Medicines optimisation in care homes via pharmacist case management: what is the impact on subsequent healthcare resource usage?

Introduction: A novel approach to improve prescribing among care home residents was recently developed in Northern Ireland, whereby a clinical pharmacist attended the care home, developed personalised pharmaceutical care plans for each resident, then case managed the resident until all interventions were completed. This care model resulted in significant improvements in prescribing appropriateness¹. However, less is known regarding the impact of such improvements on subsequent healthcare resource utilisation.

Aim: To examine primary and secondary healthcare usage by care home residents following improved prescribing appropriateness achieved via medicines optimisation by a case management pharmacist.

Methods: Secondary data analysis was conducted on data obtained from the Medicines Optimisation in Older People (MOOP) pharmacists in two Northern Irish healthcare trusts using SPSS version 25. Pre‐ and post‐intervention data on care home residents (≥65 years, N = 1095) case managed between 2015–2016 was collected by the MOOP pharmacists. Inappropriate prescribing was assessed by the case management pharmacists using the Medication Appropriateness Index (MAI)² scored at two time points, baseline and post‐intervention. Kaplan‐Meier survival analysis was used to compare time to first unplanned hospital admission (days) between those who experienced a change in MAI total score and those who did not. Poisson regression analyses were conducted with the number of General Practitioner (GP) visits and the number of out of hours (OOH) GP visits, within both 30 and 90 days, as outcome variables. Independent variables including MAI score improvement and clinical intervention types were examined in multivariate analyses, controlling for demographics, medical history and baseline levels of healthcare resource usage.

Results: Inappropriate prescribing was highly prevalent (84.1%) at baseline (M = 14.87, SD = 13.11), and was significantly improved upon following pharmacist intervention (M = 0.70, SD = 2.04, Z = ‐25.97, p < 0.001). Time to first unplanned hospital admission was not significantly increased for those who experienced a change in total MAI score (Mdn = 42) and those who did not (Mdn = 34), Χ²(1) = 0.247, p = 0.619, n = 108. Magnitude of MAI total score change was associated with a small significant increase in the number of GP call outs within 30 days (OR = 1.02, 95%CI 1.01, 1.03), but not within 90 days. Individual clinical interventions showed no significant association with numbers of GP call outs in either period. Having at least one medication started was associated with a 61% increase in OOH GP call outs within 30 but not 90 days (OR = 1.61, 95%CI 1.03, 2.51). Having at least one blood test requested by the case management pharmacist was associated with significantly fewer OOH GP call outs within both 30 (OR = 0.34, 95%CI: 0.18, 0.63) and 90 days (OR = 0.59, 95%CI 0.39, 0.91). Previous numbers of healthcare attendances were significant predictors of both GP and OOH GP visits in both periods.

Conclusions: This is the first study to examine healthcare resource usage by care home residents in Northern Ireland following pharmacist case management. Healthcare resource usage may be invariant for those residents who exhibit greater clinical complexity. However, the results were limited by the absence of comparison to a ‘usual care’ control group. Continued pharmacist involvement within care homes is warranted, given the high baseline levels of inappropriate prescribing identified, and the significant improvements achieved following medicines optimisation.

References: 1. Miller R. Developments in Practice. Medicines optimisation in older people (MOOP); the journey from pilot to permanent service. Journal of Medicines Optimisation. 2018 Sep;4(2):27.

2. Hanlon JT, Schmader KE, Samsa GP, Weinberger M, Uttech KM, Lewis IK, et al. A method for assessing drug therapy appropriateness. Journal of Clinical Epidemiology. 1992 Oct 1;45(10):1045–51.