Comparative safety of short-acting opioid dose escalation versus long-acting opioid regimen in nursing homes

Background: In nursing homes (NH), modifications to opioid regimen, such as increasing the dosage of short acting opioids (SAO) or initiating long-acting opioids (LAO) as a switch or add-on to an initial SAO regimen, are common approaches to manage pain.

Objectives: To evaluate the comparative safety of SAO dose escalation versus LAO add-on or switch regimens among NH residents.

Methods: Using the Minimum Data Set (MDS) 3.0 linked to Fee-For-Service Medicare data from 2011-2016, we conducted a retrospective cohort analysis of long-stay nursing home residents. Residents with ≥ one month of continuous SAO were identified from Part D claims. Opioid regimen changes were: (1) SAO dose escalation: a ≥20% increase in daily oral morphine equivalent (OME) dosage of SAO, without initiating any LAOs (n=9,670); (2) LAO add-on: initiation of any LAO with continued SAO use (n=2,081); (3) LAO switch: discontinuation of SAOs and initiation of LAO (n=938). Outcomes within 3-months following the opioid regimen change included all-cause hospitalization (Part A), delirium (MDS or Part A), and falls/fractures (MDS or Part A). A propensity score method was applied to calculate inverse probability of treatment weighting (IPTW), and the IPTW was applied to balance the baseline characteristics between the three exposure groups. Fine and Gray models using doubly robust IPTW estimated hazard ratios (HRs) and 95% confidence intervals (CIs) in the presence of the competing risk of death.

Results: After IPTW, standardized mean differences in covariates were well-balanced except for the baseline OME and the OME at the start of follow-up. The rate of all-cause hospitalization was lower in LAO add-on user in relation to SAO dose escalation (HR LAO add-on versus SAO dose escalation: 0.88, 95% CI: 0.79-0.97), but similar for residents who switched to LAO (HR LAO switch versus SAO dose escalation: 0.93; 95% CI: 0.81-1.07). Compared to SAO dose escalation, delirium rates were higher in residents with LAO add-on (HR LAO add-on versus SAO dose escalation: 1.20, 95% CI: 1.03-1.38) and residents switching to LAO (HR LAO switch versus SAO dose escalation: 1.29, 95% CI: 1.06-1.56). Similarly, falls/fracture rates were elevated among LAO add-on and switchers relative to SAO dose escalation (HR LAO add-on versus SAO dose escalation: 1.41, 95% CI: 0.84-2.36; HR LAO switch versus SAO dose escalation: 2.86, 95% CI:1.64-4.99).

Conclusions: Rates of delirium and falls/fractures were higher among residents who initiated an LAO as an add-on to SAO or a switch from SAO. These findings suggest a need for further investigation into the mechanisms underlying the increased risk of delirium or falls/fractures associated with LAO initiation or switching from SAO.