Study in Participants With Early Stage Coronavirus Disease 2019 (COVID-19) to Evaluate the Safety, Efficacy, and Pharmacokinetics of Remdesivir Administered by Inhalation: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study in Participants With Early Stage Coronavirus Disease 2019 (COVID-19) to Evaluate the Safety, Efficacy, and Pharmacokinetics of Remdesivir Administered by Inhalation (NCT04539262), a randomized phase 1, phase 2 study. The primary outcome is Time-weighted Average Change From Baseline in Nasopharyngeal Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Viral Load Through Day 7.

2. Clinical Question

3. Trial Design

This was a randomized, double-blind, parallel-group Phase 1b/2a study of inhaled remdesivir in participants with early-stage COVID-19. The registry lists the study as completed, with enrollment of 156. The study began on 2020-09-14, and primary completion was 2021-02-26. Gilead Sciences was the sponsor.

The intervention groups were organized into Parts A, B, and C, with separate placebo groups. In Part A, participants could receive inhaled RDV 31 mg daily for 5 days, RDV 31 mg daily for 3 days followed by placebo for 2 days, or placebo for 5 days. Part B used RDV 62 mg daily for up to 5 days, a 3-day RDV regimen followed by placebo for 2 days, or placebo for 5 days. Part C compared inhaled RDV 39 mg daily for 5 days with matching placebo for 5 days.

The primary endpoint analysis included participants who were randomized, received at least one dose, and had a positive baseline nasopharyngeal SARS-CoV-2 viral load, according to the registry description. The listed analysis group sizes differ across arms and parts, so the total enrollment should not be read as the number contributing to each primary comparison.

The registry record does not report the original sample-size justification, the planned statistical power, or the complete protocol-level analysis plan. It reports several secondary outcomes, including adverse events, medically attended visits or death, hospitalization, and pharmacokinetic measures. The posted results therefore describe multiple study questions, but the primary viral-load endpoint remains the central efficacy-related measure in the 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.

Before assignment: 168 participants were screened.

Only the first 6 of 8 groups are shown.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Participants in the Nasopharyngeal Modified Full Analysis Set (mFAS) (all participants who were randomized into the study, had received at least 1 dose of study treatment and had positive SARS-CoV-2 viral load using nasopharyngeal swab sample at baseline \[the result of 'No SARS-CoV-2 detected' was considered as negative\]) with available data were analyzed.

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.

Population: The Safety Analysis Set included all participants who were randomized into the study and received at least 1 dose of study treatment.

Only the first 6 of 9 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Time-weighted Average Change From Baseline in Nasopharyngeal Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Viral Load Through Day 7, measured over Baseline, Day 7.

The primary endpoint was the time-weighted average change from baseline in nasopharyngeal SARS-CoV-2 viral load through Day 7. Viral load was measured in log10 copies/mL. The registry defines the time-weighted average as the area under the curve of viral-load change over time, divided by the time from baseline through Day 7. This uses the measurements over the interval rather than relying only on a single measurement at the end of follow-up.

A change in viral load is a continuous outcome: each participant’s result can take a range of values, and group summaries describe the average change and its variability. The primary endpoint is intended to characterize the effect of inhaled RDV on viral load during the early period studied. Viral load is a laboratory measure, not itself a direct measure of how a participant feels or functions. The registry also lists clinical outcomes such as medically attended visits, death, and hospitalization, but those are separate secondary endpoints.

The endpoint is defined through Day 7, while some secondary outcomes are assessed through Day 28. These outcomes address different time windows and different questions, so findings from one should not be treated as direct substitutes for findings from another. The exact handling of all endpoint measurements, including any missing observations, is not reported in the registry record.

7. Results Posted in the Registry

• RDV, Part A: mean -2.04 versus -1.51 on control; difference -0.53 (95% CI -1.29 to 0.23), standardized effect d = -0.58; z-test p = 0.172.
• Placebo, Part A: mean -1.24 versus -1.51 on control; difference 0.27 (95% CI -0.47 to 1.01), standardized effect d = 0.30; z-test p = 0.477.
• RDV + Placebo, Part B: mean -1.42 versus -1.51 on control; difference 0.09 (95% CI -0.68 to 0.86), standardized effect d = 0.10; z-test p = 0.819.
• Placebo, Part B: mean -1.40 versus -1.51 on control; difference 0.11 (95% CI -0.54 to 0.76), standardized effect d = 0.14; z-test p = 0.739.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Number of Participants With All-Cause Medically Attended Visits (MAVs) or Death by Day 28

• 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:
  • RDV, Part A: 8.3% (1/12) versus 8.3% (1/12) on control; risk difference 0.0 percentage points (95% CI -22.1 to 22.1), risk ratio 1.00 (0.07–14.21); unadjusted two-proportion z-test p = 1.000.
  • Placebo, Part A: 8.3% (1/12) versus 8.3% (1/12) on control; risk difference 0.0 percentage points (95% CI -22.1 to 22.1), risk ratio 1.00 (0.07–14.21); unadjusted two-proportion z-test p = 1.000.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between RDV, Part A and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Placebo, Part A and the control. The interval gives the range of effects that remain compatible with the data.

Number of Participants With COVID-19 Related MAVs or Death by Day 28

• 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:
  • RDV, Part A: 0.0% (0/12) versus 8.3% (1/12) on control; risk difference -8.3 percentage points (95% CI -24.0 to 7.3); unadjusted two-proportion z-test p = 0.307.
  • Placebo, Part A: 8.3% (1/12) versus 8.3% (1/12) on control; risk difference 0.0 percentage points (95% CI -22.1 to 22.1), risk ratio 1.00 (0.07–14.21); unadjusted two-proportion z-test p = 1.000.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between RDV, Part A and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Placebo, Part A and the control. The interval gives the range of effects that remain compatible with the data.

Number of Participants With Hospitalization by Day 28

• 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:
  • RDV, Part A: 0.0% (0/12) versus 8.3% (1/12) on control; risk difference -8.3 percentage points (95% CI -24.0 to 7.3); unadjusted two-proportion z-test p = 0.307.
  • Placebo, Part A: 8.3% (1/12) versus 8.3% (1/12) on control; risk difference 0.0 percentage points (95% CI -22.1 to 22.1), risk ratio 1.00 (0.07–14.21); unadjusted two-proportion z-test p = 1.000.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between RDV, Part A and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Placebo, Part A and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes cover several distinct areas: treatment-emergent adverse events and laboratory abnormalities, treatment discontinuations, medically attended visits or death, COVID-19-related medically attended visits or death, hospitalization, and pharmacokinetic measures. The posted count summaries show that the event patterns vary across arms and study parts. Several groups have no recorded events for particular binary outcomes, while other groups have one or more.

These secondary measures complement the primary viral-load endpoint but do not measure the same thing. Viral load is a laboratory outcome through Day 7; clinical event outcomes extend through Day 28, and pharmacokinetic outcomes describe drug and metabolite exposure. Consequently, an observed pattern in one category does not directly confirm or contradict a result in another. The record does not report a combined analysis that would integrate these endpoints.

Many secondary comparisons involve small numbers of events or small groups. In general, results based on sparse events are imprecise and can change substantially with only a few observations. The registry lists no sponsor-reported inferential analyses for the displayed secondary binary comparisons. Any simple comparisons should therefore be treated as descriptive and exploratory rather than as definitive tests.

Unadjusted secondary results carry less weight when many outcomes and comparisons are examined because the chance of finding apparently notable patterns by chance increases with the number of tests. The registry record does not report a multiplicity adjustment or a hierarchy for interpreting the secondary endpoints. These outcomes can provide context about what was observed, but they should not be treated as independent confirmation of the primary result.

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

• Percentage of Participants Experiencing Treatment Emergent Adverse Events (TEAEs)
• Percentage of Participants With Treatment-Emergent Laboratory Abnormalities as Per Severity Grade
• Percentage of Participants With Treatment-Emergent Adverse Events Leading to Study Treatment Discontinuation
• Number of Participants With All-Cause Medically Attended Visits (MAVs) or Death by Day 28
• Number of Participants With COVID-19 Related MAVs or Death by Day 28
• Number of Participants With Hospitalization by Day 28
• Pharmacokinetic (PK) Parameter: AUC0-24h of Remdesivir (RDV) and Its Metabolites (GS-441524 and GS-704277) in Parts A and B
• PK Parameter: AUClast of RDV and Its Metabolites (GS-441524 and GS-704277) in Parts A and B

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is continuous, and the registry reports it as a mean-based measure with standard deviation. A mean summarizes the central tendency of measurements in a group; the standard deviation describes how spread out participant values are around that mean. These are different quantities: the standard deviation reflects participant-to-participant variability, while a confidence interval around a treatment contrast reflects uncertainty in the estimated group difference.

What does ANCOVA contribute?

The registry names ANCOVA, or analysis of covariance, for the primary comparisons. In general, ANCOVA estimates a group contrast while accounting for specified covariates. This can produce a least-squares mean difference rather than a simple subtraction of the raw group means. The registry record does not identify all covariates used in the model, so the adjusted estimates cannot be reconstructed from the posted group means and standard deviations alone.

The registry reports least-squares mean differences by Day 7 for comparisons within Parts A, B, and C. These estimates are the sponsor-reported inferential analyses for the primary endpoint. The simple recalculations use the posted means and standard deviations to form unadjusted mean differences and approximate confidence intervals and p-values. They are useful as a transparent check on the raw summaries, but do not reproduce the covariate-adjusted ANCOVA.

How should the difference scale be read?

The endpoint is measured in log10 copies/mL, and the estimates are contrasts on the stated endpoint scale. A contrast of zero would mean no difference between the compared group summaries on that scale. Negative and positive estimates indicate opposite directions according to the comparison ordering used for the analysis. Since the registry does not give a full model specification or all details of the contrast coding, the safest interpretation is to retain the reported contrast labels and estimates rather than infer more about the calculation.

What do p-values describe?

A p-value describes how compatible the observed data are with a specified null model, under the assumptions of the analysis. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. The reported primary p-values come from the ANCOVA analyses. The recalculated p-values are unadjusted and based on the posted summary statistics; they answer a different, simpler comparison. The registry record does not report whether the primary p-values were adjusted for multiple comparisons.

Why do small groups matter?

Several primary analysis groups contain relatively few participants. In general, estimates based on small groups are sensitive to individual observations and tend to have greater uncertainty than estimates based on larger groups, all else being equal. The standard deviations also indicate meaningful variation among observed values. The study has multiple parts and treatment regimens, so comparing every arm as though it belonged to one common experiment would overlook the design structure.

For binary secondary outcomes, the posted results are participant counts with and without an event. In general, when events are uncommon, estimated group differences and relative measures can be unstable, especially when one group has no events. A confidence interval communicates this uncertainty; a point estimate alone can appear more conclusive than the data warrant. The registry record contains no sponsor-reported statistical analyses for the listed secondary outcome comparisons.

The registry record does not report the complete prespecified statistical analysis plan, all covariates, missing-data methods, interim analyses, or a multiplicity strategy. These details should not be inferred from the reported results. General statistical interpretation depends on such design and analysis choices, particularly when several comparisons and endpoints are reported.

11. Confidence Intervals and Precision

A confidence interval describes a range of values reasonably compatible with the data and the statistical model, while reflecting sampling uncertainty. It is not a guarantee that the true effect lies in the interval, and it is not a range of individual participant outcomes. Wider intervals indicate less precision; narrower intervals indicate greater precision, conditional on the model and the information available.

The registry-reported primary estimates all have 95% confidence intervals that cross zero. For example, the Part A RDV versus placebo estimate is -0.66, with an interval from -1.49 to 0.16. The Part C estimate is 0.20, with an interval from -0.40 to 0.81. These intervals include values on either side of zero, so the posted analyses do not pin down the direction of the contrast. The point estimates should be read alongside the full intervals, not in isolation.

Precision varies across comparisons. The observed group sizes differ by study part, and the intervals reflect both the amount of information and the variability in the outcome. The registry reports standard deviations for the group summaries, with values ranging across the arms. Variation among participants contributes to uncertainty in estimates, while limited group sizes can further widen intervals.

The recalculated intervals are unadjusted and are not interchangeable with the sponsor-reported ANCOVA intervals. Each addresses its own analysis. In general, confidence intervals depend on the chosen model and assumptions, and the registry record does not provide enough detail to independently verify every element of the ANCOVA specification. The posted interval widths are therefore best interpreted as the precision reported for the stated analyses, not as a complete description of all possible sources of uncertainty.

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 sample-size figures are planning illustrations based on the observed effect, not the trial’s original design assumptions. For the continuous endpoint comparisons, the figures are participants per arm. The planning assumptions listed in the registry record are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; the target effect is the stated fraction of the observed effect. The trial’s own sample-size justification is not reported in the registry record.

The 80% power column gives the number of participants per arm estimated for a design with an 80% chance of detecting the stated effect under those assumptions, if that effect and model assumptions were correct. The 90% power column gives the corresponding per-arm planning figure for a 90% chance. These are design calculations, not results observed in this trial. They do not show whether the trial had a particular probability of success after the data were collected: this is not observed power.

Rows for 100%, 75%, and 50% of the observed effect illustrate how planning requirements change with the target effect. A smaller effect is harder to distinguish from random variation, so it requires more participants per arm under the same assumptions. The observed effects are themselves estimates from limited data and may not represent the effect a future study would target.

The displayed sample-size curve should be read as a visual presentation of the same planning relationship, not as an additional observed outcome. These illustrations do not account for every possible feature of a future study. In particular, the record does not report the original design rationale or enough information to treat these calculations as a reconstruction of the trial’s planning process.

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

Worked calculation for this trial

1. Observed difference Δ = -0.530; pooled standard deviation σ = 0.910.
2. n = 2 × (1.96 + 0.842)² × 0.910² / -0.530² = 46.3, rounded up to 47 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 <- -0.53                      # observed difference in means
sigma <- 0.910314506091164                     # 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(4539262)
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 = 47 (80% power), 62 (90% power)
 75% of observed effect: n per arm = 83 (80% power), 111 (90% power)
 50% of observed effect: n per arm = 186 (80% power), 248 (90% power)
power.t.test: 47.3 
Simulated power at n = 47 per arm: 0.796 (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.

14. Analyses Reported by the Sponsor

The registry reports ANCOVA results for the primary endpoint as least-squares mean differences by Day 7. These are the sponsor-reported analyses and should be distinguished from the simple unadjusted recalculations based on the posted group means and standard deviations. The registry record does not flag a discrepancy between the reported and recalculated results, but that does not make the methods identical.

ANCOVA can account for specified covariates, whereas the recalculations compare the posted summaries without adjustment. Differences between their estimates, intervals, or p-values can therefore arise because they answer different statistical comparisons. The adjustment details needed to reproduce the ANCOVA estimates are not reported in the registry record.

The sponsor-reported analyses are the relevant results for describing the registry’s primary endpoint analysis. The recalculated results can provide a rough, unadjusted reference based on the arm summaries, but should not replace the reported ANCOVA or be interpreted as a correction to it. Neither set of estimates should be separated from its confidence interval, study-part context, and analysis population.

15. Multiplicity, Missing Data and Analysis Populations

The registry record lists multiple primary-arm contrasts and numerous secondary outcomes. When several statistical comparisons are conducted, the chance of at least one apparently unusual result can increase even if there is no underlying effect. A multiplicity strategy can define how the analyses are prioritized or adjust inference across tests. Such a strategy is not reported in the registry record.

The primary outcome analyses are reported separately by study part and regimen. The record does not state whether a formal testing hierarchy was prespecified, whether p-values were adjusted across those comparisons, or how confirmatory versus exploratory analyses were designated. The reported p-values should therefore be read as stated for each analysis, without assuming family-wise error control or another multiplicity procedure.

The primary population is described as a Modified Full Analysis Set consisting of randomized participants who received at least one dose and had positive baseline nasopharyngeal viral load. The registry description of the baseline positivity criterion is incomplete. The record does not report the number excluded from this population, a complete accounting of missing endpoint measurements, or the method used to handle missing data.

Missing data can affect both the estimate and its uncertainty, particularly if the reasons for missing measurements are related to participant status or treatment. General methods may make different assumptions about such missingness, but the registry record does not identify which approach was used for the primary endpoint. It also does not report sensitivity analyses assessing alternative missing-data assumptions.

The record describes randomization, double masking, and a parallel-group design, but does not supply the full protocol or statistical analysis plan. Details such as stratification, covariate selection, interim analyses, and the original sample-size rationale are not reported in the registry record. These omissions limit independent assessment of whether every analysis followed the prespecified plan.

16. 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: acute myocardial infarction (8.3%), incarcerated hernia (8.3%), covid-19 pneumonia (8.3%).

The adverse-event summaries report participants at risk and participants with serious adverse events by arm. The observed serious-event counts were zero in several arms and one or two in others. In Part A, the RDV arm had 0 of 12 participants with a serious adverse event, the RDV-plus-placebo arm had 1 of 12, and the placebo arm had 2 of 12. In Part B, the corresponding counts were 0 of 12, 1 of 12, and 1 of 13. In Part C, the RDV arm had 0 of 61 and the placebo arm had 0 of 20.

The registry also lists acute myocardial infarction, incarcerated hernia, and Covid-19 pneumonia among the top serious-event terms, each with a reported value of 8.3%. These summaries describe observed events in the study; they do not establish that an event was caused by a study treatment. With small groups and few events, comparative safety estimates would be imprecise. The record does not report deaths in the listed arms.

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

The registry results describe a randomized, double-blind, parallel-group study of inhaled RDV regimens across several study parts. The primary endpoint was change in nasopharyngeal SARS-CoV-2 viral load through Day 7. The reported ANCOVA estimates have 95% confidence intervals that include zero, and the estimates vary in direction across parts. This means the posted results leave uncertainty about the direction and size of each part-specific contrast.

The findings do not provide a single pooled estimate across all study parts. Regimens and group sizes differed across parts, and the registry does not report a combined primary analysis. The secondary outcomes address other questions and have sparse event counts in several groups. Those counts are useful descriptions of what occurred, but should not be taken as precise evidence of differences between groups.

The results also should not be interpreted as proving equivalence between any regimen and placebo. An interval that includes zero can be compatible with a range of possible effects; it does not show that the groups are identical. Nor does a p-value alone indicate whether an effect is important. The registry record does not report a threshold for a clinically meaningful viral-load difference.

More broadly, a laboratory endpoint and clinical outcomes are distinct. A pattern in viral load does not by itself establish a change in symptoms, complications, or other clinical outcomes. Conversely, a sparse set of clinical events cannot resolve the primary viral-load question. Interpretation should remain tied to the endpoint, follow-up period, analysis population, and precision of each reported result.

These statistical results are not medical advice. They describe the posted trial data and their uncertainty, and do not determine what treatment is appropriate for any individual.

18. Why This Trial Matters Statistically

This trial is statistically informative as an example of why design structure and estimation method matter. It includes multiple treatment regimens organized across study parts, rather than one simple treatment-versus-control contrast. Keeping comparisons within their parts avoids treating distinct regimens and group sizes as though they were interchangeable.

The registry presents adjusted ANCOVA estimates as well as arm-level means and standard deviations from which unadjusted comparisons can be recalculated. Putting those results in context illustrates an important distinction: a raw summary comparison and a covariate-adjusted analysis can differ because they do not necessarily estimate the same quantity. Readers need to identify which analysis is being interpreted before comparing estimates or p-values.

The primary confidence intervals also highlight the role of precision. Point estimates may look directional, but intervals that span zero preserve substantial uncertainty. Small comparison groups and variability in the continuous endpoint make it important to consider the full interval rather than relying on a threshold or a single number.

Finally, the trial reports several kinds of outcomes, from viral load to clinical events and pharmacokinetics. These outcomes answer related but distinct questions. A careful statistical reading keeps the primary endpoint central, treats secondary findings with appropriate caution, and recognizes that the registry does not report all protocol and analysis details needed for a complete audit of the design.

19. 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 features of this specific record limit the conclusions that can be drawn. The primary analysis population is described, but the final sentence defining baseline positivity is truncated. The exact complete eligibility rule for that analysis set is therefore not available in the registry record.

The study is divided into Parts A, B, and C, with different doses, treatment schedules, and group sizes. The primary analyses are presented as separate within-part comparisons, and a pooled estimate across parts is not reported. This means the results do not directly summarize one common regimen across the entire enrollment.

The observed sample sizes for primary analyses vary, and some arms are small. That limits precision and makes estimates more sensitive to individual observations. The study’s original sample-size justification and design power are not reported in the registry record. The planning illustrations are separate from those original assumptions and do not reconstruct the trial’s design.

The registry names ANCOVA but does not give the complete model specification or all covariates. It also does not report the missing-data procedures, sensitivity analyses, interim analyses, or multiplicity strategy. These details are important for evaluating whether estimates and p-values reflect a prespecified plan and how robust they are to analysis choices.

Secondary event counts are sparse in several arms, and the record lists no sponsor-reported inferential analyses for the displayed binary secondary comparisons. Safety counts are descriptive and do not alone establish treatment attribution or comparative risk. These limitations should be considered alongside the primary estimates and their intervals.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Study in Participants With Early Stage Coronavirus Disease 2019 (COVID-19) to Evaluate the Safety, Efficacy, and Pharmacokinetics of Remdesivir Administered by Inhalation — NCT04539262.
• Registry record last updated 2022-03-03; status last verified 2022-02.
• Results first posted 2022-03-03.
• 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.

23. Record Summary

NCT04539262 was a completed, randomized, double-blind, parallel-group Phase 1b/2a study sponsored by Gilead Sciences. It enrolled 156 participants with early-stage COVID-19 and evaluated inhaled remdesivir in different dose and schedule regimens across Parts A, B, and C, with placebo groups in each part.

The primary endpoint was time-weighted average change in nasopharyngeal SARS-CoV-2 viral load from baseline through Day 7. The registry-reported ANCOVA comparisons have 95% confidence intervals that include zero. The secondary outcomes cover safety, clinical events, and pharmacokinetics; several posted event counts are small.

The record does not report the original sample-size rationale, complete statistical analysis plan, missing-data approach, or multiplicity strategy. Results are therefore best understood as part-specific estimates with the uncertainty and reporting limitations.