CBT-I on Alcohol Treatment Outcomes Among Veterans: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for CBT-I on Alcohol Treatment Outcomes Among Veterans (NCT03806491), a randomized na study. The primary outcome is Insomnia Severity.

2. Clinical Question

The question concerns whether adding a structured insomnia-focused behavioral intervention to usual alcohol treatment is associated with a different course of measured insomnia severity than adding one session of sleep-hygiene education. The trial description also identifies feasibility, acceptability, and initial efficacy as aims. Those aims are not interchangeable: feasibility and acceptability concern whether the intervention can be delivered and received, while an efficacy comparison concerns outcome differences between arms. Numerical findings for those aims are not reported in the registry record.

The clinical question is limited to this trial’s population, interventions, comparator, and measurement schedule. The record does not provide enough detail to characterize which Veterans were eligible, how participants were recruited, or how adherence was assessed. It also does not provide outcome estimates that would show the direction or precision of any arm difference.

3. Trial Design

The registry record describes a randomized, double-masked, parallel-group trial with an enrollment of 67. The study is marked completed, with a recorded start date of 2019-07-15 and primary completion date of 2022-07-31. The phase is listed as not applicable. The sponsor is University of Missouri-Columbia.

Participants were assigned to one of two behavioral-treatment arms. The CBT-I + AUD-TAU arm received individual CBT-I once a week for five weeks alongside usual Alcohol Use Disorder treatment. The Sleep Hygiene + AUD-TAU arm received sleep-hygiene education once alongside usual treatment. The record identifies both added interventions as behavioral. It does not describe the content of AUD-TAU, the methods used to deliver the sleep-hygiene education, or the precise procedures used to maintain masking. Which people were masked is not reported in the registry record.

The primary outcome was insomnia severity, measured on a continuous score scale using the ISI. Its stated assessment period runs from baseline to post-treatment at week 6 and follow-up at week 12. The record also names secondary outcomes involving sleep efficiency, mental-health symptoms, treatment-related learning, alcohol use related to sleep, craving, and negative affect. Specific instruments, assessment schedules for those outcomes, and results are not reported in the registry record.

The randomized, parallel design supports a comparison between assigned groups in principle, but the design description alone does not establish how the analysis handled adherence, missing measurements, or participants who discontinued. Those analysis details are not reported in the registry record. No statistical analysis is listed, and the primary outcome was not reduced to one comparison because it includes multiple time points.

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.

Overall Study

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 Insomnia Severity, measured over Baseline to post-treatment (week 6) to follow up (week 12).

The primary endpoint is insomnia severity assessed using the Insomnia Severity Index. The record describes the ISI as a seven-item measure of insomnia severity over the past two weeks. Its items address difficulty falling or staying asleep, satisfaction with the current sleep pattern, interference with daily functioning, whether others notice sleep problems, and worry or distress related to sleep problems. Each item has response options from 0 to 4, and item scores are summed. The highest possible total is 28. Higher scores indicate more severe insomnia; participants scoring 10 or higher are classified as screening positive for insomnia.

A continuous score can retain more gradation in symptom severity than a classification alone. In general, this can help describe changes across the range of measured scores, while a screening threshold identifies a specified category. The registry record defines both the total-score interpretation and the screening threshold, but does not report how many participants met that threshold or how scores changed.

The stated primary measurement schedule includes baseline, post-treatment at week 6, and follow-up at week 12. Assessing the measure at these stages can describe both change during the treatment period and later status. However, the registry record gives no arm-specific means, standard deviations, effect estimates, or uncertainty intervals for the time points. It therefore does not establish the magnitude, direction, or precision of any difference in insomnia severity.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Random allocation is intended to make treatment assignment independent of participants’ measured and unmeasured characteristics, on average. In general, this supports a fair comparison between assigned groups, although chance differences can remain, particularly in a trial with a modest enrollment. The registry record labels this trial randomized but does not report the randomization sequence, allocation concealment procedures, or any baseline characteristics with which to assess the enrolled groups.

The parallel-group structure means that participants were assigned to one of the two intervention arms rather than crossing between interventions. The record describes the trial as double-masked, but does not say which participants, clinicians, assessors, or analysts were masked, nor how masking was maintained. These details are relevant because expectations and knowledge of treatment assignment can affect reported symptoms or their assessment. The label alone does not clarify the practical masking procedures.

The primary outcome is a continuous score. In general, a mean summarizes the average score in a group, while a standard deviation describes the spread of individual scores around that mean. A group mean can communicate the average level of measured insomnia severity, but it does not show the full distribution or the experience of every participant. The record specifies mean and standard deviation as the outcome parameter and dispersion, but does not report the corresponding values.

Repeated measurements require an analysis that accounts for the fact that measurements from the same participant are related. In general, treating such measurements as independent can misrepresent uncertainty. A statistical plan might estimate change over time, compare groups at a selected assessment, or estimate differences in trajectories. Each choice defines a different target effect. The registry record does not report the model, contrasts, covariance assumptions, or handling of within-participant correlation.

A confidence interval is a general way to express uncertainty around an estimated effect. Its width depends on factors including the amount of information in the data and the variability of measurements. A narrow interval indicates greater precision under the model assumptions; a wide interval indicates less precision. An interval that includes no difference does not, by itself, demonstrate equivalence. No confidence interval for the primary outcome is reported in the registry record, so the precision of an effect cannot be assessed.

Multiplicity arises when investigators assess multiple outcomes, time points, or comparisons. In general, conducting many statistical tests can increase the chance of at least one apparently unusual result even when there is no true effect. Approaches to address this include prespecifying a primary contrast, adjusting inference, or clearly labeling analyses as exploratory. The record lists one primary outcome with multiple assessment times and several secondary outcomes, but does not report a multiplicity strategy or which contrasts were prespecified.

Over-dispersion is a general concern when observed variability is greater than a statistical model assumes. For continuous outcomes, an incorrect variance model can make uncertainty estimates unreliable. More broadly, model assumptions about the distribution and correlation of measurements should fit the data. The registry record does not report diagnostic checks, residual assessments, or any variance model for this trial, so there is no basis to describe how these issues were evaluated.

The stated sample-size illustration uses a two-sided significance level of 0.05, equal allocation, and a normal approximation. It describes planning for a stated fraction of an observed effect, rather than reporting an empirical result. In general, a planning calculation asks what sample size might be needed under a specified effect and assumptions. It is not “observed power,” and it does not establish that the trial was designed using those assumptions. The sample-size justification and the trial’s own design calculations are not reported in the registry record.

Missing outcome data can affect both the estimate and its uncertainty. In general, the implications depend on why data are missing, how much is missing, and the assumptions used by the analysis. Methods may use available measurements, imputation, or models that make assumptions about missingness; no one approach can be inferred from the trial design label. The registry record does not report missingness by arm or time point, reasons for missing data, or the method used to address it.

Finally, an analysis population describes which participants contribute to a particular estimate. An intention-to-treat approach generally analyzes participants according to randomized assignment, while other approaches may depend on treatment received or data availability. These choices can answer different questions. The registry record does not state which analysis population was used, and no statistical analyses are listed. Consequently, no estimate can be attributed to a particular analysis strategy.

10. Multiplicity, Missing Data and Analysis Populations

The registry record names insomnia severity as the primary outcome and lists secondary outcomes relating to sleep, mental-health symptoms, alcohol-related measures, treatment-related learning, and negative affect. The primary outcome itself has assessments at baseline, post-treatment at week 6, and follow-up at week 12. Multiple outcomes and time points create choices about which comparisons are primary and how to interpret additional tests. A prespecified analysis plan can clarify those choices, but such a plan and any multiplicity adjustment are not reported in the registry record.

No statistical analyses are listed in the record. It does not report which primary contrast was tested, whether secondary outcomes were treated as confirmatory or exploratory, or whether subgroup analyses were conducted. The primary endpoint has multiple time points and is not reduced to one comparison in the posted results. Thus, there is no reported collection of tests or estimates from which to assess multiplicity in practice.

The record also does not report the number of participants with outcome data at each assessment, the reasons for missing measurements, or how missing data were handled. These details matter because the observed participants may not represent all randomized participants if missingness is related to treatment assignment or outcome. The analysis population, including whether participants were analyzed according to random assignment, is not reported in the registry record.

These omissions limit assessment of both bias and precision. They do not establish that data were missing or that analyses were inappropriate; they mean the registry record does not describe those matters. Without analysis-population definitions, missing-data information, and a stated multiplicity approach, readers cannot determine how the reported outcome results—none of which are numerically presented for the primary endpoint—would relate to the original randomized groups.

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.

The registry record reports zero serious adverse events in the CBT-I + AUD-TAU group, among 32 participants at risk, and zero in the Sleep Hygiene + AUD-TAU group, among 35 participants at risk. It also records zero other adverse events in each group, with 32 and 35 participants at risk, respectively. Deaths are recorded as zero in both groups. No serious adverse-event terms are listed.

These figures describe what the registry record reports for the groups and event categories shown. They do not establish that no adverse experiences of any kind occurred, because the record’s categories and reporting scope may not capture every experience. The record does not provide event definitions, ascertainment procedures, timing, severity grading beyond the serious-event category, or details about relatedness to an intervention. It also does not explain whether the at-risk counts represent all randomized participants throughout follow-up.

Zero recorded events should be interpreted as a count in this registry record, not as proof that the interventions have no risk. The record does not provide enough detail to assess how comprehensively events were monitored or reported. These safety figures are separate from the primary insomnia-severity outcome and do not answer the trial’s comparative outcome question.

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

The record shows that the study was designed to compare two behavioral approaches added to usual Alcohol Use Disorder treatment: individual CBT-I delivered weekly for five weeks and sleep-hygiene education delivered once. It identifies insomnia severity, measured with the ISI, as the primary outcome and sets out assessments from baseline through week 6 and week 12. It also records an enrollment of 67 and reports no serious adverse events, other adverse events, or deaths in the listed groups.

What the record does not show is a numerical comparison of insomnia severity between arms or over time. There are no reported arm-specific scores, estimated differences, confidence intervals, or p-values for that endpoint. Consequently, the posted information does not establish the direction, size, or precision of a treatment effect. It also cannot establish that the groups had equivalent outcomes or that a difference was absent.

The design description and outcome definitions indicate what the investigators intended to examine; they are not themselves evidence of an outcome. The study’s stated aim includes feasibility, acceptability, and initial efficacy, but numerical findings for those aims are not reported in the registry record. The listed secondary outcomes likewise do not indicate whether any measured outcome changed.

The reported adverse-event counts should be read within their stated categories and risk counts. They are not a comprehensive description of every possible harm, and their interpretation is constrained by the absence of event definitions and monitoring details. Similarly, enrollment is not the same as the number contributing complete measurements at each time point.

In general, a clinical-trial result is most informative when the estimated effect, its uncertainty, the analysis population, and the treatment of missing data are reported together. Those elements are not available for the primary outcome in the registry record. The appropriate conclusion is therefore limited: the record describes a randomized comparison and its planned outcome measurements, but does not report the numerical primary-outcome result needed to quantify the comparison.

13. Why This Trial Matters Statistically

This trial raises a useful statistical question about how a behavioral intervention is evaluated when the primary outcome is measured repeatedly. Insomnia severity is assessed at baseline, after treatment at week 6, and at follow-up at week 12. Those measurements can describe more than one feature of change: the starting score, change during treatment, and status at follow-up. The target comparison must be clear because different choices can yield different estimates and interpretations.

The comparison also illustrates why an active behavioral comparator matters to interpretation. The two arms both include usual Alcohol Use Disorder treatment, while the added intervention differs: a course of CBT-I in one arm and a single sleep-hygiene education session in the other. Any estimated contrast would therefore concern these treatment packages as assigned, not CBT-I in isolation from all other care. The record does not report an outcome estimate that would quantify this contrast.

Randomized assignment can support a comparative analysis, but its value for inference depends on transparent reporting of outcomes, uncertainty, analysis populations, and missing data. Without those elements, enrollment and design labels cannot reveal the precision or robustness of the estimated effect. This is especially important when an endpoint includes several measurements and there are additional secondary outcomes.

The registry record also gives a chance to distinguish safety counts from efficacy results. Zero recorded events in the listed categories is useful descriptive information, but it does not answer whether the arms differed in insomnia severity. Nor does it resolve questions about the completeness of event monitoring.

Statistically, the central lesson is that a trial’s question, design, and outcome definition are not substitutes for its numerical results. A transparent account of the planned contrast and its uncertainty is needed to understand what the data say. Those outcome analyses are not reported in the registry record, so the comparison remains unquantified.

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 most consequential limitation for interpretation is that no numerical result for the primary outcome is reported. The registry record describes insomnia severity and its assessment schedule but does not provide group scores, between-group estimates, or uncertainty measures. The primary endpoint spans multiple time points and has not been reduced to one comparison, leaving the principal estimand unclear.

Analysis details are also limited. No statistical analyses are listed, and the record does not report the model, contrasts, adjustment for baseline values, multiplicity approach, or analysis population. It does not state how missing measurements were handled or how many participants contributed data at each assessment. These omissions prevent assessment of whether estimates, if available, would represent all randomized participants or a subset.

The record gives little detail about participant eligibility and characteristics beyond the trial’s focus on Veterans with Insomnia and Alcohol Use Disorder. Further population description, recruitment procedures, and baseline clinical characteristics are not reported in the registry record. This limits assessment of how broadly the study population might represent other settings or groups.

The intervention descriptions identify session frequency and the broad behavioral components, but not treatment fidelity, adherence, clinician training, or the content of usual Alcohol Use Disorder treatment. The record labels the study double-masked but does not identify who was masked or how masking was maintained. These details constrain interpretation of how consistently the interventions were delivered and how outcomes were assessed.

Safety reporting is limited to counts in the stated categories and group risk counts. Event definitions, collection procedures, follow-up coverage, and event-level descriptions are not reported in the registry record. Accordingly, zero recorded events in those categories should not be treated as a full account of all potential adverse experiences. These limitations do not establish that the trial was poorly conducted; they define what can be evaluated from the registry information.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: CBT-I on Alcohol Treatment Outcomes Among Veterans — NCT03806491.
• Registry record last updated 2025-04-18; status last verified 2025-04.
• Results first posted 2025-04-18.
• 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

NCT03806491 is a completed, randomized, double-masked, parallel-group trial sponsored by University of Missouri-Columbia. The registry record reports an enrollment of 67 and lists the trial as not applicable for phase. Its recorded start date is 2019-07-15, and its primary completion date is 2022-07-31.

The trial compares CBT-I plus usual Alcohol Use Disorder treatment with sleep-hygiene education plus usual treatment. The primary outcome is insomnia severity measured by the ISI from baseline through week 6 and week 12. Secondary outcomes include sleep efficiency, mental-health symptoms, alcohol-related measures, treatment-related learning, and negative affect.

The registry record does not report a single numerical comparison for the primary outcome, and no statistical analyses are listed. It reports zero serious adverse events, other adverse events, and deaths in each group. Analysis methods, missing-data handling, and analysis populations are not reported in the registry record.