Background: Serotonin and norepinephrine reuptake inhibitors (SNRIs) are commonly used adjuvants to short-acting opioids among nursing home residents with non-malignant pain. Whether SNRIs used as pain adjuvants are associated with adverse outcomes in older adults remains unknown.
Objectives: To evaluate the comparative safety of adding an SNRI versus nonsteroidal anti-Inflammatory drugs (NSAIDs) to short-acting oral opioid regimens among nursing home residents with non-malignant pain. We hypothesized that SNRIs used as adjuvants to opioids may increase the rate of delirium relative to NSAIDs.
Methods:
Design: New-user, retrospective cohort, with 1-year follow-up
Setting: All nursing home residents prescribed short acting oral opioids in the United States with fee-for-service Medicare (2011-2016)
Exposure: Initiation of SNRI or NSAIDs determined by Part D claims Using 2011–2016 and Medicare claims, hospitalization, and pharmacy claims, a new-user retrospective cohort of nursing home residents with prescribed short-acting opioids initiating SNRIs or NSAIDs was identified.
Outcome: Delirium was defined from Minimum Data Set 3.0 Confusion Assessment Method from quarterly MDS 3.0 assessments and hospitalization claims with ICD9/10 delirium codes.
Analysis: Propensity score matching balanced underlying differences for initiating SNRIs versus NSAIDS on 39 demographic and clinical characteristics. Fine and Gray models with propensity score matching provided hazard ratios (HRs) and 95% confidence intervals (CIs) in the presence of the competing risk of death.
Results: In the matched cohort, median duration of follow-up was 215 (Interquartile range (IQR): 322) and 150 (IQR: 345) days for residents initiating SNRIs (n=5,343) and NSAIDs (n=5,343), respectively. Hydrocodone was the most used short-acting opioid (48%). Residents received ~26mg daily oral morphine equivalent at time of SNRIs/NSAIDs initiation. The majority were women, non-Hispanic White, and aged ≥75 years. There were no differences in any of the confounders after propensity matching. Over one year, 10.8% of SNRIs initiators and 9.2% of NSAIDs initiators developed delirium. The rate of delirium onset was similar in SNRIs and NSAID initiators (HR(delirium in nursing home or hospitalization for delirium):1.05; 95% CI: 0.93-1.18; HR(hospitalization for delirium): 1.10; 95% CI: 0.93-1.30), and were similar regardless of baseline opioid dose.
Conclusions: Rates of delirium were similar among nursing home residents who intensified short-acting opioids with SNRIs and NSAIDs. Research regarding risks and benefits of pain regimens to inform clinical decisions in a population often excluded from clinical trials and research are needed.
