Abstract
Background: Potentially inappropriate medications (PIMs) are associated with adverse outcomes and higher healthcare costs in older adults. Explicit screening criteria like the Beers Criteria, STOPP criteria, and the Malaysian Potentially Inappropriate Prescribing (MALPIP) criteria served to identify PIMs, but comparative data are scarce. Aim: To evaluate the prevalence of PIMs identified by Beers 2019, STOPP version 2 and MALPIP criteria in Malaysian older adults and examine their predictive ability for adverse outcomes and cost-saving potential. Methods: A historical cohort study was conducted among older adults aged ≥ 60 years on five or more medications in four Malaysian tertiary hospitals. PIMs were identified using Beers, STOPP, and MALPIP criteria. Sensitivity, specificity and predictive abilities of these criteria were analysed against clinical outcomes. Monthly cost savings were calculated based on hypothetical deprescribing scenarios. Results: Among 1069 patients, the prevalence of PIMs was 89.1% using MALPIP, 51.3% with Beers, and 37.0% with STOPP criteria. A moderate concordance was seen between Beers and STOPP criteria (κ = 0.437), and the lowest agreement was observed between the STOPP and MALPIP (κ = 0.131). STOPP criteria significantly predicted hospital readmissions (p = 0.003), while Beers and MALPIP did not show significant predictive abilities across all outcomes. The most common PIMs identified were proton pump inhibitors (PPIs) and nonsteroidal anti-inflammatory drugs (NSAIDs). Deprescribing scenarios based on these criteria indicated potential monthly cost savings of MYR 4.83 to MYR 44.84 per patient, with the greatest savings associated with MALPIP criteria. Conclusion: MALPIP demonstrated the highest potential for cost savings, the highest sensitivity but the lowest specificity in PIM detection. Context-specific assessments and clinical judgment are crucial in optimising medication safety and efficacy in geriatric pharmacotherapy. Further research is needed to refine PIM criteria to better predict clinical outcomes and balance the benefits and risks of deprescribing in diverse healthcare settings.