Evaluation of the Healthy Love Workshop, an HIV Prevention Intervention for African American Women: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Evaluation of the Healthy Love Workshop, an HIV Prevention Intervention for African American Women (NCT00362375), a randomized phase 1, phase 2 study. The primary outcome is Condom Use During Vaginal Sex With Any Male Partner.

2. Clinical Question

3. Trial Design

This was a randomized, parallel-group behavioral intervention evaluation listed as Phase 1 and Phase 2. The registry identifies no masking. Participants were assigned to the Healthy Love Workshop or the HIV/AIDS 101 Workshop; the record does not report further details about sequence generation, allocation concealment, or how randomization was implemented. The evaluation was sponsored by the Centers for Disease Control and Prevention and was conducted by SisterLove, Inc. In Atlanta, Georgia, as stated in the official title.

The Healthy Love Workshop was a single-session, small-group HIV prevention intervention. HIV101 was a single-session, small-group workshop providing details about HIV/AIDS. This active comparison is important when interpreting the findings: the contrast is between the workshop approaches as described in the registry, not between receiving a workshop and receiving no information or contact.

The registry lists enrollment of 313, a start date of 2006-05, primary completion in 2008-04, and a completed status. It identifies HIV infections and sexually transmitted infections as conditions. The available summary does not establish whether participants were followed for outcomes beyond the stated timeframes, nor does it give additional recruitment or site details beyond the Atlanta location in the official title.

The primary outcome concerns condom use during vaginal sex with any male partner in the past three months, assessed among women who were sexually active at three-month follow-up and had outcome data. The arm-level outcome entries list 93 for Healthy Love and 94 for HIV101. Those outcome-specific values do not replace the overall enrollment figure, and the record does not explain the relationship between the enrolled population and the analyzed endpoint population in detail. Randomization describes the planned group assignment; it does not by itself establish that the reported outcome population retained all randomized participants or that missing outcome data were handled in a particular way.

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: 28 groups did not participate for logistical reasons, including lack of follow-up by group contact (10 groups), failure of women to arrive for the workshop (7 groups), and too few women to meet minimum workshop size (6 groups). Five groups did not participate because of insufficient interest by their members in receiving either workshop.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Women who were sexually active at 3-month follow-up and who provided data on outcome measure

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 Condom Use During Vaginal Sex With Any Male Partner, measured over Past 3 months.

The primary endpoint is the percentage of women who used condoms during vaginal intercourse with any male partner during the past three months. It is restricted to women who were sexually active at three-month follow-up and had data for this measure. The reported arm values are 59.3 percent for the Healthy Love Workshop and 39.6 percent for HIV101. The endpoint therefore describes reported behavior over a specified retrospective period, rather than a laboratory-confirmed infection or a count of sexual encounters.

Condom-use behavior is relevant to an HIV prevention program because it captures a protective behavior that an intervention may aim to support. It is not equivalent to a direct measure of HIV incidence or of all sexual risk. The registry lists HIV infections and sexually transmitted infections as conditions, but the primary outcome is behavioral. It should be interpreted as the specified self-reported behavior in the defined follow-up population.

Self-report can be affected by recall and by how respondents understand or answer questions. The registry record does not report the survey instrument, response options, validation procedures, or whether outcome assessors were masked. Nor does it report whether the endpoint was defined as any condom use, consistent use, or another threshold beyond the wording that women “used condoms” during vaginal intercourse. That detail matters because a percentage can represent different behaviors depending on the precise operational definition.

Using the past-three-month timeframe gives the outcome a concrete reference period, while restricting the population to sexually active women makes the behavior relevant to participants with an opportunity to report it. The restriction also means the result should not be generalized automatically to women who were not sexually active at follow-up or lacked outcome data. The record does not report a clinical threshold for interpreting the percentages.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The study is listed as randomized, with a parallel model and no masking. General explanation: randomization is intended to make assigned groups comparable on average at the start of a study, while a parallel design assigns participants to a group that receives its assigned intervention rather than crossing participants between interventions. These design features can support causal comparisons when assignment, follow-up, measurement, and analysis are implemented appropriately. The registry record does not report enough detail to assess the randomization procedure, concealment, adherence, or all aspects of follow-up.

What does the endpoint percentage represent?

The primary measure is a percentage of women reporting condom use during vaginal sex with any male partner in the past three months. The record specifies that the population comprises women who were sexually active at three-month follow-up and had endpoint data. The arm entries list 93 for Healthy Love and 94 for HIV101, with corresponding percentages of 59.3 and 39.6. These are descriptive summaries for the endpoint population, not estimates for every person in the overall enrollment figure.

General explanation: percentages are useful for describing how common a response was within a defined group, but they do not by themselves provide a measure of uncertainty or identify the reason for a difference. Comparisons also depend on consistent outcome definitions and comparable ascertainment across groups. The registry does not report the instrument or detailed response coding for this measure.

How should the odds ratio be read?

The reported model is a generalized estimating equation, with an odds ratio of 2.40, a 95 percent confidence interval from 1.28 to 4.50, and p <0.05. The analysis lists HIV101 as the group and Healthy Love Workshop as the control. General explanation: an odds ratio compares the odds of a defined event between groups. It is not a ratio of the percentages themselves, and it is not a percentage-point difference. The event definition and the reference group determine the direction of the interpretation.

Generalized estimating equations are a modeling approach often used when observations may be correlated, such as repeated measurements or observations grouped within units. The registry does not state the reason this approach was selected, whether repeated observations were modeled, or the working correlation structure. It also does not report whether the model adjusted for covariates. These missing details prevent a full assessment of the estimate’s specification and make it important not to infer unreported model features.

General explanation: a confidence interval conveys the precision of an estimate under the model and its assumptions. A wider interval indicates greater uncertainty than a narrower interval on the same scale. The interval reported for this odds ratio does not resolve the ambiguity about outcome coding. A p-value summarizes compatibility with a null model under specified assumptions; it is not a direct measure of the size, importance, or reliability of an effect.

Why is there no simple recalculation?

The registry record indicates that the posted summaries do not support an estimable two-arm comparison for this endpoint type. Accordingly, a simple unadjusted re-calculation cannot be used to independently reproduce or check the reported GEE odds ratio from the arm percentages alone. The descriptive percentages remain available as descriptive results, but they are not enough to reconstruct the model estimate, its interval, or its p-value.

General explanation: a model-based estimate can depend on details beyond arm-level percentages, including how the outcome was encoded and how the model handled its observations. A recalculation from summary values is not a substitute for the analysis performed on participant-level data. In this case, the registry’s own analysis is the inferential result available, while the record’s missing model details constrain interpretation.

What do the planning illustrations mean?

The planning illustrations use the observed effect as a reference and specify a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. They are framed around detecting stated fractions of the observed effect. These are planning assumptions, not the trial’s reported original design justification, and the registry record does not report the original sample-size justification. Such calculations illustrate how a future study’s required enrollment could vary with the effect chosen for detection and the design assumptions.

General explanation: planning calculations estimate the sample size needed to meet a chosen design target under stated assumptions. They do not measure the probability that the completed study detected an effect, and they are not observed power. Since no planning outputs are stated in the registry record for this explanation, no numerical planning requirement can be interpreted.

The record reports five analyses, four secondary outcomes, and eight outcome measures overall. These counts alone do not identify which analyses were confirmatory, how the outcomes were ordered, or whether multiplicity adjustments were made. The statistical methods available in the registry are therefore enough to describe the reported primary model at a high level, but not enough to reconstruct a full analysis plan.

10. Analyses Reported by the Sponsor

The registry reports the GEE analysis as an odds ratio of 2.40, with a 95 percent confidence interval from 1.28 to 4.50 and p <0.05. The analysis lists HIV101 as the group and Healthy Love Workshop as the control. There is no flagged discrepancy between an independent recalculation and this result; in fact, the registry indicates that a simple two-arm recalculation is not estimable from the posted summaries for this endpoint type.

That means the reported GEE result is the inferential estimate available in the registry, rather than a result that can be checked by recalculating an unadjusted odds ratio from the arm percentages. The descriptive values—59.3 percent for Healthy Love and 39.6 percent for HIV101—should still be reported as stated, but they do not reproduce the model estimate. The record does not report outcome coding or sufficient model details to explain why the odds ratio’s apparent direction may not match those percentages under a condom-use event definition.

General statistical interpretation: when a model result and descriptive summaries seem to point in different directions, it is important to check event coding, reference categories, and model specification before assigning a substantive direction. Those details are not reported in the registry record. The GEE result should therefore be quoted with its uncertainty and significance statement, while avoiding a directional interpretation that the record cannot substantiate.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists eight outcome measures overall, four secondary outcomes, and five analyses. Those counts indicate that the primary endpoint is part of a broader set of reported measures, but they do not establish how many hypothesis tests were performed, which were prespecified as confirmatory, or whether the analyses were independent. The registry record does not report a multiplicity adjustment, a testing hierarchy, or a statistical analysis plan that would show how the risk of false-positive findings across outcomes was addressed.

General explanation: when several outcomes or analyses are examined, the chance of obtaining at least one apparently notable result can increase if each result is interpreted independently. Adjustment procedures or a prespecified hierarchy can help control that risk, but the approach needs to be documented to evaluate its role. No particular adjustment or absence of adjustment should be assumed from the reported counts alone.

The endpoint population is defined as women who were sexually active at three-month follow-up and had outcome data. The record does not report the number of women excluded from this endpoint population, the amount or pattern of missing data, or the method used to address missing outcomes. It also does not say whether the primary analysis followed intention-to-treat principles, used a complete-case approach, or applied another analysis population definition. The overall enrollment and arm-level endpoint entries do not, by themselves, explain these issues.

General explanation: missing outcome data can affect estimates if the likelihood of having data is related to assigned group or to the outcome. The impact depends on the missingness mechanism and the analytic method, which cannot be assessed without additional information. The endpoint-specific restriction to sexually active women also defines who contributes to the reported measure; it is not a statement about all enrolled participants.

The registry reports a GEE method for the primary analysis but does not state its covariate set, correlation structure, handling of missing observations, or sensitivity analyses. It also does not report interim analyses or stopping rules. These omissions limit assessment of analysis choices and robustness. They do not establish that a particular procedure was or was not used; they mean the relevant details are not reported in the registry record.

12. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

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 list 0 serious events for the Healthy Love Workshop group and 0 for HIV101. The recorded numbers at risk are 161 and 152, respectively. The registry lists no serious adverse event terms among the top reported terms. Death figures are not reported in the registry record.

These entries describe what the safety record reports; they should not be read as proof that no adverse experiences occurred. The record does not provide the full adverse-event collection method, event definitions, follow-up duration for safety, or details about whether participants were actively monitored. It also does not explain the relationship between the safety populations and the primary endpoint populations. The numbers at risk in the safety entries should therefore not be substituted for the endpoint-specific arm entries.

General explanation: a reported count of zero serious adverse events means that none were entered for that category in the registry record. It does not establish the absence of all harms, particularly when the scope and completeness of collection are not described. A safety comparison also needs context about exposure and observation, which is not available in detail.

The workshops are behavioral interventions, and the registry records no serious events in the listed groups. That is a concise description of the posted safety figures, not a conclusion about comparative safety or about events that may not have been captured or reported.

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

The result describes reported condom-use behavior among women who were sexually active at three-month follow-up and had outcome data. The arm-level percentages are 59.3 percent for Healthy Love Workshop and 39.6 percent for HIV101. The registry also reports a GEE odds ratio of 2.40, a 95 percent confidence interval from 1.28 to 4.50, and p <0.05. Those model statistics indicate a reported statistical association under the analysis, but incomplete details about outcome coding and model specification prevent a confident directional interpretation of the odds ratio.

The comparison is between the Healthy Love Workshop, a single-session, small-group HIV prevention intervention, and HIV101, a single-session, small-group workshop providing details about HIV/AIDS. The findings therefore concern these two workshop approaches in the evaluated population and setting. They do not establish what would happen with no workshop, a different intervention, or a different participant population.

The primary endpoint is a self-reported behavior measure, not a measure of HIV infection, sexually transmitted infection, or long-term clinical outcome. Although the registry lists HIV infections and sexually transmitted infections as conditions, the reported primary result does not show whether either condition changed. Nor does the behavioral measure capture every aspect of sexual risk or protection.

General explanation: a statistically notable result does not, by itself, establish that an intervention caused the observed pattern. Causal interpretation depends on how randomization was conducted, the completeness of follow-up, the validity of measurement, and whether the analysis matches the intended comparison. The record identifies the study as randomized but does not report enough detail to evaluate each of these considerations.

The result also cannot be generalized automatically beyond the defined endpoint population. Women who were not sexually active at follow-up or who lacked endpoint data are not represented in that outcome definition. The record does not report enough information to assess how those exclusions might affect interpretation. The appropriate reading is therefore specific: the registry reports these arm summaries and a model result for a behavioral endpoint, with important interpretive details not reported.

14. Why This Trial Matters Statistically

This evaluation is statistically informative because it contrasts two active behavioral workshops and reports both descriptive arm percentages and a model-based analysis. The intervention was a single-session, small-group HIV prevention workshop, while the comparison was a single-session, small-group HIV/AIDS information workshop. That design makes the question relevant to the incremental difference between two forms of workshop content, rather than the general effect of receiving any program contact.

The endpoint is also defined by both a recall period and an eligible follow-up population. It focuses on condom use during vaginal sex with any male partner in the past three months among women who were sexually active and had outcome data. Clear population and timeframe definitions are essential to interpreting a behavioral percentage. At the same time, the endpoint is not a clinical infection measure and depends on self-report.

The statistical record illustrates why descriptive summaries and model estimates should not be treated as interchangeable. The posted percentages are 59.3 and 39.6, while the reported GEE odds ratio is 2.40 with a 95 percent confidence interval from 1.28 to 4.50 and p <0.05. The outcome coding and model details are not reported, and a simple unadjusted comparison is not estimable from the posted summaries. This makes transparent reporting of model direction, event definition, and analysis choices especially important.

The registry also makes visible limits that often matter in trial interpretation: the endpoint population is narrower than overall enrollment, multiplicity procedures are not reported, and the missing-data approach is not described. These are not grounds to infer a flaw or to dismiss the result. They identify what is needed to assess precision, generalizability, and robustness more fully.

More broadly, the trial illustrates how evidence about behavior, model-based estimates, and registry reporting each answer different questions. Keeping those distinctions clear helps readers avoid treating a behavioral endpoint as a clinical outcome or interpreting an odds ratio without knowing its coding.

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 registry record does not report the survey instrument, exact response coding, or procedures used to validate the self-reported condom-use endpoint. The wording identifies condom use during vaginal intercourse with any male partner in the past three months, but it does not clarify whether the measure reflects any use, consistent use, or another threshold. Recall and response behavior may affect self-reported measures; the record does not quantify these influences.

The reported GEE odds ratio is difficult to interpret directionally alongside the arm-level percentages because the event coding and model specification are not stated. The registry identifies HIV101 as the analysis group and Healthy Love Workshop as the control, but does not report enough detail to determine how the modeled outcome was encoded, which covariates were included, or how correlation was handled. The result should be quoted as reported without assigning a direction that depends on unreported details.

The primary endpoint is restricted to women who were sexually active at three-month follow-up and had outcome data. The record does not report missing-data counts or methods, the number excluded from the endpoint population, or sensitivity analyses. Consequently, it is not possible to assess whether the endpoint population differs materially from the randomized population or how missing outcomes may affect the result.

The registry lists eight outcome measures, four secondary outcomes, and five analyses, but does not report a multiplicity plan or identify which analyses were primary, confirmatory, or exploratory. It also does not report interim analyses or a complete statistical analysis plan. These absences limit evaluation of the analysis process without establishing what procedures were used outside the registry record.

Finally, the record gives limited detail about recruitment, the randomization procedure, intervention adherence, and the extent of follow-up. The endpoint is behavioral and does not directly assess HIV infection or sexually transmitted infection outcomes. These limitations narrow the conclusions to the reported behavior and the available analysis, rather than a broader claim about clinical outcomes or all eligible women.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Evaluation of the Healthy Love Workshop, an HIV Prevention Intervention for African American Women — NCT00362375.
• Registry record last updated 2010-05-14; status last verified 2010-04.
• Results first posted 2010-05-14.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

The completed randomized, parallel-group evaluation compared the Healthy Love Workshop with HIV101, a workshop providing details about HIV/AIDS. Both interventions are described as single-session, small-group behavioral workshops. The registry lists Phase 1 and Phase 2, overall enrollment of 313, a start date of 2006-05, and primary completion in 2008-04. The sponsor was the Centers for Disease Control and Prevention.

The primary endpoint was the percentage of sexually active women with outcome data who reported condom use during vaginal intercourse with any male partner in the past three months. The posted values were 59.3 percent for Healthy Love and 39.6 percent for HIV101. A GEE analysis reported an odds ratio of 2.40, a 95 percent confidence interval from 1.28 to 4.50, and p <0.05. Outcome coding and key model details are not reported, so the odds ratio’s direction should not be inferred from the percentages alone.

The registry lists four secondary outcomes and eight outcome measures overall. It records 0 serious adverse events in each group and does not report death figures. Missing-data methods, multiplicity procedures, and a full statistical analysis plan are not reported in the registry record.