Peer PrEP Referral + HIV Self-test Delivery for PrEP Initiation in Kenya: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Peer PrEP Referral + HIV Self-test Delivery for PrEP Initiation in Kenya (NCT04982250), a randomized na study. The primary outcome is cRCT: PrEP Initiation.

2. Clinical Question

3. Trial Design

The registry describes a completed, randomized, parallel-group study with no masking. Its title identifies a pilot study and a randomized trial, and the primary outcome is labeled as part of a cRCT. A cluster-randomized design is also noted in the registry’s analysis information. The exact randomization unit, cluster structure, allocation procedures, and handling of clusters in the primary analysis are not reported in the registry record.

The study tested enhanced peer PrEP referral with HIV self-test delivery against standard peer PrEP referral. The intervention description includes training, educational materials, self-test kits, and referral cards; the comparator description uses informal word-of-mouth referral and referral cards. The registry frames the wider project as a pilot to refine the model followed by a hybrid effectiveness-implementation trial. Details of the pilot’s relationship to the randomized component and the implementation outcomes are not reported in the registry record.

Enrollment is listed as 211. The study started on 2022-02-24, reached primary completion on 2024-06-05, and has a completed status. The phase is listed as NA. The sponsor is Fred Hutchinson Cancer Center. The registry lists HIV prevention as the condition. These design details set the context for interpreting the primary result, but they do not reveal how participants or clusters were distributed across groups or how many contributed to the primary outcome.

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: In both the pilot and the 2-arm cRCT, index peers (k) referred their peers to HIV services (i.e., referred peers, n). At follow-up, index peers reported outcomes for all the referred peers (started). At follow up, some of the referred peers were also enrolled (completed). Thus, some outcomes are measured among index peers, referred peers who started (as reported by index peers), and referred peers who completed (as reported by referred peers). Enrollment totals reflect those who completed.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Outcome only applies to cRCT referred peers.

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: Of the 395 participants we have information about, only 211 were directly enrolled in the study and provided demographic information. These enrolled participants include index peers (k) and the subset of referred peers (n) we were able to enroll and follow-up, corresponding to those who completed the study in each population group.

Only the first 6 of 7 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as cRCT: PrEP Initiation, measured over 3 months.

The primary endpoint is the count of referred peers who initiated PrEP among all referred peers, as reported by index peers. It applies to referred peers in the cRCT and has a 3 months timeframe. The registry characterizes the measure as a participant count, rather than reporting an individual-level change score or a time-to-event outcome.

This endpoint addresses a central question for a peer referral strategy: whether people referred by PrEP users go on to initiate PrEP. Counting initiations among referred peers connects the intervention’s referral activity to the intended behavior, rather than measuring only whether index peers distributed materials or made referrals. The registry lists referral and HIV self-test use among pilot outcomes, and lists continuation, recent HIV testing, and adherence among secondary outcomes for the cRCT. Those measures concern other stages or aspects of prevention engagement; they are not substitutes for the primary initiation measure.

The outcome is reported by index peers, so it depends on their reports. The record does not describe verification of reported initiation, the precise denominator used for each comparison, or how missing reports were handled. It also does not provide arm-specific event counts or proportions. Consequently, the endpoint definition clarifies what was intended to be counted, but the posted result alone does not show the underlying counts or their completeness.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The reported effect measure is a risk difference. In general, a risk difference contrasts the probability of an outcome between groups by subtracting one group’s risk from the other group’s risk. A value of zero represents no difference on that scale; positive and negative values indicate opposite directions according to the subtraction order. The registry reports the estimate as -6, but does not identify the subtraction order or state the unit. The comparison should therefore be read with those omissions in mind rather than assigned to a named group.

The reported 95% confidence interval runs from -26 to 11. In general, a confidence interval conveys uncertainty around an estimate under the chosen statistical model and analysis assumptions. A narrower interval indicates greater precision than a wider interval, while an interval spanning zero leaves open the possibility of no difference and effects in either direction. The interval includes zero. It also extends across values on both sides, so the result does not pin down a single direction. Confidence intervals do not assign a probability to a particular effect being true, and they do not capture every source of uncertainty, such as limitations in outcome reporting or trial implementation.

The primary endpoint is a participant count, and the record gives the result as a risk difference. The registry does not report the method used to derive that risk difference, any model specification, adjustment variables, or details of the confidence-interval calculation. It also gives no p-value. Without these details, it is not possible to independently reproduce or fully audit the sponsor’s inference from the registry record.

The cluster-randomized design matters for analysis. General explanation: when participants are grouped into clusters, outcomes from people in the same cluster may be correlated. An analysis that treats every participant as fully independent can understate uncertainty because it ignores the shared cluster context. Appropriate analysis methods account for the design and the within-cluster dependence. The registry explicitly cautions that a simple individual-level re-analysis would be misleading and presents only the sponsor’s analysis. The randomization unit, number and sizes of clusters, correlation assumptions, and specific adjustment method are not reported in the registry record.

The registry lists planning assumptions for illustrative sample-size calculations: a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The illustrative effect is defined as a stated fraction of the observed effect. These are planning illustrations based on the observed effect, not the trial’s own design justification and not evidence of observed power. No sample-size justification, target effect used in the original design, or specific planned sample-size result is reported in the registry record. The registry’s notes distinguish participant-per-arm illustrations from total-event illustrations for time-to-event outcomes; the primary outcome is not described as time-to-event.

It is important to distinguish the sponsor’s reported analysis from an unadjusted recalculation. The record does not present a simple individual-level recalculation for this outcome, because the cluster-randomized design would make that misleading. Accordingly, the reported risk difference and confidence interval should not be mixed with estimates calculated under assumptions that disregard clustering. The method behind the sponsor’s own result remains unspecified in the registry record, limiting further methodological appraisal.

What does the interval tell the reader?

It communicates uncertainty around the reported contrast. Since it crosses zero, the result is compatible with no difference and with contrasts in either direction. It does not establish that the approaches are equal, and it does not establish which named approach is favored. The missing direction convention and unit add further limits to interpretation.

Why is the analysis method important?

The method determines how the design, outcome, and dependence among participants are represented in the estimate and its uncertainty. For a cluster-randomized study, accounting for clustering is especially relevant. Because the method is not reported, readers cannot assess whether particular modeling choices or adjustments affected the estimate, nor reproduce the interval from the registry entry alone.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("cRCT: Enhanced Group; Referred Peers (n)", "cRCT: Standard Group; Referred Peers (n)")
n      <- c(137, 104)
events <- c(41, 41)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  cRCT: Enhanced Group; Referred Peers (n)  137  41  29.9  22.4  38.3  22.9  38.1  22.9  38.1
ARM  cRCT: Standard Group; Referred Peers (n)  104  41  39.4  30.0  49.5  30.6  49.0  30.6  49.1

11. Analyses Reported by the Sponsor

The registry reports the sponsor’s own primary analysis as a risk difference of -6, with a 95% confidence interval from -26 to 11. It does not report a p-value or name the statistical method used. The groups are identified by codes in the analysis entry, but the mapping from those codes to enhanced and standard referral is not stated in the record information.

No simple individual-level recalculation is presented. The registry notes that the trial is cluster-randomized and that a simple re-analysis treating participants as independent would be misleading. That caution is important: a recalculation that ignores cluster dependence would not be a suitable check of the sponsor’s inference. The sponsor’s estimate is therefore the only reported analysis to interpret, rather than one estimate among competing recalculations.

The lack of a recalculation does not independently verify the sponsor’s methods. The confidence interval and estimate cannot be reproduced from the available entry because the analysis method, group mapping, and underlying group-specific outcome counts are not reported in the registry record. The best reading is accordingly descriptive and bounded: the sponsor reported the stated contrast and interval, but the available details do not permit a full assessment of how the estimate was obtained.

12. Multiplicity, Missing Data and Analysis Populations

The registry lists 16 outcome measures and 6 analyses. It names several secondary outcomes, including pilot referral and HIV self-test use, recent HIV testing, PrEP continuation among referred peers and index peers, and PrEP adherence. The registry record does not report whether the primary outcome was designated within a prespecified hierarchy, whether multiplicity adjustments were planned, or how results across the listed outcomes were intended to be interpreted together.

General explanation: when investigators assess multiple outcomes or perform multiple statistical tests, the chance of obtaining a seemingly notable result by chance can increase. A prespecified primary outcome and an appropriate multiplicity strategy can help clarify which findings are confirmatory and which are exploratory. Without the analysis plan or multiplicity information, the reported primary confidence interval should be interpreted as the interval reported for that analysis, not as evidence that all comparisons across the study have been accounted for.

The available registry entry does not describe missing-data procedures for the primary outcome. It does not report how unreported initiation status was classified, whether any participants or referrals lacked follow-up, or whether the analysis used imputation or another approach. The count is reported by index peers, making the handling of absent or incomplete reports relevant, but the actual extent and treatment of missing information are not reported in the registry record.

Analysis populations are also not described in enough detail to establish which participants or referrals contributed to the primary estimate. Although overall enrollment is 211, the number included in the primary analysis and the group-specific denominators are not reported. The record does not state whether analyses followed an intention-to-treat principle, were restricted to a particular population, or used another population definition. These omissions limit assessment of potential selection and missing-data effects and make the result difficult to reproduce.

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 non-serious adverse event terms

The registry lists serious adverse events as zero in each of the six reported groups: the pilot index peers, pilot referred peers, cRCT enhanced-group index peers, cRCT standard-group index peers, cRCT enhanced-group referred peers, and cRCT standard-group referred peers. Deaths are also listed as zero in every group. The numbers at risk are reported separately for each group, with the group labels distinguishing index peers from referred peers.

The cRCT index-peer groups each have one other adverse event affected; the other listed groups have zero. These counts should not be confused with serious adverse events: the record reports no serious adverse events in the listed groups, while the two index-peer cRCT groups each have one event in the other-adverse-event category. The registry does not provide further descriptions of these events or clarify their nature, timing, severity, or relationship to study activities.

These figures describe the events recorded in the registry for the listed groups. They do not, by themselves, establish that no other events occurred outside the recorded categories or observation periods. The methods for safety monitoring and event ascertainment are not reported in the registry record. The figures should therefore be read as the posted counts, not as a complete characterization of safety.

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

The primary result is a sponsor-reported risk difference of -6 for PrEP initiation among referred peers over 3 months, with a 95% confidence interval from -26 to 11. The interval spans zero and values on either side of it. This means the posted estimate does not determine a clear direction or precise magnitude for the contrast. The registry does not report a p-value, the analysis method, the unit of the estimate, or the subtraction order used for the comparison.

The outcome is narrower than the broader goals described for the study. It concerns reported initiation among referred peers; it does not directly answer whether index peers continued PrEP, whether referred peers adhered to PrEP, or whether HIV testing changed. Those are among the other outcomes listed in the registry, but their results are not included in the information available for this interpretation. Nor does the primary result establish effects on HIV incidence or other clinical outcomes.

The result should not be generalized automatically to all young women, all PrEP users, or settings beyond the study described in Kenya. The registry gives overall enrollment as 211, but does not report the primary-analysis denominator or group-specific counts. The population and outcome therefore cannot be reconstructed in greater detail from the entry.

In general, an interval that includes zero is not proof of no difference or equivalence. It indicates that the estimate is uncertain under the analysis used and that the stated interval is compatible with more than one direction. Conversely, the result is not evidence of a definite benefit or harm for either named referral approach. The direction and uncertainty cannot be fully assessed without the group mapping and method details. The study’s completed status does not remove these interpretive limits.

15. Why This Trial Matters Statistically

This trial is statistically interesting because it evaluates a peer-delivered prevention strategy in a cluster-randomized setting. The intervention combines referral with training, educational material, HIV self-test delivery, and referral cards, while the comparator uses informal word-of-mouth referral and referral cards. The primary outcome focuses on whether referred peers initiated PrEP, connecting the referral model to a measured prevention behavior.

The cluster design is central to interpreting the result, not a technical footnote. Participants within clusters may share experiences or referral networks, so their outcomes can be related. Accounting for that dependence is needed to represent uncertainty appropriately. The registry’s warning against an individual-level re-analysis is therefore important for readers assessing the reported confidence interval.

The result also illustrates why an estimate should be read together with its uncertainty and reporting detail. The point estimate alone does not settle direction or magnitude; the interval crosses zero, and the record omits the analysis method, group mapping, p-value, and group-specific outcome counts. In this setting, those absences limit independent appraisal more than a bare point estimate can convey.

Finally, the registry describes a broader research program spanning pilot refinement and a hybrid trial, with outcomes beyond initiation. The primary result addresses one part of that program. Keeping the initiation endpoint distinct from continuation, adherence, HIV testing, and implementation outcomes helps prevent a single contrast from being treated as an answer to every question the study set out to explore.

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 details needed to assess the primary result are not reported in the registry record. These include the statistical method, the randomization unit and cluster structure, the analysis group mapping, group-specific outcome counts and denominators, and the p-value. The risk-difference unit and subtraction order are also not stated in the reported result. These omissions prevent a full independent reconstruction and make the numerical direction harder to assign to the named approaches.

The endpoint is reported by index peers, but the registry does not explain whether initiation was verified, how missing reports were treated, or what analysis population was used. Overall enrollment of 211 does not identify how many participants or referred peers contributed to the outcome. The record also does not provide details about the implementation of standard referral or the fidelity of either approach during the trial.

The primary outcome has a 3 months timeframe. The record does not give longer-term primary-outcome follow-up or indicate how the primary initiation measure relates to later continuation and adherence. Although those outcomes are listed as secondary outcomes, their results are not part of the information available for this interpretation.

The registry lists multiple outcomes and analyses but does not report multiplicity handling, the statistical analysis plan, sample-size justification, or interim analyses. The planning assumptions described for illustrative sample-size calculations are not a substitute for the trial’s original design justification. Safety entries provide counts but not event descriptions or ascertainment details. Together, these limitations call for a bounded reading of the posted contrast rather than a broad conclusion about the strategy or its effects beyond the measured endpoint.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Peer PrEP Referral + HIV Self-test Delivery for PrEP Initiation in Kenya — NCT04982250.
• Registry record last updated 2025-11-03; status last verified 2025-10.
• Results first posted 2025-08-07.
• 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

NCT04982250 is a completed randomized, parallel-group study with no masking, titled “Peer PrEP Referral + HIV Self-test Delivery for PrEP Initiation in Kenya.” The registry lists enrollment of 211, a start date of 2022-02-24, and primary completion on 2024-06-05. The phase is listed as NA.

The study compares enhanced peer PrEP referral with HIV self-test delivery against standard peer PrEP referral. Its primary outcome is the number of referred peers initiating PrEP among all referred peers, as reported by index peers, over 3 months. The reported risk difference is -6, with a 95% confidence interval from -26 to 11. The registry does not report the analysis method or p-value, and notes that a simple individual-level re-analysis would be misleading because of the cluster-randomized design.