Service Dog Training Program for Military Veterans With PTSD: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Service Dog Training Program for Military Veterans With PTSD (NCT03777020), a randomized na study. The primary outcome is Post Traumatic Stress Symptom Severity.

2. Clinical Question

The registry describes a comparison between two forms of dog-training-related activity, rather than a comparison with no dog-related training. One group attended weekly sessions with a trainer and service dog; the other received online training education. The stated primary measure is post-traumatic stress symptom severity, with higher scores indicating worse PTSD symptoms. The registry lists baseline and 8 weeks as the assessment time points.

This question concerns the difference between the assigned programs as delivered in this trial. It does not, by itself, isolate which component of either program might account for any observed score change. For example, the design as described does not separate the effects of in-person contact, trainer interaction, dog interaction, or the training content. No numerical primary-outcome comparison is reported in the registry record, so the magnitude and precision of any difference between the groups cannot be characterized from that record.

3. Trial Design

This was a randomized, parallel-group trial sponsored by the University of Maryland, Baltimore. The registry lists enrollment of 60 and describes two groups: the Service Dog Training Program (SDTP) and Dog Training Education. The posted adverse-event information lists 30 participants at risk in each group. The trial status is completed. Its start date is 2019-08-22 and its primary completion date is 2023-02-28. The phase is listed as NA.

Participants assigned to the SDTP were to be paired with an experienced Warrior Canine Connection Mission Based Trauma Recovery trainer and a service dog. The described schedule was one-hour training modules once a week for 8 weeks. Participants assigned to Dog Training Education were to take one-hour online service-dog training modules once a week for 8 weeks. The online modules were described as being delivered by experienced service-dog trainers and as using parallel content to the SDTP.

The registry identifies the allocation as randomized and the model as parallel: participants were assigned to one of the two groups, rather than crossing between interventions as part of the design. Masking is listed as none. Thus, the record does not describe masking of participants or other parties. It does not report further details about sequence generation, allocation concealment, adherence, treatment fidelity, or departures from assigned groups. Those details should not be inferred from the randomized designation alone.

The study description frames the service-dog program as a rehabilitative intervention being evaluated for veterans with PTSD. The registry lists 17 outcome measures and several secondary domains, including stress biomarkers, perceived stress, positive affect, health-related quality of life, and resilience. The available record does not report a single numerical primary-outcome comparison or analyses for those measures.

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

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Veterans with PTSD

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: One participant in the dog training education group no longer met inclusion criteria at first visit and was removed from the study

6. Primary Endpoint

The registry lists the primary outcome as Post Traumatic Stress Symptom Severity, measured over Baseline, 8 weeks.

The primary endpoint is post-traumatic stress symptom severity, measured using the PTSD Checklist for DSM-5. The registry describes the score as ranging from 0 to 80 and states that higher scores indicate worse PTSD symptoms. The measure is listed at baseline and 8 weeks, giving a starting assessment and a later assessment during the intervention period.

A continuous symptom score can describe the degree of reported symptoms rather than sorting participants only into categories. In general, comparing scores over time can help assess whether the groups’ symptom measurements changed differently during follow-up. Such an interpretation depends on the specific comparison and analysis used; the registry record does not report a single reduced comparison for this outcome.

The endpoint is relevant to the stated question because it measures PTSD symptom severity, the central clinical outcome named for the trial. It should be read as a score on the specified checklist, not as a complete account of functioning or overall health. The registry also lists psychosocial and biomarker outcomes, but those are separate measures and do not replace the primary symptom endpoint. Because the record does not provide a numerical primary-outcome estimate, it is not possible to describe the observed score level, change, between-group difference, or precision from the registry information.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies the allocation as randomized, the design as parallel, and masking as none. It lists a continuous primary measure summarized by a mean and standard error. The statistical concepts below are general explanations; the registry record does not report a detailed statistical analysis plan or the analysis used for the primary endpoint.

What random allocation contributes

In general, random allocation assigns participants to groups using a chance-based process. Its purpose is to reduce systematic differences between groups at assignment, including differences that may not be known to investigators. It does not guarantee that every characteristic will be identical between groups, especially in a trial with a limited number of participants. Nor does the randomization label explain how the sequence was generated or concealed. Those implementation details are not reported in the registry record.

A parallel-group design assigns participants to one group for the trial rather than having each participant receive both programs in sequence. The resulting comparison is between groups. In general, an analysis should respect the randomized assignment when estimating the effect of being assigned to a program. The registry record does not state which analysis population was used or whether the primary analysis followed assignment regardless of participation.

Means, standard errors, and confidence intervals

A mean is an average of participant scores. It describes the group’s central score but does not show how much individual scores vary around that average. The standard error describes the statistical precision of an estimated mean; it is not the same quantity as the standard deviation, which describes spread among individual observations. The primary endpoint is listed with a mean and standard error, but the numerical values are not reported as a single primary comparison in the registry record.

In general, a confidence interval expresses uncertainty around an estimate by giving a range of values compatible with the observed data and the method’s assumptions. A narrow interval indicates greater precision than a wide interval, although the interval alone does not determine clinical importance. A p-value addresses how compatible the observed data are with a specified null hypothesis under an analysis model. It is not the probability that a treatment works, and it does not measure the size or practical importance of an effect. The registry record does not report a confidence interval or p-value for one selected primary-outcome contrast.

Repeated measurements and the comparison of interest

The primary endpoint is listed at baseline and 8 weeks. In general, repeated measurements can be analyzed in different ways. Researchers may compare scores at follow-up while accounting for baseline, compare changes from baseline, or use a model that represents measurements over time. These approaches may target different estimands and can rely on different assumptions. The record does not say which approach was used, so no particular model or contrast should be attributed to this trial.

When an outcome has multiple time points or categories, choosing the primary contrast is important. A result for one time point is not automatically interchangeable with a result for change over time. The registry specifically notes that the primary outcome was not reduced to one comparison because it has several categories or time points. That makes it inappropriate to infer a single estimate from the endpoint description alone.

Planning calculations and observed results

The planning illustration is described as using the observed effect, two-sided significance level 0.05, equal allocation between the two arms, and a normal approximation. It considers the stated fraction of the observed effect as the effect to detect. These are planning assumptions, not evidence about the trial’s actual design justification. The registry record does not report the sponsor’s sample-size justification or a planned target effect in this information.

Such calculations are general planning tools: they illustrate how a hypothetical effect size and specified assumptions relate to the number of participants needed for a chosen power target. They do not represent observed power, and they do not change the uncertainty in the observed result. In this case, the record does not give a single numerical primary effect estimate to interpret alongside the planning illustration. The calculation assumptions should therefore not be read as a substitute for a reported treatment-effect estimate.

For continuous data, general modeling considerations include the shape and spread of the outcome distribution, the relationship between baseline and follow-up measurements, and whether observations are independent in the way the model assumes. If data are more variable than a model allows, uncertainty estimates can be misleading; this is one reason distributional and model assumptions matter. The registry record does not report diagnostic checks, model details, or any adjustment for over-dispersion. No specific problem with the trial’s data can be inferred from the absence of those details.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 17 outcome measures. The secondary outcomes include cortisol, alpha amylase, IgA, heart-rate variability, perceived stress, positive affect, health-related quality of life, and resilience. The primary outcome is listed at baseline and 8 weeks, and the record notes that the outcome was not reduced to one comparison because it has several categories or time points. It lists zero analyses.

In general, testing multiple outcomes, time points, or contrasts increases the chance of obtaining at least one apparently notable result by chance if each test is interpreted separately. A prespecified hierarchy or adjustment procedure can help address this issue. The registry record does not report a multiplicity strategy, a hierarchy among outcomes, or which time-point contrast was intended to be primary for analysis. The number of listed measures alone does not show which outcomes were tested or how results were interpreted.

The record also does not report how missing outcome data were handled, whether participants were analyzed in their assigned groups, or the criteria for inclusion in an analysis population. These decisions can affect estimates when outcome data are incomplete or when participants do not follow their assigned program. Without those details, the reader cannot assess the assumptions behind any such analysis. No missing-data method, imputation procedure, or sensitivity analysis is reported in the registry record.

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 posted adverse-event information lists 30 participants at risk in each group. In the Service Dog Training Program group, the record lists 0 serious adverse events, 0 other adverse events, and 0 deaths. In the Dog Training Education group, it likewise lists 0 serious adverse events, 0 other adverse events, and 0 deaths. No top serious-event terms are listed.

These figures describe what is recorded in the registry, not a complete account of every possible safety outcome or the method used to collect events. In general, a zero count in a posted category means no events are listed for that category in the record; it does not establish that no event of any kind occurred or that an intervention is free of risk. The registry information does not describe event definitions, surveillance procedures, duration of safety follow-up, or whether all participants completed safety assessment. The figures should therefore be read as reported counts, without using them to draw broader conclusions about safety.

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

The registry record establishes that a randomized comparison was planned between an in-person service dog training program and online dog training education for veterans with PTSD. It identifies PTSD symptom severity on the PTSD Checklist for DSM-5 as the primary outcome, with assessment at baseline and 8 weeks. It also lists psychosocial and stress-related secondary measures. These details clarify the research question and the kinds of outcomes the study set out to assess.

The record does not report a single numerical primary-outcome comparison. It therefore does not show the size or direction of a difference between groups, how precisely such a difference was estimated, or whether any reported contrast was compatible with no difference. It cannot support a conclusion that the programs had the same effect, different effects, or a particular degree of change in symptoms. Those statements require numerical results and their uncertainty.

Even when results are available, an average score difference would describe a group-level comparison, not predict what will happen for an individual participant. In general, interpretation also depends on the chosen contrast, missing data, adherence, and the analysis population. The registry record does not provide those details for the primary endpoint. The listed safety counts are similarly limited to the categories and counts recorded; they do not answer the efficacy question.

The comparison is between two active training-related approaches as described, not between the SDTP and no program. Any estimate from such a comparison would address the relative outcomes under those assigned programs and trial conditions. It would not by itself identify the contribution of each program component or establish how results apply to settings and populations not represented in the trial. The posted information supports describing the design and outcome measures, but not a claim about treatment benefit or lack of benefit.

13. Why This Trial Matters Statistically

This trial poses a statistical question that is more specific than whether symptom scores changed during participation. Because both groups were assigned dog-training-related activities, the central comparison is between the service-dog program and online dog-training education. A within-group change alone would not answer that comparative question. The relevant effect would need to compare outcomes between randomized groups using a clearly defined time point or change measure.

The primary endpoint is continuous and is recorded at baseline and 8 weeks. That structure makes the selection of the analysis contrast important: a follow-up score comparison, a change-from-baseline comparison, and a repeated-measures model are not automatically identical. The registry notes multiple categories or time points and does not reduce the endpoint to one comparison. This makes the absence of a clearly posted estimate especially consequential for interpretation.

The registry also lists 17 outcomes across symptom, psychosocial, and biomarker domains. In general, a broad outcome set can provide a richer view of a program while creating choices about outcome priority and multiplicity. Clear plans for primary contrasts and missing-data handling help readers distinguish confirmatory analyses from exploratory ones. Those details are not reported in the registry record, which lists zero analyses.

The statistical importance of the record, therefore, lies in separating a well-described research question from an evaluable result. Randomization and a named primary measure describe the intended design, but a numerical estimate and its uncertainty are needed to judge the observed comparison. Without them, neither a positive nor a null interpretation is warranted from the posted outcome information.

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 statistical interpretation is that the registry record does not report a single numerical primary-outcome comparison. Although the primary measure and assessment times are identified, there is no posted estimate and uncertainty interval for a selected contrast. This prevents assessment of the magnitude, direction, and precision of the comparison.

The record lists 17 outcome measures and zero analyses, but it does not identify a multiplicity strategy or a hierarchy among outcomes. It also does not report the analysis population, missing-data approach, statistical model, or adherence to assigned programs. These omissions make it difficult to evaluate how an eventual analysis might handle repeated measurements and incomplete follow-up.

Masking is listed as none. For an intervention involving in-person training with a service dog versus online training education, participants and trainers may experience the programs differently; the record does not describe any masking procedures or how outcome assessment was conducted. The outcome is a symptom checklist score, and the record does not include other information that would allow readers to assess the context of its measurement.

Enrollment is listed as 60, with 30 participants at risk in each group in the posted adverse-event information. Without the outcome estimates and their uncertainty, the precision of the primary comparison cannot be judged. Finally, the study population is described as veterans with PTSD. The record alone does not establish how well results would apply to other populations or settings, and no conclusion about broader applicability can be drawn from the design description alone.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Service Dog Training Program for Military Veterans With PTSD — NCT03777020.
• Registry record last updated 2024-07-26; status last verified 2024-02.
• Results first posted 2024-07-26.
• 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

NCT03777020 is a completed randomized, parallel-group trial sponsored by the University of Maryland, Baltimore. It compares the Service Dog Training Program with Dog Training Education among veterans with PTSD. The registry lists enrollment of 60, no masking, a start date of 2019-08-22, and a primary completion date of 2023-02-28.

The primary outcome is PTSD symptom severity on the PTSD Checklist for DSM-5, scored from 0 to 80, with higher scores indicating worse symptoms. The stated time points are baseline and 8 weeks. The registry lists 17 outcome measures and zero analyses, and does not report a single numerical primary-outcome comparison. Posted safety information lists 30 participants at risk in each group, with zero serious adverse events, zero other adverse events, and zero deaths in each group. These figures and the design description do not establish an efficacy result.