Discontinuation of Opioids for Chronic Pain in Primary Care Associated With Subsequent Overdose Mortality

Opioid registries and other monitoring strategies seek to reduce opioid-related harms among patients who receive long-term opioid therapy (LTOT) for pain. Enhanced monitoring may result in LTOT discontinuation based on safety concerns, but what happens to patients after LTOT discontinuation is unknown. This retrospective cohort study examined the opioid registry of an urban primary care clinic, including 572 patients who received LTOT between 2010 and 2015. Mortality, prescription opioid use, and primary care visits were compared between patients who discontinued LTOT versus those who maintained it.

• Three-quarters of patients had mental health conditions, one-third had substance use disorders, and chronic medical conditions were common.
• Sixty percent of patients had LTOT discontinued over the study period.
• LTOT discontinuation was mostly initiated by providers (80%), not patients.
• Overall mortality in the cohort was 20% (4.7 per 100 person-years) with 4% dying of a definite or possible overdose.
• LTOT discontinuation was significantly associated with overdose death (hazard ratio, 2.9).
• Of patients who had LTOT discontinued, three-quarters subsequently received opioids from other prescribers and only half maintained primary care at the clinic.

Comments: In this cohort, overdose risk remained high after LTOT discontinuation. The study’s methodology cannot establish a causal relationship between LTOT discontinuation and future overdose death, but there is good reason for concern. The best strategy for responding to “red flag” behaviors uncovered by monitoring is still unclear; however, clinicians need tools, including options for opioid use disorder treatment and comprehensive pain management, to enhance patient safety.

Aaron D. Fox, MD

Reference: James JR, Scott JM, Klein JW, et al. Mortality after discontinuation of primary care-based chronic opioid therapy for pain: a retrospective cohort study. J Gen Intern Med. 2019;34(12):2749–2755.