Preventing Violence Among Veterans in Substance Use Disorder Treatment: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Preventing Violence Among Veterans in Substance Use Disorder Treatment (NCT01337973), a randomized na study. The primary outcome is Conflict Tactics Scale-Structured Interview (CTS-SI).

2. Clinical Question

The registry describes two related intervention questions: the impact of an integrated violence-prevention treatment during the early substance use disorder treatment phase, and the impact of adding continuing care after that phase. It also states an interest in whether combining early treatment and continuing care yields additional benefits. These aims do not identify one specific contrast as the only primary comparison, and the statistical analysis plan is not reported in the registry record.

The primary outcome description focuses on changes in rates of interpersonal aggression from the period before baseline to the follow-up period. Outcomes included overall physical aggression and injuring another person, with partner and nonpartner physical aggression and injury also described. The registry does not report the numerical results for these outcomes, so the observed direction and size of any differences cannot be characterized.

3. Trial Design

This was a randomized study with a listed enrollment of 180 and three behavioral-treatment arms. The registry identifies the study model as factorial and masking as single. It does not specify who was masked, how the factorial structure was implemented across the three named arms, or the exact comparisons assigned to each factor. Allocation details beyond the randomized designation are not reported in the registry record.

The MI-CBT arm received six sessions during the acute treatment phase, integrating motivational interviewing and cognitive behavioral approaches. The MI-CBT+CC arm received that acute-phase MI-CBT intervention followed by a 12-week phone-based continuing care counseling intervention. The E-TAU arm received enhanced treatment as usual, described as a brief session and provision of resources. The study summary places the early intervention in an 8-week early substance use disorder treatment phase and the continuing care intervention in the subsequent 3-month period.

The record lists the study phase as NA. The study status is completed. The recorded start date is 2012-06-20, and the primary completion date is 2016-12-01. The sponsor is the VA Office of Research and Development. These administrative details describe the registered study but do not establish how faithfully participants received their assigned interventions or how much follow-up information was available.

The primary outcome was CTS-SI interpersonal violence, assessed at baseline and at follow-up. The registry describes follow-up collections at 3 and 6 months for the prior 90 days and at 12 months for the prior 180 days. The stated outcome timeframe refers to the percentage difference between baseline and collapsed 3-, 6-, and 12-month follow-up data. Although the record describes the design and outcome schedule, it does not report enrollment by arm beyond the adverse-event risk counts, participant flow, adherence, or the analysis population for each result.

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: Of 1097 eligible for screening, 183 refused participation (17%), 75 (7%) were missed and 839 completed the screening assessment. Of those completing screening, 639 were not eligible for the RCT and 20 refused further participation. Thus, 180 out of 200 eligible veterans (90%) were randomized to one of the three treatment conditions.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Main analyses used General Estimating Equations which accounts for missing data. Results in tables below are from bivariate analyses.

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.

6. Primary Endpoint

The registry lists the primary outcome as Conflict Tactics Scale-Structured Interview (CTS-SI), measured over % difference between baseline and the collapsed 3-, 6-, and 12-month follow-up data.

The primary endpoint uses the Conflict Tactics Scale-Structured Interview (CTS-SI), a semi-structured interview assessing interpersonal violence. The record describes its content as violence severity, injury, and characteristics of interpersonal conflict incidents. It reports baseline collection covering the 180 days before enrollment, with follow-up collection at 3 and 6 months covering the prior 90 days and at 12 months covering the prior 180 days.

The stated outcome is a percentage change comparing baseline with collapsed follow-up information. The description says that means at baseline and follow-up were compared to produce a number without a measure of dispersion. It refers to monthly rates of different types of interpersonal aggression before baseline and after intervention across the follow-up period. Outcomes included overall physical aggression and injuring another person, as well as partner and nonpartner physical aggression and injury.

In general, a measure of interpersonal violence can capture outcomes that may not be represented by substance-use measures alone. A structured interview can organize information about the occurrence and characteristics of incidents. That general rationale does not establish the reliability of participants’ reports or show that the measure changed in this study. The registry record does not report the numerical endpoint estimates, their uncertainty, or a clear direction convention for interpreting a positive versus negative percentage change. The actual magnitude and precision of the observed changes therefore cannot be assessed from the reported endpoint description.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry describes randomized allocation, single masking, and a factorial study model. It names three behavioral-treatment arms, but does not give the factorial factors, assignment scheme, or planned contrasts. The distinction matters because a factorial analysis can estimate effects associated with separate intervention components and, where specified, whether their combination changes the effect. The arm descriptions alone do not establish which contrasts were planned or whether an interaction was tested. Those details are not reported in the registry record.

What does a bivariate analysis tell us?

In general, a bivariate analysis examines the relationship between two variables or compares outcomes between groups without adjusting for additional variables. It can describe an unadjusted association or contrast, but it does not by itself account for potential differences in other measured characteristics. The registry explicitly says that the results listed were from bivariate analyses. It does not provide the underlying models, estimates, or covariates, and therefore does not support a more detailed account of those analyses.

The registry also says that the main analyses used General Estimating Equations. In general, GEE is a method for estimating average patterns when observations are correlated, such as repeated measurements from the same participant. It can account for within-participant correlation through a working correlation structure while estimating population-average effects. The record states that GEE accounts for missing data, but does not identify the missing-data procedure, model specification, correlation structure, assumptions, or results. The statement should not be taken to mean that all missing information was recovered or that the estimates are unaffected by missingness.

Why do repeated measurements matter?

In general, repeated measurements from one participant are not independent observations. A person’s reports at different follow-up times can be related to one another. Statistical methods that account for this relationship can produce uncertainty estimates suited to the repeated-measures structure. The CTS-SI schedule included baseline and follow-up assessments at 3, 6, and 12 months. The registry says the outcome comparison collapsed those follow-up data, but does not explain how the observations were combined in the main GEE analyses or how participants with incomplete follow-up contributed.

The endpoint description reports means compared across baseline and follow-up and says the resulting number has no measure of dispersion. In general, dispersion describes variability in observations, while uncertainty measures describe the precision of an estimated quantity. A mean or percentage-change estimate without either kind of accompanying information provides limited evidence about variability and precision. No confidence interval or standard error for the primary endpoint is reported in the registry record.

What can a confidence interval add?

In general, a confidence interval gives a range of values compatible with an estimate under a specified statistical procedure and its assumptions. It communicates precision as well as the estimate’s direction and scale. It does not guarantee that the true value lies in the particular interval, and it does not resolve bias from study conduct or analysis choices. No confidence interval for the primary outcome is reported in the registry record, so the precision of the estimated change cannot be assessed.

The record lists 30 analyses and two outcome measures. In general, conducting multiple analyses creates more opportunities for apparently notable findings by chance, particularly when many outcomes or contrasts are considered. Methods such as prespecification, hierarchy of outcomes, or multiplicity adjustment can address different aspects of this issue. The registry record does not report whether any such method was planned or applied, and the analysis count alone does not identify which findings were considered primary.

A sample-size calculation requires assumptions about the target effect, variability, significance level, power, allocation, and analysis. The registry record does not report its original sample-size justification or the assumptions used to plan enrollment. Any sample-size illustrations associated with observed effects would be planning illustrations, not evidence that the completed trial had a particular probability of detecting an effect. The record does not report enough outcome information to evaluate such an illustration.

Finally, interpretation depends on the analysis population and missingness. In general, an analysis based on participants with observed outcomes can differ from an analysis that retains randomized participants under an explicit missing-data approach. The registry says GEE was used in main analyses to account for missing data but does not define the populations, give missingness counts, or state whether an intention-to-treat principle was followed. These omissions prevent assessment of how robust the reported comparisons were to incomplete outcome data.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 30 analyses and two outcome measures. The primary outcome description includes several types of aggression, and the study summary also names substance use and health services use as outcomes of interest. Multiple outcomes, follow-up times, and possible arm contrasts can create a multiplicity problem: the more comparisons considered, the greater the chance that some estimates may appear noteworthy by chance. In general, prespecifying a hierarchy or adjusting inference can help manage this issue. Whether either approach was used is not reported in the registry record.

The record identifies the CTS-SI as the primary outcome and lists percent change in the percentage of days of substance use for each substance as a secondary outcome. It does not specify how the 30 analyses map to the outcome families, time points, or intervention contrasts. It also does not report which comparisons were primary for inference or whether the analysis plan was finalized before outcomes were examined. These details are needed to distinguish confirmatory analyses from exploratory ones.

The registry says the main analyses used GEE, which it describes as accounting for missing data, and that the listed results were bivariate analyses. The extent and pattern of missing follow-up information, the assumptions made about missingness, and the method used to handle it are not reported in the registry record. The randomized analysis population, exclusions, and participant flow are also not reported. Without these details, readers cannot determine whether the missingness approach depended on assumptions that were plausible for the study.

The record does not state whether analyses followed an intention-to-treat principle or whether participants were analyzed according to treatment assignment. It likewise does not report a per-protocol population or other analysis-set definitions. These gaps make it difficult to assess the effect of missing observations, departures from assigned treatment, or multiple comparisons on the interpretation of the results.

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

Serious events with the highest arm-level rate: participant death (3.3%).

The adverse-event summary lists 60 participants at risk for serious adverse events in each arm. It reports serious adverse events affecting 1 participant in the MI-CBT arm, 2 participants in the MI-CBT+CC arm, and 1 participant in the E-TAU arm. Other adverse events affected 0 participants in each arm. These are counts as recorded, not comparisons adjusted for follow-up duration or other factors.

The registry lists deaths of 1 participant in MI-CBT, 2 in MI-CBT+CC, and 1 in E-TAU. It separately identifies Participant Death as the top serious-event term, with 3.3% reported. The record does not give enough detail to attribute deaths to an intervention, establish their timing, describe causes, or determine whether they were considered related to treatment. The counts should not be interpreted as evidence of causation or as a complete account of safety.

In general, adverse-event counts describe events observed and recorded during a study; they do not by themselves establish whether an intervention caused an event. Differences in counts can also be difficult to interpret without information about exposure, follow-up completeness, event definitions, and review procedures. Those details are not reported in the registry record. The listed figures therefore support only a descriptive reading of the recorded counts.

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

The registry record establishes that the study evaluated three behavioral-treatment arms in a randomized study of veterans receiving substance use disorder treatment. The intervention descriptions distinguish acute-phase MI-CBT, MI-CBT with a subsequent continuing care intervention, and enhanced treatment as usual. The primary outcome was described as percentage change in CTS-SI measures of interpersonal aggression, comparing baseline with collapsed follow-up information.

The record does not report numerical primary-outcome estimates, uncertainty intervals, or the results of the arm comparisons. It therefore does not support a claim that one intervention reduced aggression more than another, that the interventions had no effect, or that the arms were equivalent. Absence of numerical findings in the record is not evidence that the findings were null. The direction convention for the percentage-change measure is also not stated clearly enough to interpret a hypothetical value without further information.

The same caution applies to the study’s broader interests in substance use and health services use. The registry names these outcomes but does not include their numerical results in the record described. Adverse-event counts are available, but they are descriptive and do not establish whether an event was caused by a study intervention.

In general, a study’s results are interpreted by considering the estimate, its uncertainty, the analysis plan, missing data, multiplicity, and the population to which the findings may apply. Important details on these topics are not reported in the registry record. The enrollment of 180 participants and the randomized design alone cannot substitute for outcome estimates or show how precisely an effect was measured.

The appropriate conclusion is limited: the registry describes a trial designed to investigate violence, substance use, and service-use outcomes across three behavioral arms, but the posted information does not provide enough numerical outcome evidence to determine what effects were observed. It should not be read as a clinical recommendation or as evidence for or against any intervention.

13. Why This Trial Matters Statistically

This trial raises a statistical question that is broader than comparing one treatment with one control. It includes early MI-CBT, MI-CBT combined with continuing care, and enhanced treatment as usual, while the registered model is described as factorial. Evaluating separate intervention components and their combination can require clearly specified contrasts and an analysis that matches the design. The registry does not report those details, so the study’s structure highlights why a design label alone is not enough to reconstruct the statistical question.

The primary outcome also involves repeated assessments and several types of interpersonal aggression. Collapsing follow-up information may simplify interpretation, but it can obscure how outcomes vary across time unless time-specific results are also available. In general, repeated-measures analyses must address dependence among observations from the same participant. The registry’s mention of GEE indicates attention to this issue in the main analyses, but the missing model details and absence of estimates limit evaluation of the method’s implementation.

Multiplicity is another important issue because the record lists 30 analyses and two outcome measures, alongside multiple aggression categories and broader study aims. Prespecified outcomes and contrasts help readers distinguish planned tests from exploratory comparisons. The registry does not report a multiplicity strategy or the mapping between analyses and hypotheses.

Finally, the distinction between the main GEE analyses and listed bivariate results matters. Unadjusted comparisons can be informative descriptions, but they may not answer the same question as the primary model. Reporting effect estimates with uncertainty, the analysis population, missing-data handling, and multiplicity decisions would allow readers to judge both statistical precision and the connection between the analyses and the design. The record’s limited numerical results make these reporting features particularly consequential.

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 central limitation for interpreting this study from the registry record is that numerical primary-outcome estimates and uncertainty measures are not reported. The endpoint description says means were compared and the result was a number without a measure of dispersion. Without numerical arm contrasts, standard errors, confidence intervals, or another uncertainty measure, the direction, magnitude, and precision of the primary outcome cannot be assessed.

The outcome description combines follow-up assessments at 3, 6, and 12 months, but the method used to collapse those data is not detailed. The measure includes several aggression categories, yet the registry record does not provide the results for each category or clarify how the percentage-change direction should be read. This limits interpretation of both the overall measure and the specific forms of aggression described.

Although the study is identified as factorial, the factors, allocation structure, and planned comparisons are not reported. Masking is described as single, but the masked party is not identified. Details of the GEE model and the bivariate analyses are also absent. These gaps prevent readers from assessing whether the analyses align with the design and the stated study aims.

The record does not report a sample-size justification, missing-data quantities or assumptions, analysis population definitions, or a multiplicity strategy. It lists 30 analyses, but does not explain how these relate to the outcomes, time points, or contrasts. Numerical results for the broader substance-use and health-services aims are not reported in the registry record. Finally, adverse-event counts are not accompanied by details on event timing, causes, attribution, or completeness of follow-up. These limitations restrict the conclusions that can be drawn from the registered information.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Preventing Violence Among Veterans in Substance Use Disorder Treatment — NCT01337973.
• Registry record last updated 2019-09-06; status last verified 2019-07.
• Results first posted 2019-09-06.
• 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

NCT01337973 is a completed randomized study with a listed enrollment of 180 and three behavioral arms: MI-CBT, MI-CBT+CC, and E-TAU. The registry describes the design as factorial and single-masked. The recorded study start date is 2012-06-20, and the primary completion date is 2016-12-01. The sponsor is the VA Office of Research and Development.

The primary outcome was CTS-SI percentage change in measures of interpersonal aggression, comparing baseline with collapsed follow-up information. The registry says main analyses used GEE to account for missing data and that the listed results were bivariate analyses. Numerical primary-outcome findings and measures of uncertainty are not reported in the registry record. The record lists two outcome measures and 30 analyses, but does not describe a multiplicity approach, sample-size justification, or analysis populations. Serious adverse-event counts are reported by arm, but the details needed for causal interpretation are not reported.