 (67.36%) patients were PPI us ers . Res ults from multivariate time varying Cox regression model revealed that 53% increased risk of pneumonia (HR, 1.53; 95% CI, 1.35-1.75) was associated among PPI users compared to non-users. The findings were consistent across multiple sensitivity analyses. Exposure of PPIs on a daily basis was associated with 32% (HR, 1.32; 95% CI, 1.18-1.47) increased risk and cumulative duration of PPI use found to have a 37% (HR, 1.37; 95% CI, 1.20-1.56) increased risk. Categorizing PPI users into current user, former user and non-user of PPI showed a 2.44 times (HR, 2.44; 95% CI, 1.96- 3.03) increased risk of pneumonia among current users compared to non-users. For the third objective, after controlling for risk factors, overall use of PPIs was not significantly associated with risk of C. difficile infection (Odds ratio (OR): 1.24, 95% Confidence Interval (CI): 0.57-2.68). Previous hospitalization (OR: 46.00, 95% CI: 9.26-228.59), previous antibiotic exposure (low-risk (OR: 1.07, 95% CI: 1.02-1.12), medium-risk (OR: 1.22, 95% CI: 1.10-1.37) and high-risk (OR: 1.28, 95% CI: 1.11-1.47)), abdominal pain (OR: 4.10, 95% CI: 1.74-9.69) and disease severity (OR: 1.22, 95% CI: 1.02-1.46) showed increased risk of C. difficile infection.

Conclusion The study found that several predisposing, enabling and need factors were associated with PPI persistence in GERD patients. PPI use was found significantly associated with pneumonia infection, however the medication use was not significantly associated with C. difficile infection.



PPI, Gastroesophageal reflux disease (GERD)