Healthy Mothers: an Intervention to Support Perinatal Women Living With HIV in Kenya: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Healthy Mothers: an Intervention to Support Perinatal Women Living With HIV in Kenya (NCT05219552), a non-randomized na study. The primary outcome is Maternal Mental Health.

2. Clinical Question

The study’s official title frames the work around food insecurity, depression and non-adherence to Option B+ using a syndemics approach. The registered primary outcome, however, is maternal mental health; the listed secondary outcomes include relationship status, women’s empowerment, HIV-related stigma and social support. No estimates of between-group differences, uncertainty intervals or hypothesis-test results are reported in the registry record. The design can therefore be described, but the direction and size of any observed group differences cannot be established from the recorded information.

3. Trial Design

The registry describes a parallel-group, nonrandomized intervention study with no masking. Its status is completed, and the recorded start date is 2022-02-23, with primary completion on 2023-03-09. The registry lists enrollment as 40. The brief summary describes 20 women in the intervention group and 20 in the control group, but these figures do not establish how many participants completed follow-up or contributed to each outcome analysis.

The intervention combines personalized lactation support with monthly unconditional cash transfers. The summary describes support and transfers beginning at approximately 30 weeks of pregnancy and continuing to approximately 6-months postpartum. The arm description says a professional lactation specialist met with intervention participants during pregnancy and at 2-weeks, 4-weeks, 6-weeks and 3-months postpartum. The control group received standard care at a similar but distinct clinic. The registry identifies personalized lactation support as a behavioral intervention.

Because allocation was nonrandomized, group membership may be related to characteristics that also relate to the outcomes. The registry explicitly cautions that the arms are not a like-for-like control comparison. Differences in setting, participant characteristics, care access or other factors could contribute to observed differences, although the registry record does not report whether such differences occurred or how they were addressed.

The record does not report a protocol for assignment, a rationale for the enrollment target, an interim-analysis plan or a statistical analysis plan. It lists 10 outcome measures and 0 analyses. Completion status indicates that the study is recorded as completed; it does not, by itself, mean that results or participant-level follow-up information have been reported.

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: The analysis population is the same as the assignment in Participant Flow. Baseline data for participants is reported under outcome meassures.

6. Primary Endpoint

The registry lists the primary outcome as Maternal Mental Health, measured over baseline (20-35 weeks pregnant), 6-weeks postpartum and 6-months postpartum.

The primary outcome is maternal mental health, measured using the Patient Health Questionnaire-9 (PHQ-9). The registry describes it as a score on a scale ranging from 0-27, with higher scores indicating greater depressive symptoms. It specifies the median as the summary measure and the inter-quartile range as the measure of dispersion. Assessments are scheduled at baseline, defined as 20-35 weeks pregnant, and at 6-weeks postpartum and 6-months postpartum.

A symptom questionnaire provides a structured way to record responses across participants and assessment times. In general, a score summarizes the responses captured by the instrument; it is not itself a diagnosis. Reporting the median and inter-quartile range emphasizes the middle of the score distribution and its spread, rather than assuming that an average alone adequately describes the data. The inter-quartile range describes the central portion of the observed scores.

Repeated assessments can describe how scores vary across pregnancy and postpartum follow-up, but a meaningful comparison requires the actual score distributions, participant counts at each time point and an analysis that accounts for the study design. Those results are not reported in the registry record. The outcome definition and schedule are available, but no group medians, inter-quartile ranges or estimates of change are reported. Consequently, the endpoint’s measurement can be explained, but the results cannot be interpreted as evidence for a particular direction or magnitude of difference.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies a parallel-group design: participants are described in an intervention group and a standard-care group. Parallel-group comparisons ask whether outcomes differ between groups assessed over the same general period. In this study, allocation is explicitly nonrandomized. The distinction matters because random allocation, when used and implemented as intended, helps balance both measured and unmeasured characteristics on average. Nonrandom allocation does not provide that design-based protection. Statistical adjustment may address measured differences, but cannot automatically remove bias from unmeasured or poorly measured factors.

What does nonrandomized allocation mean for comparison?

In general, an observed difference between nonrandomized groups can reflect the intervention, pre-existing group differences, differences in care settings, or a combination of these. The registry says the control clinic was similar to, but distinct from, the intervention site, and cautions that the arms are not a like-for-like comparison. That makes it especially important to know how participants entered each group, which baseline characteristics were recorded and whether adjustment was planned. The assignment process, adjustment variables and analysis model are not reported in the registry record. Without those details, readers cannot judge how the analysis might have addressed confounding.

Why report the median and inter-quartile range?

The PHQ-9 outcome is described using the median and inter-quartile range. In general, the median is the middle of an ordered set of observations, while the inter-quartile range describes the spread of the central part of that distribution. These summaries can communicate a typical score and variation without relying on a symmetric, bell-shaped distribution. They do not, on their own, provide an estimate of the uncertainty in a between-group comparison. A confidence interval, for example, expresses a range of values compatible with the data and a specified statistical procedure. No confidence interval is reported in the registry record.

When an outcome is summarized by medians, analysis methods should fit the question and data structure. Investigators might compare distributions or estimate a location difference using a method suited to the outcome and design. If repeated scores from the same participant are analyzed together, the method also needs to account for the fact that measurements from one person are related. General methods can address these issues, but the registry record does not identify the method used or report whether repeated observations were modeled jointly.

How should the repeated assessment schedule be understood?

The primary measure is scheduled at baseline, 6-weeks postpartum and 6-months postpartum. These are repeated observations over time, not three independent samples by definition. In general, a longitudinal analysis can distinguish differences between groups from changes over time and can use information from multiple assessments, subject to its assumptions. A simpler approach might report each time point separately. The registry record does not say which approach was used, whether baseline scores were used for adjustment, or whether a group-by-time contrast was specified. It also does not report how many participants provided scores at each time point.

Attrition and missing observations matter for longitudinal results. If the participants with recorded follow-up scores differ from those without them, the observed score distribution may not represent the group originally enrolled. Statistical methods for missing data rely on assumptions about why data are missing; those assumptions should be described and considered alongside sensitivity analyses when appropriate. No missing-data method, missingness summary or sensitivity analysis is reported in the registry record. The amount and pattern of missing outcome data are therefore unknown.

What can be concluded from the analysis record?

The registry lists 10 outcome measures and 0 analyses. It does not report an outcome model, an effect estimate, uncertainty intervals, p-values or an analysis population. It also does not report a statistical plan or an interim-analysis plan. These omissions mean that there is no reported inferential result to evaluate for model fit, precision or robustness. The absence of a reported analysis is not evidence that a particular result was null or that no analysis was conducted outside the information in the registry record.

The record describes sample-size illustrations based on an observed effect, with a two-sided significance level of 0.05, equal allocation, and a normal approximation; the effect to detect is described as a fraction of the observed effect. These are planning illustrations, not the trial’s own design justification. They are not observed power. The registry record does not report the trial’s sample-size justification, and without a reported observed effect there is no basis to interpret the illustrations as a measure of the study’s ability to detect a clinically meaningful effect. For time-to-event outcomes, the described convention is a total number of events; for other outcomes, it is participants per arm. The primary endpoint is a score, not a time-to-event outcome.

In general, over-dispersion means that variability is greater than a particular statistical model expects. When present, it can make standard errors and confidence intervals too small if the model does not account for it. The registry’s choice to describe the PHQ-9 outcome with a median and inter-quartile range does not establish whether over-dispersion was assessed or whether any model requiring such an assessment was used. No model or diagnostic results are reported. Thus, general statistical considerations can guide what to look for in a full analysis report, but they cannot substitute for details absent from the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 10 outcome measures, including maternal mental health as the primary outcome and relationship status, women’s empowerment, HIV-related stigma and social support among the secondary outcomes. When several outcomes or time points are examined, the chance of at least one apparently notable result can increase. In general, investigators may define a primary comparison in advance and use a multiplicity strategy for additional tests, or treat secondary analyses as exploratory. The registry record does not report a multiplicity plan, a hierarchy of secondary comparisons or which comparisons, if any, were formally tested.

The record lists 0 analyses and does not report results for the primary or secondary outcomes. It also does not identify analysis populations, such as whether the reported analysis was intended to include everyone enrolled or only participants with available outcome data. The number of participants contributing data at baseline, 6-weeks postpartum and 6-months postpartum is not reported in the registry record.

Missing data can affect both precision and the interpretation of group comparisons. In general, complete-case summaries may be misleading if availability of an outcome is related to its value or to participant characteristics. Other approaches require assumptions about the missingness process. The registry does not report how missing values were handled, whether reasons for missingness were recorded, or whether sensitivity analyses were planned or conducted. These gaps prevent assessment of how much the reported outcome, if analyzed, might depend on follow-up completeness.

The registry record also does not report a statistical analysis plan, interim analyses or a sample-size justification. The completed status does not fill in those methodological details. Without results, an analysis population, missing-data information and a multiplicity approach, readers cannot assess the scope of testing or the potential influence of incomplete follow-up.

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 serious adverse event entries record 0 affected participants among 20 at risk in the intervention group and 0 affected participants among 20 at risk in the control group. Deaths are recorded as 0 in both groups. The entries for other affected participants are also 0 in each group. These are the counts recorded for the groups and categories specified in the registry record.

A zero count in a registry entry should be interpreted narrowly. It means that the record reports no affected participants for that category; it does not, on its own, establish that no adverse events of any kind occurred, that every event was collected, or that the intervention has no safety risks. The record does not give event descriptions, ascertainment procedures, follow-up coverage for adverse events or further detail about reporting. The listed counts should not be treated as a comprehensive safety assessment.

The entries also do not allow a precise comparison of safety between the groups. In general, rare events can be difficult to characterize in small studies, and the absence of recorded events is not proof of absence of risk. Because the study is nonrandomized, any comparison would also need to account for differences between groups and their care settings. No comparative safety analysis is reported in the registry record.

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

The registry record establishes that this completed study evaluated a combined intervention of personalized lactation support and monthly unconditional cash transfers against standard care for perinatal women living with HIV in Kenya. It also specifies maternal mental health, measured with the PHQ-9 at baseline, 6-weeks postpartum and 6-months postpartum, as the primary outcome. The stated score range is 0-27, with higher scores indicating greater depressive symptoms.

Those design details do not establish what the study found. No PHQ-9 medians, inter-quartile ranges, group contrasts, confidence intervals or p-values are reported in the registry record. There is no basis to say whether scores were higher or lower in either group, whether they changed over time, or how precise a difference might have been. The same limitation applies to the listed secondary outcomes.

The nonrandomized design also limits causal interpretation. Even if group differences were reported, an association between group membership and outcome would not by itself show that the intervention caused the difference. Participants may differ for reasons related to group placement, and the clinics were distinct sites. The registry record does not report whether baseline imbalance was assessed, which factors were adjusted for or how residual confounding was considered.

The record’s serious adverse event entries are zero for affected participants in each group, but those counts should be read within the scope of the registry’s reporting. They do not demonstrate that all possible harms were absent. Likewise, a completed study status does not mean that its outcome findings have been posted.

The defensible conclusion is therefore about the study’s question and methods, not an intervention effect. The registry describes an intervention and a planned mental health measure, but it does not report the results needed to determine the direction, magnitude or precision of any difference. It also does not report enough analysis and follow-up detail to assess missing data, confounding control or the handling of multiple outcomes.

13. Why This Trial Matters Statistically

This study addresses a combined behavioral and material-support approach in a population and setting identified in its title and summary: perinatal women living with HIV in Kenya. Statistically, the design raises a central question about how to interpret group differences when allocation is nonrandomized. A reader must distinguish a descriptive comparison from an estimate that can be attributed to an intervention. The registry itself cautions that the arms are not a like-for-like comparison, making that distinction especially important.

The repeated PHQ-9 assessments also make the timing and completeness of measurement relevant. Baseline and postpartum observations can describe different points in the perinatal period, but changes over time require clear definitions and suitable methods for correlated measurements. Reporting participant counts at each time point, missing-data methods and the analysis model would help readers assess whether the outcome comparisons reflect the enrolled groups and how uncertainty was handled. Those details are not reported in the registry record.

The study also lists several secondary outcomes alongside its primary mental health measure. That breadth can help describe related dimensions of participants’ lives, but it makes a prespecified analysis strategy important. Without a reported approach to multiple outcomes or posted estimates, readers cannot determine whether findings were consistent across outcomes or how to distinguish confirmatory from exploratory comparisons.

The registry record is useful for identifying the clinical question, intervention components, comparator, design and planned outcome schedule. Its importance as a statistical case lies partly in the information still needed to interpret findings: results, uncertainty, analysis populations, missingness and methods for addressing nonrandom allocation. Since those results are not reported in the registry record, no conclusion about the intervention’s effect can be drawn from this record alone.

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 trial-specific limitation for interpreting the primary outcome is that the registry record does not report any PHQ-9 results. It gives the measurement schedule and summary approach, but no group medians, inter-quartile ranges, estimates of differences or uncertainty. It also lists 0 analyses. Consequently, the clinical outcome cannot be assessed from the registry record.

Nonrandomized allocation is a design limitation for causal comparisons. The registry explicitly notes that the intervention and control arms are not a like-for-like comparison, and the control clinic was similar to but distinct from the intervention site. The record does not describe the assignment process, baseline comparability, adjustment strategy or other measures that might address confounding. It is not possible to determine whether observed differences, if reported elsewhere, would reflect the intervention or other differences between groups.

The record describes planned repeated assessments but does not report how many participants contributed outcome data at each time point, the extent or reasons for missing follow-up, or the method used to handle missing observations. It does not identify an analysis population or report sensitivity analyses. These omissions limit evaluation of whether follow-up availability could have influenced the observed score distributions.

The registry lists 10 outcome measures but does not report a multiplicity strategy, statistical analysis plan, interim-analysis plan or sample-size justification. Its sample-size description concerns planning illustrations based on an observed effect, not the trial’s own design rationale. The available adverse-event entries are limited to the recorded categories and counts; fuller ascertainment and event details are not reported in the registry record. Together, these gaps constrain interpretation of both efficacy-related outcomes and safety reporting.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Healthy Mothers: an Intervention to Support Perinatal Women Living With HIV in Kenya — NCT05219552.
• Registry record last updated 2025-08-14; status last verified 2025-07.
• Results first posted 2025-08-14.
• 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.

18. Record Summary

The registry describes a completed, parallel, nonrandomized study with no masking, evaluating personalized lactation support and monthly unconditional cash transfers against standard care for perinatal women living with HIV in Kenya. It lists enrollment as 40, and the summary describes 20 women in each group. The recorded start date is 2022-02-23 and primary completion date is 2023-03-09.

The primary outcome is maternal mental health measured with the PHQ-9 at baseline, 6-weeks postpartum and 6-months postpartum. The registry specifies medians and inter-quartile ranges, but does not report outcome values or analyses. It lists 10 outcome measures and 0 analyses. Serious adverse event entries record 0 affected participants and 0 deaths in each group, with 20 at risk in each group.

Because allocation was nonrandomized and the record cautions that the arms are not like-for-like, causal comparisons require particular care. Results, uncertainty estimates, analysis populations, missing-data methods and a multiplicity plan are not reported in the registry record. The design and outcome definitions can be described, but the intervention’s effects cannot be determined from the available registry information.