Reducing the Burden of Chronic Psychotic Disorders in Tanzania (CAPACITY): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Reducing the Burden of Chronic Psychotic Disorders in Tanzania (CAPACITY) (NCT04327843), a na phase 3 study. The primary outcome is Tablets Routine Questionnaire (TRQ).

2. Clinical Question

This is a question about whether adherence could be assessed during a combined care approach, not a controlled test of whether the approach caused a change. The registry record does not provide an analyzed estimate of TRQ change or a comparison with another treatment strategy.

3. Trial Design

The study, titled “Reducing the Burden of Chronic Psychotic Disorders in Tanzania (CAPACITY),” is registered as NCT04327843. It is marked completed and designated phase 3. The registry record gives a start date of 2019-11-05, a primary completion date of 2020-07-30, and enrollment of 22.

The design is listed as single group, with allocation recorded as not applicable and masking listed as none. One group is described: CAE + LAI. The intervention listings identify Customized Adherence Enhancement as behavioral and haloperidol decanoate as a drug. There is no concurrent comparator in the record, so this is not a randomized, like-for-like comparison. The phase designation does not change that design feature.

The project description presents the intervention as a care approach combining an injectable antipsychotic with a behavioral program focused on medication adherence. It describes a manualized curriculum and use of existing injection-clinic health workers. Those descriptions explain the intervention’s intended structure; they do not establish how consistently each element was delivered or how participants engaged with it. Such implementation details are not reported in the registry record.

The primary outcome was change from baseline to the 6 month visit on the TRQ. The record also refers to a 25 week follow-up when explaining why a mean change was not calculated. It lists secondary measures of medication attitudes, psychiatric symptoms, clinical global impressions, social and occupational functioning, body mass index, and movement-related adverse effects. In total, the record lists 11 outcome measures and 8 analyses, but does not provide a complete account of how each analysis was conducted.

Because the design has one group, any change over time would not by itself distinguish an intervention effect from other explanations, such as natural fluctuation, concurrent care, or changes in who remained available for follow-up. In this case, the registry record does not report the primary change estimate, so even a within-group description of that endpoint is unavailable.

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: 38 individuals were assessed for preliminary eligibility. 16 people were excluded, 5 were unable to contact, 2 were not interested, 3 did not meet criteria, and 6 were no-shows for screening. 22 provided informed consent initially. 2 of these people failed the oral tolerance test and 1 did not complete baseline measures. 19 participants enrolled and attended visits 1-6 and 18 attended all 7 visits.

Baseline Assessment

Week 13 Assessment

Week 25 Assessment

Analysis population for the primary outcome, as defined in the registry: Since participants stopped using oral medications once they were stable on LAI, there were only two participants remaining at 25 Week visit and therefore analysis was not done. As only 2 individuals were on concommittent oral CPD medication at 25 week follow up, mean change in TRQ was not calculated. This was just done to enter into clinical trials.

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 Tablets Routine Questionnaire (TRQ), measured over Change from Baseline to 6 month visit.

The primary endpoint is the Tablets Routine Questionnaire (TRQ), recorded as percentage change in adherence from baseline to the 6 month visit. The registry description says the TRQ is a brief self-report instrument for medication adherence. It was validated in populations with bipolar disorder medication adherence and is intended to identify people who miss 20-30% or more of their medication in the last week or month.

The score ranges from 0 to 100. The direction matters for interpretation: higher TRQ scores indicate greater nonadherence, and therefore worse adherence. A change in score should be interpreted using that direction rather than assuming that a higher score represents improvement. The registry’s outcome label refers to “percentage of change in adherence,” while its description explains that the scale itself records nonadherence. No analyzed change estimate is reported, so the direction and magnitude of any observed change cannot be established from this outcome record.

A self-report adherence measure can capture reported medication-taking behavior, which is relevant when the intervention is intended to address adherence. It is not the same as a direct count of doses taken, and the record does not report a corroborating adherence measure for the primary endpoint. Nor does it report a mean, standard deviation, confidence interval, or statistical test for TRQ change.

The record explains that only two participants remained at the 25 week visit after stopping oral medication once stable on LAI, and that only two individuals were on concomitant oral chronic psychotic disorder medication at that follow-up. It consequently states that the mean change was not calculated. This missing primary estimate limits what can be concluded about adherence change.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical interpretation depends first on the design and then on what outcome data are actually reported. This record describes a single group receiving CAE + LAI, with no randomized comparison and no masking. A single-group study can document outcomes among participants over time, but it cannot use a concurrent control group to distinguish intervention-related change from other influences. That is a general limitation of the design, not a claim about what happened to any individual participant.

What does a change score represent?

In general, a change score compares a participant’s value at follow-up with their value at baseline. A group mean change summarizes those participant-level changes for the participants included in the analysis. The interpretation depends on the measure’s scoring direction, the follow-up time, and which participants have usable values at both times. Here, the intended measure is TRQ change from baseline to the 6 month visit. The record says higher TRQ scores indicate greater nonadherence, so the scale direction must be kept in view when describing a change.

The record does not report the mean change or its standard deviation. It says the mean change was not calculated because only two participants remained at the 25 week visit after stopping oral medication once stable on LAI, and only two individuals were on concomitant oral chronic psychotic disorder medication at that follow-up. Without a reported estimate, there is no primary effect size or precision measure to interpret. The reason stated in the record should not be expanded into an unreported statistical procedure or assumed analysis population.

What would a confidence interval tell a reader?

As a general statistical concept, a confidence interval describes the range of effect estimates compatible with the data and the method used, under the assumptions of that method. A narrower interval indicates greater precision than a wider interval, but an interval does not by itself establish causation or clinical importance. No confidence interval for the primary TRQ change is reported in the registry record, so the precision of that outcome cannot be assessed from the record.

How does the single-group design affect inference?

In general, repeated measurements in one group can describe how measured outcomes vary over time. They do not control for changes unrelated to the intervention. For example, time trends, concurrent care, or differences between participants who do and do not return for assessment can complicate interpretation. These are general possibilities, not documented explanations for this trial’s outcomes. With no concurrent comparison group, an observed before-and-after change would not be an estimate of the difference between treatment strategies.

The intervention is a combination of a behavioral program and an injectable medication. In general, when components are given together to one group, the design cannot identify the separate contribution of each component. The registry record does not describe a design that varies or compares the components independently.

What do the planning assumptions mean?

The record describes a sample-size planning illustration based on an observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It says the target effect is a stated fraction of the observed effect. For time-to-event tables, the planning quantity is the total number of events; for other tables, it is participants per arm. These are planning assumptions, not observed study outcomes and not a report of the trial’s original design justification. The registry record does not report the trial’s own sample-size justification.

In general, a sample-size illustration estimates how much information a hypothetical design might require under its assumptions. It does not establish that the actual study had a particular probability of detecting an effect, and it is not observed power. An illustration using two equal arms is especially distinct from this record’s single-group design. Any planning value should therefore be read as a hypothetical design exercise under the stated assumptions, not as a result of the completed study.

Why can over-dispersion matter?

As a general concept, over-dispersion occurs when variability in observed data is greater than a statistical model expects. If a model assumes less variability than the data contain, standard errors and tests can be misleading. Whether over-dispersion is relevant depends on the outcome and model. The registry record does not report enough detail about the analyses to establish that an over-dispersion assessment was performed or that a particular model was used. It would be inappropriate to infer either from the outcome list.

What is known about the analysis strategy?

The record lists 11 outcome measures and 8 analyses, but does not report the full statistical analysis plan, the model specifications, covariate adjustments, handling of repeated measurements, or the planned testing hierarchy. It also does not report whether estimates were adjusted or unadjusted. These details affect how estimates should be interpreted, but cannot be reconstructed from the study title or endpoint descriptions.

The primary outcome is described as a mean and standard deviation, yet the record states that the mean change was not calculated. This indicates that the planned summary cannot be read as a posted result. No p-value, confidence interval, or calculated mean change is reported for the primary endpoint. Statistical significance and precision therefore cannot be assessed for that outcome using this registry record.

10. Multiplicity, Missing Data and Analysis Populations

The registry record lists 11 outcome measures and 8 analyses. It identifies the TRQ as the primary outcome and names secondary measures covering medication attitudes, psychiatric symptoms, global clinical impressions, functioning, body mass index, and movement-related effects. However, it does not report a testing hierarchy, a multiplicity adjustment, or a family of outcomes defined for a single confirmatory claim. Whether or how multiple testing was addressed is not reported in the registry record.

In general, examining multiple outcomes or multiple analyses can increase the chance of at least one apparently notable result arising by chance. Adjustment methods and prespecified priorities can help readers understand how a study addressed that issue. The available record does not permit an assessment of such procedures, so the listed number of measures and analyses should not be taken as proof that testing was or was not adjusted.

The primary outcome has a specific missing-result explanation: the mean TRQ change was not calculated. The record says only two participants remained at the 25 week visit after stopping oral medication once stable on LAI, and only two individuals were on concomitant oral chronic psychotic disorder medication at that follow-up. It does not provide participant-level missingness, reasons for every missing assessment, or a complete account of the primary analysis population.

It is also not reported whether any imputation, sensitivity analysis, or alternative analysis was performed. The record does not provide a full flow of participants from enrollment through each assessment. The enrollment of 22 should therefore not be treated as the denominator for every outcome, nor should the two people described at follow-up be assumed to define every analysis population. These gaps limit assessment of how missing observations may affect the interpretation of outcomes.

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: substance use relapse and hospitalization (5.3%).

The safety summary describes one group, CAE + LAI Treatment. It lists 19 participants at risk for serious adverse events, with 1 affected; 19 at risk for other adverse events, with 13 affected; and 19 at risk for deaths, with 0 deaths. The most frequently listed serious event term is “Substance use relapse and hospitalization,” reported as 5.3%.

These are descriptive counts for the group recorded in the registry, not a comparison between treatment strategies. The serious-event and other-event summaries are separate categories in the record; they should not be assumed to be mutually exclusive or to count distinct people across categories. The record does not provide event timing, duration, severity beyond the serious classification, or details about how events were adjudicated.

With no concurrent comparator, the counts cannot indicate whether events were more or less common than they would have been under another approach. The enrollment figure is 22, while the safety summaries use 19 as the number at risk; the registry record does not explain the difference. These safety figures should therefore be read using their stated at-risk denominators rather than substituting total enrollment.

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

The record establishes that a completed, single-group study evaluated a combined behavioral and injectable-medication approach in Tanzania. It specifies a primary measure of adherence, describes the TRQ’s scoring direction, and reports selected safety counts. It does not report a calculated mean change in the primary outcome. As a result, the record does not establish whether TRQ scores changed, how large any change was, or how precisely such a change could be estimated.

The missing primary estimate is especially important because the endpoint is the study’s stated measure of adherence change. The registry explanation connects the absence of a mean change calculation to the small number remaining at the 25 week follow-up and the small number on concomitant oral medication at that time. It does not provide an alternative analysis that resolves the missing estimate. The result cannot be interpreted as evidence of either improvement or worsening.

Even if a within-group change had been reported, the design would limit causal interpretation. There is no concurrent randomized comparator, and the group received both CAE and LAI. A single-group result could not isolate intervention effects from other changes over time, and the combination design could not separate the behavioral component from the injectable medication.

The safety numbers describe events recorded among 19 participants at risk, including 1 participant affected by a serious adverse event, 13 affected by other adverse events, and 0 deaths. These observations provide a descriptive account for this group, not a comparative estimate of risk. The 5.3% listed for the most frequent serious event term is also descriptive and does not supply a controlled comparison.

In short, the registry record supports an account of study design, intended measurement, and selected safety observations. It does not support a quantified conclusion about primary adherence change, its precision, or a causal effect of the combined approach. The record also does not report enough statistical-plan detail to resolve questions about missing data handling or multiplicity.

13. Why This Trial Matters Statistically

This study raises an important statistical issue: a clearly specified primary outcome is not the same as an available primary result. The registry record identifies TRQ change from baseline to the 6 month visit, describes the instrument, and explains how its score direction should be read. It also states that the mean change was not calculated because only two participants remained at the 25 week visit after stopping oral medication once stable on LAI. The absence of that estimate constrains interpretation more than a small or imprecise estimate would, because no effect size is available to evaluate.

The record also illustrates why the design must be considered alongside the intervention. CAE and LAI were delivered as a combined approach to one group. This makes the study different from a randomized comparison and prevents attribution of any observed change to one component. The single-group design may describe experience over time, but it does not furnish a concurrent counterfactual for what would have happened under another treatment strategy.

Another statistical point is the difference between enrollment and outcome-specific availability. Enrollment is 22, while the safety summaries list 19 at risk and the primary follow-up explanation refers to two people remaining at the 25 week visit. These figures describe different aspects of the record and should not be collapsed into a single denominator. A complete participant flow and analysis-population description would be needed to understand the availability of observations across outcomes.

Finally, the record names a range of secondary outcomes and analyses, but does not describe a multiplicity strategy or a complete analysis plan. This makes it difficult to assess how confirmatory or exploratory results were distinguished. The case is a useful reminder that endpoint definition, follow-up availability, design, and reporting detail jointly determine what a reader can infer from clinical-trial data.

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 interpretation is that the registry record does not report the primary TRQ mean change. It states that this mean was not calculated and gives a follow-up explanation involving only two participants remaining at 25 weeks after stopping oral medication once stable on LAI, and only two individuals on concomitant oral chronic psychotic disorder medication at that follow-up. No alternative primary estimate or uncertainty interval is reported.

The study is single group, with allocation not applicable and no masking. It has no concurrent control group, so its design cannot establish how outcomes would compare with another care strategy. Because the intervention combines CAE and LAI, the record also cannot distinguish the effects of those components from each other.

The primary outcome timing is described as change from baseline to the 6 month visit, while the explanation for not calculating change refers to a 25 week visit. The record does not explain the relationship between those time descriptions or provide a more complete schedule of assessments. It also does not report a full participant flow to clarify how enrollment of 22 relates to outcome-specific follow-up and the 19 participants at risk in the safety summaries.

Other specific limitations include the absence of reported primary outcome estimates, confidence intervals, p-values, detailed model specifications, analysis populations, missing-data methods, and multiplicity procedures. The record lists secondary measures and a total number of outcomes and analyses, but does not provide enough detail to evaluate how all analyses were selected or conducted. Implementation fidelity and participant engagement with the behavioral program are not reported in the registry record.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Reducing the Burden of Chronic Psychotic Disorders in Tanzania (CAPACITY) — NCT04327843.
• Registry record last updated 2021-10-12; status last verified 2021-09.
• Results first posted 2021-10-12.
• 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

NCT04327843 is a completed, phase 3, single-group study of CAE combined with LAI in Tanzania. The registry record lists enrollment of 22 and identifies the primary outcome as TRQ change from baseline to the 6 month visit. Higher TRQ scores indicate greater nonadherence. The record states that the mean change was not calculated, citing the small number of participants remaining at the 25 week follow-up and on concomitant oral medication.

There is no concurrent comparator, and no primary adherence estimate or confidence interval is reported. The safety summary lists 19 participants at risk, with 1 affected by a serious adverse event, 13 affected by other adverse events, and 0 deaths. The record also lists 11 outcome measures and 8 analyses, but does not report a complete analysis plan, missing-data strategy, or multiplicity procedure.