Comparative Safety and Efficacy of Two Antiviral Treatments in the Treatment of Recurrent Herpes Simplex Labialis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Comparative Safety and Efficacy of Two Antiviral Treatments in the Treatment of Recurrent Herpes Simplex Labialis (NCT02265913), a randomized phase 3 study. The primary outcome is Lesion Healing.

2. Clinical Question

3. Trial Design

The registry identifies NCT02265913 as a phase 3 randomized, double-masked, parallel-group trial. It is listed as completed, with an enrollment of 4076. The registry record gives a start date of 2014-10-06 and a primary completion date of 2016-04-27. Padagis LLC is listed as the sponsor.

The study included a test acyclovir cream, a reference acyclovir cream, and placebo cream. The primary outcome entry reports separate group means and standard deviations for the per protocol population. The reported group sizes for that outcome are 633 for the test product, 652 for the reference product, and 641 for placebo. These are outcome-specific counts and should not be treated as a reconstruction of the overall enrollment or of participant flow.

Randomization means allocation was described as randomized; double masking means the registry describes the trial as masked for two parties, but it does not identify those parties in the record. A parallel design means participants were assigned to a group rather than described as crossing between treatments. These are general explanations of the design terms, not additional details about how this trial was conducted.

The registry record does not report the allocation sequence method, concealment procedures, masking details, treatment instructions, or the trial’s original sample-size justification. The record also does not give the original statistical analysis plan or interim-analysis arrangements. Those details cannot be inferred from the posted outcome values.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Analysis population for the primary outcome, as defined in the registry: per protocol population

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 Lesion Healing, measured over up to 21 days.

The primary endpoint is lesion healing, recorded as the time to complete healing in hours from the time of first dosing. The timeframe is up to 21 days, and the reported summary measure is the mean, accompanied by a standard deviation. The endpoint was reported for the per protocol population.

Time to healing represents the duration until a specified event—in this case, complete lesion healing. Expressing the duration in hours gives a continuous measure that can be summarized with a group mean and standard deviation. In general, the mean describes the average observed duration, while the standard deviation describes how spread out individual durations are around that average. Neither statistic alone describes the full distribution of healing times.

The endpoint is relevant to the stated comparison of antiviral treatments because it quantifies the time course of lesion healing. However, interpretation depends on how healing was assessed, how participants with incomplete or unobserved healing were handled, and the analysis population used. The registry record does not report those assessment and missing-outcome procedures. It also does not provide other posted efficacy outcomes, so the endpoint should be read as one specific measure rather than as a complete account of all possible clinical outcomes.

7. Results Posted in the Registry

• Test Product: mean 107.47 versus 110.08 on control; difference -2.61 (95% CI -11.55 to 6.33), standardized effect d = -0.03; z-test p = 0.567.
• Placebo Product: mean 112.41 versus 110.08 on control; difference 2.33 (95% CI -6.74 to 11.40), standardized effect d = 0.03; z-test p = 0.614.

8. Statistical Interpretation

9. Statistical Methods Explained

The posted results combine a registry-reported equivalence analysis with simple unadjusted recalculations from the group summary statistics. These answer related but distinct statistical questions. The registry-reported analysis is the one to consult for the sponsor’s stated equivalence result; the recalculations provide a transparent comparison of the posted group means and their uncertainty.

What does a mean difference describe?

For a continuous outcome, a mean difference expresses the separation between two group averages in the outcome’s original units. Here, the units are hours. The recalculated test-versus-reference estimate is -2.61 hours, based on means of 107.5 and 110.1 hours. The placebo-versus-reference estimate is 2.33 hours, based on means of 112.4 and 110.1 hours. The sign indicates which group mean is higher under the comparison’s stated ordering; it does not, by itself, determine clinical importance.

General explanation: a mean is useful because it condenses a set of continuous observations into a familiar summary, but it can conceal skewness, unusual observations, or differences in distribution shape. The standard deviations reported for the test, reference, and placebo groups are 80.66, 82.84, and 83.47 hours, respectively. These describe spread within groups rather than uncertainty in the difference between group means.

What does the interval add?

The unadjusted recalculated interval for test product versus reference product runs from -11.55 to 6.33 hours. For placebo versus reference product, it runs from -6.736 to 11.4 hours. General explanation: a confidence interval is a range produced by a statistical procedure intended to convey uncertainty about an estimate. Its width reflects precision under the method’s assumptions. When an interval includes no difference, the data do not distinguish the estimate from no difference at the interval’s stated confidence level. The confidence level for these recalculated intervals is not reported in the registry record, so it is not specified.

The calculations use the posted group means and standard deviations. The pooled standard deviations are 81.77 hours for the test-versus-reference comparison and 83.15 hours for the placebo-versus-reference comparison. The reported standard errors for the means are 3.244 and 3.206 hours for reference and test, and 3.244 and 3.297 hours for reference and placebo. A standard error describes uncertainty in an estimated mean, rather than the variation among individual observations. These summaries do not reveal all features of the underlying data.

How should the p-values be read?

The recalculated unadjusted p-values are 0.567 for test product versus reference product and 0.614 for placebo versus reference product. General explanation: a p-value describes how unusual data at least as incompatible with a specified null model would be if that model and the analysis assumptions held. It is not the probability that the null hypothesis is true, and it does not measure the probability that a treatment works. Because these values are unadjusted recalculations, they should not be mistaken for the p-value from a complete prespecified analysis plan.

Testing for a difference and testing for equivalence are not the same task. A conventional difference test asks whether the data are inconsistent with a null of no difference. Equivalence analysis instead asks whether the uncertainty is sufficiently contained within prespecified bounds for differences considered acceptable. The record reports an equivalence ratio and interval, but does not report the bounds. Thus the ratio result cannot be converted into a definitive equivalence conclusion from the information available.

Why does the registry use Fieller’s method?

The registry names Fieller’s method for the equivalence ratio. General explanation: ratio estimates can have uncertainty that is not well described by simply applying the methods used for differences. Fieller’s method is one approach for constructing an interval for a ratio of quantities. The reported value is 98.0, and its 90% confidence interval is 92 to 105. The registry does not provide further method details, such as the exact ratio definition or equivalence margins.

The result is also tied to the per protocol population for the primary endpoint. In general, analysis populations can change what question an estimate addresses. The registry record does not explain how the per protocol population was defined, how participants were excluded from it, or how missing outcome measurements were handled. No further details about adjustment for other variables or multiplicity are reported.

10. Confidence Intervals and Precision

The uncertainty around the recalculated mean differences is substantial relative to the point estimates. For test product versus reference product, the estimate is -2.61 hours and the interval extends from -11.55 to 6.33 hours. For placebo versus reference product, the estimate is 2.33 hours and the interval extends from -6.736 to 11.4 hours. Both intervals include no difference and encompass values on either side of it.

General explanation: a confidence interval communicates precision by showing a range of estimates compatible with the data and the calculation method. A narrower interval indicates greater precision than a wider interval, all else equal. An interval should not be read as a list of effects that are equally likely, nor as proof that the true value lies within its endpoints. The confidence level for the recalculated intervals is not reported in the registry record.

The registry-reported equivalence result has a different scale and stated confidence level: an equivalence ratio of 98.0 with a 90% confidence interval from 92 to 105. It should not be combined with the difference-in-hours intervals. Whether that interval demonstrates equivalence depends on the equivalence bounds, which are not reported in the registry record.

Precision also has limits beyond interval width. The posted means and standard deviations do not show the full distribution of healing times or how incomplete observations were treated. The per protocol designation identifies the reported analysis population, but its definition is not available in the record. These omissions constrain how confidently the estimates can be applied to populations or analyses beyond those described.

11. 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 sample-size illustrations are planning calculations based on the observed effects, not the trial’s original design assumptions. They give participants per arm for each comparison. The calculations assume a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect equal to the stated fraction of the observed effect. The trial’s own original sample-size justification and design assumptions are not reported in the registry record. These figures are not observed power.

For test product versus reference product, planning for 100% of the observed effect gives 15,410 participants per arm for 80% power and 20,629 participants per arm for 90% power. Planning for 75% of the observed effect gives 27,395 participants per arm for 80% power and 36,674 participants per arm for 90% power. Planning for 50% of the observed effect gives 61,637 participants per arm for 80% power and 82,515 participants per arm for 90% power.

For placebo versus reference product, the corresponding figures are 19,993 participants per arm for 80% power and 26,765 participants per arm for 90% power when planning for 100% of the observed effect; 35,543 and 47,581 participants per arm when planning for 75%; and 79,970 and 107,057 participants per arm when planning for 50%.

General explanation: the 80% and 90% power columns describe the planned probability of detecting the specified effect under the calculation assumptions if that effect were present. The 90% target requires more participants per arm than the 80% target in these illustrations. Smaller effects need more participants because they are harder to distinguish from sampling variation. The sample-size curve should be read as a planning aid under these assumptions, not as a forecast of the trial’s actual results.

12. 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, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -2.610; pooled standard deviation σ = 81.773.
2. n = 2 × (1.96 + 0.842)² × 81.773² / -2.610² = 15,409.2, rounded up to 15,410 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
delta <- -2.61                      # observed difference in means
sigma <- 81.7734051712228                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^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 (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(2265913)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 15,410 (80% power), 20,629 (90% power)
 75% of observed effect: n per arm = 27,395 (80% power), 36,674 (90% power)
 50% of observed effect: n per arm = 61,637 (80% power), 82,515 (90% power)
power.t.test: 15410.2 
Simulated power at n = 15,410 per arm: 0.809 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-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.

13. Analyses Reported by the Sponsor

The registry-reported analysis for test product versus reference product uses Fieller’s method and reports an equivalence ratio of 98.0, with a 90% confidence interval from 92 to 105. This is distinct from the unadjusted recalculation, which estimates a difference in mean healing time of -2.61 hours, with an interval from -11.55 to 6.33 hours and p = 0.567.

The analyses use different scales and address different statistical questions: the recalculation compares means in hours, while the reported analysis concerns an equivalence ratio. The recalculated p-value is not the p-value for the equivalence analysis; no p-value is reported for that registry analysis. The equivalence bounds needed to interpret the ratio interval against a formal equivalence criterion are not reported in the registry record.

No discrepancy is flagged between the posted summary statistics and the recalculation. The registry-reported analysis should therefore be read on its own terms alongside the recalculated mean comparison, rather than treating one as a direct replacement for the other. The registry record does not provide enough detail to reconstruct the full equivalence analysis or its prespecified decision rule.

14. Multiplicity, Missing Data and Analysis Populations

The registry record lists one posted outcome measure and one analysis. The primary outcome is lesion healing, reported as a mean in hours for the per protocol population. No secondary outcomes are posted. The absence of additional posted outcomes does not show whether other outcomes were collected; it means they are not reported in the registry record.

Multiplicity arises when several outcomes, comparisons, time points, or analyses are considered. General explanation: conducting multiple statistical tests can increase the chance of finding a nominally notable result by chance, so an analysis plan may specify how to control or interpret that issue. The registry record does not report a multiplicity adjustment, a hierarchy among outcomes, or a plan for interpreting the test-versus-reference and placebo comparisons together. The recalculated p-values are unadjusted and should be understood in that context.

The primary outcome is described as being measured up to 21 days, but the registry record does not report how observations were handled for participants whose lesions had not healed by then, who stopped participating, or whose outcome measurements were unavailable. It does not report an imputation method, a censoring procedure, or sensitivity analyses for missing data. These are material details for a time-to-healing outcome, and their absence limits assessment of the analysis.

The reported population is per protocol. The registry record does not define that population or describe exclusions from it. It does not report an intention-to-treat analysis or explain whether one was conducted. As a result, the posted estimate should be interpreted specifically as the result reported for the named population, without assuming how it relates to other possible analysis populations.

The record also does not report covariate adjustment, subgroup analyses, interim analyses, or a full statistical analysis plan. None of these methods can be inferred from the presence of group means, confidence intervals, or the reported equivalence result.

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 5% reporting threshold.

Most frequent serious adverse event terms

Serious events with the highest arm-level rate: colon cancer (0.1%), atrial fibrillation (0.1%), thyroid cancer (0.1%), worsening of hypertension (0.1%), dehydration (0.1%).

The serious adverse event summaries list 2 affected participants among 1357 at risk for the test product, 4 among 1361 for the reference product, and 0 among 1358 for placebo. The summaries report 0 other adverse events and 0 deaths in each group. These are the registry’s posted counts, not estimates adjusted for differences between groups.

The registry also lists serious-event terms with percentages of 0.1%: Colon Cancer, Atrial Fibrillation, Thyroid Cancer, Worsening Of Hypertension, Dehydration, and Kidney Stone. The record does not connect these listed terms to specific groups in the displayed summary or provide enough detail to attribute them to a treatment. They should be read as reported event terms, not as evidence of causation.

General explanation: serious adverse event counts describe events meeting a seriousness definition, which is distinct from the overall frequency of symptoms or other events. A count alone does not establish whether an event was treatment-related, and small or zero counts do not establish the absence of risk. The registry record does not report event-level timing, severity details, causal assessments, or a statistical comparison of safety outcomes. The figures are therefore descriptive and limited to the posted summaries.

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

The posted primary outcome results describe mean time to complete lesion healing in hours. The test-product mean was 107.5 hours, the reference-product mean was 110.1 hours, and the placebo mean was 112.4 hours. The recalculated test-versus-reference and placebo-versus-reference mean differences have intervals that include no difference. These data do not establish a clear difference in average healing time through the unadjusted recalculations.

The registry-reported equivalence analysis is a separate result: an equivalence ratio of 98.0 with a 90% confidence interval from 92 to 105. Without the equivalence margins, the interval cannot be judged against the criterion needed for a formal equivalence conclusion. A non-significant difference test is not proof of equivalence, and a reported ratio interval is not self-interpreting without its decision boundaries.

These results do not establish that the treatments are identical, that any observed difference is clinically important, or that the products have the same effects on outcomes not included in the posted results. The primary outcome is restricted to the per protocol population, and the registry record does not define that population or describe the handling of missing outcomes. Those details affect how the estimates should be generalized.

The safety summaries show the serious adverse event counts recorded by group, but they do not establish whether events were caused by treatment or whether groups differed in risk. The registry record provides no adjusted safety analysis or event-level causal assessment.

Overall, the posted findings are best understood as estimates of average healing time with substantial uncertainty, accompanied by a registry-reported equivalence analysis whose margins are unavailable. They are statistical descriptions of the recorded trial results, not advice about treatment choice.

17. Why This Trial Matters Statistically

This trial illustrates why the statistical question matters as much as the numerical result. A comparison of mean healing times estimates a difference in hours, whereas an equivalence analysis evaluates a ratio relative to an acceptable range. The two analyses can be complementary, but they cannot be treated as interchangeable. In particular, a non-significant difference test does not establish equivalence.

The posted mean differences are small relative to the uncertainty intervals, which extend across values indicating differences in either direction. That makes precision central to interpretation. A point estimate alone can make a comparison appear more decisive than the data support; intervals show the range of uncertainty that accompanies the estimate.

The registry-reported equivalence interval also highlights the importance of transparent decision criteria. The ratio is reported with a 90% confidence interval, but the equivalence boundaries are not reported. Without those boundaries, readers cannot determine from the interval alone how the formal equivalence question was judged.

Finally, outcome populations and missing-data methods shape what a trial result means. The endpoint is reported for the per protocol population, while the population definition and missing-data handling are unavailable in the registry record. Clear reporting of these details helps readers understand which participants contribute to an estimate and how incomplete outcomes affect it. These issues matter for interpreting both efficacy estimates and safety summaries.

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 limitations are specific to the registry results available for this trial. The primary outcome is reported for the per protocol population, but the record does not define that population or explain how it was formed. The posted group counts for the outcome therefore cannot be fully reconciled with overall enrollment from the information reported.

The endpoint is time to complete healing, measured from first dosing and assessed up to 21 days. The registry record does not describe the assessment schedule, who assessed healing, how healing was defined operationally, or what happened to observations when healing was not recorded within the timeframe. It also does not report missing-data procedures or sensitivity analyses.

The registry reports one outcome and one analysis, with no secondary outcomes posted. It does not provide a complete statistical analysis plan, the original sample-size justification, multiplicity procedures, or details of any interim analyses. The equivalence ratio and interval are reported, but the definition of the ratio and the equivalence margins are absent. Those omissions prevent a complete appraisal of the equivalence claim.

Safety information is limited to group counts and listed serious-event terms. The record does not provide causal assessments, event-level timing, or comparative statistical analyses. The summaries should not be expanded into claims about treatment attribution or the full safety profile.

These reporting gaps do not change the posted estimates, but they limit independent assessment of how the analyses were conducted and how broadly the results apply. The registry record does not report further detail that would resolve these limitations.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Comparative Safety and Efficacy of Two Antiviral Treatments in the Treatment of Recurrent Herpes Simplex Labialis — NCT02265913.
• Registry record last updated 2021-10-22; status last verified 2021-10.
• Results first posted 2021-01-05.
• 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

NCT02265913 is listed as a completed phase 3, randomized, double-masked, parallel-group trial sponsored by Padagis LLC. The registry gives enrollment as 4076, a start date of 2014-10-06, and a primary completion date of 2016-04-27. The study compared test acyclovir cream, reference acyclovir cream, and placebo cream for herpes labialis.

The primary outcome was mean time to complete lesion healing in hours from first dosing, up to 21 days, in the per protocol population. The registry-reported test-versus-reference equivalence ratio was 98.0 with a 90% confidence interval from 92 to 105. Unadjusted recalculations of the posted means give uncertain differences for both test versus reference and placebo versus reference. Serious adverse event counts were 2, 4, and 0 across the test, reference, and placebo groups, respectively. The registry record does not report the equivalence margins, missing-data procedures, or a full statistical analysis plan.