Predictors for inappropriate proton pump inhibitor use: observational study in primary care

Abstract

Background Proton pump inhibitor (PPI) indications are limited to gastrointestinal disorders and ulcer prophylaxis. However, PPIs are among the most frequently prescribed drugs. Aim To evaluate the appropriateness of PPI prescriptions and identify predictive factors for inappropriate PPI use. Design and setting Observational study using a Dutch primary care database with all new PPI prescriptions between 2016 and 2018. Method Individual patient data and details on PPI use were collected. The appropriateness of initiation and continuation of PPI prescriptions was evaluated using the applicable guidelines. Results In total, 148 926 patients (aged ≥18 years) from 27 general practices were evaluated. A total of 23 601 (16%) patients started PPI therapy (mean age 57 [SD 17] years, 59% female). Valid PPI indications at initiation were seen in 10 466 PPI users (44%). Predictors for inappropriately initiated PPI use were older age (odds ratio [OR] 1.03, 95% confidence interval [CI] = 1.03 to 1.03), and use of non- selective non-steroidal anti-inflammatory drugs (OR 5.15, 95% CI = 4.70 to 5.65), adenosine diphosphate receptor inhibitors (OR 5.07, 95% CI = 3.46 to 7.41), COX-2 inhibitors (also known as coxibs) (OR 3.93, 95% CI = 2.92 to 5.28), and lowdose aspirin (OR 3.83, 95% CI = 3.07 to 4.77). Despite an initial valid indication, PPI use was inaccurately continued in 32% of patients on short-course therapy for dyspepsia and in 11% of patients on ulcer prophylaxis. Conclusion More than half of PPI users in primary care were found to have an inappropriate indication, with unnecessary ulcer prophylaxis related to drug use being one of the leading causes. Future initiatives to reduce PPI use for unnecessary ulcer prophylaxis and timely deprescription if PPI is no longer indicated, are needed.