Reducing Stigma Among Healthcare Providers to Improve Mental Health Services: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Reducing Stigma Among Healthcare Providers to Improve Mental Health Services (NCT02793271), a randomized na study. The primary outcome is Social Distance Scale (SDS) - Baseline.

2. Clinical Question

The central study question concerns whether adding RESHAPE’s social-contact component to standard training and supervision can be assessed in a cluster randomized pilot. In particular, the protocol describes assessing trainees’ views of service users’ participation in training and supervision, fidelity to the intervention, possible contagion between conditions, randomization, and the feasibility and psychometric properties of outcome measures. These are feasibility and measurement questions, distinct from establishing an effect on provider or patient outcomes.

The listed primary measure warrants careful interpretation: its title identifies a baseline Social Distance Scale assessment, not a reported between-arm change or treatment-effect estimate. The description says the scale was collected before training and at follow-up, but no outcome analysis is posted. Consequently, the record does not answer whether scores changed, whether the arms differed, or how precise any such difference might be.

3. Trial Design

RESHAPE-mh is registered as a randomized, parallel, triple-masked trial and described in its official title as a feasibility and acceptability pilot cluster randomized control trial. Cluster randomization means allocation is organized around groups rather than treating every participant’s assignment as independent. The record does not identify the cluster unit, the number of clusters, or who was masked. Its masking label should therefore not be taken to establish which participants, trainers, assessors, or analysts were unaware of allocation.

The registry records enrollment of 88. The trial status is completed. The recorded start date is 2016-02-01 and the primary completion date is 2018-08-31. Those dates describe the registry record; they do not establish when every participant was enrolled or completed follow-up.

The PRIME arm received PRIME/mhGAP training. The PRIME+RESHAPE arm received that training with the RESHAPE adjunct, described as social contact with mental health service users participating as training co-facilitators. The comparator is described as existing mental health training and supervision for non-specialists, including mhGAP and psychosocial treatments. The record’s descriptions are truncated, so further details of the delivery, intensity, or implementation cannot be established from them.

The stated primary objectives focus on trainees’ perspectives on acceptability, intervention fidelity and contagion, assessment of randomization, and feasibility and psychometric properties of outcome measures in a cluster design. Change in provider and patient outcomes is described as secondary. This emphasis matters: a feasibility pilot is intended to address whether a larger study’s procedures and measures can be used, not necessarily to deliver a definitive estimate of clinical effectiveness.

Although the registry lists a baseline Social Distance Scale measure as primary and describes collection at follow-up, it reports no outcome analyses. The number of clusters, allocation procedures in detail, follow-up completeness, and analysis plan are not reported in the registry record.

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.

Population: 17 health facilities assigned in each group. Health workers from those health facilties are participants.

6. Primary Endpoint

The registry lists the primary outcome as Social Distance Scale (SDS) - Baseline, measured over Baseline (Pre-Training).

The listed primary measure is the Social Distance Scale (SDS) at baseline, before training. Its description identifies a 12-item measure of healthcare providers’ willingness to interact with persons with mental illness. Scores range from 12 to 72, with higher scores described as a worse outcome. The registry says the measure was collected at baseline and at 4 months and 16 months after training.

Social distance is relevant to this intervention because RESHAPE adds contact with people who have lived experience of mental illness to provider training. A measure of willingness to interact can therefore capture an attitude related to the intervention’s stated focus on stigma. As a continuous outcome, it can describe score distributions and average scores, rather than simply classifying providers into categories. The registry specifies a mean and standard deviation as the measure’s summary format.

However, the primary outcome title specifically names the baseline assessment. A baseline score describes participants before training; on its own, it is not evidence of a post-training change or a difference attributable to the intervention. While the description mentions later assessments, the record does not post analyzed follow-up scores, an estimated change, an arm contrast, or a confidence interval. The registry record also does not report a threshold for interpreting the importance of a score difference.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Because the record describes a cluster randomized pilot, design and analysis need to be read together. Cluster randomization assigns groups, rather than independently assigning each person. This can be useful when an intervention is delivered through shared training or supervision settings, but the registry does not say what the randomized cluster was in this trial. It also does not report the number of clusters or the allocation procedure in detail.

Why does clustering matter?

General statistical explanation: people in the same cluster can share providers, training, local services, or other conditions. Their responses may therefore be more alike than responses from people in different clusters. This similarity is called within-cluster correlation. If an analysis ignores it and treats each person as independent, the uncertainty around a group comparison can be understated. A cluster-aware analysis accounts for this dependence, for example by modeling clusters or using an appropriate variance adjustment. The registry does not report the analysis method used for any outcome.

The registry specifically notes that a simple individual-level re-analysis would be misleading and that only the sponsor’s own analysis is shown. This is a relevant warning against reconstructing an independent-participant comparison from summary information. At the same time, the record lists 0 analyses, so no sponsor analysis is available among the posted results to serve as an estimate. There is no reported cluster-adjusted effect, standard error, confidence interval, or test result.

What does the listed endpoint represent?

The Social Distance Scale is described as a continuous measure summarized by a mean and standard deviation. A mean represents the average score in a group, while a standard deviation describes the spread of individual scores around that average. General statistical explanation: these summaries can help characterize a score distribution, but they do not by themselves establish a difference between randomized groups. To assess a comparison, an analysis would need to define the contrast and account for the design, with uncertainty reported in a suitable form.

The endpoint title specifies baseline, before training. The description also names assessments at 4 months and 16 months after training. A baseline summary can help describe the study population before intervention, but a baseline value is not a change score. A follow-up contrast might compare groups at a specified time, compare changes from baseline, or use a model of repeated measurements. These choices answer related but not identical questions. The statistical plan and the actual approach are not reported in the registry record, and no follow-up result is posted.

How should a confidence interval be read?

General statistical explanation: a confidence interval expresses the range of effect estimates reasonably compatible with the observed data under a specified statistical model and its assumptions. A narrow interval indicates greater precision than a wide interval, while an interval spanning values on both sides of no difference may leave the direction uncertain. Its interpretation depends on the outcome, the comparison, the design, and the model. No confidence interval is reported for the outcomes in this registry record, so precision cannot be assessed from posted analyses.

What are the sample-size illustrations?

General statistical explanation: a planning illustration based on an observed effect asks what sample size might be needed under stated assumptions to detect an effect of a chosen fraction of that observed effect. Such calculations are prospective planning aids, not observed trial outcomes and not evidence that a completed study had a particular probability of detecting an effect. The record describes its illustrations as using a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation; the target effect is expressed as a fraction of the observed effect. For non-time-to-event outcomes, the stated convention is participants per arm for 80% power or 90% power; time-to-event illustrations use a total event count. These assumptions do not describe the trial’s own design rationale, which is not reported. In particular, the cluster design makes a simple individual-level interpretation inappropriate.

General statistical explanation: over-dispersion arises when observed variability exceeds what a chosen model expects. If present, it can make standard errors and confidence intervals from an overly restrictive model too optimistic. The registry does not report whether over-dispersion was assessed or how any model handled it. No claim about its presence in this trial is warranted.

Finally, the registry labels the design triple-masked but does not identify the masked roles. Masking can reduce the influence of expectations on care, measurement, or analysis, depending on who is masked and how it is maintained. The label alone does not establish which protections applied in practice. Procedures for masking and the statistical analysis plan are not reported in the registry record.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("PRIME", "PRIME+RESHAPE")
n   <- c(45, 43)
m   <- c(33.8, 36.5)
sp  <- c(12.1, 12.5)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  PRIME  45  33.800  12.100  30.165  37.435
CARM  PRIME+RESHAPE  43  36.500  12.500  32.653  40.347

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 15 outcome measures and names several secondary outcomes assessed at baseline or follow-up. It does not report an analysis plan describing which comparisons were primary for inference, how repeated measures were handled, or whether adjustment for multiple comparisons was planned. The record also lists 0 analyses, so there are no posted estimates or tests from which to evaluate multiplicity in practice.

General statistical explanation: when many outcomes, time points, or contrasts are examined, the chance of obtaining at least one apparently notable result by chance can increase. A prespecified hierarchy, adjustment procedure, or clear exploratory framing can help readers interpret a family of tests. Without a reported plan and analyzed results, it is not possible to determine how multiplicity was addressed or whether any result was selected from a broader set of comparisons.

Missing-data methods and analysis populations are also not reported in the registry record. The record does not state the amount or pattern of missing outcome data, the number assessed at each follow-up, the rules for inclusion in each analysis, or whether an intention-to-treat approach was used. These omissions prevent assessment of how missing observations might affect estimates or whether the analyzed participants represented the randomized groups.

General statistical explanation: missingness can be related to observed characteristics, unobserved outcomes, or both. The implications depend on the reasons for missing data and the method used to address them. Without follow-up completeness and the analysis approach, no assumption about the likely direction or size of bias is justified. The registry’s cluster design also makes it important to know whether missingness occurred at participant or cluster level and how that structure was handled; those details are not reported.

12. 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 1% reporting threshold.

The registry’s adverse-event entries list the PRIME arm with 45 participants at risk and 0 serious adverse events, 0 other adverse events, and 0 deaths. For PRIME+RESHAPE, the record lists 43 participants at risk and 0 serious adverse events, 0 other adverse events, and 0 deaths. No serious adverse-event terms are listed.

These entries are descriptive registry counts, not a comparison of safety between arms. The record does not provide event definitions, collection procedures, observation time, or a narrative explaining the absence of reported events. The participants listed as at risk should not be assumed to be the full enrollment or the number assessed for every outcome. The registry records enrollment of 88, but no total should be inferred by combining the arm entries.

General statistical explanation: an entry of zero reported events does not establish that an event is impossible or that the intervention has no risk. The information is limited to what the registry records, and the record does not provide an uncertainty estimate for these counts. These figures should therefore be read as reported event counts in this pilot, not as proof of a safety difference or of absence of risk.

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

This record describes a completed feasibility and acceptability pilot of a cluster randomized training intervention. It documents the comparison of PRIME/mhGAP training with that training plus RESHAPE’s social-contact component, and it identifies feasibility questions alongside secondary provider and patient outcomes. It also records a baseline Social Distance Scale measure and names further outcome measures.

What the record does not establish is an effect on social distance, provider knowledge, patient outcomes, or any other measured outcome. No outcome analyses are listed, and no estimates, confidence intervals, or hypothesis tests are posted. The baseline measure describes an assessment before training; it cannot alone establish change after training or a causal difference between arms. The stated follow-up time points do not supply results when no analyzed values are reported.

General statistical explanation: a randomized design can support causal comparisons when allocation, follow-up, measurement, and analysis are implemented appropriately. Cluster randomization additionally requires analysis that respects dependence within clusters. The registry’s design label and intervention descriptions do not, by themselves, demonstrate that a comparison was estimated or that any particular analysis was appropriate. The analysis plan, cluster details, and outcome results are not reported in the registry record.

The registered focus on feasibility is important when interpreting the study’s intended contribution. A pilot can examine whether training components, randomization, and outcome measures are workable, and the registry lists these as objectives. However, no numerical results for these feasibility objectives are posted. The stated aims should not be mistaken for findings that the procedures were feasible or acceptable.

The appropriate conclusion is narrow: the registry records the design, intervention arms, outcome definitions, and selected adverse-event counts, but does not provide analyzed evidence about outcome effects. It neither demonstrates that RESHAPE improved outcomes nor demonstrates that it did not. Readers should distinguish the protocol’s rationale and objectives from results actually reported.

14. Why This Trial Matters Statistically

This trial is statistically notable because its cluster randomized design changes how evidence must be analyzed. Participants within a shared training or service setting may have correlated responses. An analysis that ignores that structure can give misleadingly precise comparisons, which is why the registry cautions against a simple individual-level re-analysis. The cluster unit and number of clusters are not reported, limiting assessment of how the design was implemented or how a suitable analysis would be specified.

The record also illustrates the distinction between a protocol’s aims and a results record. The study’s objectives include feasibility, acceptability, fidelity, and assessment of outcome measures, while its listed primary measure is a baseline Social Distance Scale score. These are not interchangeable claims: a measure can be registered without a reported analysis, and a planned feasibility objective is not a finding. The record lists 0 analyses, so it cannot support a statistical conclusion about changes or arm differences.

For clinical researchers, this underscores the value of reporting the analysis population, follow-up completeness, cluster handling, effect estimates, and uncertainty. For biostatisticians, it emphasizes that design information is necessary before interpreting sample-size illustrations or reconstructing contrasts from arm summaries. General statistical explanation: the amount of information in a cluster trial depends not only on the number of individuals but also on how many independent clusters contribute and how strongly outcomes are correlated within them. Those details are not reported.

The trial remains informative as a registered example of a stigma-focused training intervention and a pilot design with explicit implementation and measurement objectives. Its posted record, however, is not a source of analyzed efficacy results. That distinction is central to responsible statistical interpretation.

15. 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.

Several trial-specific limitations in interpretation follow from what is absent in the registry record. The cluster unit, number of clusters, allocation details, and cluster-level characteristics are not reported. Without them, readers cannot assess the distribution of randomized units, potential imbalance between clusters, or the information available for cluster-adjusted estimation.

The primary outcome is titled as a baseline Social Distance Scale measure, while its description notes collection at 4 months and 16 months after training. The record does not clarify which follow-up contrast was intended as the main inferential target, nor does it report analyzed scores at any time point. As a result, the primary endpoint designation does not yield an interpretable treatment-effect estimate.

The record does not report outcome completion, missing-data patterns, analysis populations, or methods for handling missingness. It also does not report the statistical analysis plan, any adjustment for repeated measures or multiple outcomes, or the approach to within-cluster dependence. These details matter for evaluating bias and uncertainty, but their absence does not establish that the trial did or did not use a particular procedure.

The arm descriptions and brief summary are truncated in the record, limiting detail about intervention delivery and implementation. The masking label is recorded as triple, but the roles masked and procedures used are not specified. The adverse-event entries are counts without further collection or observation details. Finally, the registry lists 0 analyses. Consequently, the record supports description of the registered design and outcome definitions, not claims about the magnitude, direction, or precision of outcome effects.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Reducing Stigma Among Healthcare Providers to Improve Mental Health Services — NCT02793271.
• Registry record last updated 2026-07-22; status last verified 2026-06.
• Results first posted 2026-07-22.
• 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.

19. Record Summary

NCT02793271 is titled “Reducing Stigma Among Healthcare Providers to Improve Mental Health Services.” The official title describes the RESHAPE-mh protocol for a feasibility and acceptability pilot cluster randomized control trial. The record lists randomized allocation, triple masking, a parallel model, and enrollment of 88. Its status is completed, with a start date of 2016-02-01 and primary completion date of 2018-08-31.

The comparison is PRIME/mhGAP training versus PRIME/mhGAP plus RESHAPE, which adds social contact through mental health service users participating as training co-facilitators. The primary measure is the baseline Social Distance Scale, described as a 12-item scale with higher scores indicating a worse outcome. Follow-up assessments are described at 4 months and 16 months after training. The registry lists 15 outcome measures and 0 analyses.

The record’s stated aims emphasize feasibility and acceptability, but no analyzed outcome estimates are posted. The participant and cluster details, analysis plan, missing-data handling, and results for the listed outcomes are not reported in the registry record.