Intervention to Improve Developmental and Health Outcomes for Female Adolescents: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Intervention to Improve Developmental and Health Outcomes for Female Adolescents (NCT04231669), a randomized na study. The primary outcome is Intention to Migrate.

2. Clinical Question

The study asks whether participation in the Anzansi Family Program is associated with a change in adolescent girls’ intention to migrate, compared with bolstered care. The registry describes a pilot study intended to assess the program’s feasibility and acceptability and to explore preliminary effects on child development outcomes. Its stated context is concern about poor female youth’s rural-to-urban migration for child labor in low- and middle-income countries.

The primary outcome is a score based on the adolescent girl’s answer to a question about how likely she sees herself migrating. The question uses a scale from 1 (very unlikely) to 5 (very likely). Measurements are listed at baseline, post-test (9 months), and 6-months follow-up. The reported analysis is a linear mixed-effects regression comparing the two groups, with a p-value of 0.05. The registry record does not report the estimated group effect, its direction, or a confidence interval, so the p-value alone does not show how large or precisely estimated any difference was.

The intervention combines Family Economic Empowerment with Multiple Family Groups and is offered to adolescent girls and caregivers. The comparator is usual school services bolstered with notebooks and lunch; the record says this bolstered care is also provided to the treatment group. The study is described as randomized, single-masked, and parallel, while a methodological note identifies a cluster-randomized design. The details needed to understand the randomization unit and masking are not reported in the registry record.

3. Trial Design

The registry describes a completed, randomized, single-masked, parallel-group pilot study sponsored by Washington University School of Medicine. The listed enrollment is 200. The phase is recorded as NA. The study start date is 2020-07-20, and the primary completion date is 2022-06-30. The record identifies Independent Child Migration as the condition.

The two groups are Control: Bolstered Care and Anzansi Family Program. Participants in the control group receive services and education as usual in their schools, with school notebooks and lunch added as bolstered care. The description says the bolstered care is also provided to the treatment group. In addition, the intervention group receives the Anzansi Family Program, which combines Family Economic Empowerment with Multiple Family Groups for adolescent girls and caregivers.

A note about the analysis identifies the design as cluster-randomized and cautions that a simple individual-level re-analysis would be misleading. The registry’s design fields also identify random allocation and a parallel model. The cluster unit, number of clusters, allocation process, and reason for single masking are not reported in the registry record. These details matter because observations within a cluster may be related, and because the interpretation of masking depends on who was unaware of group assignment.

The study’s stated aims extend beyond the primary outcome. They include piloting feasibility and acceptability, comparing groups on child development outcomes, and exploring individual-, family-, and program-level factors that affect participation and experience. Secondary outcome names include attitudes towards school, school attendance, self-concept, multidimensional student life satisfaction, adolescent stress, family cohesion, child-caregiver relationship, and social support. The record lists 16 outcome measures overall, but does not provide a full account of their timing or the analysis plan for each.

The study is characterized as a pilot, and the registry does not report a sample-size justification or a detailed statistical analysis plan. The listed enrollment should therefore be read as a registry fact, not as evidence that the study was designed to detect a particular effect with a specified precision.

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

Analysis population for the primary outcome, as defined in the registry: Adolescent girls. One participant was lost to follow-up at 9 months

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 Intention to Migrate, measured over Baseline, post-test (9 months), 6-months follow-up.

The primary endpoint is intention to migrate, recorded as a score on a scale. The adolescent girl is asked how likely she sees herself migrating. Response options range from 1 (very unlikely) to 5 (very likely), and the registry describes the outcome as a mean with standard deviation. Assessments are specified at baseline, post-test (9 months), and 6-months follow-up.

This outcome is relevant to the study’s stated concern about rural-to-urban migration for child labor. It measures an expressed intention, however, rather than a recorded migration event. A response on a likelihood scale can capture a participant’s stated outlook at an assessment time; by itself, it does not establish whether migration later occurred or why a participant gave that response.

Repeated assessments can describe how scores change over time and whether the pattern differs between groups. The registry identifies a linear mixed-effects regression for the reported comparison, a method that can account for correlated repeated measurements when appropriately specified. The model’s details, including its treatment of time and group-by-time differences, are not reported in the registry record.

The record does not provide the group means, estimated change, contrast between groups, or confidence interval for this endpoint. Consequently, readers cannot determine the magnitude or direction of the posted comparison from the reported p-value alone. The registry also notes that one participant was lost to follow-up at 9 months, but does not report the missing-data method or the number of observations contributing to each assessment.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

General statistical explanation follows. For this trial, the registry names a linear mixed-effects regression and identifies the design as cluster-randomized. The specific model formula and estimation details are not reported in the registry record.

What does a mixed-effects model do?

A mixed-effects model can represent both average patterns shared across participants and variation associated with particular participants or groups. For repeated measurements, it can account for the fact that observations from the same person are related rather than independent. The exact structure matters: a model may represent participant-level correlation, cluster-level correlation, or both, depending on its specification. The registry record does not state which random effects or covariance structure were used.

For a score measured at baseline and later follow-ups, a model can estimate changes over time and compare those patterns across groups. A group contrast could be defined at one assessment or as a pattern across assessments. Those are not interchangeable questions. Because the registry does not report the model terms or its estimated coefficients, the particular contrast represented by the reported p-value is not clear from the record.

Why does cluster randomization matter?

General statistical explanation: when allocation occurs by cluster, outcomes within a cluster can be correlated. This reduces the amount of independent information compared with an analysis that assumes every participant contributes an unrelated observation. An analysis should account for the design when estimating uncertainty. Ignoring clustering can make standard errors too small and confidence intervals too narrow. The degree of this effect depends on features such as within-cluster similarity and cluster sizes, which are not reported in the registry record.

The methodological note explicitly cautions that a simple individual-level re-analysis would be misleading and says only the sponsor’s own analysis is shown. This is a reason not to derive a new treatment comparison from individual-level summary information. The record does not give cluster counts, allocation details, or enough model output to independently verify the sponsor’s cluster-adjusted analysis.

What does the p-value communicate?

General statistical explanation: a p-value describes how compatible the observed data are with a specified null model, under the analysis assumptions. A smaller value can indicate greater incompatibility with that null model, but the p-value does not measure the size of an effect or its importance. Nor is it the probability that the null hypothesis is true.

The registry reports p = 0.05 for the stated superiority analysis. Without the estimated contrast, confidence interval, and clearer identification of the tested contrast, that value is not enough to tell whether scores were higher or lower in either group or how much they differed. A confidence interval would show a range of effect values compatible with the data and model at its stated confidence level. No confidence interval is reported for the primary comparison.

What would the mean and standard deviation tell readers?

General statistical explanation: a mean describes the average score in a group, while a standard deviation describes the spread of individual scores around that average. Reporting both can help readers understand the scale and variability of a continuous outcome. For a repeated-measures comparison, group means at each time point are useful descriptive information, but they do not by themselves replace a model-based contrast that accounts for the design and repeated observations. The registry identifies the primary outcome parameter as a mean and its dispersion as a standard deviation, but does not report those outcome values.

How should planning illustrations be read?

General statistical explanation: sample-size calculations are conditional on assumptions about the target effect, variability, allocation, significance level, and analysis. The registry’s stated planning illustration uses a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; the effect to detect is a stated fraction of the observed effect. For non-time-to-event outcomes, the illustration is expressed as participants per arm for 80% power or another stated target. These are planning illustrations, not observed power and not a record of the trial’s original planning rationale. The registry does not report that rationale.

For cluster-randomized studies, planning also depends on clustering. The assumptions summarized in the registry description do not provide the cluster-specific information needed to assess how any planning illustration incorporates the design. Any sample-size illustration should therefore be read as conditional on its stated assumptions, rather than as a direct measure of the actual study’s ability to detect an effect.

10. Analyses Reported by the Sponsor

The registry identifies the sponsor’s primary analysis as a linear mixed-effects regression and reports a p-value of 0.05 for a superiority comparison. It does not report an estimated effect, confidence interval, or the group-level outcome values for intention to migrate. The reported p-value therefore cannot be compared with an independently recalculated effect estimate or used to establish its direction and size.

The methodological note says the study was cluster-randomized, warns that a simple individual-level re-analysis would be misleading, and specifies that only the sponsor’s own analysis is shown. Accordingly, no simple re-analysis should be used to replace or challenge the sponsor’s design-aware analysis. The model’s cluster adjustment and other specification details are not reported in the registry record, so the analysis cannot be independently checked from the posted summary.

The registry record does not identify a discrepancy between a sponsor analysis and a recalculation. More importantly, it does not provide the summary statistics required for an unadjusted recalculation of the primary comparison. The sponsor’s reported p-value can be described as posted, but the magnitude, direction, and precision of the underlying comparison remain unavailable in the record.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 16 outcome measures and names secondary outcomes that include attitudes towards school, school attendance, self-concept, multidimensional student life satisfaction, adolescent stress, family cohesion, child-caregiver relationship, and social support. Testing several outcomes or time points can create a multiplicity issue: when many hypotheses are examined, the chance of at least one apparently notable result can increase. General statistical practice may address this through prespecification, adjustment, or cautious interpretation. The registry record does not report a multiplicity strategy.

The record lists one analysis for the reported results and identifies the primary comparison as a superiority analysis. It does not explain whether that count represents one model, one hypothesis, or a broader analysis plan. It also does not report which outcomes or time-point contrasts were prespecified, or whether additional analyses were conducted. The full statistical analysis plan is not reported in the registry record.

For missing data, the primary outcome description notes that one participant was lost to follow-up at 9 months. The record does not describe the reason, whether that participant contributed data at other assessment times, or how missing observations were handled. Imputation methods, assumptions about missingness, sensitivity analyses, and outcome-specific analysis counts are not reported in the registry record. A mixed-effects model can use available repeated observations under particular assumptions, but the model name alone does not establish how missing data were addressed.

The analysis population is also not reported in the registry record. The record does not specify whether the primary analysis followed intention-to-treat principles, how participants were assigned to analysis groups, or whether exclusions occurred. These omissions limit assessment of whether the analyzed population corresponds to the randomized population and how robust the result may be to missing outcomes or alternative analysis choices.

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 0% reporting threshold.

The registry’s adverse-event figures list 0 serious adverse events among the 100 participants at risk in Control: Bolstered Care and 0 among the 94 participants at risk in the Anzansi Family Program group. Deaths are listed as 0 in both groups. Other adverse events are also listed as 0 among the 100 participants at risk in control and 0 among the 94 at risk in the program group.

These are the posted counts and at-risk group figures; they should not be extended beyond the categories and reporting in the registry. Zero reported events in these fields does not, by itself, establish that no health problems occurred or that the groups had equivalent safety. The record does not provide event definitions, collection procedures, observation time for safety, or a detailed account of how events were assessed. It also does not list serious adverse-event terms.

The listed at-risk figures should be read as the registry records them, without combining them or inferring additional participant totals. The safety information is descriptive and does not provide a comparative statistical analysis. It is therefore not possible to assess the precision of a between-group safety difference from these figures alone.

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

The registry documents a randomized pilot study of the Anzansi Family Program compared with bolstered care, with intention to migrate as its primary outcome. The posted analysis is identified as a linear mixed-effects regression and reports a p-value of 0.05. The outcome is a self-reported likelihood score measured at baseline, post-test (9 months), and 6-months follow-up.

The record does not report the estimated group difference or change, its direction, or a confidence interval. Thus, the result cannot establish how large or precise the comparison was. A p-value by itself does not show the scale of an effect, prove that the program caused a change, or determine whether a finding is important in the study’s context. The registry record also does not provide enough detail to identify exactly which model contrast the p-value represents.

The primary outcome measures intention, not migration itself. The study’s stated motivation concerns rural-to-urban migration for child labor, but a stated intention is not the same as a subsequently observed migration event. The primary endpoint alone therefore cannot answer whether participants migrated or what circumstances shaped migration.

The pilot aims include feasibility, acceptability, preliminary developmental outcomes, and factors affecting participation and experience. The registry record does not report detailed results for these aims. Its secondary outcome list indicates a broad range of social, school, family, and wellbeing measures, but the listed outcomes do not establish that every measure had a reported result.

The record supports a limited conclusion: a sponsor analysis was registered and a p-value was posted, but the information needed to judge the size, direction, precision, and robustness of the primary comparison is not reported in the registry record. Neither a claim of benefit nor a claim of no effect follows from the p-value alone.

14. Why This Trial Matters Statistically

This trial is statistically notable because its methodological note identifies cluster randomization and cautions against a simple individual-level re-analysis. That warning highlights a central design principle: the analysis must reflect how participants were allocated. When allocation is clustered, the dependence among participants in the same cluster can affect uncertainty, so an analysis that ignores clustering can give misleading precision.

The endpoint also illustrates the distinction between a proximal, self-reported measure and a downstream event. Intention to migrate is relevant to the study’s stated question, but it is a score based on a likelihood response rather than a direct record of migration. Keeping that distinction clear helps prevent an outcome from being interpreted as measuring more than it does.

The posted analysis illustrates why a p-value is not a complete results report. The registry reports p = 0.05, but does not report an effect estimate or confidence interval. Without those, readers cannot assess the direction, scale, or precision of the group comparison. Statistical interpretation depends on the estimate and its uncertainty, as well as the model and design.

The study is described as a pilot with aims that include feasibility and acceptability as well as preliminary outcome comparisons. Pilot findings and planning calculations serve different purposes. A sample-size illustration based on an observed effect is conditional on its assumptions and is not observed power or evidence of the original planning rationale. The registry does not report a sample-size justification.

Finally, the broad set of listed outcomes makes prespecification and multiplicity relevant. The record names 16 outcome measures but does not report a multiplicity strategy or a complete analysis plan. This reinforces the importance of reporting which questions were primary, how repeated assessments were modeled, and how uncertainty was handled.

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 details needed to assess the primary result are not reported in the registry record. Although the analysis is named as a linear mixed-effects regression and a methodological note identifies cluster randomization, the cluster unit, cluster counts, allocation procedure, and model specification are not described. The note cautions against individual-level re-analysis, but the posted model output is not detailed enough for independent verification.

The primary outcome description gives the scale and assessment schedule, but the record does not report group means, an estimated contrast, or a confidence interval. The p-value of 0.05 is not sufficient to determine the direction or magnitude of the comparison. Nor does the record clarify which time-related contrast the reported analysis tests.

The registry describes the study as a pilot but does not report a sample-size justification. The stated planning illustration assumptions are not a substitute for the original design rationale, and the cluster-specific assumptions needed to assess the design’s influence on planning are not reported. The listed enrollment of 200 should not be treated as proof that a particular effect could be detected with a specified precision.

One participant was lost to follow-up at 9 months, but the reason, handling of missing outcomes, and sensitivity analyses are not reported. The analysis population and treatment of participants with incomplete follow-up are also not reported in the registry record.

Finally, the record lists 16 outcome measures and several secondary outcomes but does not describe a multiplicity strategy, a complete analysis plan, or detailed results for the study’s feasibility and acceptability aims. Single masking is recorded, but who was masked is not reported. These gaps limit interpretation of both the primary comparison and the broader study aims.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Intervention to Improve Developmental and Health Outcomes for Female Adolescents — NCT04231669.
• Registry record last updated 2026-02-13; status last verified 2026-02.
• Results first posted 2023-03-14.
• 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

The registry record describes a completed, randomized, single-masked, parallel-group pilot sponsored by Washington University School of Medicine. It lists enrollment of 200 and identifies Independent Child Migration as the condition. The study compares Anzansi Family Program plus bolstered care with bolstered care alone.

The primary outcome is an intention-to-migrate score on a scale from 1 (very unlikely) to 5 (very likely), assessed at baseline, post-test (9 months), and 6-months follow-up. The reported method is linear mixed-effects regression, with p = 0.05. The estimated effect, its direction, and its confidence interval are not reported in the registry record. A methodological note identifies cluster randomization and says a simple individual-level re-analysis would be misleading.

The registry lists 16 outcome measures and records 0 serious adverse events, 0 other adverse events, and 0 deaths in each group’s adverse-event fields. One participant was lost to follow-up at 9 months. The missing-data approach, analysis population, multiplicity strategy, and detailed statistical plan are not reported in the registry record.