Introduction: Medication use has been recognised as a potentially modifiable risk factor for poor quality and safety outcomes in long-term care facilities (LTCFs).

Research Question or Hypothesis: To examine medication-related predictors of poor quality and safety outcomes post-LTCF admission.

Study Design: Review of prognostic model development studies using the Registry of Senior Australians Historical National Cohort, which contains integrated health and aged care information for Australians living in LTCFs.

Methods: Five studies examining older (≥65 years) LTCFs new residents (01/01/2009-31/12/2016) were included. The five cohort sizes varied between 32,316 and 206,540 depending on selection criteria, study time periods, and data availability. Prognostic models for six outcomes were of interest: fall-related hospitalisations, ED presentations, unplanned hospitalisations, fractures, hospitalisations with pressure injury, and mortality. Medication-related predictors in these studies were extracted.

Results: In the six prognostic models presented in five studies, 36 medication classes independently predicted the six outcomes examined. Sedative load and warfarin were predictors of falls, ED presentations, unplanned hospitalisations, fractures and mortality, while sulfonamide diuretics were predictors of ED presentations, unplanned hospitalisations, pressure injuries and mortality. Selective serotonin reuptake inhibitors and organic nitrates were predictors of three outcomes. Systemic glucocorticoids, natural opium alkaloids, selective beta blockers, proton pump inhibitors, inhaled adrenergics or anticholinergics, dopa and dopa derivatives, gabapentinoids, statins and anticholinesterases were predictors of two outcomes. Medications predicting only one outcome include uric acid production inhibitors, antiepileptics, dihydropyridine derivatives, fluoroquinolones, anilide class analgesics, benzodiazepine derivatives, propulsives and aldosterone antagonists.

Conclusion: Medications are common predictors of poor quality and safety outcomes in LTCFs and targeted medication reviews might improve poor outcomes in these settings. Additionally, several medication-related exposures are common predictors of multiple outcomes. Therefore, strategies to optimise these common elements will contribute to improvement in more than one area.