Frailty is a predictor of medication-related harm requiring healthcare utilisation: a multicentre prospective cohort study

Stevenson, J. M.; Parekh, N.; Chua, K.-C.; Davies, J. G.; Schiff, R.; Rajkumar, C.; Ali, K.

Objectives: To determine the association between frailty and medication-related harm requiring healthcare utilisation. Design: Prospective observational cohort study. Setting: 6 primary and 5 secondary care sites across South East England, September 2013 to November 2015. Participants: 1280 participants, 65 years old or over, who were due for discharge from general medicine and older persons wards following an acute episode of care. Exclusion criteria were limited life expectancy, transfer to another hospital and consent not gained. Main outcome measures: Medication-related harm requiring healthcare utilisation, including adverse drug reactions, non-adherence and medication error determined via triangulation of: patient/carer reports gathered through a structured telephone interview; primary care medical record review; and prospective consultant-led review of re-admission to recruiting hospital. Frailty was measured using a Frailty Index, developed using a standardised approach. Marginal estimates were obtained from logistic regression models to examine how probabilities of healthcare service use due to medication-related harm were associated with increasing polypharmacy and frailty. Results: Healthcare utilisation due to medication-related harm was significantly associated with frailty (OR = 10.06, 95% CI 2.06 to 49.26, p = 0.004), independent of age, gender, and polypharmacy. With increasing polypharmacy, the need for healthcare use as a result of MRH increases from a probability of around 0.2 to 0.4. This is also the case for frailty. Conclusions: Frailty is as important as polypharmacy in its association with MRH. Reducing the burden of frailty through an integrated health and social care approach, alongside strategies to reduce inappropriate polypharmacy, may reduce MRH related healthcare utilisation. Trial registration: Approved by the National Research Ethics Service, East of England (REC Reference 13/EE/0075) Keywords: adverse drug reactions; non-adherence; aged