Comparative Evaluation of STOPPFrail and STOPP Criteria for Detecting Potentially Inappropriate Medications in Severely Frail Older Inpatients

Background: Severely frail older adults are highly vulnerable to potentially inappropriate medications (PIMs) due to multimorbidity and polypharmacy. Although the STOPP and STOPPFrail criteria were developed to identify PIMs in older patients, evidence comparing their utility in severely frail populations remains limited. This study aimed to compare PIMs identified by both criteria and to examine their associated clinical factors in severely frail older inpatients. Methods: This cross-sectional study enrolled 188 severely frail older inpatients (Clinical Frailty Scale ≥ 7) at a medical center in southern Taiwan. PIMs were identified using the STOPPFrail and STOPP criteria. Clinical variables collected included frailty status, cognitive function, nutritional status, comorbidity burden, disease complexity, history of falls, admission source, and number of medications. Results: STOPPFrail identified significantly more PIMs per patient than STOPP (mean of 2.2 vs. 1.4; p < 0.001) and detected at least one PIM in a greater proportion of patients (81.9% vs. 71.8%; p = 0.020). STOPPFrail predominantly flagged preventive medications, including antidiabetic agents, antihypertensives, lipid-lowering agents, and gastric acid suppressants, while STOPP more frequently identified sedatives, antipsychotics, antihypertensives, and anticoagulants. A higher number of regularly prescribed medications was associated with a higher PIM burden under both criteria. Female sex and a history of falls were additionally associated with STOPP-identified PIMs, whereas poorer nutritional status, admission from non-geriatric outpatient settings, and the CFS 9 subgroup were associated with a higher burden of STOPPFrail-identified PIMs, while facility residence and malignancy were associated with a lower burden. Conclusions: PIMs were highly prevalent among severely frail older inpatients. STOPPFrail primarily identified preventive medications with limited expected benefits, whereas STOPP focused on medications posing immediate safety risks. The two criteria sets identified different patterns of PIM use to guide individualized deprescribing in this vulnerable population.