Comparing the efficacy and safety of medications to prevent nonsteroidal anti-inflammatory drug-induced ulcers in high-risk patients: A network meta-analysis.

The optimal strategy for preventing nonsteroidal anti-inflammatory drug-induced peptic ulcers in patients with prior peptic ulcers remains uncertain. We performed a systematic review and network meta-analysis of 22 randomized controlled trials (7,768 patients) on nonsteroidal anti-inflammatory drug users with a documented ulcer history to compare the efficacy and safety of medications for the prevention of these ulcers. The eligible interventions included proton pump inhibitors, potassium-competitive acid blockers, cyclooxygenase-2 inhibitors, prostaglandin analogs, histamine-2 receptor antagonists, and gastric mucosal protective agents. The primary outcomes were ulcer recurrence and bleeding. The safety outcomes included treatment-emergent adverse events and discontinuation owing to adverse events. Random-effects network models estimated the relative and absolute risks, numbers needed to treat, and rankings. Compared with placebo, cyclooxygenase-2 inhibitor plus proton pump inhibitor, potassium-competitive acid blocker, cyclooxygenase-2 inhibitor alone, and proton pump inhibitor monotherapy markedly reduced ulcer recurrence (numbers needed to treat = 5-6). Similar patterns were observed for bleeding prevention. Prostaglandin analogs increased adverse events and discontinuations. Proton pump inhibitor monotherapy remains the most evidence-based first-line strategy for these patients. Cyclooxygenase-2 inhibitor plus proton pump inhibitors or potassium-competitive acid blockers may provide additional benefits in very high-risk patients, but safety and cost considerations warrant caution.