Frequently Asked Questions

Any illicit opioid use in the first 2 weeks met the study’s optimal threshold for early nonresponse. In this analysis, the weeks 1–2 threshold was ≥1 day of opioid use out of 14 days, with sensitivity 0.747, specificity 0.688, positive predictive value 52%, and negative predictive value 86% for predicting later sustained opioid use during weeks 5–12.

Two or more days of opioid use in the first 3 weeks was the optimal early nonresponse threshold for that period. The weeks 1–3 threshold was ≥2 days of use out of 21 days, and it predicted later sustained opioid use with sensitivity 0.649, specificity 0.810, positive predictive value 61%, and negative predictive value 83%.

No. In this study, opioid use frequency during week 1 alone did not show acceptable discrimination for later sustained opioid use, whereas opioid use frequency during weeks 1–2, weeks 1–3, and weeks 1–4 did.

Early opioid use was strongly associated with later sustained opioid use. Early nonresponse in weeks 1–2 was associated with 6.43 times greater odds of sustained opioid use during weeks 5–12, and early nonresponse in weeks 1–3 was associated with 7.54 times greater odds of sustained opioid use.

Patients meeting early nonresponse thresholds had fewer opioid-free weeks during weeks 5–12. Early nonresponse in weeks 1–2 was associated with approximately 3.12 fewer opioid-free weeks, and early nonresponse in weeks 1–3 was associated with approximately 3.70 fewer opioid-free weeks.

Yes. Early nonresponse in weeks 1–2 was associated with approximately 1.11 fewer weeks retained in treatment, and early nonresponse in weeks 1–3 was associated with approximately 1.46 fewer weeks retained.

Not necessarily. In exploratory analyses, participants with exactly 1 day of opioid use during weeks 1–3 did not have significantly different outcomes based on whether that day occurred in weeks 1–2 or in week 3: sustained opioid use (OR = 0.80, 95% CI = 0.26–2.42), opioid-free weeks (b [SE] = 0.09 [0.69], P = .896), and retention weeks (b [SE] = −1.04 [0.85], P = .220).

Among participants with 1 day of use during weeks 1–3, 73.2% used during the first 2 weeks, and 71.1% of those who used 1 day in weeks 1–2 did not report use in week 3.

The weeks 1–2 threshold may be more useful for broad early risk detection, while the weeks 1–3 threshold may be more useful for more targeted identification of higher-risk patients. The weeks 1–2 cutoff (≥1 day of use) had higher sensitivity than specificity (0.747 vs 0.688), whereas the weeks 1–3 cutoff (≥2 days of use) had higher specificity than sensitivity (0.810 vs 0.649).

The authors suggest these different performance characteristics may make the earlier threshold more suitable for lower-burden interventions and the 3-week threshold more suitable for intensive or costly interventions.

Sustained opioid use was defined as 4 consecutive weeks with either an opioid-positive or missing urine drug screen during weeks 5–12. This definition was used as the main binary outcome when evaluating how early opioid use frequency predicted later treatment course.

Yes. This study identified clinically useful early-response thresholds using self-reported daily opioid use collected with a calendar-based measure. The authors note that self-reported opioid use frequency may have practical advantages over urine toxicology, including less burden on patients and clinicians, easier use in remote visits, and potentially greater trust in the patient-provider relationship.

The findings should be interpreted with several limitations in mind. The data were collected before synthetic fentanyl dominated the drug supply, so thresholds may differ in people primarily using fentanyl. The sample was largely non-Hispanic and white, and the analysis used the first 12 weeks of treatment even though one trial lasted 34 weeks.

The authors also note that generalizability to naturalistic clinical settings needs further study, and one included trial used POATS phase 2 participants who had already returned to opioid use after an earlier taper.