Improving Health Outcomes in Young Cisgender Men and Transgender Women (mLab App): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Improving Health Outcomes in Young Cisgender Men and Transgender Women (mLab App) (NCT03803683), a randomized na study. The primary outcome is Total Number of Youth Tested for HIV.

2. Clinical Question

The central question is whether more youth were tested for HIV by 6 months in the group assigned to the mLab App than in the standard-of-care HIV information group. A second comparison asks whether the HIV home-test group differed from standard of care on the same endpoint. These are separate comparisons with a shared control group; the registry record does not report a direct comparison between the mLab App and HIV home-test groups.

3. Trial Design

The registry describes a randomized, single-masked, parallel-group trial with an enrollment of 525. It is marked completed. The start date is 2020-01-21, and primary completion is recorded as 2023-10-02. The phase is listed as NA. The identity of the party or parties masked is not reported in the registry record.

Three arms are described: mLab App intervention, standard-of-care HIV information control, and HIV home tests. The primary endpoint is reported at 6 months. The primary outcome population is described as participants who completed the 6-month visit and from whom data were collected: 175 in the mLab App arm, 189 in the standard-of-care arm, and 94 in the HIV home-test arm. These analyzed group sizes are not the same as the trial's overall enrollment, and the record does not provide a full account of participant flow or the reasons that participants were not included in this endpoint population.

The two posted primary comparisons use the standard-of-care arm as the reference. This permits a separate assessment of each intervention arm against standard of care, but does not establish a randomized direct contrast between the two intervention arms in the results described. The registry record does not report the original sample-size justification, the detailed allocation procedure, or the full protocol for masking. Further operational details of the arm descriptions are incomplete in the registry record, so they should not be inferred.

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: 525 participants enrolled, however 523 were randomized as two participants withdrew prior to randomization.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: 175 in Arm1, 189 in Arm 2, and 94 in Arm 3 analyzed as this was the number of participants that completed the 6 month visit and from which data was collected.

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 Total Number of Youth Tested for HIV, measured over 6 months.

The primary endpoint is the total number of youth tested for HIV at 6 months, reported as a count of participants. For the primary outcome, the registry describes the analysis population as people who completed the 6-month visit and from whom data were collected. The posted counts are 154 in the mLab App arm, 123 in the standard-of-care arm, and 68 in the HIV home-test arm.

This endpoint is useful for assessing testing uptake: it records whether participants reached the testing milestone during the stated follow-up period. It is a behavioral or service-use outcome, rather than a measure of whether the tests were correct or whether an individual had HIV. The endpoint also does not, by itself, establish whether someone who tested received follow-up care. Linkage to HIV care appears among the secondary outcomes, but numerical results for that outcome are not posted in the record.

A count alone can be hard to compare when analyzed group sizes differ. The unadjusted comparisons therefore also express the contrast between the count of participants tested and the number analyzed in each arm. These comparisons describe the observed groups; they do not remove possible effects of missing follow-up information or establish that the same pattern would occur in other populations.

7. Results Posted in the Registry

• mLab App Intervention: 88.0% (154/175) versus 65.1% (123/189) on control; risk difference 22.9 percentage points (95% CI 14.6 to 31.2), risk ratio 1.35 (1.20–1.52); unadjusted two-proportion z-test p = <0.001.
• HIV Home Tests: 72.3% (68/94) versus 65.1% (123/189) on control; risk difference 7.3 percentage points (95% CI -4.1 to 18.6), risk ratio 1.11 (0.94–1.31); unadjusted two-proportion z-test p = 0.219.

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.

Number of Sexual Partners in the Past 3 Months

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • mLab App Intervention: mean 2.12 versus 1.69 on control; difference 0.43 (95% CI -0.35 to 1.21), standardized effect d = 0.11; z-test p = 0.279.
  • HIV Home Tests: mean 1.78 versus 1.69 on control; difference 0.09 (95% CI -0.55 to 0.73), standardized effect d = 0.03; z-test p = 0.782.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between mLab App Intervention and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between HIV Home Tests and the control. The interval gives the range of effects that remain compatible with the data.

Number of Condomless Anal Sex Acts

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • mLab App Intervention: mean 7.35 versus 8.21 on control; difference -0.86 (95% CI -4.70 to 2.98), standardized effect d = -0.05; z-test p = 0.660.
  • HIV Home Tests: mean 7.56 versus 8.21 on control; difference -0.65 (95% CI -4.33 to 3.03), standardized effect d = -0.03; z-test p = 0.730.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between mLab App Intervention and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between HIV Home Tests and the control. The interval gives the range of effects that remain compatible with the data.

The numerical secondary results cover two areas: number of sexual partners in the past 3 months and number of condomless anal sex acts, measured at 6 months. Across both outcomes and both intervention-versus-control comparisons, the unadjusted confidence intervals include no difference. Their estimates point in different directions across outcomes, and the posted results do not establish a consistent secondary pattern.

These measures are related to sexual behavior, whereas the primary endpoint measures HIV testing uptake. A difference in testing does not imply a difference in these behaviors, and a lack of clear secondary evidence does not negate the primary testing result. The outcomes capture different aspects of prevention and care. The registry lists HIV results and linkage-to-care outcomes as secondary outcomes as well, but numerical results for them are not posted in the registry record.

Secondary findings generally carry less weight when many outcomes or comparisons are examined without a reported multiplicity strategy. Repeated testing increases the chance that some results may appear unusual by chance. Here, the registry record does not report an adjustment for multiple comparisons, a prespecified hierarchy among secondary endpoints, or the corresponding sponsor analyses. The unadjusted p-values are therefore best treated as exploratory evidence rather than definitive confirmation.

There is also incomplete reporting across the listed outcomes: repeated labels appear in the record, while only some outcomes have numerical group summaries and comparisons. The absence of posted estimates is not evidence that an outcome showed no effect. It means the direction, magnitude, and uncertainty for those outcomes cannot be evaluated from the registry record.

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

• Total Number of HIV Negative/ Positive Results
• Total Number of HIV Negative/ Positive Results
• Total Number of Youth Who Link to HIV Care Services
• Total Number of Youth Who Link to HIV Care Services
• Number of Sexual Partners in the Past 3 Months
• Number of Sexual Partners in the Past 3 Months
• Number of Condomless Anal Sex Acts
• Number of Condomless Anal Sex Acts

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is a participant count: whether a participant was tested by the 6-month assessment. For statistical comparison, the count is interpreted relative to the number analyzed in each group. This makes the risk difference a contrast between the proportions tested, rather than a subtraction of raw counts from groups of different sizes. The recalculated mLab App and home-test contrasts are unadjusted comparisons with the standard-of-care group.

What does a risk difference express?

A risk difference expresses the difference in the observed probability of an outcome between two groups. A positive value for this endpoint indicates a higher observed testing probability in the intervention group than in standard of care; a negative value indicates a lower one. A risk difference is on the probability scale and is often easier to relate to an absolute change than a relative measure. The exact recalculated risk-difference estimates are based on the posted group counts and analyzed sizes.

How do relative risk and odds ratio differ?

The relative risk compares the probability of testing in one group with that in the reference group. The odds ratio compares the odds, where odds are the probability of the event relative to the probability of no event. These measures answer related but distinct questions, so their numerical magnitudes are not interchangeable. Both are recalculated unadjusted summaries. They do not account for potential differences between groups beyond the randomized assignment, nor do they replace the absolute group counts.

What do the p-values say?

The posted p-values are unadjusted. In general, a p-value measures how compatible the observed data are with a specified no-difference model, under the assumptions of the analysis. It is not the probability that the no-difference model is true, and it does not measure the size or practical importance of an effect. For the primary mLab App comparison, the unadjusted p-value is <0.001. For the home-test comparison, it is 0.219. The registry record does not report sponsor analyses for these results, so the estimates and tests discussed are recalculations from the posted summary statistics.

How are the secondary outcomes summarized?

The two secondary measures with numerical results are means and standard deviations: number of sexual partners in the past 3 months, and number of condomless anal sex acts, each reported at 6 months. A mean difference compares group averages; a positive difference means the intervention-arm average is higher than the control average, while a negative difference means it is lower. The standard deviation describes variation among observations within a group. It is not the uncertainty around the mean difference; that uncertainty is reflected in the confidence interval.

For the mLab App comparison, the recalculated mean differences are 0.43 partners and -0.86 condomless anal sex acts. Their confidence intervals include zero, and the unadjusted p-values are 0.279 and 0.66, respectively. For the home-test comparison, the corresponding mean differences are 0.09 and -0.65, with unadjusted p-values of 0.782 and 0.73. These unadjusted comparisons do not provide strong evidence of group differences on those outcomes. The record does not report model-based analyses or adjustment for covariates.

General statistical note: outcomes that are counts or averages can have distributions that differ from the assumptions used in simple comparisons. Over-dispersion, for example, means that count outcomes vary more than a simple model expects; it can make standard errors too small if not addressed. The registry record does not report whether such distributional features were assessed or whether alternative models were used. The recalculated summaries should therefore be understood as simple comparisons of the posted group statistics, not as a complete reconstruction of the trial's planned analysis.

The registry lists additional secondary outcomes, including HIV-negative or HIV-positive results and linkage to HIV care services, but numerical results for these outcomes are not posted. The record also lists repeated outcome labels. The existence of an outcome listing alone does not reveal its direction, magnitude, or statistical uncertainty.

11. Confidence Intervals and Precision

A confidence interval describes a range of effect values compatible with the observed data under the statistical method used. In general, narrower intervals indicate greater statistical precision, while wider intervals leave more uncertainty about the magnitude and sometimes the direction of an effect. An interval that includes the no-difference value does not establish equivalence; it indicates that the data do not clearly distinguish the comparison from no difference at the interval's stated confidence level.

For the primary mLab App comparison, the recalculated risk-difference interval lies entirely above zero. The relative-risk interval lies entirely above 1, and the odds-ratio interval also lies entirely above 1. These intervals align with the direction of the posted counts and the unadjusted p-value <0.001. They indicate that the primary contrast is not merely a small, imprecisely estimated direction in these recalculations. They do not address unmeasured outcomes or guarantee the same effect in a different setting.

For the home-test comparison, the risk-difference interval crosses zero; the relative-risk and odds-ratio intervals cross their respective no-difference values. The estimates are therefore less precise about whether the contrast is positive or negative. The same general pattern applies to the posted secondary mean differences: their confidence intervals include zero. No confidence intervals are reported for the group means themselves in the registry record.

Precision should also be interpreted alongside the endpoint population. The primary analysis uses participants with 6-month visit data, and the registry record does not provide enough information to assess the outcomes of those without such data or the effect of missingness. An interval can quantify sampling uncertainty under its calculation assumptions, but it does not account automatically for selection or measurement bias.

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 illustrative sample sizes, not the trial's original design justification. The planning assumptions use a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The target effect is set to a stated fraction of the observed effect. The registry record does not report the trial's original sample-size assumptions.

For these non-time-to-event outcomes, the figures are participants per arm, not a total across arms. The 80% power column gives the number of participants per arm the planning calculation indicates would be needed to have an 80% chance of detecting the chosen effect under the stated assumptions, if that effect and the assumptions were correct. The 90% power column uses a higher chance of detection and therefore generally requires more participants per arm.

The rows that plan for 75% or 50% of the observed effect represent smaller target effects than planning for the full observed effect. Smaller effects are harder to distinguish from random variation, so they require more participants under these assumptions. The figures for the primary outcome differ substantially between the two comparisons because the observed contrasts used as planning inputs differ. They should not be read as evidence about the actual trial's power or as a post hoc measure of whether the trial succeeded in detecting an effect.

This is not observed power. It is a planning illustration built from the observed effect and specified assumptions; it does not replace the confidence intervals, which describe the uncertainty in the observed contrasts.

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 = 0.651; treatment proportion p₁ = 0.880; average p̄ = 0.765.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5993 = 1.1746.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5769 = 0.4856.
4. n = (1.1746 + 0.4856)² / (0.2292)² = 52.5, rounded up to 53 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 <- 0.650793650793651                          # control proportion
p1 <- 0.88                          # 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(3803683)
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 = 53 (80% power), 70 (90% power)
 75% of observed effect: n per arm = 102 (80% power), 136 (90% power)
 50% of observed effect: n per arm = 246 (80% power), 329 (90% power)
power.prop.test: 52.5 
Simulated power at n = 53 per arm: 0.823 (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 record lists multiple secondary outcomes, including HIV test results, linkage to HIV care, number of sexual partners, and condomless anal sex acts. It does not report a multiplicity adjustment, a testing hierarchy, or a plan identifying which secondary comparisons should be considered confirmatory. Without such information, unadjusted secondary p-values should be interpreted cautiously: when many hypotheses are assessed, chance alone can produce apparently notable findings.

The registry lists 10 outcome measures and reports numerical results for some secondary measures, but it does not post sponsor analyses for the secondary results described in the record. Repeated outcome labels are present. The record does not explain whether repeated labels refer to separate assessments, repeated entries, or another reporting structure. It is therefore not possible to reconstruct the full family of planned tests or determine how the posted outcomes relate to any prespecified analysis hierarchy.

For the primary endpoint, the stated analysis population consists of participants who completed the 6-month visit and from whom data were collected. This is a follow-up-based analysis population. The registry record does not report a complete participant-flow account, the number or outcomes of participants missing the visit, reasons for missing data, imputation methods, or sensitivity analyses. Consequently, the possible influence of missing outcome data cannot be assessed from the posted information.

The record also does not report whether the primary analysis followed an intention-to-treat principle, whether all randomized participants were included in any analysis, or whether additional adjusted models were prespecified. Randomization is recorded, but that fact alone does not answer how missing observations were handled in the outcome analysis. These gaps limit assessment of the robustness of the estimates and of how closely the reported analysis followed the original plan.

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.

The registry reports zero serious adverse events and zero deaths in each of the three safety groups. The counts at risk are 211 for the mLab App intervention, 208 for standard-of-care HIV information, and 104 for HIV home tests. The registry also reports zero other adverse events in each group.

These figures describe the events recorded in the registry, not proof that no adverse events occurred outside the reporting definitions, observation period, or available follow-up. The safety counts at risk differ from the analyzed group sizes reported for the 6-month primary endpoint. The registry record does not explain that difference or provide event descriptions, severity details for other events, or additional safety analyses. The figures should therefore be read as the posted safety summary rather than as a complete account of all possible harms.

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

The primary posted result is an unadjusted difference in the number and proportion of analyzed participants tested for HIV by 6 months. The mLab App group had more participants tested than the standard-of-care group, and the recalculated interval for that contrast does not include no difference. This is evidence about the recorded testing endpoint in the analyzed groups.

It does not show that the app improved test accuracy, changed HIV status, or increased linkage to care. Those are different outcomes. Although linkage to care is listed among the secondary outcomes, numerical results are not posted in the registry record. The primary endpoint also does not establish whether the observed difference arose through reminders, app information, home-test feedback, or another element of the intervention package.

The home-test arm's comparison with standard care is less conclusive in the posted unadjusted analysis: its interval includes no difference. This does not demonstrate that the groups are equivalent or that a meaningful difference is impossible. The observed data and their uncertainty permit more than one interpretation.

The results do not establish a direct comparison between the mLab App and home-test arms, since the posted contrasts use standard care as the reference. Nor do the results establish what would happen in populations or settings beyond those represented by the trial. The record's incomplete information on missing data, analysis plans, and some secondary outcomes also constrains broader interpretation. Statistical evidence about one endpoint should not be extended to claims about all aspects of HIV prevention or care.

17. Why This Trial Matters Statistically

This trial illustrates why outcome definition matters. HIV testing uptake is a concrete service-use endpoint, and a participant count paired with the analyzed group size allows an absolute comparison between groups. That comparison answers a narrower question than whether the intervention changed health status or improved every step along a care pathway.

The three-arm design also highlights the importance of identifying the actual contrast being estimated. The posted results compare each intervention arm with standard of care. They do not report a direct test of the mLab App against home testing. A reader should not infer that one intervention outperformed the other merely by comparing their separate estimates against the shared control.

The results also show why effect estimates and confidence intervals are needed alongside p-values. The primary comparison is statistically clear in the recalculated unadjusted analysis, whereas the home-test and secondary comparisons have more uncertainty about direction. A non-significant result is not equivalent to evidence of no effect, and a small p-value alone does not describe clinical or practical importance.

Finally, this record underscores the value of complete reporting. The listed outcomes extend beyond testing, but numerical findings for some are absent, and details of missing-data handling and multiplicity are not reported. Such information is important for assessing whether an endpoint was emphasized in advance, how robust results are, and how much confidence to place in secondary patterns. These considerations are part of statistical interpretation, not merely administrative detail.

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.

Several trial-specific limitations affect interpretation of the posted results. The primary analysis population is described as participants who completed the 6-month visit and from whom data were collected, rather than a clearly documented analysis of all randomized participants. The registry record does not explain missing follow-up, reasons for missingness, or methods used to address it. Selection into the observed endpoint population could therefore affect how well the posted contrast represents the randomized groups.

The posted primary comparisons are unadjusted recalculations from group counts. No covariate-adjusted sponsor analysis is reported. The registry record also does not report the original sample-size justification, detailed allocation methods, the identity of the masked party, or the trial's full statistical analysis plan. The planning sample sizes are illustrative calculations from observed effects, not evidence about the original trial design.

Outcome reporting is incomplete for some listed secondary measures, including HIV results and linkage to HIV care. The record contains repeated outcome labels and does not explain their structure. It also does not report a multiplicity strategy, so unadjusted secondary results warrant caution. These gaps prevent a full assessment of the planned outcome family and the strength of secondary evidence.

Finally, the primary endpoint is testing uptake at a specified follow-up point. It does not directly evaluate test accuracy, HIV status, or care linkage. The posted comparisons use standard care as the reference and do not provide a direct mLab App versus home-test comparison. Conclusions should remain limited to the outcomes and contrasts that are actually reported.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Improving Health Outcomes in Young Cisgender Men and Transgender Women (mLab App) — NCT03803683.
• Registry record last updated 2024-11-26; status last verified 2024-11.
• Results first posted 2024-11-26.
• 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

NCT03803683 is a completed randomized, single-masked, parallel-group trial of the mLab App, standard-of-care HIV information, and HIV home tests. It enrolled 525 and assessed its primary endpoint, the total number of youth tested for HIV, at 6 months. The stated primary analysis population consists of participants with 6-month visit data.

The mLab App group had 154 participants tested among 175 analyzed, compared with 123 among 189 in standard care. The unadjusted recalculated comparison favors a higher observed testing rate in the mLab App group, with p < 0.001. The home-test comparison with standard care is less conclusive. Numerical secondary results are posted for sexual partners and condomless anal sex acts; other listed secondary outcomes lack numerical results in the registry record.

Interpretation is limited by unreported details on missing data, multiplicity, and the original analysis plan. The planning sample-size figures are illustrations based on observed effects, not the trial's original design assumptions.