The Safety of Molnupiravir (EIDD-2801) and Its Effect on Viral Shedding of SARS-CoV-2 (END-COVID): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for The Safety of Molnupiravir (EIDD-2801) and Its Effect on Viral Shedding of SARS-CoV-2 (END-COVID) (NCT04405739), a randomized phase 2 study. The primary outcome is Number of Participants That Achieve Virologic Clearance After Oral Administration of EIDD-2801.

2. Clinical Question

3. Trial Design

This was a phase 2, randomized, double-blind, parallel-group, placebo-controlled study of EIDD-2801 for SARS-CoV-2. The registry describes a multicenter study in newly hospitalized adults with PCR-confirmed COVID-19. Its stated purpose was to assess the efficacy and safety of EIDD-2801 on SARS-CoV-2 virus shedding. Enrollment is listed as 71. The record gives a start date of 2020-06-16, a primary completion date of 2022-02-21, and a status of completed.

The outcome arms are listed as molnupiravir 200 mg, 400 mg, and 800 mg, alongside placebo. The intervention description characterizes active treatment as oral administration twice daily for 5 days; placebo was also described as oral administration twice daily for 5 days and matched for size and appearance. The arm description for active treatment additionally refers to Dose A through Dose F, but the registry’s posted primary outcome results are organized by the three stated milligram-dose arms. The record does not explain this difference in wording.

The primary result is a participant count for virologic clearance by day 5. For this endpoint, the posted arm counts are 0 of 4 in the 200 mg arm, 4 of 16 in the 400 mg arm, 10 of 25 in the 800 mg arm, and 7 of 24 in the placebo arm. These are the endpoint analysis counts, not a replacement for the record’s overall enrollment figure. Safety populations have their own reported denominators, which differ for some arms. The registry record does not report the original sample-size justification or a detailed protocol for allocation, concealment, or dose assignment beyond its designation as randomized and parallel-group.

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: EAS 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 Number of Participants That Achieve Virologic Clearance After Oral Administration of EIDD-2801, measured over 28 days.

The primary endpoint counts participants whose SARS-CoV-2 RNA became undetectable, meaning below the assay’s limit of detection, in nasopharyngeal swabs by day 5. Measurement was by quantitative polymerase chain reaction. The record labels the measure as the number of participants achieving virologic clearance and gives a timeframe of 28 days, while its endpoint description specifies achievement by day 5. Both details are relevant when interpreting what the outcome represents.

A binary endpoint records whether each participant met the stated clearance criterion or did not. Reporting the number who met it alongside the arm denominator gives the observed frequency in each group. The endpoint is intended to capture virologic shedding in the specified sample and at the specified time; it is not itself a measure of symptoms, hospitalization duration, or longer-term clinical outcomes. The registry record does not report a posted secondary outcome analysis.

For the endpoint population, participants had to receive at least one dose of study drug and have at least one post-baseline quantitative PCR assessment from a nasopharyngeal swab. Thus, the EAS definition depends on treatment and post-baseline measurement, rather than simply including everyone enrolled. The record gives no further account of how missing day-5 measurements were handled, how measurements outside the exact target day were treated, or whether any assay-related threshold decisions were made beyond the stated limit of detection.

7. Results Posted in the Registry

• Molnupiravir 400 mg: 25.0% (4/16) versus 29.2% (7/24) on control; risk difference -4.2 percentage points (95% CI -32.1 to 23.8), risk ratio 0.86 (0.30–2.46); unadjusted two-proportion z-test p = 0.772.
• Molnupiravir 800 mg: 40.0% (10/25) versus 29.2% (7/24) on control; risk difference 10.8 percentage points (95% CI -15.6 to 37.3), risk ratio 1.37 (0.62–3.01); unadjusted two-proportion z-test p = 0.426.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary at the participant level: each person either met the stated virologic-clearance criterion by day 5 or did not. For a binary outcome, a count and its denominator describe how many participants experienced the event among those included in the relevant analysis population. Here, the event is clearance, not an adverse event. For example, the posted primary endpoint result for the 800 mg group is 10 of 25, while the placebo result is 7 of 24. These arm-specific denominators should be retained when interpreting the comparison.

What does a risk difference represent?

A risk difference expresses the difference between the event frequencies in two groups, with the direction depending on which group is subtracted from which. The registry’s comparisons use placebo as the reference and report the active-arm estimate relative to it. A value of zero would correspond to equal observed event frequencies under that contrast. A value above or below zero indicates the estimated direction and size of the difference on the risk-difference scale. The estimate is not a probability that the treatment works, and it does not by itself account for uncertainty.

For molnupiravir 400 mg, the reported risk difference is -0.042. For 800 mg, it is 0.108. These are unadjusted estimates derived from the posted arm-level counts. An unadjusted comparison uses the reported group results without adjustment for other variables. The registry record does not report covariate-adjusted estimates for these comparisons.

How should the p-values be read?

The unadjusted p-values are 0.772 for the 400 mg comparison and 0.426 for the 800 mg comparison. A p-value describes how unusual data at least as incompatible with a specified null model would be, assuming that model and the test’s other conditions. It is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of a difference. The registry record does not specify the statistical test used for these posted p-values or the analysis significance threshold for the primary comparisons. The planning illustration separately states a two-sided significance level of 0.05; that assumption should not be mistaken for a documented analysis rule for the trial.

The comparison fields also include relative-risk and odds-ratio estimates with intervals. These are alternative scales for expressing a binary-outcome contrast. A relative risk compares event frequencies as a ratio; an odds ratio compares event odds as a ratio. These measures can differ in numerical value and interpretation from a risk difference, especially when the event frequency is not rare. The posted relative-risk interval for 400 mg is 0.299 to 2.457, and for 800 mg it is 0.624 to 3.012. The corresponding odds-ratio intervals are 0.193 to 3.395 and 0.493 to 5.319. These intervals include the ratio value of one, consistent with considerable uncertainty about the direction and magnitude of the relative contrasts.

Why does the analysis population matter?

The endpoint is specified for the Efficacy Analysis Set, defined as participants treated with at least one dose and with at least one post-baseline SARS-CoV-2 RNA assessment by quantitative PCR. A population defined this way does not necessarily match all enrolled participants. Participants who were untreated or lacked a post-baseline assessment would not meet the stated definition. The record does not report how many participants were excluded from the EAS, give a participant-flow account, or describe how missing endpoint data were handled.

In general, missing outcome data can affect a binary comparison if the likelihood of missingness differs between groups or is related to the outcome. Different analysis choices can produce different estimates, so transparent reporting of missingness and its handling is important. The registry record does not report imputation, sensitivity analyses, or another missing-data procedure for the primary endpoint.

What can be inferred from the posted comparisons?

The intervals quantify uncertainty around the estimated contrasts under the method used to construct them. Because the 400 mg and 800 mg risk-difference intervals extend on both sides of zero, the observed data do not pin down the direction of the underlying difference. Their breadth reflects limited precision. A lack of a small p-value is not evidence of equivalence; equivalence requires a prespecified margin and an analysis designed to assess it. The registry record does not report an equivalence or non-inferiority framework, a prespecified margin, or a multiplicity adjustment for the two posted comparisons.

Finally, the primary endpoint is a measure of viral RNA detection in nasopharyngeal swabs at a specified time. It should not be treated as interchangeable with clinical outcomes or interpreted as an individual prediction. The registry record does not report a detailed statistical analysis plan, the trial’s original design assumptions, the named tests underlying the comparisons, or the confidence level used for the intervals. Those omissions limit how fully the posted estimates can be audited or connected to the intended analysis strategy.

10. Confidence Intervals and Precision

The risk-difference estimate for molnupiravir 400 mg is -0.042, with a reported interval from -0.321 to 0.238. For 800 mg, the estimate is 0.108, with an interval from -0.156 to 0.373. Each interval includes zero, meaning the posted data are compatible with no difference as well as with differences in either direction on the risk-difference scale. The intervals also extend substantially beyond their point estimates, indicating limited precision.

A confidence interval is generally a range of values produced by a method that reflects sampling uncertainty; repeated use of the method has a defined long-run coverage property. The registry record does not state the confidence level or describe the interval-construction method, so neither should be inferred. The interval endpoints are not guarantees about where the true effect lies, and they do not summarize uncertainty from every possible source, such as outcome measurement, missing data, or choices among analyses.

The small reported endpoint denominators are relevant to precision: the 200 mg, 400 mg, and 800 mg arms have denominators of 4, 16, and 25, respectively, while placebo has a denominator of 24. With relatively few observations, a change in the event status of a participant can have an appreciable influence on an observed frequency. The record’s displayed estimates therefore should not be read as highly stable measurements. The primary endpoint counts for the 200 mg arm are reported, but a comparative interval is not reported for that arm. No interval should be inferred for it.

The relative-risk and odds-ratio intervals are also broad: the relative-risk interval is 0.299 to 2.457 for 400 mg and 0.624 to 3.012 for 800 mg; the odds-ratio interval is 0.193 to 3.395 and 0.493 to 5.319, respectively. These ranges reinforce that the posted data do not precisely determine the comparative effect. They are not measures of certainty that the treatment has a particular effect.

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 planning figures are participants per arm, not total events. They are illustrations based on the observed effect, not the trial’s original design assumptions. The stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; the effect to detect is the stated fraction of the observed effect. The registry record does not report the trial’s original sample-size justification or design assumptions.

For the comparison using the observed 400 mg effect, planning for 100% of the observed effect gives 1,785 participants per arm for 80% power and 2,389 participants per arm for 90% power. Planning for 75% gives 3,212 participants per arm for 80% power and 4,299 participants per arm for 90% power. Planning for 50% gives 7,311 participants per arm for 80% power and 9,786 participants per arm for 90% power.

For the comparison using the observed 800 mg effect, planning for 100% gives 302 participants per arm for 80% power and 404 participants per arm for 90% power. Planning for 75% gives 527 participants per arm for 80% power and 705 participants per arm for 90% power. Planning for 50% gives 1,161 participants per arm for 80% power and 1,554 participants per arm for 90% power.

The 80% and 90% power columns describe the probability, under the stated planning model and assumed effect, of detecting a difference at the specified significance level. The curve should be read as showing how the planning requirement changes as the assumed detectable effect changes. Smaller effects need more participants to detect under the same assumptions. These are planning illustrations, not observed power.

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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.292; treatment proportion p₁ = 0.250; average p̄ = 0.271.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6285 = 1.2318.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6278 = 0.5283.
4. n = (1.2318 + 0.5283)² / (-0.0417)² = 1,784.4, rounded up to 1,785 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 <- 0.291666666666667                          # control proportion
p1 <- 0.25                          # 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(4405739)
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 = 1,785 (80% power), 2,389 (90% power)
 75% of observed effect: n per arm = 3,212 (80% power), 4,299 (90% power)
 50% of observed effect: n per arm = 7,311 (80% power), 9,786 (90% power)
power.prop.test: 1784.4 
Simulated power at n = 1,785 per arm: 0.800 (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.

13. Multiplicity, Missing Data and Analysis Populations

The registry record lists three outcome measures and no posted secondary analyses. It reports unadjusted comparisons for the 400 mg and 800 mg groups against placebo, but does not report a comparative analysis for 200 mg. Because multiple dose comparisons are present, multiplicity is relevant: when several hypotheses are tested, the chance of at least one apparently positive result can increase if no adjustment or other prespecified strategy is used. The record does not report whether the trial’s analysis plan used a multiplicity adjustment, hierarchical testing, or another approach. The unadjusted p-values should therefore be interpreted as individual comparisons, not as results adjusted across the dose comparisons.

The primary outcome description defines the EAS as participants treated with at least one dose and having at least one post-baseline quantitative PCR assessment from a nasopharyngeal swab. The record does not give the numbers screened, randomized, treated, included in the EAS, or excluded from each analysis, beyond the arm-level denominators reported for the outcomes. It also does not describe the handling of missing day-5 measurements, withdrawals, or other unavailable endpoint assessments.

Missingness matters because an analysis restricted to participants with an observed post-baseline assessment may differ from an analysis including everyone randomized. If missingness is related to treatment group or likelihood of clearance, the observed counts may not represent the same target population as a full randomized-group analysis. This is a general statistical concern, not evidence that missingness affected this trial’s results. The registry record does not report imputation methods, sensitivity analyses, or an intention-to-treat analysis for the primary endpoint.

The record also does not report the statistical analysis plan, the confidence level for the intervals, the method used for the unadjusted p-values, or a prespecified rule for handling multiple comparisons. These omissions prevent a full assessment of whether the posted contrasts match the trial’s planned primary analysis. They also make it important to keep the scope of interpretation narrow: the posted data describe the specified endpoint and reported comparisons, while unreported analyses should not be assumed to have been performed.

14. 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

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: hypotension (4.2%), non-cardiac chest pain (4.2%), bradycardia (4.2%), dehydration (4.2%), acute respiratory failure (4.0%).

The serious adverse event counts are reported separately from the virologic-clearance endpoint. The safety denominators at risk are 4 for molnupiravir 200 mg, 18 for 400 mg, 25 for 800 mg, and 24 for placebo. Serious adverse events are recorded for 0, 0, 1, and 3 participants, respectively. Deaths are recorded for 0, 0, 1, and 0 participants, respectively. These are counts in the registry record, not comparative safety estimates adjusted for differences between groups.

Other adverse events are also counted, with 4 of 4 participants in the 200 mg group, 11 of 18 in the 400 mg group, 8 of 25 in the 800 mg group, and 12 of 24 in the placebo group reported in the corresponding fields. The listed serious-event terms and percentages include hypotension, non-cardiac chest pain, bradycardia, and dehydration at 4.2%, and acute respiratory failure and hypovolemic shock at 4.0%. The registry record does not provide enough detail to establish the relationship of each term to treatment or to reconstruct event timing and recurrence.

These counts should be interpreted descriptively. Safety denominators differ from some endpoint denominators, and the record does not provide adjusted comparisons or a detailed account of exposure duration and missing safety follow-up. The figures alone do not establish whether a difference is attributable to treatment.

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

The posted primary outcome describes the number of participants in each arm who had SARS-CoV-2 RNA below the assay’s limit of detection in nasopharyngeal swabs by day 5. It shows 0 of 4 in the molnupiravir 200 mg arm, 4 of 16 in the 400 mg arm, 10 of 25 in the 800 mg arm, and 7 of 24 in the placebo arm. The corresponding unadjusted risk-difference estimates are -0.042 for 400 mg and 0.108 for 800 mg, and both reported intervals include zero.

The estimates do not establish that the groups are the same, because the intervals are compatible with differences in either direction. Nor do the counts establish that a treatment caused the observed difference. Randomization is part of the trial design, but the limited arm-specific endpoint counts and the unreported details of missing-data handling and the analysis plan constrain interpretation. The comparisons concern a single virologic measure at a specific time, rather than symptoms, clinical recovery, or other outcomes.

The endpoint should also be understood in the context of its analysis population. The EAS required at least one dose and at least one post-baseline swab assessment. The record does not report a full participant-flow description or enough information to determine how people without a qualifying assessment were handled in all analyses. The posted safety counts are separate descriptive results and do not resolve the uncertainty in the primary endpoint comparisons.

In practical statistical terms, the results are imprecise estimates from relatively small outcome groups. The 200 mg group has a reported event count but no posted comparative estimate. The registry record does not report a detailed sponsor analysis, a multiplicity strategy, or the original sample-size rationale. The evidence should therefore be described as the posted endpoint counts and unadjusted comparisons, with uncertainty, rather than as a definitive finding about clinical benefit or harm.

16. Why This Trial Matters Statistically

This trial illustrates several recurring issues in the analysis of clinical data. A binary endpoint may be easy to state, but its interpretation depends on a precise event definition, a clearly described analysis population, and complete reporting of denominators. Here, virologic clearance is defined by an assay threshold in nasopharyngeal swabs by day 5, while the primary outcome timeframe is also labeled as 28 days. Keeping both details visible avoids treating the endpoint as broader or more clinically direct than the record describes.

The trial also shows why point estimates should be read alongside intervals. The risk-difference estimates differ in direction between the 400 mg and 800 mg comparisons, but the reported intervals for both extend across zero. The observed directions alone do not settle the question; uncertainty is central to what can be concluded. Small arm-specific endpoint counts contribute to that uncertainty and make precise estimation difficult.

Multiple dose comparisons raise another statistical issue. Results for more than one active dose can create a multiplicity problem if several tests are interpreted without a prespecified strategy. The registry record does not report whether an adjustment or other plan was used. This makes transparent reporting of the analysis plan especially important when readers assess individual p-values.

Finally, the record separates the EAS endpoint population from safety populations, and those denominators are not identical in every arm. That distinction reinforces a general lesson: different outcomes can involve different analysis sets, and counts should not be combined or assumed to refer to the same participants. The posted results are useful for understanding the design and uncertainty of the virologic comparison, but the missing analysis details limit deeper appraisal.

17. 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 reporting gaps limit interpretation. The registry record gives an endpoint timeframe of 28 days, while the endpoint description defines clearance by day 5. It does not explain the relationship between these two time references. The active-treatment arm description mentions Dose A through Dose F, while the posted outcome arms are identified by molnupiravir 200 mg, 400 mg, and 800 mg; the record does not reconcile those descriptions.

The outcome counts are not accompanied by a full participant-flow account. Although enrollment is listed as 71, the record does not detail how participants moved from enrollment and randomization into the EAS or safety analysis populations. The endpoint denominators differ from some safety denominators, and no explanation of these differences is reported. The registry record does not describe missing-data handling, the assessment schedule beyond the stated endpoint, or sensitivity analyses.

The registry record does not report the original sample-size justification, a detailed statistical analysis plan, the confidence level or construction method for the intervals, or the tests used to produce the unadjusted p-values. It does not report how multiplicity across dose comparisons was addressed. There is no posted comparative estimate for the 200 mg arm, despite its endpoint count being listed. No adjusted estimates are reported.

The primary outcome is a virologic measure, and the posted results do not establish corresponding effects on other clinical outcomes. The record states the study’s general aim of assessing safety and viral shedding, but the available results do not support conclusions beyond the reported measures. These limitations are reporting limitations; they should not be treated as proof that a particular analysis or procedure was absent from the full trial documentation.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: The Safety of Molnupiravir (EIDD-2801) and Its Effect on Viral Shedding of SARS-CoV-2 (END-COVID) — NCT04405739.
• Registry record last updated 2023-03-23; status last verified 2023-03.
• Results first posted 2023-03-23.
• 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.

21. Record Summary

The registry identifies a completed phase 2 randomized, double-blind, parallel-group, placebo-controlled study of EIDD-2801 for SARS-CoV-2, with enrollment listed as 71. The study description concerns newly hospitalized adults with PCR-confirmed COVID-19. Active arms are listed as molnupiravir 200 mg, 400 mg, and 800 mg, with placebo as comparator; the intervention description specifies oral dosing twice daily for 5 days.

The primary endpoint counts participants achieving undetectable SARS-CoV-2 RNA in nasopharyngeal swabs by day 5. Posted counts are 0 of 4, 4 of 16, 10 of 25, and 7 of 24 for the 200 mg, 400 mg, 800 mg, and placebo arms, respectively. The unadjusted risk-difference estimates are -0.042 for 400 mg and 0.108 for 800 mg, with both intervals spanning zero. The registry record does not report secondary analyses, the original sample-size justification, or key details of the statistical analysis plan and missing-data handling.