• Monika Sarkar, Sean Hennessy, and Yu-Xiao Yang.
• University of Pennsylvania, Philadelphia, Pennsylvania 19104-6021, USA.
• Ann. Intern. Med. 2008 Sep 16; 149 (6): 391-8.

BackgroundRecent studies suggest that proton-pump inhibitors (PPIs) may increase the risk for community-acquired pneumonia (CAP).ObjectiveTo examine the association between PPI use and CAP in adults followed in general practices in the United Kingdom.DesignNested case-control study.SettingThe General Practice Research Database (1987 to 2002) in the United Kingdom.ParticipantsPatients age 18 years or older with at least 6 months of initial pneumonia-free follow-up in the database. Case patients (n = 80 066) were defined as those who received an incident diagnosis of CAP. Control participants (n = 799 881) were selected by using incidence density sampling, matching on practice site, calendar period, and follow-up duration.MeasurementsUse of PPIs within 30 days before the index date. Adjusted odds ratios (ORs) were estimated by using conditional logistic regression, adjusting for potential confounders.ResultsOverall, current PPI use was not associated with an increased risk for CAP (adjusted OR, 1.02 [95% CI, 0.97 to 1.08]) or risk for CAP that required hospitalization (adjusted OR, 1.01 [CI, 0.91 to 1.12]). There was a strong increase in risk for CAP associated with current use of PPI therapy that was started within the previous 2 days (adjusted OR, 6.53 [CI, 3.95 to 10.80]), 7 days (adjusted OR, 3.79 [CI, 2.66 to 5.42]), and 14 days (adjusted OR, 3.21 [CI, 2.46 to 4.18]), but there was no statistically significant association for longer-term current PPI therapy. A separate matched case-control analysis, which included the 3 strongest confounders as additional matching factors, yielded similar results as the primary analysis (adjusted OR, 0.96 [CI, 0.91 to 1.02]).LimitationsAdherence to PPI prescription was assumed to be 100%. No radiographic evidence was available to corroborate a diagnosis of CAP.ConclusionProton-pump inhibitor therapy started within the past 30 days was associated with an increased risk for CAP, whereas longer-term current use was not.

18 keywords...

hide…