Did a 326-Person Treatment Trial Obtain Urine Outcomes From All 326 Participants?
Randomizing 326 participants did not produce urine follow-up results for all 326. Harada and colleagues compared a residential cognitive-behavioral program added to therapeutic-community treatment with therapeutic-community treatment alone. Their different outcome measures had different follow-up samples and did not support one uniform treatment-effect claim.
The randomized treatment contrast
The Philippines trial enrolled 162 men in the intervention group and 164 in the control group. The intervention added a six-month CBT program to the existing treatment. It was therefore an add-on comparison, not CBT versus no care.
The investigators assessed methamphetamine re-use by urine testing and self-report three months after discharge. They also collected measures of well-being, coping and related psychological outcomes at defined time points.
The urine denominator was 54
The urine comparison contained 27 participants in each group. Three intervention participants and one control participant had positive results. The reported difference was not statistically significant, and its confidence interval included differences in either direction.
Those observations cannot be presented as a urine-verified result in the full randomized sample. The article identified substantial follow-up dropout, which reduced power. Participants who returned for testing may differ from those who did not.
Self-report gave a different result
The self-report comparison involved 118 participants: eight of 57 in the intervention group reported re-use, compared with none of 61 controls. That comparison was statistically significant in the reported analysis.
The discrepancy should not be erased by summarizing the entire trial as showing either successful abstinence or a proven harmful effect. Urine and self-report were obtained from different-sized follow-up samples and address different observation windows. The study did not establish why every difference between the measures occurred.
Psychological improvement was a separate outcome
The investigators reported small benefits on selected pre-discharge health and coping measures. Those findings do not turn the nonsignificant urine comparison into evidence of reduced drug re-use. Equally, the urine result does not invalidate every distinct well-being outcome.
Interpretation requires the specific outcome, time point and denominator. A pre-discharge improvement is not automatically a sustained benefit after returning to the community.
What remains uncertain?
The authors described substantial dropout, a short follow-up period, variation in treatment delivery and restriction to adult men. The trial did not establish that the same program would have the same effects in other populations or settings.
The defensible reading preserves both the original randomized enrollment and the much smaller observed follow-up groups. It does not treat missing urine results as negative tests or infer continuous abstinence from the participants who were tested.
Source
Harada T, Shirasaka T, Baba T, and colleagues. Residential cognitive–behavioral therapy versus therapeutic community for patients with methamphetamine use disorders in the Philippines: A randomized controlled trial. Addictive Behaviors Reports. 2025;21:100605. DOI: 10.1016/j.abrep.2025.100605. Supplied PDF: PMC12032182.



