Co-prescription of gastroprotective agents and their efficacy in elderly patients taking nonsteroidal anti-inflammatory drugs: a systematic review of observational studies.

BACKGROUND & AIMS: Guidelines recommend prescribing gastroprotective agents (proton pump inhibitors, misoprostol) to older patients (primarily ≥65 years old) taking nonsteroidal anti-inflammatory drugs (NSAIDs) to prevent gastrointestinal ulcers. Older individuals are underrepresented in clinical trials of these agents. We systematically reviewed evidence from observational studies on the use of gastroprotective agents in elderly patients and their ability to prevent NSAID-related ulcers in this population.

METHODS: We performed a systematic search of Embase and MEDLINE and identified 23 observational studies that focused on elderly patients and reported data on co-prescription of gastroprotective agents and NSAIDs and/or the effectiveness of the agents in preventing gastrointestinal events in NSAID users. We collected data on rates of co-prescription and NSAID-related gastrointestinal events in patients with and without gastroprotection.

RESULTS: A median of 24% (range, 10%-69%) of elderly patients taking NSAIDs received a co-prescription for gastroprotective agents; this percentage was only slightly higher in the oldest age groups. All studies of efficacy showed a positive effect of gastroprotection. However, the adjusted results were not suitable for synthesis, and the 5 studies reporting unadjusted results were too heterogeneous for meta-analysis (I(2) = 97%). The studies differed in outcomes, definitions of co-prescription, and differences in baseline risk factors between patients with and without gastroprotection. None of the studies assessed adverse effects of gastroprotective agents. The 2 cost-effectiveness studies reached opposing conclusions.

CONCLUSIONS: In a systematic review, the observational evidence for the efficacy of gastroprotective agents in preventing NSAID-associated gastrointestinal events was in agreement with results of randomized controlled trials. However, because of heterogeneity of included studies, it is not clear what the effect would be if more patients were treated, or at what age gastroprotection should be recommended. We offer suggestions to facilitate comparison with other work and address the questions of risk and benefit in relation to age.