Anxiety disorders are common among people receiving treatment for opioid use disorder, and the combination tends to make recovery harder. McHugh and colleagues set out to build a therapy that treats both at once, then tested whether it helped more than existing care.
The team developed a 12-session individual cognitive behavioural therapy protocol, refining it through patient interviews and expert feedback. They tried it first in a small open trial with 5 people, then ran a small randomised trial with 32 participants comparing it against 12 sessions of individual drug counselling. Everyone in both groups also received medication for opioid use disorder.
The practical signs were good. People enrolled, stayed, and reported being satisfied with the treatment. Anxiety severity fell by 11 points on a 0 to 56 scale, and the improvement held three months later. Opioid outcomes were encouraging too, with 85% of participants abstinent in the month before treatment ended.
The new protocol did not outperform the comparison treatment on anxiety or on opioid use. In a trial this small, that result cannot settle the question either way, but the authors report it plainly rather than dressing it up.
If anxiety and substance use are tangled together for you, help is available for both.
The problem being tackled
Anxiety disorders occur at high rates among people with opioid use disorder, and their presence is associated with a worse course and poorer treatment outcomes. That has prompted interest in treating both problems within a single protocol rather than sequentially or in separate services, which asks a great deal of someone already managing a demanding treatment regime.
This study was a Stage 1A/1B behavioural treatment development trial. That naming refers to a staged model for developing psychological treatments: the early stages are about building the protocol and establishing that it can be delivered and tolerated, not about proving it works. Reading such a trial as a test of effectiveness misunderstands what it was designed to do.
What the researchers did
Development came first. The manual was written and revised iteratively, drawing on interviews with patients and feedback from content experts, so that the protocol reflected what the intended recipients said they needed.
Testing came in two steps. An open pilot trial, meaning everyone knew what treatment they were receiving and there was no comparison group, ran with 5 participants. That was followed by a small randomised controlled trial with 32 participants, comparing the new 12-session individual CBT protocol against 12 sessions of manualised Individual Drug Counselling, an established psychosocial treatment for substance use. Every participant in both arms also received medication for opioid use disorder, so the comparison was between two psychosocial treatments added on top of medication, not between treatment and nothing.
What it found
Feasibility and acceptability were supported on three grounds: recruitment, retention, and patient satisfaction ratings. In plain terms, people were willing to join, they stayed, and they rated the treatment positively. For a Stage 1 trial, that is the primary result.
Within-subjects results, meaning change measured in the same people over time rather than between groups, showed an 11-point reduction in anxiety symptom severity on a 0 to 56 point scale, and those gains were sustained through 3 months of follow-up. For opioid outcomes, 85% of participants were abstinent in the prior month at the end of treatment.
The comparison told a different story. The changes in anxiety did not differ between the randomised conditions, and neither did the opioid use outcomes. The authors conclude that in this small pilot the new protocol did not show an initial benefit over an evidence-based treatment targeting opioid use alone, alongside medication.
Reading a null result properly
Two errors are easy here. The first is to treat the within-group improvement as evidence the new therapy worked, when both groups improved and time, medication and attention are all plausible contributors. The second is to conclude the protocol failed. A trial of 32 people is not built to detect a difference between two active treatments, and an absence of evidence for an advantage is not evidence there is none.
What can be said is narrower and still useful. Delivering combined CBT for anxiety and opioid use disorder is feasible, people accept it, and a larger trial would be needed to determine whether it adds anything beyond good existing care.
If anxiety and substance use are affecting each other in your life, both can be treated, and speaking with a psychologist or your doctor is a reasonable place to start.