Mortality After Discontinuation of Primary Care–Based Chronic Opioid Therapy for Pain: a Retrospective Cohort Study

Abstract

Background: Despite known risks of using chronic opioid therapy (COT) for pain, the risks of discontinuation of COT are largely uncharacterized. Objective: To evaluate mortality, prescription opioid use, and primary care utilization of patients discontinued from COT, compared with patients maintained on opioids. Design: Retrospective cohort study of patients with chronic pain enrolled in an opioid registry as of May 2010. Participants: Patients with chronic pain enrolled in the opioid registry of a primary care clinic at an urban safety-net hospital in Seattle, WA. Main Outcomes and Measures: Discontinuation from the opioid registry was the exposure of interest. Pre-specified main outcomes included mortality, prescription and primary care utilization data, and reasons for discontinuation. Data was collected through March 2015. Key Results: The study cohort comprised 572 patients with a mean age of 54.9 ± 10.1 years. COT was discontinued in 344 patients (60.1%); 254 (73.8%) discontinued patients subsequently filled at least one opioid prescription in Washington State, and 187 (54.4%) continued to visit the clinic. During the study period, 119 (20.8%) registry patients died, and 21 (3.7%) died of definite or possible overdose: 17 (4.9%) discontinued patients died of overdose, whereas 4 (1.75%) retained patients died of overdose. Most patients had at least one provider-initiated reason for COT discontinuation. Discontinuation of COT was associated with a hazard ratio for death of 1.35 (95% CI, 0.92 to 1.98, p = 0.122) and for overdose death of 2.94 (1.01–8.61, p = 0.049), after adjusting for age and race. Conclusions: In this cohort of patients prescribed COT for chronic pain, mortality was high. Discontinuation of COT did not reduce risk of death and was associated with increased risk of overdose death. Improved clinical strategies, including multimodal pain management and treatment of opioid use disorder, may be needed for this high-risk group.