PhD Candidate School of Population Health, RCSI University of Medicine and Health Sciences, Dublin 2, Ireland. Dublin, Ireland
Background: Drug-drug interactions (DDIs) can lead to medication-related harm, and the older population is at greatest risk. To date, in the older population, the average causal effect of DDI exposure on an adverse drug reaction (ADR)-related hospital admission is unknown. Moreover, few studies have examined the effect of DDI exposure on health-related quality of life (HRQoL) and length of stay (LOS) in hospital.
Objectives: To estimate the prevalence of potentially clinically important DDIs, the average causal effect of DDI exposure on ADR-related hospital admission, and to examine the effect of DDI exposure on HRQoL and LOS in hospital in a population of older (≥65 years) adults admitted acutely to a large academic teaching hospital in Ireland.
Methods: Prospectively collected (2016-17) data from the Adverse Drug Reactions in an Ageing PopulaTion cohort were used (N=798). Medication (current/recently discontinued/over-the-counter) and objective clinical data (e.g., renal function) were available. DDIs (which may result in a life-threatening event/permanent detrimental effect) were identified using the BNF and Stockley’s drug interactions. Causal inference models were developed using directed acyclic graphs to eliminate or reduce confounding and bias in the relationship between DDI exposure and ADR-related hospital admission. HRQoL was measured using the EQ-5D. LOS was measured as the number of days in hospital from date of admission to discharge. Multivariable logistic regression was used. DDI prevalence, odds ratios (OR), and 95% confidence intervals (CIs) are reported. Analysis was performed using SAS(v9.4).
Results: N=782 patients using ≥2 drugs were included, mean age=80.9(±7.5); 52.2% female; 36.2% polypharmacy (5-9 drugs), 56% excessive polypharmacy (≥10 drugs); 45.1% (n=353) had an ADR-related hospital admission. N=316 (40.4% [95%CI: 37.0-43.9]) patients were potentially exposed to at least one (range: 1-15) DDI at hospital admission. After controlling for confounding, the average causal effect of DDI exposure on ADR-related hospital admission was OR=1.2 [95%CI: 0.9-1.6]. This risk was significantly increased for patients exposed to DDIs which increase bleeding risk (OR=2.0[1.3-3.1]); and for aspirin-warfarin (OR=2.8[1.4-5.7]) and esomeprazole-escitalopram (OR=3.2[1.1-10.3] DDIs. Patients with at least one DDI had a lower HRQoL (mean EQ-5D=0.49[±0.39]) compared to the non-DDI exposed group (mean EQ-5D=0.57[±0.41]) (p=0.03). Patients with at least one DDI had a greater LOS in hospital (8 [IQR5-16] days) compared to those with no DDI (7 [(IQR 4-14) days] (p=0.04).
Conclusions: DDIs are prevalent among older adults on acute admission to hospital. These DDIs carry an increased likelihood of ADR-related hospitalisation, which is significantly increased (2-fold) for patients exposed to DDIs that increase bleeding risk. Patients with a DDI had lower HRQoL, and prolonged LOS in hospital.
