Long-Term Opioid Therapy Tapering and Risk of Substance Use Disorder and Overdose: Differences by Opioid Dose and Duration

BACKGROUND: Tapering from long-term opioid therapy (LTOT) for chronic pain is increasingly common. However, there is inadequate evidence regarding the possibility of adverse effects, particularly across heterogeneous LTOT courses. OBJECTIVE: To estimate associations of tapering from various LTOT doses and durations with risk of substance use disorder (SUD) and overdose. DESIGN: Within-individual cohort study. PARTICIPANTS: US adults tapering from at least 3 months of stable LTOT, obtained from Optum's de-identified Clinformatics® Data Mart Database for 2010-2021. MAIN MEASURES: Dose tapering was defined as 2 consecutive months of ≥ 15% lower dose within 6 months of LTOT baseline without abrupt discontinuation. Outcomes were assessed from emergency department or inpatient events with recorded SUD or overdose diagnoses. We compared risk of SUD/overdose across up to 1 year following tapering onset with risk during stable LTOT using conditional logistic regression with time-varying covariates. KEY RESULTS: Of 732,038 adults who experienced an LTOT taper, 7.6% had at least 1 SUD/overdose event during LTOT baseline or up to 1 subsequent year. Relative to LTOT baseline, odds of SUD/overdose were comparably greater in the month before taper onset (odds ratio [OR], 1.36; 95% CI, 1.32, 1.40) and the first month of tapering (OR, 1.38; 95% CI, 1.34, 1.43). Odds decreased partially across the subsequent year (e.g., month 12: OR, 1.11; 95% CI, 1.07, 1.15). This pattern was consistently observed across groups defined by baseline LTOT dose and duration, although odds returned to baseline levels within approximately 6 months of taper onset for those tapering from higher dose (≥ 90 MME/day) and/or longer duration (≥ 12 months) LTOT. CONCLUSIONS: This within-individual study found that LTOT recipients who experienced a dose taper had greater risk of SUD or overdose around the onset of the taper. Much of that risk escalation may precede-rather than result from-changes in monthly dose.