A Randomized Trial of ImpACT+, a Coping Intervention for HIV Infected Women With Sexual Trauma in South Africa: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Randomized Trial of ImpACT+, a Coping Intervention for HIV Infected Women With Sexual Trauma in South Africa (NCT04793217), a randomized na study. The primary outcome is Number of Participants With Viral Suppression at the 12-month Assessment.

2. Clinical Question

The trial is framed around both clinical and mental-health aims. Its stated purpose was to examine ImpACT+ in relation to clinical outcomes after ART initiation and to explore mental-health and behavioral mechanisms relevant to viral suppression. The primary outcome posted in the registry is a count of participants meeting the viral-suppression threshold at the 12-month assessment. The record also names measures of care engagement and ART adherence as secondary outcomes.

This comparison addresses the relative outcomes of two behavioral interventions in the population described; it is not a comparison with no intervention. The registry record does not report enough detail to determine how the results apply to women outside the stated population, to other settings, or to other care circumstances. The eligibility criteria and the full protocol are not reported in the registry record.

3. Trial Design

The registry describes Project Someleze as a randomized, double-masked, parallel-group trial. The two behavioral interventions were ImpACT+ and adapted problem-solving therapy. The recorded enrollment is 350. The trial status is completed, with a start date of 2021-02-18 and a primary completion date of 2025-07-03. The phase is listed as not applicable. Columbia University is named as the sponsor.

In a parallel design, participants are assigned to one of the intervention groups and the groups are compared over the study period. Random allocation can help balance measured and unmeasured participant characteristics across groups on average, although the record does not report the sequence-generation method, allocation concealment, or the details of how masking was implemented. The record labels the trial double-masked but does not identify which people were masked or whether masking was maintained throughout follow-up.

The ImpACT+ description identifies individual coping support addressing traumatic stress and HIV-care engagement. The control intervention is described as a brief adapted PST based on Problem Management Plus, with a psychoeducational focus. Full descriptions of either intervention, including the number and length of sessions, are not reported in the registry record. The ImpACT+ arm description ends mid-sentence, so the complete registered description is not available.

The primary outcome was assessed at 12 months. The registry says the analytic sample included participants who completed that assessment and participants who missed the final assessment but had a 12-month viral-load result in the National Health Laboratory Service database. It also states that 250 participants completed the assessment. The relationship between this statement and the arm-specific primary outcome denominators is not explained in the registry record. The enrollment count should therefore not be treated as the denominator for every outcome: the posted counts differ across outcomes.

The registry lists additional outcomes for missed visits, visit adherence, gaps in care, visit constancy, and measures of ART adherence. Numerical comparisons are posted for some, but not all, of these listed outcomes. The record does not report a full statistical analysis plan, the original sample-size justification, or the details of interim analyses.

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: 350 participants were enrolled. One participant withdrew from the study prior to randomization and thus is not counted here. Randomization occurred at the participant level. No units other than participants were assigned in this study.

Overall Study

Analysis population for the primary outcome, as defined in the registry: The primary outcome was only assessed at the 12-month assessment. Those 250 that completed the assessment and those individuals who did not attend their final assessment but had viral loads at 12 months in the National Health Laboratory Service database were included in the analytic sample.

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: One participant was never randomized to a condition, and two from each condition withdrew from the study and are not counted in any analyses or descriptions of the cohort.

6. Primary Endpoint

The registry lists the primary outcome as Number of Participants With Viral Suppression at the 12-month Assessment, measured over Up to 12 months.

The primary endpoint is the number of participants with viral suppression at the 12-month assessment. Viral suppression was defined as HIV-1 RNA below 50 copies/ml, measured using the COBAS AmpliPrep/COBAS TaqMan HIV-1 Test, version 2.0, from Roche. In this endpoint, each participant is classified according to whether their viral load meets the registered threshold, and the outcome is reported as a count within each arm.

Viral load is a clinical measure of HIV activity. A threshold-based endpoint makes the result interpretable as the number of participants meeting a specified laboratory criterion at the assessment. It does not describe the full range of viral-load values: participants on either side of the threshold may have different values, and those differences are not represented by the binary count alone.

The registry description also mentions continuous viral-load measures modeled using a log10 transformation. However, the posted primary result is the count meeting the suppression threshold, and a continuous viral-load analysis is not reported in the registry record. The reason for selecting the 12-month assessment as the primary time point and the original rationale for the endpoint are not reported in the registry record.

The stated analytic population includes participants with a completed 12-month assessment and some who did not attend the final assessment but had a 12-month viral-load result in a national laboratory database. This approach means that the primary endpoint is not described as limited only to participants who attended the final assessment. The registry does not report how missing viral-load information was handled beyond that description.

7. Results Posted in the Registry

• Adapted Problem-Solving Therapy: 84.3% (107/127) versus 93.9% (124/132) on control; risk difference -9.7 percentage points (95% CI -17.2 to -2.2), risk ratio 0.90 (0.82–0.98); unadjusted two-proportion z-test p = 0.012.

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.

HIV Care Engagement- Number of Participants With Gaps in Care

• 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:
  • Adapted Problem-Solving Therapy: 8.1% (14/172) versus 11.0% (19/173) on control; risk difference -2.8 percentage points (95% CI -9.0 to 3.4), risk ratio 0.74 (0.38–1.43); unadjusted two-proportion z-test p = 0.369.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Adapted Problem-Solving Therapy and the control. The interval gives the range of effects that remain compatible with the data.

Number of Participants With ART Adherence - Dried Blood Spots

• 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:
  • Adapted Problem-Solving Therapy: 71.2% (84/118) versus 76.2% (93/122) on control; risk difference -5.0 percentage points (95% CI -16.2 to 6.1), risk ratio 0.93 (0.80–1.09); unadjusted two-proportion z-test p = 0.375.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Adapted Problem-Solving Therapy and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists secondary outcomes covering missed visits, visit adherence, gaps in care, visit constancy, and ART adherence. Numerical comparisons are posted for gaps in care and for ART adherence measured with dried blood spots. Additional listed measures include pharmacy visits and total pills dispensed, but their results are not reported in the registry record.

Across the two numerical secondary comparisons, the estimates do not establish a clear difference between groups: both confidence intervals include zero, and both unadjusted p-values are greater than 0.05. This does not demonstrate that the interventions have identical outcomes. It indicates that these results, as reported, leave uncertainty about whether a group difference exists and about its direction and size.

The primary viral-suppression comparison has a different pattern: its posted interval lies below zero and its unadjusted p-value is 0.012. Secondary outcomes can add context to a primary outcome, because care engagement and adherence are related areas of measurement, but they should not be treated as direct confirmation of the primary result. They measure different behaviors or indicators, have different denominators, and may be based on different available observations. The registry does not report analyses linking individual secondary outcomes to viral suppression or testing a mediation pathway.

Unadjusted secondary findings carry less weight when multiple outcomes are considered and no multiplicity strategy is reported. When many comparisons are examined, some may appear notable by chance. The registry record does not identify which secondary outcomes were prespecified as confirmatory, whether any testing sequence was planned, or whether adjustment was used. The secondary results therefore provide descriptive evidence about the posted measures, rather than a firm basis for claims about a broad intervention effect across all care-engagement and adherence outcomes.

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

• HIV Care Engagement - Number of Participants With Missed Visits
• HIV Care Engagement- Visit Adherence
• HIV Care Engagement- Number of Participants With Gaps in Care
• HIV Care Engagement- Visit Constancy
• Number of Participants With ART Adherence - Dried Blood Spots
• ART Adherence - Pharmacy Visits
• ART Adherence - Total Number of Pills Dispensed

9. Statistical Interpretation

10. Statistical Methods Explained

Outcome scale and arm comparison. The primary outcome is binary at the participant level: a participant either meets the viral-suppression threshold or does not. The registry presents the numbers meeting the threshold and the numbers assessed in each arm. The comparison is reported as a risk difference, which is the difference between the proportions meeting the endpoint in the two groups. A value below zero for adapted PST versus ImpACT+ indicates a lower observed proportion in the adapted PST group. This is a description of the direction of the posted contrast, not an explanation of its cause.

Risk differences and relative measures. A risk difference expresses a contrast in absolute proportion units. A relative risk compares the proportions as a ratio, while an odds ratio compares the odds of the outcome. Each scale answers the comparison in a different way. Relative measures can appear more or less pronounced than absolute differences, especially when the outcome is common or uncommon. They are not separate outcomes, and agreement among them does not amount to replication because they are derived from the same group data. The registry posts all three forms for the primary endpoint, but does not report the specific statistical model used to obtain them.

Confidence intervals. A confidence interval describes the uncertainty around an estimated contrast under the assumptions of the method used to calculate it. A narrower interval indicates greater precision; a wider interval leaves a broader range of values compatible with the data and method. For a risk difference, zero represents no difference between groups. An interval that excludes zero is consistent with a nonzero contrast at the interval's stated confidence level, while an interval that contains zero includes no difference among the values compatible with the interval. The registry does not identify the confidence level or calculation method for the posted intervals, so those details should not be inferred.

p-values. A p-value measures how surprising data at least as discrepant as those observed would be under a specified null model. It is not the probability that the null hypothesis is true, and it does not measure the size or practical importance of a contrast. The registry reports unadjusted p-values for the comparisons. An unadjusted value does not account for the number of outcomes or comparisons considered. The analysis plan, the test used, and any prespecified threshold for interpreting trial results are not reported in the registry record.

Randomization and masking. Random allocation is intended to make treatment groups comparable on average at the start of a trial. Its value depends on implementation and on how outcomes and missing data are handled. The registry labels the design randomized and double-masked, but does not report sequence generation, concealment, masking roles, or masking success. The design label alone cannot answer those implementation questions. A parallel-group structure means the groups receive their assigned intervention rather than crossing over between interventions, as described in the registry.

Analysis population and denominators. The primary description includes participants who completed the 12-month assessment and those who missed the final assessment but had a viral-load result in the national laboratory database. It does not state a general intention-to-treat rule, explain the treatment of all missing measurements, or identify a per-protocol population. Denominators vary across the posted outcomes. In general, outcome-specific denominators can reflect different availability of measurements or analysis populations, but the reason for each difference in this trial is not reported in the registry record.

Binary outcomes and dispersion. For a binary endpoint, the relevant uncertainty depends on the number of participants with and without the event in each group and on the analysis method. The registry does not post a dispersion estimate for these outcomes. Over-dispersion is a general concern when observed variability exceeds what a model assumes; it can make standard errors too small if ignored. That issue is most directly relevant to certain count or clustered-data models, and should not be presumed to apply to this trial's posted binary comparisons. The registry does not describe a model for clustering or over-dispersion.

Continuous viral-load description. The registry says continuous viral-load measures were modeled using a log10 transformation, but does not post a continuous-outcome estimate or model details. In general, transforming a skewed measurement can make its distribution more suitable for a statistical model and can change how effects are interpreted. A log transformation does not itself establish that model assumptions were met. The data, diagnostic checks, and resulting continuous viral-load analyses are not reported in the registry record.

Secondary outcomes and multiplicity. The registry names several measures of HIV care engagement and ART adherence and posts numerical comparisons for selected outcomes. Each additional comparison creates another opportunity for a result to appear notable by chance if no multiplicity strategy is used. This is a general statistical concern, not evidence that any specific result is false. The registry does not report whether outcomes were hierarchically tested, adjusted for multiplicity, or designated as exploratory. The secondary p-values are unadjusted, so they should be interpreted with that limitation.

Planning illustrations. The registry's planning illustrations use a two-sided significance level of 0.05, equal allocation, and a normal approximation. They consider effects defined as stated fractions of the observed effect. These are planning calculations based on the observed contrast, not documentation of how the trial was originally designed. The original sample-size justification and its assumptions are not reported in the registry record.

11. Confidence Intervals and Precision

The primary comparison has a confidence interval that lies entirely below zero for the risk difference of adapted PST versus ImpACT+. This means the posted interval does not include an equal observed proportion under the method used to form it. The interval still expresses uncertainty about the size of the group contrast. The registry does not report the confidence level or the method used to calculate the interval, so the interval should be interpreted without assuming those details.

Precision is distinct from statistical significance. A result can have an interval that excludes zero while still allowing uncertainty about the magnitude of the contrast. The arm counts are 124 of 132 for ImpACT+ and 107 of 127 for adapted PST; the interval applies to the contrast between groups, not to each count separately. No new estimate should be inferred from those counts beyond the registry's posted comparison.

The secondary intervals are less conclusive in relation to a zero difference: both include zero. Their ranges therefore allow both a difference in one direction and no difference, and in some cases a difference in the other direction. This is a statement about uncertainty, not proof that the groups are identical. The registry does not report additional analyses that could narrow or otherwise clarify these estimates.

Precision depends on the amount of outcome information, the balance and size of the groups, the event frequency, and the method of analysis. Those factors should be considered alongside the interval rather than treating a p-value as a complete summary. Denominators differ across the trial's outcomes, which affects how much information is available for each comparison. The registry does not give a full account of why outcome-specific data are unavailable for some participants.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are expressed as participants per arm, not as the total number across both arms. They are illustrations computed from the observed effect, using a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The calculation asks how many participants per arm would be needed to detect a specified fraction of the observed effect under those assumptions.

For the primary comparison, planning for 100% of the observed effect gives 162 participants per arm for 80% power and 216 participants per arm for 90% power. Planning for 75% of the observed effect gives 260 participants per arm for 80% power and 347 participants per arm for 90% power. Planning for 50% of the observed effect gives 519 participants per arm for 80% power and 694 participants per arm for 90% power.

In plain terms, the 80% power column gives the planning sample size associated with an 80% chance of detecting the specified effect under the calculation's assumptions, if that effect and model assumptions hold. The 90% power column describes the corresponding planning target for a higher chance of detection. These are planning concepts, not guarantees about the results of a completed study. Smaller effects require more participants to detect under the same assumptions, which is why the figures increase as the target effect is reduced.

These values are not observed power. They are not the trial's original design justification: the registry's own sample-size assumptions and rationale are not reported in the registry record. Other outcome-specific illustrations are based on their respective observed contrasts, so their planning figures should not be substituted for those of the primary outcome. The registry's stated assumptions do not include a reported allowance for missing outcome data or other adjustments.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.939; treatment proportion p₁ = 0.843; average p̄ = 0.891.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.4408 = 0.8640.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.4354 = 0.3665.
4. n = (0.8640 + 0.3665)² / (-0.0969)² = 161.3, rounded up to 162 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
p0 <- 0.939393939393939                          # control proportion
p1 <- 0.84251968503937                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(4793217)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 162 (80% power), 216 (90% power)
 75% of observed effect: n per arm = 260 (80% power), 347 (90% power)
 50% of observed effect: n per arm = 519 (80% power), 694 (90% power)
power.prop.test: 161.3 
Simulated power at n = 162 per arm: 0.817 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists multiple primary and secondary outcomes, but it does not report a strategy for controlling multiplicity. This leaves open whether the primary endpoint was intended as the sole confirmatory test, whether secondary outcomes were exploratory, or whether a testing hierarchy or adjustment procedure was planned. The reported p-values are unadjusted. They should not be interpreted as accounting for the full set of outcomes or comparisons considered in the trial.

The registry does not report a complete statistical analysis plan, the model used for the primary comparison, the prespecified confidence level, or the planned handling of all missing observations. It also does not report interim analyses or stopping rules. These omissions limit the ability to distinguish analyses specified before looking at outcomes from analyses chosen after data were available. Absence of a description in the registry is not evidence that a method was or was not used outside the record.

The primary population description is specific: participants who completed the 12-month assessment and participants who missed the final assessment but had a 12-month viral-load result in the National Health Laboratory Service database were included. It also states that 250 completed the assessment. The posted arm denominators for the primary endpoint do not match that number when considered together, and the registry does not explain how the assessment-completion statement relates to those denominators. The record also does not describe the handling of participants without a qualifying laboratory result.

Different denominators across the primary, secondary, and safety results indicate that the posted measures do not all use an identical count of participants at risk or assessed. Reasons for those differences are not fully described. The record does not provide a complete flow from enrollment through assignment, follow-up, and analysis, nor does it identify a general intention-to-treat or per-protocol analysis population. As a result, the extent and consequences of missing data cannot be fully assessed from the registry record.

15. 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

Serious events with the highest arm-level rate: hospitalization - pregnancy related (6.9%), hospitalizations (unrelated to pregnancy) (2.3%), sexual trauma (1.2%), community violence (1.2%), intimate partner violence (0.6%).

The safety figures report serious adverse events in the two intervention groups, with 173 participants at risk in the ImpACT+ group and 172 in adapted PST. The registry records 21 serious adverse events for ImpACT+ and 7 for adapted PST. It records zero other adverse events in each group and deaths of 1 and 2, respectively. These are counts as posted; the registry does not explain whether the counts represent unique participants, repeated events, or how relatedness to an intervention was assessed.

The registry's listed serious-event terms include pregnancy-related hospitalization at 6.9%, hospitalization unrelated to pregnancy at 2.3%, sexual trauma at 1.2%, community violence at 1.2%, intimate partner violence at 0.6%, and suicide attempt at 0.6%. The record does not provide a denominator or arm-specific breakdown for these listed percentages, so they should not be assigned to an arm or converted into participant counts.

Event counts alone do not establish that an intervention caused an event or that differences between groups are meaningful. The registry does not report a safety-analysis method, a comparison with uncertainty intervals, or details of event adjudication. These figures describe what is recorded, not a causal assessment.

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

The posted primary data describe a difference between the two randomized groups in the number meeting the viral-suppression threshold at the 12-month assessment. The risk-difference interval lies below zero for adapted PST compared with ImpACT+, and the unadjusted p-value is 0.012. This is a group-level statistical result within the registered trial; it does not establish the mechanism behind the difference or how an individual participant would fare under either intervention.

The result does not show that every participant in one group had a better or worse outcome than every participant in the other group. It also does not describe changes in viral load over time, how quickly suppression occurred, or whether it was sustained. The primary endpoint is a threshold-based count, while the registry's mention of continuous viral-load modeling is not accompanied by posted continuous-outcome results.

The comparison is between two behavioral interventions. It does not establish the effect of ImpACT+ relative to no support or usual care. The adapted PST group is an active comparator, and the registry describes it as a psychoeducational intervention. The reported contrast therefore answers a narrower question about outcomes in the two assigned groups.

The secondary results do not establish clear group differences in the measures with posted comparisons. They also do not establish that no difference exists. Their uncertainty, unadjusted status, and outcome-specific denominators all matter. The registry does not report a multiplicity procedure or detailed analysis plan, so the secondary findings should not be taken as definitive evidence for or against effects across the full set of care and adherence outcomes.

Finally, the registered population, follow-up point, and analytic description define the scope of the result. The registry does not report complete eligibility criteria, all missing-data procedures, or enough details to assess generalizability to populations and settings beyond those described. Statistical evidence about this trial should be kept distinct from clinical recommendations, which cannot be derived from these aggregate registry results alone.

17. Why This Trial Matters Statistically

This trial raises a statistical question that is common in behavioral and clinical research: how to interpret a threshold-based clinical outcome when participants are assigned to two active interventions and not all outcomes share the same denominator. Viral suppression is represented as a count meeting a laboratory threshold, while the trial also names behavioral and care-engagement measures. Keeping those outcome types distinct helps prevent a result on one measure from being generalized to every aspect of care.

The randomized design is important because it is intended to support a group comparison, but design labels are not a substitute for reporting methods. Sequence generation, masking implementation, the analysis plan, and missing-data procedures all affect how confidently readers can understand the estimates. The registry record identifies the trial as randomized and double-masked but leaves those operational and analytic details unreported.

The record also illustrates why effect estimates and intervals matter alongside p-values. A p-value alone does not show the size or precision of a contrast. For the primary endpoint, the posted interval is wholly on one side of zero; for the posted secondary comparisons, the intervals include zero. The distinction helps readers recognize that the primary and secondary results carry different levels of statistical evidence without turning uncertainty into certainty.

Multiplicity and planning are also relevant. The registry includes several secondary outcomes, but does not describe a strategy for interpreting them collectively. Its sample-size illustrations are explicitly based on observed effects and specified assumptions; they are not evidence of the trial's original design rationale or its realized ability to detect an effect. Together, these features make transparent reporting especially important for evaluating the limits of the findings.

18. 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 leaves important details unresolved for interpreting this particular trial. The primary analytic description says that 250 participants completed the 12-month assessment and that people with a laboratory viral-load result could also be included if they missed the final assessment. The posted primary arm denominators are 132 and 127, and the record does not reconcile those counts with the completion statement. A full participant-flow account is not reported.

The denominators vary across outcomes. For example, the primary viral-suppression comparison and the posted care-engagement, dried-blood-spot, and safety results use different counts. The record does not explain all differences or state whether the same eligibility and missing-data rules were applied to each outcome. This limits direct comparison across outcomes and prevents a complete assessment of potential selection from missing measurements.

The trial description identifies the interventions only briefly. The ImpACT+ arm description is incomplete, and the full intervention schedule, delivery procedures, fidelity assessment, and participant adherence to sessions are not reported in the registry record. The complete eligibility criteria, recruitment details, and number assigned to each group at randomization are also not reported.

Analytic details are limited. The record does not report the original sample-size justification, a full statistical analysis plan, the method used for the confidence intervals, or a multiplicity strategy. Although the registry mentions continuous viral-load measures modeled with a log10 transformation, corresponding results and model checks are not reported. The mechanisms named in the trial summary are not accompanied by posted mediation analyses.

Safety reporting is also limited to counts and listed serious-event terms. The registry does not provide arm-specific breakdowns for the listed terms, event attribution, or an analysis that quantifies uncertainty around safety contrasts. These reporting gaps do not determine what occurred outside the registry; they define what can be evaluated from the record.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: A Randomized Trial of ImpACT+, a Coping Intervention for HIV Infected Women With Sexual Trauma in South Africa — NCT04793217.
• Registry record last updated 2026-08-28; status last verified 2026-08.
• Results first posted 2026-08-28.
• 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.

22. Record Summary

The record describes a completed, randomized, double-masked, parallel-group trial sponsored by Columbia University, with a recorded enrollment of 350. It compares ImpACT+, an individual coping intervention, with adapted problem-solving therapy among South African women with HIV and sexual trauma histories. The primary endpoint is the count with viral suppression at the 12-month assessment, defined as HIV-1 RNA below 50 copies/ml.

The posted primary comparison favors ImpACT+ in the direction of the observed group proportions: the risk-difference interval for adapted PST versus ImpACT+ lies below zero, and the unadjusted p-value is 0.012. Posted secondary comparisons do not establish clear differences. The record lists additional secondary outcomes and safety counts, but does not provide a complete analysis plan, multiplicity approach, original sample-size rationale, or full explanation of outcome denominators and missing data.