Design: Secondary analysis of a pragmatic multisite randomised comparative effectiveness trial
Population: 778 US veterans on long-term opioid therapy for chronic pain; 250 with possible substance use disorder
Primary outcome: 12-month change in pain interference, pain intensity and opioid dose
Effect: Both groups improved; no adjusted difference between those with and without possible substance use disorder

This secondary analysis drew on a 12-month pragmatic trial in US veterans on long-term opioid therapy for chronic pain, randomised to either an interdisciplinary pain team or pharmacist-led collaborative management. The arms were combined to ask whether a possible substance use disorder at baseline changed outcomes.

Of 778 participants, 250 (32%) screened positive for possible substance use disorder. They were younger and more often had mental health diagnoses, but started with similar pain and opioid doses. Both groups had reductions in pain interference, pain intensity and opioid dose over 12 months, and adjusted analyses found no difference between them.

The practical message for general physicians is that a substance use history is not a reason to exclude someone from structured pain care or a supported opioid taper.

Screen for substance use disorder in patients on long-term opioids, and treat it alongside pain.
Do not exclude patients with a substance use history from collaborative pain care or a supported taper.
Involve a pharmacist where available; lower-intensity pharmacist-led care was one of the trial arms.
Co-prescribe naloxone for patients on long-term opioids, particularly with substance use.

Why it matters

It challenges the reflex to refer these patients away rather than include them.

The statistics, in plain English

The finding is an absence of difference between groups after adjustment. That does not prove the groups respond identically; a modest difference could exist and not be detected in a secondary analysis.