Enhancing patient stratification methods for medication review during acute admissions.

Medication review in the hospital is an effective tool for identifying inappropriate prescribing, but not every patient can receive one due to resource constraints. Thus, it is important to identify patients who stand to benefit most from medication review. Patients are typically identified for medication review based on number of medications and age, but this may not be the most efficient strategy for identifying patients at high risk of future adverse clinical outcomes. In this exploratory study of 27 873 acutely admitted medical patients (≥18 years) at Copenhagen University Hospital Hvidovre, Denmark, we investigated whether other clinical factors could improve the prediction of adverse clinical outcomes and, therefore, be incorporated into patient stratification for medication review. The outcomes of interest were acute readmission or mortality within 90 days post-discharge. Medication risk score (MERIS), renal function and anticholinergic cognitive burden did not improve prediction of adverse clinical outcomes compared to number of medications and age. However, addition of a frailty index based on routine biomarkers (FI-OutRef) or plasma soluble urokinase plasminogen activator receptor (suPAR) did improve prediction of outcomes. Use of these practical proxies of disease burden may improve the identification of acutely hospitalized patients who would benefit most from a medication review.