The Risk of HIV Acquisition Among Traditional Healers in South Africa: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for The Risk of HIV Acquisition Among Traditional Healers in South Africa (NCT04440813), a randomized na study. The primary outcome is Blood Exposure.

2. Clinical Question

3. Trial Design

The registry describes a randomized, parallel-group study with no masking. Its listed enrollment is 136, and the status is completed. The study began on 2021-08-23, with primary completion on 2023-01-25. The listed phase is NA. Vanderbilt University Medical Center is identified as the sponsor.

The two arms compared implementation strategies for PPE education and use. The clinician training arm was the control. It included PPE education and training, general HIV prevention education and skill building, and educational outreach and coaching visits. The other arm involved training and education from traditional healers and clinicians, in addition to those shared educational components. The description of the latter strategy emphasizes healers who had adopted and used PPE. The registry descriptions of the arm-specific visits are truncated, so further details of their delivery are not reported in the registry record.

Random allocation supports a comparison of assigned strategies, but the registry record does not report the allocation sequence method or concealment procedures. Masking is listed as none; further details about who knew assignment are not reported in the registry record. The parallel design means participants were assigned to one of the strategies rather than crossing between them, according to the listed design classification. The record does not describe whether participants could receive other training outside the assigned strategy.

The primary outcome was assessed at 7 months. Although the study title and rationale concern HIV acquisition, the posted primary outcome is a measure of glove use relative to razor blade use. The registry record does not report an HIV acquisition outcome among the posted measures. It lists one outcome measure and no posted analysis records.

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

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Some of the traditional healers did not conduct any traditional vaccinations during the study period, thus our numbers are smaller than the total number of participants

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 Blood Exposure, measured over 7 months.

The primary endpoint is the number of glove pairs used per razor blade used, measured at 7 months. It is a continuous measure, and the reported statistic is the mean with standard deviation. The measure relates PPE use to razor blade use in the traditional practice described in the registry. It can therefore represent glove use in relation to an activity that may involve exposure to patient blood, rather than simply counting glove use without regard to razor use.

As an implementation outcome, the measure is relevant to whether the training strategies are associated with protective behavior during the practice being studied. It is not itself a measure of blood exposure, HIV infection, or the occurrence of an injury. Nor does the posted endpoint establish whether gloves were used correctly or consistently in every relevant circumstance. The registry does not report further validation details or a threshold defining adequate use.

The registry notes that some traditional healers did not conduct traditional vaccinations during the study period, and that the outcome numbers were consequently smaller than the total participant count. This means the primary-outcome analysis population does not match the full enrollment description. The precise reasons for each participant's lack of an outcome, and the handling of missing outcome data, are not reported in the registry record.

7. Results Posted in the Registry

• Healer + Clinician Training: mean 1.90 versus 1.90 on control; difference 0.00 (95% CI -0.28 to 0.28), standardized effect d = 0.00; z-test p = 1.000.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous, so the registry reports an average for each arm and a standard deviation describing variation among observations. The mean summarizes the center of the observed values. The standard deviation describes how dispersed those values are around the mean; it is not the uncertainty in the mean itself. The posted means are 1.9 for each arm, with standard deviations of 0.79 for clinician training and 0.73 for healer-plus-clinician training.

What does the mean difference represent?

The comparison is expressed as the mean in the healer-plus-clinician arm minus the mean in the clinician training arm. Its reported estimate is 0.0. A positive value would indicate a higher average number of glove pairs per razor blade in the healer-plus-clinician arm; a negative value would indicate a lower average. Here the reported arm means are equal at the displayed precision, yielding a zero estimated difference. This is a comparison of group averages, not a statement that every participant used the same number of gloves.

What does the p-value describe?

The unadjusted p-value is 1.0. In general, a p-value measures how unusual the observed data, or data more extreme, would be under a specified null hypothesis and statistical model. It is not the probability that the null hypothesis is true, and it does not measure the size or practical importance of a difference. The record does not give enough detail to identify all assumptions or the exact test used to calculate this p-value. No adjustment is reflected in the posted comparison.

Why does the confidence interval matter?

A confidence interval gives a range of values that are compatible with the data under the method used to construct it. For this mean difference, the interval includes zero and extends in both directions. That indicates uncertainty about whether the underlying average difference favors one strategy or the other. It also helps show how precisely the difference is estimated: an interval can include zero while still allowing differences of some size. Interpretation should consider the endpoint's meaning and what differences would matter in practice. A threshold for meaningful change is not reported in the registry record.

What is the role of standard deviation and sample size?

Standard deviation describes participant-level variation, whereas the standard error describes uncertainty in an estimated mean. The registry lists standard errors for the two arm means as 0.108 and 0.094. In general, estimates tend to be more precise when there is less variability or more information, although the appropriate calculation also depends on the design and analysis. The outcome counts differ by arm, and both are smaller than the total enrollment. The registry attributes the reduced outcome numbers to healers who did not conduct traditional vaccinations during the study period. It does not report how missing observations were handled.

What can planning calculations tell us?

The record describes a planning illustration based on the observed effect, with a two-sided significance level of 0.05, equal allocation, and a normal approximation. It says the effect to detect is expressed as a stated fraction of the observed effect. Such calculations describe hypothetical future planning under assumptions; they are not observed power and do not change the uncertainty in the trial's result. The record does not report its own design justification or sample-size rationale. The calculated planning counts are not reported in the registry record.

The registry lists one outcome measure and no posted analysis records. It therefore does not report the complete statistical analysis plan, the model used, adjustment covariates, the approach to outliers or distributional concerns, or sensitivity analyses. General statistical considerations, such as whether a continuous outcome is well represented by its mean and standard deviation, cannot be resolved from the posted summaries alone. The reported summary comparison should be read as the analysis information available in the registry, not as evidence that unreported methods were or were not used.

10. Confidence Intervals and Precision

The estimated difference in mean glove pairs used per razor blade is 0.0. The reported confidence interval runs from -0.282 to 0.282 when expressed to three decimal places. It includes zero and permits differences in either direction, so the estimate does not precisely establish which strategy has the higher underlying mean. The interval is centered on the reported estimate and provides a range of values compatible with the comparison under the method used.

The interval's width is important alongside the point estimate. Identical arm means alone can make the comparison seem conclusive, but the interval expresses sampling uncertainty around the mean difference. Whether the range excludes differences large enough to matter depends on a threshold for meaningful change, and no such threshold is reported in the registry record. Thus the interval should not be read as evidence of equivalence.

For context, the posted standard deviations are 0.79 and 0.73, while the standard errors of the arm means are 0.108 and 0.094. These statistics describe different features: standard deviations characterize spread among observations, while standard errors characterize uncertainty in estimated means. The endpoint counts are 53 and 60. The smaller endpoint population relative to total enrollment affects how broadly the result can be generalized to all enrolled participants, but the record does not report the missing-data mechanism or methods for addressing it.

The registry does not report the confidence-interval construction method or a prespecified precision target. The interval is therefore best interpreted as the posted uncertainty estimate for the mean difference, without assuming details about its derivation beyond what is recorded.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists one outcome measure, identified as Blood Exposure, and no posted analysis records. The primary outcome is specified as glove pairs used per razor blade at 7 months. No secondary outcomes are posted. With one reported primary outcome, the registry does not document a family of multiple outcome tests, but it also does not report a multiplicity strategy, such as adjustment across outcomes or time points. The analysis plan and any additional analyses are not reported in the registry record.

The record describes random allocation, but does not report the allocation sequence method, concealment, or details of the analysis population. It lists outcome counts of 53 and 60 for the two arms; these are smaller than the total enrollment of 136. The registry explains that some healers did not conduct traditional vaccinations during the study period, resulting in smaller outcome numbers. It does not report the full accounting of missing observations by reason, whether those without outcome data differed from those included, or how missing values were handled.

Consequently, it is not possible to determine from the registry whether the comparison followed an intention-to-treat approach, a complete-case approach, or another specified population definition. The record does not report imputation, sensitivity analyses for missing data, or procedures for handling protocol deviations. These omissions limit assessment of whether the posted comparison could be affected by differential availability of the endpoint.

Randomization can support causal comparisons when assignment and follow-up are appropriately handled, but the summaries alone do not establish that all analysis assumptions were met. No adjustment variables, model specification, or interim analysis plan are reported in the registry record. The posted unadjusted comparison should therefore be interpreted narrowly as the result recorded for the endpoint, rather than as a complete account of all analyses that may have been planned or conducted.

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 reports no serious adverse events in either arm: 0 of 69 at risk in clinician training and 0 of 67 at risk in healer-plus-clinician training. It also reports no other adverse events in either group and no deaths. These figures describe the safety entries in the registry; they do not establish that no adverse experiences occurred outside the categories or reporting procedures represented by the record.

The at-risk counts differ from the endpoint counts, which are 53 and 60. These are distinct reported denominators for different parts of the record and should not be substituted for one another. The registry does not explain the difference in safety and endpoint populations or give additional details about adverse-event collection, follow-up, definitions, or attribution.

Zero recorded events means no events were entered for these categories in the registry. It does not prove that the underlying risk is zero, especially when the record does not describe the ascertainment process or provide further safety analyses. There are no serious adverse-event terms listed. These entries should be read as reported counts, not as a comprehensive guarantee about safety.

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

The posted primary result compares mean glove pairs used per razor blade at 7 months. Both arms have a reported mean of 1.9, and the estimated mean difference is 0.0. The unadjusted p-value is 1.0, and the confidence interval includes zero. Taken together, these statistics show no difference between the reported means while retaining uncertainty about the underlying difference.

The result does not demonstrate that the strategies are equivalent, because the registry does not report an equivalence margin or a non-inferiority design. It also does not establish that either strategy increased PPE use compared with no training: both arms received training and education, and the posted comparison is between the two strategies. The record does not include a no-intervention comparator for this outcome.

The endpoint measures glove use relative to razor blade use. It does not directly measure blood exposure, correct glove use, injuries, or HIV acquisition. The study title and rationale address HIV acquisition risk, but an HIV acquisition result is not among the posted outcomes. It would therefore be incorrect to translate the endpoint comparison into a conclusion about HIV incidence or prevention.

The result applies to the observations contributing to the endpoint, not necessarily every participant enrolled. The registry notes that some healers did not conduct traditional vaccinations during the study period, leading to outcome counts below total enrollment. Missing-data handling and broader analysis-population details are not reported in the registry record.

In general, a null estimated difference can reflect genuinely similar means or limited ability to distinguish a difference. The confidence interval helps describe this uncertainty, but the registry does not state what difference in glove use would be meaningful for implementation or exposure reduction. The most defensible interpretation is limited to the recorded endpoint, comparison, and follow-up time, without extending it to clinical outcomes or unreported analyses.

14. Why This Trial Matters Statistically

This trial raises a statistical question about how implementation strategy can be evaluated when both groups receive active education and support. The comparison is between clinician-led training and training involving traditional healers and clinicians, rather than between training and no training. That distinction matters: the estimate addresses the relative performance of the two approaches on the recorded endpoint, not the absolute effect of receiving education.

The endpoint connects reported glove use to razor blade use, a practice described in the registry as relevant to blood exposure. Framing the measure as a ratio-like count per blade gives the behavior a practice-specific context. At the same time, it remains a behavioral outcome and should not be treated as a direct measurement of infection or exposure. The separation between an implementation measure and a clinical outcome is central to interpreting the study's result.

The posted means are identical at 1.9, with an estimated mean difference of 0.0. The confidence interval shows uncertainty around that estimate, so the result illustrates why point estimates and precision need to be considered together. A lack of observed separation between group means does not by itself establish that the strategies are interchangeable, particularly without a stated threshold for a meaningful difference.

The registry also highlights a common issue in outcome analysis: the number contributing to a particular endpoint may be smaller than total enrollment. Here, the record links the reduced outcome numbers to healers who did not conduct traditional vaccinations during the study period. Clear accounting of eligible participants, missing outcomes, and analysis populations is essential for judging how representative an endpoint comparison may be.

Finally, the record lists no analysis records despite posting summary comparisons. This leaves important methodological details, including model specification and missing-data handling, unreported. Statistical interpretation must therefore be calibrated to the information available rather than assuming a complete analysis plan.

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.

The primary outcome population is smaller than total enrollment, and the registry attributes this to some healers not conducting traditional vaccinations during the study period. The record does not give a detailed flow of participants or explain the outcome availability for each arm beyond the reported counts. It also does not report whether missing outcome data were related to assignment or observed behavior, or how missingness was addressed. This limits assessment of selection and missing-data bias.

The endpoint is a proxy for protective behavior: glove pairs used per razor blade used. It does not directly record blood exposure, adherence in every procedure, HIV infection, or whether protective equipment was used correctly. No threshold for adequate glove use or a clinically meaningful difference is reported. The posted outcome therefore cannot answer whether the intervention strategies changed HIV acquisition risk.

The registry record does not provide detailed information about randomization procedures, allocation concealment, the statistical model, analysis population, adjustment factors, or sensitivity analyses. Arm descriptions are truncated, limiting assessment of exact delivery and fidelity. Masking is listed as none, but further details about awareness of assignment are not reported in the registry record.

The result is a comparison of two active strategies. Without a no-training control, the posted means do not estimate the effect of training compared with no training. The record also lists no posted analysis records and does not report an equivalence or non-inferiority margin. Thus the identical reported means cannot establish that the approaches are equivalent, and unreported details should not be inferred.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: The Risk of HIV Acquisition Among Traditional Healers in South Africa — NCT04440813.
• Registry record last updated 2024-06-18; status last verified 2024-06.
• Results first posted 2023-03-22.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

NCT04440813 is a completed randomized, parallel-group study sponsored by Vanderbilt University Medical Center. It enrolled 136 traditional healers and compared clinician training with training and education from traditional healers and clinicians. Masking is listed as none, and the listed phase is NA.

The primary outcome was mean glove pairs used per razor blade at 7 months. The reported means were 1.9 in each arm, with an estimated mean difference of 0.0 and an unadjusted p-value of 1.0. The confidence interval includes zero. Outcome counts were 53 and 60; the registry notes that some healers did not conduct traditional vaccinations during the study period.

The registry lists one outcome measure and no posted analysis records. No HIV acquisition outcome is posted. Further analysis-plan and missing-data details are not reported in the registry record.