HIV Self Testing of Male Partners of Women in PMTCT: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for HIV Self Testing of Male Partners of Women in PMTCT (NCT03484533), a randomized na study. The primary outcome is Proportion of Men Who Test for HIV in the Self-test Arm to Men Who Test for HIV in the Standard of Care Arm.

2. Clinical Question

The question concerns whether offering a route to self-testing, through a woman enrolled in care, changes men's uptake of HIV testing relative to a clinic invitation. The trial description also connects testing with linkage to HIV care or prevention, including confirmatory testing, counseling, PrEP for men who test negative and ART for men who test positive. The primary endpoint, however, is testing; it does not itself measure linkage, prevention uptake, or treatment use.

The intervention was directed to women as the means of reaching their male partners. The registry describes training, test kits, service information, and counseling intended to reduce social harms. The comparison was an invitation for fast-track clinic testing. Thus, the registered contrast is between two approaches to offering testing, rather than a comparison of different HIV tests or clinical treatments.

The record also identifies postpartum viral suppression among HIV-infected women and acceptability of self-testing as secondary outcomes. Those outcomes address related but distinct questions. The primary outcome concerns men's testing, while viral suppression concerns women and acceptability concerns perceptions of the self-testing approach.

3. Trial Design

The registry describes a completed, randomized, parallel-group trial with no masking. It lists an enrollment of 736 and identifies the University of Washington as sponsor. The recorded start date is 2018-05-17, and the primary completion date is 2021-11-22. The trial is listed as completed and its phase as not applicable.

Participants were assigned to one of two approaches. In the HIV self-test kit arm, women were given kits to use or deliver to male partners; in the standard-of-care arm, women were given an invitation letter for fast-track HIV testing at a clinic. The registry describes the self-test as an oral-fluid test and says women were trained in use and interpretation. Men were offered confirmatory testing regardless of self-test result, counseling, and services appropriate to their result: PrEP if negative and ART if positive. The description also includes counseling to minimize social harms and referral to social support services if harms occurred.

The parallel design means the two assigned approaches were compared across groups, rather than each participant receiving both approaches in sequence. Random allocation is the method stated for forming the groups; masking is listed as none. The record does not report further details about the randomization procedure, allocation concealment, or whether outcome assessment was blinded.

The total enrollment is not the same thing as the number represented in every endpoint. The primary outcome describes 236 enrolled male partners tested for HIV, and the posted viral-suppression outcome has different arm-level participant counts. The registry does not report a complete account of how participants moved from enrollment to each analysis population. No protocol-level sample-size justification or detailed statistical analysis plan is reported in the registry record.

Although the study is marked completed, completion status alone does not establish that all planned outcomes or analyses were posted. The registry lists five outcome measures and zero reported analyses. It names two secondary outcomes, while only the viral-suppression outcome has arm-level results and a comparison in the posted results. Results for acceptability 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.

Before assignment: Data provided is for the women who were enrolled and randomized to the intervention of secondary distribution of HIV self-test kits to their male partner or standard of care delivery of invitation letter to their male partner to test for HIV at the clinic and for male partners who were consented later in the study.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Overall 236 enrolled male partners tested for HIV

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 baseline characteristics of the women who were enrolled and randomized to the intervention or control standard of care and their male partners who enrolled in the study.

6. Primary Endpoint

The registry lists the primary outcome as Proportion of Men Who Test for HIV in the Self-test Arm to Men Who Test for HIV in the Standard of Care Arm, measured over Up to 12 months post-partum.

The primary endpoint is described as the proportion of male partners who test for HIV in the self-test arm compared with the proportion in the standard-of-care arm, over a period of up to 12 months postpartum. Testing is a direct measure of uptake: it records whether the intended recipients completed testing, rather than whether they linked to care or prevention after receiving a result.

This endpoint is relevant to the stated trial question because the intervention is an alternate way of reaching male partners with an offer of testing. A test-kit offer and a clinic invitation may create different practical routes to testing. Counting men who test can therefore describe uptake under each approach, although interpretation of a proportion also requires knowing the number eligible or assigned in each arm.

The registry labels the primary outcome as a number of participants and reports 161 in the HIV self-test kit arm and 75 in the invitation-letter arm. Its population description says 236 enrolled male partners tested for HIV. The comparison fields list each arm's proportion as 1 and the risk difference as 0. The record does not report separate primary-outcome denominators beyond the displayed arm-level participant counts, so the posted counts and proportions need to be read with that limitation in mind.

Testing is an intermediate outcome in the broader program described by the investigators. It does not establish the result of testing, confirmatory diagnosis, linkage, PrEP initiation, ART use, disclosure, or post-partum ART continuation. Those require separate outcome measures and cannot be inferred from a count of men tested.

7. Results Posted in the Registry

• HIV Self-test Kit: 100.0% (161/161) versus 100.0% (75/75) on control; risk difference 0.0 percentage points (95% CI 0.0 to 0.0), risk ratio 1.00 (1.00–1.00); unadjusted two-proportion z-test p = n/a.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Viral Suppression at 12 Months Post-partum in HIV-infected Ugandan Women as Measured by Viral Load Testing

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • HIV Self-test Kit: 84.2% (239/284) versus 76.4% (110/144) on control; risk difference 7.8 percentage points (95% CI -0.4 to 15.9), risk ratio 1.10 (0.99–1.22); unadjusted two-proportion z-test p = 0.050.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between HIV Self-test Kit and the control. The interval gives the range of effects that remain compatible with the data.

The registry identifies viral suppression at 12 months postpartum and acceptability of HIV self-testing as secondary outcomes. Viral suppression has arm-level counts, denominators, proportions, and an unadjusted comparison. Acceptability is named as an outcome, but results are not reported in the registry record. The record therefore supports a limited overall reading across the named secondary outcomes rather than a complete assessment of all secondary questions.

Viral suppression addresses a different population and question from the primary endpoint. The primary outcome concerns male partners' HIV testing; the secondary viral-suppression outcome concerns HIV-infected women. It can inform a related part of the broader PMTCT program context, but it does not directly explain or verify the primary testing result. The named acceptability outcome could describe how participants viewed self-testing, but without posted results it cannot be used to interpret uptake or social harms.

The viral-suppression comparison is unadjusted, and its interval includes zero for the risk difference. It should consequently be treated as an uncertain secondary finding, not as a stand-alone confirmation. The record does not report whether outcomes were prioritized in advance, how multiplicity was handled, or whether any secondary analysis was prespecified.

In general, when several outcomes are examined, the chance of finding an apparently notable result by chance can increase unless the analysis plan accounts for the number and structure of comparisons. This does not make secondary outcomes uninformative; it means their role, prespecification, and adjustment strategy matter. Here, the registry lists five outcome measures, but reports zero analyses in its analysis listing. That limits the ability to establish how the posted secondary comparison fits into a planned family of tests.

Overall, the secondary record is incomplete for a broad interpretation: one named outcome has numerical results and an unadjusted comparison, while the acceptability outcome has no result posted. The secondary results should be considered supportive descriptive context, with less inferential weight than a prespecified primary analysis whose population, denominators, and methods are fully documented.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Viral Suppression at 12 Months Post-partum in HIV-infected Ugandan Women as Measured by Viral Load Testing
• The Acceptability of HIV Self-testing Among Pregnant Women Taking Part in PMTCT B+, and Their Male Partners

9. Statistical Interpretation

10. Statistical Methods Explained

For a binary outcome, each participant is classified according to whether the event occurred. In this trial, HIV testing and viral suppression are described as participant outcomes. A count records how many participants had the event; a proportion relates that count to the number of participants in the relevant group. Counts and proportions answer different descriptive questions, so a proportion cannot be interpreted fully without its denominator.

What does a risk difference describe?

In general, a risk difference is the proportion with the event in one group minus the proportion in the comparison group. It is expressed on the same scale as the proportions. A value of zero indicates equal observed proportions; a positive or negative value indicates which group has the larger observed proportion, according to the subtraction order used. The registry lists the invitation-letter arm as the control for the comparisons.

For the primary outcome, the comparison fields report a proportion of 1 in each arm and a risk difference of 0. The associated p-value is 1. The arm-level primary entries are counts, and the record does not identify separate denominators for those proportions. This matters because the proportion comparison cannot be reconstructed or assessed from counts alone when the group sizes at risk for the outcome are not clear.

What does the confidence interval add?

In general, a confidence interval describes the range of effect estimates reasonably compatible with the observed data under the analysis assumptions. It communicates uncertainty about the estimated contrast; it is not a range of individual participant outcomes and does not guarantee that the true effect lies within its bounds. A wide interval signals less precision than a narrow one, while an interval that spans zero for a risk difference includes no difference among the values compatible with the data at that confidence level.

For the secondary viral-suppression comparison, the posted risk difference is 0.078 and its confidence interval extends from -0.004 to 0.159. Because the interval includes zero, the data are compatible with no difference as well as with a positive difference. The estimate is unadjusted, and the record does not report an adjusted model or covariate set.

How should a p-value be read?

In general, a p-value describes how unusual data at least as incompatible with a specified null hypothesis would be if that hypothesis and the analysis assumptions were true. It is not the probability that the null hypothesis is true, nor does it measure the size or practical importance of an effect. A p-value must be considered alongside the estimate, interval, design, and limitations of the outcome data.

The primary comparison lists a p-value of 1 alongside equal displayed proportions. The secondary comparison lists an unadjusted p-value of 0.05. That value should not be treated as a stand-alone conclusion, particularly because the interval spans zero, multiple outcomes are identified, and no multiplicity strategy is reported. The registry record does not report whether any planned interim analyses or other hypothesis-testing procedures were used.

Why can binary outcomes be difficult to compare?

General statistical interpretation of binary outcomes depends on clear definitions of the event, the observation period, and who is included in each group. Missing outcomes, eligibility rules, and varying follow-up can change the denominator and therefore the estimated proportion. The record gives a primary timeframe of up to 12 months postpartum but does not provide a detailed account of missing data or the analysis population for the primary comparison.

For the secondary outcome, the registry gives arm-level counts and denominators, making the reported proportions easier to interpret descriptively. However, the record does not explain whether every participant with a posted denominator had viral-load testing, how missing results were handled, or whether the analysis included all randomized participants. These details are important for assessing how the observed proportions relate to the groups originally assigned.

What do the planning illustrations mean?

The registry includes sample-size planning illustrations based on fractions of the observed viral-suppression effect. They list participants per arm for 80% power and participants per arm for 90% power under a two-sided significance level of 0.05, equal allocation, and a normal approximation. The rows describe planning for 100%, 75%, and 50% of the observed effect. These are illustrations for a hypothetical future study, not observed power and not evidence that the trial was designed to meet those targets.

The record explicitly says the illustration is not the trial's own design. A protocol sample-size justification, the assumptions used to plan this trial, and any interim-analysis plan are not reported in the registry record. Planning estimates based on an observed effect are also sensitive to uncertainty in that effect, so they should not be read as a guarantee about the size needed in another study.

The registry does not report a detailed statistical analysis plan, a complete set of analysis populations, or reported analyses beyond the comparisons embedded in the results. This limits assessment of whether the posted comparisons followed prespecified methods, how missingness was addressed, and whether all planned outcomes are represented. The posted values can still be described, but their inferential interpretation should remain bounded by these reporting gaps.

11. Confidence Intervals and Precision

Precision is about how tightly the data locate an estimated effect. In general, a confidence interval gives a range of values compatible with the data under the analysis assumptions. A narrow interval indicates a more precise estimate than a wide one, but precision does not by itself establish that an estimate is unbiased or important.

For viral suppression, the posted unadjusted risk difference is 0.078, and its interval runs from a value just below zero to 0.159. The lower end lies below zero, so the interval includes no difference; the upper end allows a positive difference. The estimate therefore does not pin down the contrast precisely enough to exclude no difference under the stated interval. The corresponding unadjusted p-value is 0.05, which should be read with the interval and with the fact that this is a secondary outcome.

The registry also posts relative measures for the viral-suppression comparison: a risk ratio of 1.102 with an interval from 0.993 to 1.222, and an odds ratio of 1.642 with an interval whose lower end is just below 1. These are different effect scales. A risk ratio compares proportions as a ratio, while an odds ratio compares odds; neither should be substituted for the risk difference without explaining the scale. The intervals for both relative measures include values close to the no-association value of 1.

Precision for the primary endpoint is harder to assess from the registry record. The arm-level entries report counts, and the listed comparison reports proportions of 1 in both arms and a risk difference of 0, but separate denominators are not identified. A confidence interval for the primary contrast is not posted. The absence of an interval does not mean there is no uncertainty; it means the registry record does not provide a precision estimate for that comparison.

These intervals describe sampling uncertainty under the analysis used. They do not account automatically for missing data, selection, measurement problems, or unreported analysis choices. Since the record does not explain the analysis populations or missing-data handling, the intervals should not be read as resolving those concerns.

12. Multiplicity, Missing Data and Analysis Populations

The registry lists five outcome measures and names two secondary outcomes. It does not report a multiplicity plan explaining how inference across multiple outcomes was managed. In general, multiplicity matters because examining several outcomes or contrasts can increase the chance that at least one result appears notable even when effects are absent. Prespecified outcome hierarchy, adjustment methods, or a clear distinction between confirmatory and exploratory analyses can help readers interpret that risk. Such details are not reported in the registry record.

The analysis listing contains zero reported analyses. Numerical comparisons are present with the outcome results, but the record does not provide a full statistical analysis plan or state whether the displayed analyses were prespecified. It also does not report interim analyses. The absence of these details means readers cannot assess the timing and scope of hypothesis testing from the registry record.

Missing-data handling is also not reported. The primary endpoint is described as a proportion, but the posted results are counts and the record does not identify separate arm-specific denominators for the primary comparison. The secondary viral-suppression results include denominators, but the registry does not explain how participants without a measured outcome were handled or whether all randomized participants were included in that analysis.

Analysis populations are not described in enough detail to establish whether the primary and secondary comparisons follow an intention-to-treat approach, a per-protocol approach, or another rule. The overall enrollment of 736 should not be assumed to be the denominator for every endpoint. The primary outcome refers to 236 enrolled male partners tested, while the secondary outcome has different counts in its arms. The record does not provide a participant flow account that reconciles enrollment with endpoint-specific populations.

These gaps do not erase the posted values, but they restrict conclusions about completeness, potential selection due to missing outcomes, and whether statistical testing followed an a priori plan. They also make it difficult to judge how much weight to place on the p-value attached to the secondary result. The record does not report a protocol-level sample-size justification, so the trial's intended power for its primary endpoint cannot be assessed from the registry.

13. 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: hospitalization (1.5%).

The registry posts serious adverse event figures for women in each arm. In the HIV self-test kit group, 332 women are listed at risk, with 5 serious events and 2 deaths. In the invitation-letter group, 168 women are listed at risk, with 1 serious event and 2 deaths. Hospitalization is the leading listed serious-event term, reported as 1.5%.

These are descriptive counts. The registry record does not provide confidence intervals or a formal statistical comparison for the serious adverse event figures. The posted numbers therefore do not establish whether the groups differ in safety, and the death counts should not be interpreted as caused by either assigned approach. The record does not supply further information about event relatedness, timing, severity grading, or the methods used to collect and classify events.

The groups' listed numbers at risk differ, so raw event counts alone are not a like-for-like measure of event frequency. The 1.5% hospitalization figure is reported as the leading serious-event term, but the record does not give an arm-specific breakdown for that term. No conclusion about comparative safety can be drawn from these figures alone.

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

The posted results describe a randomized comparison of two ways to offer HIV testing to male partners: oral-fluid self-test kits delivered through women in PMTCT B+ programs, and invitation letters for fast-track clinic testing. The primary endpoint is intended to measure men's testing uptake. The registry reports counts by arm and a comparison with equal displayed proportions and a risk difference of 0, but does not identify separate primary-outcome denominators.

The result does not establish that one testing approach is better or worse. The primary comparison's listed proportions are equal, but the denominator limitation prevents a fuller assessment of the stated proportion endpoint. The record also does not provide a primary confidence interval or a detailed analysis report. The appropriate reading is limited to the figures posted, not a broader claim about testing behavior in other programs or populations.

The secondary viral-suppression outcome is related to the program's postpartum context, but it is a distinct endpoint among women rather than a measure of male-partner testing. Its unadjusted estimate has an interval spanning zero, so the posted result remains uncertain about the size and direction of the contrast. Acceptability results are not reported in the registry record.

Testing alone does not establish confirmatory diagnosis, linkage to care, PrEP uptake, ART use, disclosure, or continued treatment. The trial description includes these broader services and aims, but the primary measure does not capture them. Similarly, the viral-suppression figures do not explain adherence, service access, or the causes of observed outcomes.

The trial's randomized design is relevant to evaluating the assigned approaches, but design labels alone do not resolve missing information about outcome denominators, missing data, analysis populations, or multiplicity. The registry record does not report enough detail to determine how those issues were handled. The findings should therefore be understood as a registry-level account of posted outcomes, with inferential claims constrained by the available reporting.

15. Why This Trial Matters Statistically

This trial raises an important statistical distinction between an intervention's intended mechanism and the outcome used to assess it. The primary question concerns a way of reaching male partners with an offer of HIV testing, while the posted primary endpoint is a count described as a proportion. To interpret a proportion, the number of participants who could have experienced the event must be clear. The registry reports the number of men tested by arm but does not separately identify the primary denominators, making outcome definition and denominator reporting central to interpretation.

The trial also shows why endpoint boundaries matter. Testing is an early step in a broader sequence that the study description connects to confirmatory testing, linkage, prevention, and treatment. A testing count cannot establish whether those subsequent steps occurred. Separating outcomes prevents an uptake measure from being mistaken for evidence about later service use.

The secondary viral-suppression result illustrates the value of reading an estimate together with its confidence interval. The unadjusted point estimate suggests a positive risk difference, while the interval includes zero. The resulting uncertainty cannot be reduced to the p-value alone. The registry also lists acceptability as a secondary outcome but does not report its results, emphasizing that completion of a trial does not guarantee complete outcome reporting in the record.

Finally, the planning illustrations are explicitly based on the observed effect and are not the trial's own design. This distinction prevents retrospective calculations from being mistaken for evidence of the study's original sample-size rationale or its observed power. The registry does not report the trial's own sample-size justification or a complete analysis plan, so design intent and inferential procedures remain unclear.

These features make the record useful for statistical learning: it demonstrates how randomization, outcome definitions, denominators, confidence intervals, secondary outcomes, and reporting completeness each shape what can be concluded from trial results.

16. 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 limitations are specific to the posted record. The primary endpoint is titled as a proportion of men who test, yet the primary arm entries are counts and the comparison lists proportions of 1 in both groups. The registry does not identify separate arm-specific denominators for that primary comparison or provide a confidence interval. This limits interpretation of the primary estimate beyond the displayed fields.

The overall enrollment is 736, while the primary outcome description refers to 236 enrolled male partners tested. These figures relate to different descriptions of the trial population, and the registry does not provide a flow account explaining their relationship. The trial also includes women and male partners, with different populations represented in its outcomes. Enrollment should therefore not be treated as the denominator for any one endpoint without additional information.

The viral-suppression outcome has arm-specific counts and denominators, but the record does not explain missing viral-load results, analysis-population rules, or whether all randomized participants were included. Its comparison is unadjusted. The acceptability outcome is named but has no numerical result reported in the registry record, and the registry lists five outcome measures but zero reported analyses.

The record does not report a detailed randomization procedure, allocation concealment, a protocol-level sample-size justification, an interim-analysis plan, or a complete statistical analysis plan. It also does not report a multiplicity strategy. These omissions restrict evaluation of prespecification and the treatment of missing outcomes and multiple comparisons.

Safety figures are descriptive and lack a formal comparison or uncertainty estimates. The registry does not detail event relatedness, timing, or arm-specific hospitalization results. These limitations apply to interpretation of the registry record and do not by themselves establish problems in trial conduct.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: HIV Self Testing of Male Partners of Women in PMTCT — NCT03484533.
• Registry record last updated 2025-03-27; status last verified 2024-06.
• Results first posted 2025-03-27.
• 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.

20. Record Summary

NCT03484533 is a completed randomized, unmasked, parallel-group trial sponsored by the University of Washington. It compared HIV self-test kits delivered through women with invitation letters for fast-track clinic testing for male partners in PMTCT B+ programs in Uganda. The registry lists enrollment of 736, a start date of 2018-05-17, and primary completion on 2021-11-22.

The primary outcome concerns men's HIV testing up to 12 months postpartum. The registry reports 161 men in the self-test arm and 75 in the invitation-letter arm; its comparison lists equal proportions of 1 and a risk difference of 0. Separate primary denominators and a primary confidence interval are not reported. The secondary viral-suppression outcome has an unadjusted comparison with an interval that spans zero. Acceptability results are not reported in the registry record.

The record does not include a detailed analysis plan, missing-data methods, multiplicity strategy, or trial-specific sample-size justification. Planning illustrations are based on the observed secondary effect and are not the trial's own design.