Evaluation of Rapid HIV Self-testing Among MSM (eSTAMP): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Evaluation of Rapid HIV Self-testing Among MSM (eSTAMP) (NCT02067039), a randomized na study. The primary outcome is Frequency of HIV Testing by Internet-recruited MSM..

2. Clinical Question

3. Trial Design

The registry describes a randomized, open-label, parallel-group study with enrollment of 2665 and a completed status. The listed study start is 2015-03-25 and the listed primary completion is 2016-10-21. These dates identify the registry milestones; they do not by themselves describe the timing or duration of recruitment for each participant.

The study summary describes four consecutive parts. The first three were formative and intended to guide development and implementation of Part 4. The primary outcome and arm descriptions concern the randomized controlled trial. The record does not give further details about the formative parts, such as their participants, procedures, or findings.

Random allocation places the study within a design intended to make the intervention and comparison groups comparable on average, while open-label status means masking was listed as none. The registry does not report the sequence-generation method, allocation concealment procedures, or whether outcome assessment was masked. The parallel model means the listed groups were compared as separate arms rather than as a crossover sequence.

The HIV self-testing arm was described as receiving four kits and completing short follow-up surveys at three-month intervals during the 12-month period. The information-only group also completed surveys at three-month intervals. The registry does not fully state the procedures at each survey or the details of any testing access offered after follow-up. It also does not report the trial's original sample-size justification or its statistical analysis plan.

The record lists 2665 enrolled participants, while the posted primary outcome counts are 1014 and 977. Those figures describe different recorded quantities; the registry does not explain the relation between the overall enrollment and the primary outcome denominators. They should not be treated as interchangeable.

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: Participants who did not provide contact information, complete and submit their baseline interview online, provided fraudulent information or later reported an HIV diagnoses before study enrollment were excluded.

Overall Study

Analysis population for the primary outcome, as defined in the registry: percent of MSM who reported testing at least three times over the course of the RCT

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 Frequency of HIV Testing by Internet-recruited MSM., measured over 12 months.

The primary endpoint is the count of participants who reported HIV testing at least three times over a 12-month follow-up period. The registry describes it as frequency of HIV testing by internet-recruited men who have sex with men, with the population measure expressed as the percent reporting testing at least three times during the randomized trial.

This endpoint addresses a behavioral question: whether participants reported repeated HIV testing during follow-up. A binary outcome reduces each participant's report to whether the stated threshold was met, which permits the groups to be compared by the proportion meeting that definition. It does not preserve how many tests beyond the threshold an individual reported, when testing occurred, or whether all testing was verified independently. The registry record does not report verification procedures or the exact survey questions used to elicit testing frequency.

The endpoint is relevant to the study's stated public health aim of assessing whether providing self-test kits affects testing frequency. It is distinct from newly identified HIV infections, linkage to services, and infections among social network associates, which are listed as secondary outcomes. A reported difference in testing frequency therefore describes the primary behavioral outcome, not all aspects of testing access, diagnosis, or subsequent care.

7. Results Posted in the Registry

• Information Only: 22.0% (215/977) versus 76.6% (777/1014) on control; risk difference -54.6 percentage points (95% CI -58.3 to -50.9), risk ratio 0.29 (0.25–0.32); unadjusted two-proportion z-test p = <0.001.

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.

Newly Identified HIV Infections

• 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:
  • Information Only: 0.8% (11/1340) versus 1.9% (25/1325) on control; risk difference -1.1 percentage points (95% CI -1.9 to -0.2), risk ratio 0.44 (0.21–0.88); unadjusted two-proportion z-test p = 0.017.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Information Only than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The secondary outcomes broaden the questions beyond the primary measure of reported testing frequency. The registry lists linkage to HIV testing and care services after a positive rapid test, newly identified HIV infections, and HIV infections among social network associates. These outcomes concern downstream services and infection identification, rather than the behavior captured by the primary endpoint.

Numerical results are posted for newly identified HIV infections. The other listed secondary outcomes do not have numerical results in the registry record. As a result, the secondary evidence cannot be treated as a complete set of findings across all the questions named in the record. The available result concerns a different outcome from the primary one, so it should not be used as a direct confirmation or contradiction of the testing-frequency contrast.

Secondary outcomes can provide useful context by exploring possible consequences related to the main study question. However, results from multiple outcomes are more vulnerable to chance findings when many comparisons are examined. An unadjusted secondary p-value does not account for the number of outcomes tested, unless a suitable plan says otherwise. The registry does not report a multiplicity adjustment, a hierarchy for interpreting secondary endpoints, or a prespecified strategy for controlling the overall false-positive risk.

For that reason, the secondary findings generally carry less confirmatory weight than a primary outcome that was designated in advance, particularly when the analysis plan and multiplicity procedures are unavailable. This does not make secondary outcomes irrelevant; it means they should be interpreted as supporting or exploratory evidence rather than as a standalone basis for broad conclusions. The record's lack of posted results for the remaining secondary outcomes further limits any assessment of the overall pattern.

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

• Linkage to HIV Testing and Care Services Following a Positive Rapid Test Result.
• Newly Identified HIV Infections
• HIV Infections Among Social Network Associates

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is binary: each participant either did or did not report testing at least three times during the 12-month follow-up. For a binary endpoint, the basic group summary is a count and a proportion. The registry reports both the counts and denominators for the primary outcome, along with the proportions used in the comparison. This makes the outcome definition and the denominator visible, although the record does not explain why the outcome denominators are smaller than overall enrollment.

What does the risk difference express?

The risk difference subtracts one group's event proportion from the other group's event proportion. In this comparison, the reported value is -0.546 for Information Only relative to HIV Self-testing. Its negative direction indicates a lower proportion meeting the endpoint in the information-only group under that comparison order. The risk difference is an absolute measure, expressed on the proportion scale. It is distinct from a relative measure, which describes one group in relation to another.

How should the ratios be read?

The risk ratio compares event proportions, while the odds ratio compares event odds. Both can help express the direction and relative size of a contrast, but their numerical values are not interchangeable. Odds and probabilities are different quantities, so an odds ratio should not be read as though it were a direct ratio of proportions. The record reports risk ratio and odds ratio estimates and intervals, but does not state the interval confidence level or the exact method used to calculate them.

What does a p-value tell the reader?

The primary unadjusted p-value is <0.001. A p-value is a measure of how compatible the observed data are with a specified null model, under the statistical procedure used. It is not the probability that the null hypothesis is true, and it does not measure the size or importance of an effect. The registry does not report the test procedure or full analysis plan, so interpretation should stay with the value as posted and not assume details of the test.

The word “unadjusted” matters. It indicates that the reported p-value is not presented as adjusted for multiple comparisons or covariates. The registry contains no reported analyses section, and the record does not say whether other analyses were conducted but not posted. Consequently, the primary result cannot be characterized as adjusted for prognostic factors or as protected by a stated multiplicity procedure.

Why do intervals add information?

An interval around an estimate represents a range of values that are compatible with the data under a chosen statistical method. A narrow interval conveys greater precision than a broad interval, all else being equal. It does not show the full range of possible outcomes or account automatically for systematic error, missing data, or limitations in the way the outcome was measured. Because the registry does not identify the confidence level or interval method, the reported limits should be described as intervals rather than assigned an unreported confidence level.

For the primary risk difference, the interval is -0.583 to -0.509. Its endpoints are both below zero, consistent with the direction of the reported contrast. This does not resolve questions about bias, generalizability, or outcome completeness. Statistical intervals quantify uncertainty under a model; they do not capture every source of uncertainty in a study.

What is the role of randomization?

Randomization is intended to distribute both measured and unmeasured participant characteristics across study arms by chance. That supports interpreting a between-group contrast as an estimate of the effect of assignment, provided the analysis and follow-up preserve the benefits of random allocation. The registry confirms randomized allocation but does not report the sequence-generation method, concealment, or how participants with missing outcomes were handled. The arm counts therefore remain important, and the record does not support assumptions about an intention-to-treat analysis or any other analysis population.

The trial was open label, with masking listed as none. Participant awareness of receiving test kits could affect reported behavior, survey responses, or participation in follow-up. That is a general consideration in open-label studies, not evidence that reporting was biased in this particular trial. The registry does not report validation of self-reported testing or masked outcome assessment.

Why can over-dispersion matter?

General explanation: over-dispersion arises when observed variability is greater than a statistical model expects. It can make uncertainty estimates too small if a model assumes less variability than the data contain. For this primary endpoint, the posted summaries are counts for a binary outcome, and the registry does not report a model-based analysis or an over-dispersion assessment. It would therefore be inappropriate to claim that over-dispersion was assessed or affected the result.

The secondary outcome of newly identified HIV infections is also presented as participant counts and proportions, with a risk difference, ratios, intervals, and an unadjusted p-value. The same broad interpretation of binary outcomes applies. The outcome definitions, analysis methods, and missing-data handling for secondary outcomes are not detailed in the registry record. Other listed outcomes do not have posted numerical results, so their methods or estimates cannot be evaluated from this record.

Finally, the comparison should be distinguished from the planning illustrations. The posted observed outcome contrast describes the recorded trial data. The sample-size illustrations use stated assumptions to show how participant requirements would vary for selected target effects; they are not a reanalysis of the trial design and do not establish what assumptions the investigators originally used.

11. Confidence Intervals and Precision

The primary risk difference is -0.546, with reported interval -0.583 to -0.509. The interval does not cross zero, and the unadjusted p-value is <0.001. In the comparison's stated direction, the estimate and its interval indicate a lower proportion meeting the frequent-testing endpoint in the information-only group than in the HIV self-testing group. The width of an interval is one way to assess statistical precision: a narrower interval indicates less uncertainty about the estimate under the method used.

The registry also reports a risk ratio of 0.287, with interval 0.254 to 0.325, and an odds ratio of 0.086, with interval 0.070 to 0.106. The different measures describe the same group contrast on different scales. Their intervals likewise remain on one side of the relevant null value, but they should not be interpreted as interchangeable measures.

The confidence level and interval calculation method are not reported in the registry record. The displayed limits can therefore be discussed as reported intervals, but should not be labelled with a specific confidence level. Precision is also not the same as validity: these intervals do not account automatically for self-report error, missing outcomes, selection, or other systematic limitations. The primary outcome denominators and the overall enrollment are not the same recorded quantity, and the record does not explain the difference.

For newly identified HIV infections, the reported risk difference is -0.011, with interval -0.019 to -0.002 and unadjusted p-value 0.017. That interval is comparatively close to zero on the risk-difference scale. The estimate should be read with its own event counts and endpoint definition; its precision cannot be judged solely by comparison with the primary behavioral endpoint, which measures a different outcome.

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 participant counts per arm, not total events. The listed assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The illustrated target effect is a stated fraction of the observed effect. These are planning illustrations based on the observed contrast, not the trial's original design assumptions; the original sample-size justification is not reported in the registry record.

The 80% power column gives the illustrative number of participants per arm for a design with an 80% chance of detecting the selected target effect under the stated assumptions, if that effect and assumptions were correct. The 90% power column uses a higher target probability of detecting the same effect, and therefore generally calls for more participants per arm. The columns are planning quantities, not descriptions of what happened to participants in the completed trial.

The rows consider planning for 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect. Smaller effects are harder to distinguish from random variation and therefore require more participants for the same target power under the same assumptions. For the primary outcome, the registry lists 12 and 16 participants per arm for 80% and 90% power when planning for 100% of the observed effect; for 75%, it lists 22 and 29; and for 50%, it lists 48 and 64.

These figures are not observed power. They do not establish that the original trial was adequately or inadequately sized, because the trial's own assumptions and justification are not reported. Nor do they indicate the probability that the intervention's effect is real. The sample-size curve should be read as a visual representation of the same assumption-based planning relationship, rather than as additional evidence about the trial result.

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 = 0.766; treatment proportion p₁ = 0.220; average p̄ = 0.493.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.7070 = 1.3858.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5922 = 0.4984.
4. n = (1.3858 + 0.4984)² / (-0.5462)² = 11.9, rounded up to 12 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 <- 0.766272189349112                          # control proportion
p1 <- 0.220061412487206                          # 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(2067039)
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 = 12 (80% power), 16 (90% power)
 75% of observed effect: n per arm = 22 (80% power), 29 (90% power)
 50% of observed effect: n per arm = 48 (80% power), 64 (90% power)
power.prop.test: 11.9 
Simulated power at n = 12 per arm: 0.840 (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 a primary endpoint and three secondary outcome topics, but does not report the statistical analysis plan or the reported analyses. It does not state whether a hierarchy was used to test outcomes, whether p-values were adjusted for multiple comparisons, or whether any outcomes were designated exploratory. The posted p-values are described as unadjusted. Thus, the record does not establish control of the chance of a false-positive finding across the set of outcomes.

Missing-data handling is also not reported. The registry does not say how participants who missed follow-up surveys were classified, whether outcome data were imputed, or whether analyses used complete cases. It does not provide the number of missing primary outcome reports or reasons for missingness. The overall enrollment is 2665, while the primary outcome counts use denominators of 1014 and 977. The record does not explain why those denominators differ from enrollment or describe the population used for the posted primary comparison.

The record also does not identify an intention-to-treat population, a per-protocol population, or another analysis set. It does not describe adherence to kit use, crossover between groups, or treatment exposure beyond the arm descriptions. Without this information, the posted counts can be reported as registry outcome data, but the analysis population and the role of participants with incomplete follow-up cannot be fully assessed.

Randomization is an important design feature, but its protection in the final comparison depends partly on the analysis and follow-up. The sequence-generation and allocation concealment methods are not reported, nor are sensitivity analyses for missing outcomes or deviations from assigned group. These absences do not show that such procedures were not performed; they mean their details are not reported in the registry record.

The registry does not report interim analyses, stopping rules, or a sample-size justification for the original trial. No assumptions about those procedures should be inferred from the separate planning illustrations, which are based on the observed effects and stated illustrative assumptions.

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 serious adverse event summary lists 1325 participants at risk in the OraQuick in Home & Sure Check HIV Tests group and 1340 in the Information Only group. It records zero serious adverse events in each group and zero deaths in each group. The other adverse events fields also show zero affected participants in both groups.

These figures should be read as the registry's posted safety counts for the listed groups and denominators. They do not mean that every possible harm was ruled out, nor do they describe the nature, severity, timing, or relatedness of individual events. No serious-event terms are listed. The record does not explain how adverse events were collected, which events met the study's definitions, or whether the safety populations differ from the primary outcome populations.

Absence of a recorded event in these counts is not proof that an event could not occur. The figures provide a bounded account of what the registry reports, rather than a complete safety assessment or a basis for medical advice.

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

The primary result describes a difference in reported frequent HIV testing between randomized groups. In the posted counts, 777 of 1014 participants in the HIV self-testing group and 215 of 977 in the information-only group reported testing at least three times during the 12-month follow-up. The reported risk difference and its interval quantify the contrast on the proportion scale. This is evidence about the study's stated behavioral endpoint in the recorded outcome data.

It does not establish that every participant used all test kits, that every reported test was independently verified, or that all testing occurred at a particular time or location. The registry record does not report those details. It also does not show whether the observed pattern would be the same in populations recruited or supported differently from the internet-recruited men described by the study.

The outcome is not a measure of HIV status alone. Newly identified infections are reported as a secondary outcome, while linkage to services and infections among social network associates are also listed but lack posted numerical results. The primary result should not be extended to claims about those outcomes or broader health consequences. The record also does not provide the information needed to assess long-term outcomes beyond the 12-month timeframe.

Randomization supports comparison of assigned groups, but the available registry information leaves important analytic questions unresolved, including missing outcomes, analysis populations, and the prespecified statistical plan. The reported unadjusted estimates should therefore be interpreted in light of both the randomized design and the limits of the reported methods.

Finally, statistical evidence is not the same as a clinical or public health recommendation. The posted numbers describe outcomes and uncertainty for this trial record; they do not establish how a particular person should test or what care they should seek. Decisions about testing and care require appropriate clinical and public health guidance, which is outside the scope of these statistical results.

17. Why This Trial Matters Statistically

This trial is statistically informative because it links a randomized comparison to a clearly stated behavioral endpoint: whether participants reported testing at least three times during follow-up. The outcome is expressed as both counts and proportions, and the registry supplies absolute and relative measures of the group contrast. Reading those measures alongside their intervals helps distinguish the estimated size and direction of a difference from the uncertainty around it.

The open-label design also illustrates why outcome type matters. The primary endpoint relies on participant reports about behavior, so interpretation concerns reported testing, not a directly verified count of tests. General methodological considerations about awareness and self-report should be kept separate from evidence of actual bias; the registry does not report validation procedures that would resolve this issue.

The record provides a useful contrast between a posted primary result and a thinner secondary evidence base. One secondary outcome has numerical results, while other named outcomes do not. The absence of a reported multiplicity plan means that unadjusted secondary results warrant caution and cannot be treated as a coordinated set of confirmatory tests.

It also demonstrates the importance of transparent reporting. The registry confirms randomized allocation and gives outcome denominators, but does not explain the relation between those denominators and overall enrollment or report how missing outcomes were handled. Such details affect how readers assess completeness and the population represented by the estimates.

Finally, the sample-size illustrations are useful for understanding how target effect size and desired power influence planning needs, but they are not a record of the investigators' original design assumptions. Keeping that distinction clear prevents planning calculations from being mistaken for observed trial evidence.

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 limits affect interpretation of the posted results. The primary endpoint is based on participants reporting testing frequency. The registry does not report whether tests were independently verified, how reports were elicited, or whether the same measurement procedures were applied consistently across groups. Since the study was open label, participants knew whether they had received test kits; any influence of that awareness on reporting or follow-up is a possibility to consider, not a demonstrated bias.

The registry lists overall enrollment of 2665, but the primary outcome denominators are 1014 and 977. It does not explain the relationship between these figures, identify the analysis population, or report how missing outcome data were handled. This limits assessment of how complete the primary endpoint data were and how the analyzed participants relate to everyone enrolled.

The record does not provide the statistical analysis plan, original sample-size justification, randomization procedures, or details of any interim analyses. It also does not report adjustment for multiple outcomes. These omissions make it difficult to determine whether the posted analyses followed a prespecified plan or how strongly the secondary results should be viewed as confirmatory.

Generalizability is also not fully assessable from the registry description. Participants were internet-recruited men who have sex with men, and the record does not provide enough information about recruitment reach or participant characteristics to establish how well the findings apply beyond the study population. The intervention involved specific test kits, including a test described as approved for professional use rather than home use; results should not automatically be assumed to apply to other testing approaches.

Finally, numerical results are not posted for all named secondary outcomes. That limits evaluation of whether the primary behavioral result coincided with changes in linkage or network outcomes. The reported safety counts are also limited to the registry summary and do not provide event descriptions or collection methods.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Evaluation of Rapid HIV Self-testing Among MSM (eSTAMP) — NCT02067039.
• Registry record last updated 2020-03-31; status last verified 2020-03.
• Results first posted 2020-03-31.
• 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

NCT02067039, Evaluation of Rapid HIV Self-testing Among MSM (eSTAMP), is listed as a completed randomized, open-label, parallel-group study sponsored by the Centers for Disease Control and Prevention. Enrollment is recorded as 2665. The study compared provision of rapid HIV self-test kits with information only among men who have sex with men.

The primary endpoint was self-reported HIV testing at least three times over 12 months. The registry reports 777 of 1014 participants in the HIV self-testing group and 215 of 977 in the information-only group meeting that outcome, with a risk difference of -0.546 for Information Only relative to HIV Self-testing and an unadjusted p-value of <0.001.

Newly identified HIV infections are the only secondary outcome with numerical results in the record. Serious adverse event and death counts are recorded as zero in both listed groups. Details of the analysis plan, missing-data handling, and original sample-size justification are not reported in the registry record.