3055 — Proton Pump Inhibitor Discontinuation in VA Long-Term Care
Linsky A (COLMR), Hermos JA
(MAVERIC), Lawler EV
(MAVERIC), Rudolph JL
Proton pump inhibitors (PPIs) are commonly prescribed and may contribute to overmedication. We sought to determine factors associated with PPI discontinuation in Veterans Affairs (VA) long-term care patients.
We performed a retrospective cohort analysis using national VA administrative and pharmacy data linked with Minimum Dataset patient assessments from 2005. We identified a study population of 10,371 patients admitted to VA long-term care facilities who were prescribed a PPI within the first seven days of admission, stayed a minimum of seven days, and were not receiving hospice care. Time to PPI discontinuation was calculated from admission day 7, with censoring at death, discharge or 180 days. Patient characteristics recorded were age, sex, pre-admission PPI prescriptions and hospitalizations, number of medications on day 7, prevalent and incident histamine-2 receptor antagonist (H2RAs) prescriptions, functional status, and comorbidities. Descriptive statistics summarized baseline characteristics, prevalence of PPI discontinuation, and H2RA prescription patterns. Survival curves displayed time to and hazard of discontinuation. Cox proportional hazard ratios associated patient characteristics with early (before day 28) and late PPI discontinuation.
The study patients were predominantly male (97%) with a median age of 73 (IQR 60-81). Pre-admission PPI use (67%), hospitalizations (65%), and comorbidities were common. There were 2,749 (27%) patients with PPI discontinuation, with higher hazards earlier in admission. Hospitalizations (HR 1.22; 95% CI, 1.01-1.46), pre-admission PPI use (HR 1.35; 95% CI, 1.17-1.57), and worse functional status (HR 1.22; 95% CI, 1.03-1.45) were associated with early PPI discontinuation in adjusted models. Factors associated with decreased risk of early discontinuation were six or more non-acid suppressing medications (HR 0.78; 95% CI, 0.66-0.92), gastric acid related disease (HR 0.53; 95% CI, 0.46-0.61), and diabetes (HR 0.82; 95% CI, 0.72-0.94). New H2RA prescriptions in those with PPI discontinuation were uncommon (9%).
More than one in four patients admitted to VA long-term care facilities prescribed a PPI on admission had that medication discontinued, with less discontinuation in those prescribed six or more medications.
Patients with polypharmacy may be prime targets for focused attention to medication appropriateness in an effort to reduce adverse drug effects in VA long-term care.