Previous observational studies have found an association between proton pump inhibitors (PPIs), a class of medications used to treat gastric conditions, and the risk of community acquired pneumonia. These studies, however, had important limitations including confounding by indication and protopathic bias. The purpose of this study is to determine whether PPIs increase the risk of hospitalization for community-acquired pneumonia (HCAP). To overcome the limitations of previous studies that examined this issue, this study will be conducted in a cohort of new users of non-steroidal anti-inflammatory drugs (NSAIDs), in whom PPIs are often prescribed prophylactically to prevent dyspepsia and other gastric side effects.
The investigators will use a common-protocol approach to conduct retrospective cohort studies using administrative health care data from eight jurisdictions (the Canadian provinces of Alberta, Saskatchewan, Manitoba, Ontario, Quebec, and Nova Scotia, as well as the United States (US) MarketScan, and the United Kingdom (UK) General Practice Research Database \[GPRD\]). Briefly, the Canadian databases include population-level data on physician billing, diagnoses and procedures from hospital discharge abstracts, and dispensations for prescription drugs. Alberta, Nova Scotia, Ontario, and Quebec data will be restricted to patients aged 65 years and older as prescription data are not available for younger patients. For Quebec, a 10% random sample of eligible patients will be used. The GPRD is a clinical database that is representative of the UK population and contains the records for patients seen at over 680 general practitioner practices in the UK. US MarketScan includes individuals and their dependents covered by large U.S. employer health insurance plans, and government and public organizations. As Medicare eligibility begins for those above the age of 65, the US MarketScan data will be restricted to patients aged 40 to 65 years in order to ensure complete data capture.
In each jurisdiction, the investigators will assemble a source population that includes all patients with a prescription for an oral NSAID (WHO Anatomical Therapeutic Chemical (ATC) Code M01A). From this source population, a study cohort will be created including all patients who received a prescription for an oral NSAID of ≥ 30 days duration, as patients receiving short duration prescriptions are unlikely to be prescribed a PPI for prophylactic reasons. The date of prescription (for the GPRD) or dispensation (for all other sites) of the oral NSAID will define the date of study cohort entry.
Patients will be followed from the date of study cohort entry until an event (defined below) or censoring due to death, departure from the database, end of follow-up (180 days), or the end of the study period (September 30, 2011 or the last date of data availability at that site), whichever occurs first. Patients will be permitted to enter the cohort multiple times provided that all inclusion criteria are met.
The investigators will create three mutually exclusive exposure categories: 1) PPI users, 2) H2RA users, and 3) unexposed patients. The investigators will use an analysis analogous to an intention-to-treat approach. Exposure to PPIs will be defined as a prescription for a PPI (esomeprazole, omeprazole, pantoprazole, lansoprazole, rabeprazole) on the same day as their cohort entry defining prescription for an NSAID. Exposure to H2RAs will be defined as a prescription for a H2RA (cimetidine, ranitidine, famotidine, nizatidine, niperotidine, roxatidine, ranitidine bismuth citrate, lafutidine, cimetidine combinations, and famotidine combinations) on the same day as their cohort entry defining prescription for an NSAID. Patients that are considered to be unexposed will be defined as patients not prescribed a PPI or H2RA on the same day as their cohort entry defining prescription for an NSAID. The primary outcome will be defined as incident HCAP during the 6 months following initiation of NSAID therapy.
Multiple logistic regression will be used to estimate site-specific adjusted odds ratios (aORs) and corresponding 95% confidence intervals (CIs) for the association of incident HCAP at 6 months and PPI exposure. This is considered the primary analysis. Several sensitivity analyses will be performed to assess the robustness of study results. Such analyses include: restricting analyses to a single, random, observation per patient; excluding patients who received a prescription for a PPI, H2RA, or NSAID in the 12 months before cohort entry; and excluding crossovers between PPI and H2RAs. High dimensional propensity scores will be estimated for all patients in the cohort using logistic regression. Finally, all site-specific estimates will be meta-analyzed using fixed models with inverse variance weighting. The amount of between-site heterogeneity will be estimated using the I-squared statistic.
