Family Telemental Health Intervention for Veterans With Dementia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Family Telemental Health Intervention for Veterans With Dementia (NCT03116464), a na na study. The primary outcome is Healthy Aging Brain Center Monitor (HABC Monitor):.

2. Clinical Question

3. Trial Design

This was a single-group behavioral study of a telemental health family intervention for Veterans with dementia and their caregivers. The registry lists allocation as NA and masking as NONE. It describes the intervention group as caregiver and patient-with-dementia dyads who receive the family intervention; no separate control group or randomized comparison is reported. The study is listed as phase NA.

The project’s stated aim was developmental: to create a video assessment clinicians could use to identify interpersonal difficulties in caregiver–patient interactions, and a related family therapy addressing difficulties identified in the assessment. The registry rationale describes behavioral problems such as wandering, agitation, and sleep difficulties as stressful aspects of caregiving. It also states that caregivers may not recognize how their interactions contribute to behavioral problems and may not seek help to improve interpersonal skills. These are the project’s motivating statements, not findings from an outcome analysis.

Enrollment is listed as 4. The study started on 2017-09-18, had a primary completion date of 2022-12-30, and is listed as TERMINATED. The registry explains that insufficient recruitment meant phase I was not completed; consequently, a treatment protocol could not be developed and implemented. No intervention-phase data were obtained. Thus, the intended treatment evaluation did not proceed to a stage where the planned outcome could be analyzed.

The record lists the VA Office of Research and Development as sponsor. It also lists the Kansas Marital Conflict Scale, Patient Health Questionnaire - 9, and Zarit Burden Scale as secondary outcomes. Their listing does not indicate that results were collected. The registry does not report a completed protocol, a comparator, or analysis results.

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: Not applicable as due to insufficient recruitment, phase I of the project was not completed and so a treatment protocol could not be developed and implemented. No data was obtained as the intervention phase of the study did not begin.

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: The Overall Number of Baseline Participants reflects individual participants, not dyads.

6. Primary Endpoint

The registry lists the primary outcome as Healthy Aging Brain Center Monitor (HABC Monitor):, measured over Post Treatment, an average of 12-18 weeks.

The planned primary endpoint was the Healthy Aging Brain Center Monitor (HABC Monitor), described as a 31-item caregiver assessment of dementia severity, caregiver stress, and mood. Its stated domains include patient cognition, function, and behavioral or psychological symptoms, together with caregiver quality of life. The registry describes the measure as having good internal consistency (0.73-0.92) and construct validity.

A measure spanning patient symptoms and caregiver quality of life is relevant to the project’s stated focus on caregiver–patient interactions and caregiving challenges. As a general statistical point, a prespecified endpoint gives an evaluation a defined outcome to assess; a multidomain measure can represent several related aspects of a clinical or caregiving context. Those features do not establish that the measure changed in this study, nor do they show which domain would have driven any change.

The planned timeframe was post-treatment, an average of 12-18 weeks. However, the registry says the intervention phase did not begin and no data were obtained. There is therefore no endpoint value, change estimate, uncertainty interval, or treatment result to interpret. The registry does not report how the HABC Monitor would have been scored or analyzed.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies the study as single group and does not report a randomized control comparison. In general, a single-group design measures outcomes in one participant group, rather than comparing outcomes between separately assigned treatment and control groups. Such a design can be useful for developmental work, feasibility questions, or descriptive assessment, depending on its objectives and implementation. Its interpretation is limited when the question is whether an intervention caused an observed change, because other influences on outcomes may not be separated from treatment effects without a suitable comparison or other design features.

What an endpoint estimate would represent

In general, an endpoint estimate is a numerical summary of the outcome being studied, such as a score at a specified time or a change between measurements. The estimate’s interpretation depends on the outcome definition, timing, population, and analysis method. A confidence interval is a range of values generated by a statistical procedure to express uncertainty around an estimate. A narrow interval generally indicates greater precision than a wide interval, while an interval alone does not resolve design limitations or establish causation.

For this study, the HABC Monitor was the planned primary outcome, measured post-treatment at an average of 12-18 weeks. The registry does not report a score, an estimate, a confidence interval, or an analysis. It says the intervention phase did not begin and that no data were obtained. Consequently, statistical methods that would summarize observed outcome values cannot be applied to results that are not reported.

What hypothesis tests can and cannot say

In general, a hypothesis test assesses how compatible data are with a specified statistical model or null hypothesis. A p-value is not the probability that a treatment works, nor is it a measure of the size or practical importance of an effect. Its interpretation depends on the analysis plan, data quality, design, and assumptions. A p-value without an effect estimate and uncertainty interval would also provide an incomplete description of the result.

No analysis results or p-values are reported in this registry record. The statistical plan, including any prespecified hypothesis tests, is not reported in the registry record. It is therefore not possible to identify a planned model, adjustment strategy, or testing procedure beyond the assumptions attached to the separate planning illustration.

How to read the planning illustration

The registry describes a planning illustration using a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is described as a stated fraction of an observed effect. The illustration is explicitly characterized as planning rather than the trial’s own design. Its assumptions do not change the actual single-group design listed for this study and should not be mistaken for an observed treatment comparison.

In general, a sample-size calculation estimates how much information a proposed study might need under chosen assumptions. It is sensitive to the effect size, variability, significance level, allocation, and analysis method assumed. If those assumptions do not match the eventual study or the outcome data, the calculation may not describe the study’s actual ability to estimate an effect. A planning illustration is not observed power: observed power is not a substitute for reporting an effect estimate and its uncertainty. The registry’s planning description also distinguishes total event counts for time-to-event illustrations from participants per arm for other illustrations. Those conventions describe how a hypothetical calculation is expressed, not what happened in this single-group project.

Over-dispersion is a general issue in count-data analysis: it occurs when variability exceeds what a simple model assumes, and ignoring it can make uncertainty appear too small. It is not possible to determine whether over-dispersion was relevant here, because no count outcome analysis or statistical model is reported. More broadly, a model should match the outcome type and data structure, but the registry record does not describe a completed model-based analysis.

What is not available for reanalysis

A recalculation from summary statistics requires posted outcome summaries and a clearly defined comparison. Neither is available for the primary endpoint. The record reports no analysis results, and the study had no like-for-like control comparison. It therefore does not support an unadjusted recalculation of treatment effect, an uncertainty interval, or a statistical test. The absence of such results is not evidence of no effect; it means the intervention outcome question was not evaluated in the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists the HABC Monitor as the primary outcome and names the Kansas Marital Conflict Scale (KMCS), Patient Health Questionnaire - 9 (PHQ), and Zarit Burden Scale as secondary outcomes. It does not report results for any of them. Because the intervention phase did not begin and no data were obtained, there are no reported outcome analyses across which to assess multiplicity.

In general, multiplicity arises when a study conducts several statistical tests or examines multiple outcomes, time points, subgroups, or analysis choices. When many tests are considered, the chance of at least one apparently notable result can increase. A prespecified testing strategy or adjustment can address some multiplicity concerns, while a clear distinction between primary and exploratory analyses helps readers interpret results. The registry does not report a multiplicity strategy, and it does not report whether one was planned.

The record also does not report a missing-data method, a method for defining analysis populations, or an intention-to-treat approach. These items are not reported in the registry record. In this case, the stated lack of data is linked to the intervention phase not beginning, rather than to a described pattern of missing outcome measurements after treatment. There are consequently no reported participant-level outcome data from which to assess missingness, compare analysis populations, or evaluate how an imputation method might affect results.

The listed enrollment of 4 does not supply an outcome analysis population or indicate how many dyads completed any particular assessment. The registry does not provide participant flow or endpoint-specific denominators. Those details cannot be inferred from the outcome names or the enrollment entry.

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

The record documents the development aims, planned intervention, intended primary measure, and reason the project did not reach intervention testing. It does not report evidence that the family intervention changed caregiver stress, mood, dementia-related symptoms, interpersonal conflict, or caregiver burden. The absence of results should not be read as a finding that outcomes did not change; the registry says the intervention phase did not begin and no data were obtained.

The project’s stated premise is that caregiver–patient interactions may contribute to behavioral difficulties and that some caregivers may not recognize interpersonal skills as an area where support could help. That premise motivates the planned video assessment and family therapy. It is not confirmed or refuted by outcome analyses in this record. Likewise, the registry’s description of the HABC Monitor’s measurement properties is not a report of how participants scored on it.

Because there was no comparator group, the design as registered would not have posted a like-for-like control comparison. In general, comparisons between groups can help separate changes associated with an intervention from changes that might occur for other reasons, though they too require suitable design and analysis. In this study, the more immediate constraint is that the intervention protocol was not developed and implemented, leaving no intervention outcome data to describe.

The adverse-event entries are not evidence that treatment caused no harm: the registry lists zero participants at risk in the intervention group, consistent with its statement that the intervention phase did not begin. Similarly, the planning illustration does not establish that the actual study had sufficient or insufficient statistical power. The registry record does not report an effect estimate, precision measure, hypothesis test, or completed treatment analysis.

The appropriate reading is narrow. This is a terminated, single-group project with enrollment listed as 4 and no reported intervention-phase results. It records a proposed assessment and treatment-development objective, but cannot establish the intervention’s effects or provide a statistical estimate of those effects.

12. Why This Trial Matters Statistically

This record is statistically instructive because it separates a research question from evidence capable of answering it. A primary measure, several secondary measures, a named intervention, and planning assumptions can all appear in a registry without any completed outcome analysis. Readers need to distinguish those planned elements from collected data and estimated effects.

The study also illustrates how recruitment and development stages can determine whether a planned efficacy or outcome question becomes analyzable. The registry states that insufficient recruitment prevented completion of phase I, so a treatment protocol could not be developed and implemented. The reported enrollment alone does not remedy that limitation: without intervention delivery and endpoint observations, there is no treatment estimate to calculate.

Design labels matter as well. A single-group study is not a randomized comparison, and the registry explicitly lists no control comparison. Statistical methods cannot manufacture a counterfactual comparison that the design did not include. Even if observations had been collected, interpretation would need to account for the limits of a one-group design. In this record, no endpoint observations are reported, so the question of estimating change does not arise.

The listed planning assumptions are another useful distinction. A hypothetical sample-size exercise expresses what a study might require under chosen conditions; it is not a description of realized study performance or a substitute for observed results. The status and data-collection statements are therefore essential context for reading any planning display associated with the record.

Finally, an empty result record and a null result are not the same thing. A null estimate would still summarize observed data and have some degree of precision. No such estimate is reported. Statistical interpretation must therefore focus on what was and was not observed, rather than treating the absence of estimates as evidence about the intervention’s effect.

13. 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 is that the intervention phase did not begin. The registry says insufficient recruitment prevented completion of phase I, so a treatment protocol could not be developed and implemented. No primary or secondary outcome results are reported. This prevents estimation of change, assessment of precision, and evaluation of the intended treatment question.

Enrollment is listed as 4, but the registry does not report participant flow, the number of dyads completing assessments, or endpoint-specific denominators. It also does not report participant characteristics or any completed assessment data. The listed enrollment therefore cannot be used to infer how many participants contributed to an outcome.

The single-group design has no like-for-like control comparison. This limits what could be concluded about treatment-related change even if outcome results were available. In this record, the absence of intervention-phase data is the more fundamental constraint. The registry also does not report a completed protocol, a statistical analysis plan, an analysis population, a missing-data method, or a multiplicity strategy.

The primary outcome is described, but its scoring and analytic handling are not reported. The secondary outcomes are named without results. The adverse-event record lists zero participants at risk and no reported events in the intervention group; given that the intervention phase did not begin, those entries do not characterize safety during intervention exposure.

The registry lists a primary completion date of 2022-12-30 and status TERMINATED, but does not report further detail about the termination beyond insufficient recruitment and failure to complete phase I. No additional reason should be inferred from the record.

14. Related Tutorials

15. Related Statistical Calculators

16. Sources

• ClinicalTrials.gov: Family Telemental Health Intervention for Veterans With Dementia — NCT03116464.
• Registry record last updated 2025-06-10; status last verified 2025-06.
• Results first posted 2025-06-10.
• 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.

17. Record Summary

NCT03116464 is a terminated, single-group behavioral study sponsored by the VA Office of Research and Development. It planned to develop a video assessment of caregiver–patient interaction and a related telemental health family intervention for Veterans with dementia and their caregivers. The primary outcome was the HABC Monitor, planned post-treatment at an average of 12-18 weeks; three secondary measures are listed.

Enrollment is listed as 4. The registry states that insufficient recruitment prevented completion of phase I, so a treatment protocol could not be developed and implemented. The intervention phase did not begin, and no data were obtained. No outcome analyses or treatment-effect estimates are reported. Consequently, the record describes the project’s aims and planned measures but does not provide results evaluating the intervention.