Respiratory Syncytial Virus (RSV) Human Challenge Study of Molnupiravir in Healthy Participants (MK-4482-017): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Respiratory Syncytial Virus (RSV) Human Challenge Study of Molnupiravir in Healthy Participants (MK-4482-017) (NCT05559905), a randomized phase 2 study. The primary outcome is Panel A vs Panel C: Peak Viral Load (PVL) Determined by Viral Quantitative Culture.

2. Clinical Question

This was an experimental challenge study in healthy participants, rather than a study of treatment in a naturally occurring RSV infection. Its primary question concerns a laboratory measure during a specified post-inoculation period. The registry reports 116 enrolled participants overall, but that enrollment total is not the same as the analysis population for the primary endpoint: the reported primary groups contained 32 participants in Panel A and 34 in Panel C. The trigger-based Panel B regimen addresses a different treatment timing question and was excluded from the primary model according to the registry description.

3. Trial Design

The registry describes a Phase 2a, randomized, double-blind, placebo-controlled, parallel study of molnupiravir in healthy participants experimentally inoculated with RSV-A Memphis 37b. The study status is completed, and the registry lists enrollment as 116. The primary endpoint was limited to Panel A, the prophylaxis regimen, and Panel C, matched placebo. The trial also included Panel B, in which molnupiravir began after a positive RSV test; this panel was not part of the primary endpoint model.

The planned regimens distinguish treatment timing. Panel A began molnupiravir on Day -1 and continued it for 5 days, then switched to placebo. Panel B began with placebo and started molnupiravir after testing positive, for 5 days. Panel C received placebo throughout the stated period. All panels were inoculated on Day 0. These schedules create distinct comparisons, so results from the prophylaxis-versus-placebo primary analysis should not be read as a direct estimate of the triggered-treatment regimen.

For the primary analysis, the registry describes a linear model with treatment group as a fixed categorical effect. The target population was restricted to randomized participants in Panel A or Panel C who received one dose of the correct clinical material for their assigned group, received the viral inoculation, and had available peak viral load data from viral quantitative culture. The reported arm summaries are based on 32 participants in Panel A and 34 in Panel C. Safety summaries use different group sizes: 39 in Panel A, 36 in Panel B, and 40 in Panel C. These differences reflect different reported analysis populations and endpoints; the registry record does not give a fuller account of why the analysis counts differ.

Although the study is described as randomized and double-blind, further details about the randomization procedure, allocation concealment, or how masking was maintained are not reported in the registry record. A sample-size justification and the trial’s complete statistical analysis plan are also not reported in the registry record.

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

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All randomized participants in Panel A or Panel C who received 1 dose of the correct clinical material corresponding to the treatment group the participants were randomized into, who received the viral inoculation, and had available PVL viral quantitative culture data. Per protocol, Panel B was not included in this analysis.

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 Panel A vs Panel C: Peak Viral Load (PVL) Determined by Viral Quantitative Culture, measured over From Day 2 up to Day 12.

The primary endpoint was peak viral load (PVL) measured by viral quantitative culture, also described as a plaque assay. PVL was defined as the maximum viral load during the specified period, from Day 2 up to Day 12, the end of participant quarantine. The measure is expressed on a log10 scale in plaque forming units (PFU)/mL. The primary analysis compared Panel A, molnupiravir prophylaxis, with Panel C, matched placebo.

A maximum over a defined period gives a single summary of the highest measured viral burden during that interval. Viral quantitative culture measures virus capable of forming plaques under the assay conditions; this makes it distinct from an assay that detects viral genetic material. The registry also lists qRT-PCR outcomes among the secondary measures, but those are not the primary endpoint considered in the sponsor’s main comparison.

Using a continuous endpoint allows the analysis to compare mean levels on the log10 scale. The registry reports least-squares means for each arm and a difference in least-squares means from an analysis of variance (ANOVA). The arm estimates were 2.55 for Panel A and 2.84 for Panel C. These are model-based summaries for the specified analysis population and measurement window. They do not describe every participant’s viral-load trajectory, nor do they alone establish how much the observed difference matters outside this experimental setting.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is continuous: each included participant contributes a peak viral-load measurement on the log10 PFU/mL scale. The registry says that the sponsor used a linear model with treatment group as a fixed categorical effect and reports least-squares means. A least-squares mean is a model-based group estimate; in this description, treatment group is the stated model effect. The contrast is the difference between the estimated means for Panel A and Panel C.

The sponsor reports an ANOVA comparison and a 90% confidence interval for that difference. A confidence interval describes a range of values reasonably compatible with the observed data under the model and its assumptions. Its width is a measure of precision: a wider interval leaves more uncertainty about the possible size and direction of the contrast. Because the reported interval includes zero, the posted results do not rule out no difference under this analysis. This is a statistical interpretation, not a statement about clinical importance.

The arm summaries also have confidence intervals, but those are 95% intervals for each arm’s least-squares mean. They should not be substituted for the confidence interval of the between-arm contrast. The contrast’s interval directly addresses the primary comparison, while the arm intervals describe uncertainty in each group estimate. The registry does not state why the reported contrast interval uses a different confidence level from the arm-level intervals.

The posted calculations for some secondary binary outcomes are labelled as recalculated and are simple, unadjusted comparisons from the reported counts. A risk difference expresses the difference between two observed proportions. Its confidence interval describes uncertainty around that difference. The unadjusted p-value tests a null contrast under the assumptions of the recalculation; it is not the same as a sponsor-reported analysis unless the registry separately lists it as such. For the primary outcome, the registry reports a sponsor analysis; there is no listed recalculation to use in its place.

For binary outcomes with few events, estimates can be unstable. When event counts are sparse, small changes in the observed counts can have a large influence on relative measures such as risk ratios or odds ratios, and confidence intervals can be wide. This is a general statistical feature, not a conclusion that a particular estimate is wrong. In these data, several adverse-event outcomes have very few reported events, so their uncertainty deserves particular attention.

The planning illustrations associated with recalculated binary comparisons are not observed trial results and are not the trial’s original sample-size calculation. They use the observed effect as a planning input, with stated assumptions of a two-sided significance level of 0.05, equal allocation, and a normal approximation. The illustrations ask how many participants per arm might be required to detect a stated fraction of the observed effect at a specified power level. They are sensitive to the observed effect, which can be especially unstable when event counts are sparse. They should not be described as observed power, and they do not show that the study was designed with those assumptions. The trial’s own sample-size justification is not reported in the registry record.

For continuous endpoints, the spread of participant-level measurements matters to precision. The registry lists confidence intervals for the arm means and the sponsor’s contrast, but does not provide a separate measure of participant-level dispersion in the posted primary summaries. A confidence interval around a mean is not a description of the full distribution of individual values. The registry does not report further details about model diagnostics, transformations beyond the stated log10 scale, handling of influential observations, or departures from model assumptions.

Randomization can support a fair comparison when the analysis preserves the assigned groups and follows the planned procedures, but the registry’s primary population also requires correct clinical material, inoculation, and available endpoint data. The record does not fully describe the consequences of exclusions, how missing endpoint measurements were handled, or whether any sensitivity analyses were conducted. It also does not report interim analyses, a multiplicity strategy, or a complete protocol-level statistical plan. Those omissions limit assessment of how all reported analyses relate to the originally planned inferential framework.

10. Analyses Reported by the Sponsor

The sponsor’s own primary analysis is the ANOVA comparison of the Panel A and Panel C least-squares means. It reports a difference of -0.29 log10 PFU/mL, a 90% confidence interval from -1.16 to 0.58, and p = 0.578. The registry does not flag a discrepancy between the sponsor result and a recalculated primary result. No recalculated primary contrast is listed, so there is no alternate primary estimate to set alongside the sponsor’s reported analysis.

Where recalculated secondary comparisons are available, they are simple unadjusted calculations based on posted summary counts. Such calculations can be useful for understanding the scale and uncertainty of an observed contrast, but they may not reproduce a sponsor’s prespecified model or analysis choices. If the registry had flagged a discrepancy, the sponsor’s own analysis would take precedence; no such discrepancy is flagged for this record.

The primary comparison should therefore be read from the sponsor’s reported estimate and its confidence interval, not inferred from the separate arm-level intervals or from secondary calculations. The interval’s inclusion of zero indicates uncertainty about the direction and magnitude of the primary contrast. The registry does not report enough detail to reconstruct every aspect of the sponsor’s analysis independently.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 33 outcome measures, 22 analyses, and 31 posted secondary outcomes. It does not report whether the primary endpoint was designated as the sole confirmatory test, whether a hierarchy was used for secondary outcomes, or whether p-values or confidence intervals were adjusted for multiple comparisons. Without that information, secondary analyses should not be interpreted as though they collectively formed one prespecified confirmatory test.

Multiplicity matters because examining numerous outcomes and comparisons creates more opportunities for apparently unusual results to arise by chance. Adjustment methods can address some of this issue, but no such method is reported in the registry record. The primary comparison has its own sponsor-reported estimate and interval; the multiplicity information does not change the numerical result, but it affects how much weight to assign to patterns across the broader set of analyses.

The primary analysis population required available PVL viral quantitative culture data, as well as randomization, receipt of the correct clinical material, and viral inoculation. The registry does not report a detailed accounting of missing primary measurements, reasons for missingness, imputation, or sensitivity analyses. It is therefore not possible to assess from the posted description whether missingness could have affected the estimated contrast. Requiring available outcome data identifies who contributed to the reported analysis, but does not by itself explain why other randomized participants may not have contributed.

The safety summaries use group sizes of 39, 36, and 40, while the primary endpoint analysis includes 32 and 34 participants in Panels A and C. These are not identical populations or denominators. The registry record does not provide a complete explanation of the analysis-set definitions across all outcomes, so differences in counts should be kept in mind when comparing primary and safety results.

Interim analyses, stopping rules, the trial’s sample-size justification, and the complete statistical analysis plan are not reported in the registry record. Their absence prevents a full assessment of whether every reported analysis followed a prespecified plan. This is a limitation of the available registry reporting, not evidence that those procedures were or were not used.

12. 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 non-serious adverse event terms

The registry reports zero serious adverse events in each panel: the at-risk groups were 39 in Panel A, 36 in Panel B, and 40 in Panel C. It also reports zero deaths in all three panels. These are observed counts in the registry’s safety summaries; they do not establish that serious events could not occur, particularly because the groups were limited in size and the study involved an experimental viral challenge.

For other adverse events, the registry lists 1 in Panel A, 2 in Panel B, and 0 in Panel C. The broader secondary outcomes also include viral challenge-related adverse events, but those are a distinct classification. Counts should be read with their stated group sizes and outcome definitions, rather than treated as a complete account of all possible safety risks. The registry record does not provide detailed serious-event terms, since none are listed among the top serious terms.

These summaries describe what was recorded during the study; they do not support a broad safety conclusion beyond the observed participants and follow-up reported in the registry. They also should not be used alone to compare safety across regimens, given the small event counts and the differing treatment schedules.

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

The result estimates a specific contrast: peak RSV viral load measured by viral quantitative culture from Day 2 up to Day 12 in participants in the prophylaxis and matched-placebo panels who met the stated primary analysis criteria. The sponsor’s estimated difference was -0.29 log10 PFU/mL, with a 90% confidence interval from -1.16 to 0.58. The point estimate is lower for Panel A, but the interval includes zero and extends in both directions. Thus, the estimate is uncertain in both magnitude and direction.

This result does not show that molnupiravir works or does not work. A nonsignificant p-value does not prove equality, and the confidence interval does not establish that every value within its range is equally plausible. The statistical result describes how compatible the observed data are with a range of contrasts under the reported analysis. It does not by itself determine whether a difference would be important in another setting.

The finding also does not directly answer the triggered-treatment question. Panel B began molnupiravir after testing positive, whereas the primary comparison concerns treatment beginning before inoculation in Panel A versus placebo in Panel C. The primary endpoint cannot be generalized directly to natural RSV infection: participants were healthy and received an experimental inoculation. Further, a culture-based peak viral-load result is not the same outcome as infection status by qRT-PCR, symptoms, or clinical outcomes not reported as the primary endpoint.

Safety counts are likewise bounded by the registry’s reported groups and observations. Zero reported serious adverse events does not establish absence of risk. The results are best understood as a description of outcomes observed in this completed experimental study, with uncertainty and applicability constrained by its population, regimen, endpoint, and analysis set.

14. Why This Trial Matters Statistically

This trial illustrates why the estimated effect and its precision should be read together. The primary point estimate is numerically lower in the prophylaxis panel, but its confidence interval includes zero and covers a broad range of possible contrasts. The point estimate alone would give an incomplete account; the interval shows that the data do not pin down the effect closely.

It also demonstrates the importance of distinguishing endpoints and treatment strategies. The primary outcome is a culture-based peak viral-load measure, while secondary outcomes include qRT-PCR infection status and safety events. The trial’s prophylaxis and triggered-treatment schedules ask different questions. Keeping these distinctions clear prevents a result from one panel or measurement method from being generalized to another.

The study is a useful example of how experimental challenge designs can evaluate a biological measurement under controlled inoculation conditions. At the same time, the registry’s primary population is selected by receipt of the assigned material, inoculation, and availability of endpoint data. The distinction between overall enrollment and the endpoint analysis population matters when judging the scope and precision of the estimate.

Finally, multiple secondary outcomes and sparse event counts make disciplined interpretation important. Secondary patterns may add context, but without a reported multiplicity strategy they should not be treated as an array of independent confirmations. The statistical lesson is not to reduce the trial to a p-value: the endpoint definition, analysis population, contrast, confidence interval, and limits of generalization all shape what the posted results can support.

15. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The primary outcome is a laboratory measurement in an experimental RSV challenge, and the participants were described as healthy. The result therefore has a narrower context than a clinical outcome measured in naturally acquired infection. The primary analysis also concerns only prophylaxis versus matched placebo; it does not include Panel B’s triggered-treatment regimen.

The primary analysis includes participants with available viral quantitative culture PVL data and specified treatment and inoculation characteristics. The registry does not report detailed missing-data reasons or sensitivity analyses, so the potential effect of unavailable measurements cannot be assessed from the record. The primary endpoint group sizes differ from the safety group sizes, and the record does not fully explain the analysis-set differences.

The sponsor’s primary interval is a 90% confidence interval, while arm-level summaries use 95% confidence intervals. The record does not explain this difference or provide the complete statistical analysis plan. It also does not report a sample-size justification, interim analyses, or multiplicity control. These reporting gaps limit assessment of how the analyses were planned and how secondary results should be weighed.

The registry describes the primary model as a linear model with treatment group as a fixed categorical effect, but does not report diagnostics or further sensitivity analyses. Sparse event counts limit the precision of several secondary safety comparisons. These limitations do not negate the posted estimates, but they constrain the conclusions that can be drawn from them.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Respiratory Syncytial Virus (RSV) Human Challenge Study of Molnupiravir in Healthy Participants (MK-4482-017) — NCT05559905.
• Registry record last updated 2025-07-18; status last verified 2025-07.
• Results first posted 2024-05-20.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

NCT05559905 is a completed, randomized, double-blind, parallel Phase 2a study of molnupiravir in healthy participants experimentally inoculated with RSV-A Memphis 37b. The registry lists enrollment of 116. The primary endpoint was peak viral load by viral quantitative culture from Day 2 up to Day 12, comparing Panel A molnupiravir prophylaxis with Panel C matched placebo.

The sponsor reported a difference in least-squares mean of -0.29 log10 PFU/mL, with a 90% confidence interval from -1.16 to 0.58 and p = 0.578. The interval includes zero, leaving the direction and magnitude of the primary contrast uncertain. Panel B’s triggered-treatment regimen was not included in the primary model.

The record also reports secondary viral and safety outcomes. No serious adverse events or deaths were recorded in the reported safety groups. Multiplicity procedures, detailed missing-data handling, and the trial’s sample-size justification are not reported in the registry record.