Screen, Treat and Retain Meth-using People With Opioid Use Disorders at Methadone Clinics: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Screen, Treat and Retain Meth-using People With Opioid Use Disorders at Methadone Clinics (NCT04706624), a randomized na study. The primary outcome is Increase in HIV Viral Suppression for HIV-positive Participants: High- vs. Low-Intensity Frontline Intervention With SMS in Stage 2.

2. Clinical Question

3. Trial Design

This was a randomized, unmasked, sequential trial with an enrollment of 667. The registry lists the phase as not applicable and the status as completed. The study was conducted in Hanoi and Ho Chi Minh City, Vietnam, according to the trial summary. Its stated aim was to evaluate combinations of behavioral interventions for people with opioid use disorder who use methamphetamine while receiving methadone maintenance therapy, with HIV outcomes and methamphetamine use among the concerns.

The design began with two frontline strategies. The high-intensity frontline intervention was 12 weeks of contingency management. The low-intensity frontline intervention combined 6 weeks of contingency management with 6 weeks of group education. The record describes later stages in which participants could receive SMS reminders, Matrix alone, or Matrix with contingency management. The intervention options are part of a sequential, adaptive structure; the record does not describe the allocation probabilities or every decision rule in sufficient detail to reconstruct how each participant proceeded.

The primary endpoint concerns HIV viral-suppression status at baseline and week 24. The registry states that viral load was assessed at baseline, week 13, and week 24, but the primary measure reports baseline and week 24 status. It limits the analysis population to Stage 1 responders with completed week 24 viral-load assessments. Thus, the endpoint does not describe all enrolled participants without qualification.

There are 16 outcome measures and 36 analyses listed. The posted adverse-event groups include the frontline arms and several later intervention combinations. Those group counts are not a substitute for the primary efficacy results. A detailed account of randomization procedures, allocation concealment, and the statistical analysis plan is not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: After enrollment, 2 participants withdrew consent and declined to participate in the intervention. Thus, they were not randomized. This study employed a Sequential Multiple Assignment Randomized Trial (SMART) design. In the first phase, participants were randomized to one of two frontline interventions for 12 weeks. At Week 13, they were re-assigned to one of three adaptive strategies for an additional 12 weeks. We report outcomes stratified by the initial frontline intervention groups.

Only the first 6 of 8 groups are shown.

Stage 1 Intervention (Weeks 0 to 12)

Stage 2 Intervention (Weeks 13 to 26)

Analysis population for the primary outcome, as defined in the registry: Only participants classified as Stage 1 responders with completed viral load assessments at week 24 were included in this analysis.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Population: During the frontline intervention stage (Weeks 1-12), participants were randomized into two groups. Therefore, we analyzed baseline data according to these two groups. In the second phase, participants were re-assigned to one of three adaptive intervention groups based on their response in the first phase, with each original group branching into all three adaptive arms.

6. Primary Endpoint

The registry lists the primary outcome as Increase in HIV Viral Suppression for HIV-positive Participants: High- vs. Low-Intensity Frontline Intervention With SMS in Stage 2, measured over Baseline (start of Stage 1) and week 24 (end of Stage 2).

The primary endpoint records HIV viral-suppression status at baseline and at week 24, the end of Stage 2. Viral load was assessed at baseline, week 13, and week 24. The measure classifies results as at least 200 copies/mL, from at least 20 to less than 200 copies/mL, or undetectable at less than 20 copies/mL. It is expressed as a count of participants in the categories, not as a continuous viral-load value.

Using ordered viral-load categories allows the record to describe suppression status at two points while preserving distinctions among detectable viral-load levels and undetectable status. Comparing baseline with week 24 can describe status across the full intervention period. The primary endpoint title specifies a comparison of high- versus low-intensity frontline intervention with SMS in Stage 2, but the posted outcome definition includes multiple categories and time points. The registry record does not report a single numerical contrast or statistical estimate for this endpoint, so the direction, size, and precision of any between-strategy difference cannot be assessed.

The analysis population is restricted to Stage 1 responders with completed week 24 viral-load assessments. This restriction is important when interpreting the measure: it does not, by itself, establish what the result would have been for all enrolled participants or for those without a completed week 24 assessment.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Randomization and sequential structure. Randomization means that assignment to intervention strategies is determined using a chance process rather than solely by clinician or participant choice. This can support a fair comparison at the point of randomization, subject to the actual allocation procedure and the analysis. The registry identifies the trial as randomized and sequential, but the randomization method, allocation concealment, and exact probabilities are not reported in the registry record. In a sequential design, later intervention decisions may depend on earlier response status. The record describes Stage 1 responders and later-stage options, but a complete decision algorithm is not reported.

Intervention comparisons. The named frontline comparison is 12 weeks of contingency management versus 6 weeks of contingency management followed by 6 weeks of group education. Later strategies include SMS reminders and Matrix-based approaches. A comparison in this setting should be interpreted with its stage and eligible population in view. Comparing groups formed at different stages may involve different participant sets and decision points. The registry information does not state the analysis model used to account for those features, so one should not assume a particular weighting, adjustment, or causal estimand.

Categorical endpoint. The primary outcome is viral-suppression status, organized into three viral-load categories, at baseline and week 24. For categorical outcomes, analyses can compare category distributions, changes in category, or the probability of a particular category. Those approaches answer different questions. The registry record does not report which statistical model or contrast was used. Nor does it report a single numerical result that would allow the reader to assess the observed category pattern.

Precision and uncertainty. In general, a confidence interval describes a range of values compatible with the observed data and the statistical model, while conveying uncertainty in an estimate. A narrow interval indicates greater precision than a wide interval, considered in context. The primary result's confidence interval is not reported in the registry record. Without an estimate and its uncertainty, readers cannot judge whether the data are compatible with a broad range of effects or concentrated around a particular value. A p-value, if reported, would address compatibility with a specified null model; it would not by itself describe effect size or practical importance. A primary p-value is not reported.

Missing observations and analysis population. The registry explicitly limits the primary endpoint population to Stage 1 responders with completed week 24 viral-load assessments. This makes the endpoint population narrower than the full enrollment. In general, incomplete follow-up can affect an analysis if participants with missing measurements differ from those with observed measurements. Methods such as imputation, weighting, or sensitivity analysis may be considered in some studies, but the registry record does not report which, if any, were used. It also does not report the number of missing week 24 measurements or the reasons for missingness.

Multiplicity. The record lists 16 outcome measures and 36 analyses. When many outcomes or comparisons are examined, the chance of finding an apparently notable result by chance can increase if no adjustment or hierarchy is used. This is a general statistical concern, not a claim that a particular false-positive result occurred. The registry record does not report a multiplicity strategy, prespecified testing hierarchy, or adjustment method. The planned status of each analysis and the relationship among the listed tests are also not described in the record.

Sample-size illustrations. Any planning illustrations shown with the page are not observed power and are not the trial's sample-size justification. The stated illustration assumptions are a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect defined as a fraction of the observed effect. For non-time-to-event outcomes, the illustration is expressed as participants per arm for the stated power; time-to-event illustrations use total events. These assumptions describe the illustration, not necessarily the design decisions used for this study. The trial's own sample-size justification is not reported in the registry record. An illustration based on an observed effect is conditional on that effect and should not be interpreted as evidence that the trial had adequate power.

Model assumptions. Statistical models require assumptions suited to the outcome and design. For example, in general, over-dispersion occurs when outcome variability is greater than a simple model expects; failing to account for it can make uncertainty appear too small. This is a general concept and does not imply that over-dispersion applies to this trial. The registry record does not report the primary outcome model, its assumptions, diagnostic checks, covariate adjustments, or interim analyses. Those details should not be inferred from the fact that outcomes and analyses are listed.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 16 outcome measures and 36 analyses, including the primary viral-suppression measure and secondary outcomes relating to heroin-negative tests, ART adherence, HIV testing frequency, and opioid overdose. With multiple outcomes and comparisons, statistical interpretation depends partly on whether the analyses were prioritized in advance and whether any adjustment for multiple testing was planned. In general, testing many hypotheses without an appropriate strategy can increase the chance of apparently notable findings arising by chance. The registry record does not report a testing hierarchy or a multiplicity-adjustment procedure, so it is not possible to determine how this issue was handled.

The primary endpoint population is restricted to Stage 1 responders with completed week 24 viral-load assessments. The record does not report how many participants lacked the required assessment, why observations were missing, or how missing values were treated. It does not report whether the analysis included sensitivity analyses for missing outcomes. These omissions matter because the reported endpoint applies to participants meeting the stated response and completion criteria, not automatically to all enrolled participants.

The record also does not report a detailed analysis population definition for each of the 36 analyses, nor does it specify whether analyses followed initial random assignment or another rule. It does not describe interim analyses or stopping rules. These are absences in the registry record, not evidence that particular methods were or were not used outside the posted information. The primary result itself is not reported in the information, so neither missing-data handling nor multiplicity can be evaluated against an estimate and its uncertainty.

11. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: sent to compulsory drug rehabilitation programs (7.5%), incarceration (5.0%), hospitalization (2.5%).

The registry reports serious-event counts and other-event counts for eight intervention groups, along with deaths. For example, the high-intensity frontline group lists 328 participants at risk, 7 with serious events, 2 with other events, and 2 deaths; the low-intensity frontline group lists 337 at risk, 9 with serious events, 4 with other events, and 1 death. The later-stage groups have their own denominators and counts. These figures should be read as posted group-level counts, not as proof that an intervention caused an event.

The groups are not all interchangeable. They include frontline strategies as well as combinations involving SMS or Matrix interventions, and the trial is sequential. Later-stage groups may therefore represent participants at a different point in the study pathway. Comparing their counts directly as if they arose from one simple, fully randomized comparison can be misleading. The record does not report detailed event definitions, event-specific follow-up time, causal attribution, or a statistical comparison of safety outcomes.

The listed serious-event terms include being sent to compulsory drug rehabilitation programs, incarceration, and hospitalization, with percentages shown in the registry. Those labels describe reported events; they do not establish a mechanism or imply intervention-related harm. Death counts are also reported by group, but the record does not provide further context for attribution. Safety interpretation should remain descriptive unless the registry reports the relevant comparative analyses.

12. What the Result Does, and Does Not, Mean

This trial addresses an important statistical challenge: evaluating interventions that may change over time according to participant response. Its randomized, sequential structure and range of behavioral strategies make the study question more complex than a single comparison of two fixed treatments. The primary endpoint also concerns an objectively classified measure, HIV viral load, across baseline and week 24. The registry record therefore identifies a question that combines intervention stage, response status, and repeated categorical measurement.

However, the posted information does not include numerical primary outcome results. It does not show the distribution of viral-suppression categories, the estimated contrast between the named strategies, or an uncertainty interval. It cannot show whether any apparent difference is large or small, precise or uncertain. The fact that an endpoint was selected and defined does not tell us what the observed endpoint values were.

The primary endpoint population is limited to Stage 1 responders with completed week 24 assessments. Any interpretation must preserve that restriction. It should not be generalized automatically to all 667 enrolled participants, to participants who did not respond at Stage 1, or to those whose week 24 assessment was incomplete. The record does not report enough information about missing outcomes to assess how that restriction might affect interpretation.

The listed secondary outcomes and adverse-event counts add context about the scope of the study, but they do not fill the gap in primary efficacy results. Nor do the counts alone establish that an intervention caused or prevented a safety event. The registry record also does not report the analysis plan details needed to evaluate the handling of multiple comparisons, the sequential design, or missing data.

Thus, the defensible conclusion is limited: the record describes a completed randomized sequential trial and states what the primary endpoint was intended to measure, but the outcome contrast and its precision are not reported. This is a limit on what can be concluded from this record, not a conclusion that the strategies had no effect.

13. Why This Trial Matters Statistically

Statistically, this trial is notable because it brings together a sequential intervention structure, multiple behavioral strategies, and outcomes relevant to HIV and substance use. The registry describes frontline interventions followed by later options, with Stage 1 response informing the population considered for the primary analysis. This structure makes the estimand—the exact treatment effect a comparison is intended to estimate—especially important. A reader needs to know which participants, intervention pathways, and outcome time points the estimate represents.

The primary outcome also illustrates why a clear endpoint definition matters. Viral load is grouped into three categories, and status is considered at baseline and week 24, while an additional assessment is described at week 13. A report of category counts, transition patterns, and an appropriately defined contrast could answer different questions. The registry record identifies the measurement framework, but does not report a single numerical primary comparison in the information.

The trial also lists many analyses and several secondary outcomes. This breadth can be useful for understanding a complex program, while creating a need for clear prioritization and transparent multiplicity methods. Likewise, the responder-based analysis population makes reporting of missing assessments and the treatment of incomplete follow-up important. Those methods are not reported in the registry record.

Finally, the safety tables show why denominators and study stages matter. Counts from groups at different points in a sequential study do not automatically form directly comparable risks. The data are most useful when paired with clear event definitions, follow-up descriptions, and suitable comparative analyses. The posted information supplies a framework for asking these questions, but not all the estimates required to resolve them.

14. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The main limitation for interpreting this record is that numerical primary outcome results are not reported in the information. The endpoint has multiple categories and time points and is not reduced to one comparison, so the posted material does not support an estimate of the primary contrast or its precision. The registry record also does not report the analysis model, effect measure, confidence intervals, or p-values for that endpoint.

The primary analysis population is restricted to Stage 1 responders with completed week 24 viral-load assessments. Counts of eligible participants and missing assessments are not reported, and missing-data methods are not reported in the registry record. Without this information, readers cannot assess the extent of incomplete outcome data or how the restriction may shape interpretation.

The sequential design is only partly described in the record. It identifies frontline strategies and later intervention options, but does not report a complete decision algorithm, allocation probabilities, or the statistical method used to account for response-based pathways. The record also does not report the trial's sample-size justification, any interim analyses, or a multiplicity strategy, despite listing 36 analyses.

Safety counts are presented by group, but detailed event definitions, follow-up periods, attribution, and comparative analyses are not reported. Because the groups include both frontline and later-stage combinations, their counts should not be treated as directly comparable without further design and analysis information. These limitations concern what can be concluded from the registry record; they do not establish that unreported methods were not used elsewhere.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Screen, Treat and Retain Meth-using People With Opioid Use Disorders at Methadone Clinics — NCT04706624.
• Registry record last updated 2026-01-23; status last verified 2026-01.
• Results first posted 2026-01-23.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

18. Record Summary

The registry record describes a completed, randomized, unmasked, sequential trial with an enrollment of 667, conducted in Hanoi and Ho Chi Minh City. It evaluates behavioral intervention strategies for people with opioid-use disorder who use methamphetamine and receive methadone maintenance therapy. The two frontline approaches were 12 weeks of contingency management and 6 weeks of contingency management followed by 6 weeks of group education, with later options including SMS reminders and Matrix-based interventions.

The primary endpoint concerns viral-suppression status at baseline and week 24, among Stage 1 responders with completed week 24 viral-load assessments. The outcome uses three viral-load categories. Numerical primary results, the analysis model, and the handling of missing outcomes and multiple analyses are not reported in the registry record. Serious-event and other-event counts are listed by intervention group, but do not establish causation or a comparative safety effect.