A Study of Ad26.COV2.S in Adults (COVID-19): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Ad26.COV2.S in Adults (COVID-19) (NCT04436276), a randomized phase 1, phase 2 study. The primary outcome is Cohorts 1a and 1b: Number of Participants With Solicited Local (Injection Site) Adverse Events (AEs) for 7 Days After Vaccination 1 in the Primary Regimen.

2. Clinical Question

3. Trial Design

NCT04436276 is a completed, randomized, double-blind, placebo-controlled Phase 1/2a study with a parallel-group design. The registry records enrollment of 1085 participants. The study began on 2020-07-15, and primary completion was recorded as 2023-02-21. Janssen Vaccines & Prevention B.V. is listed as the sponsor.

The broad study purpose was to assess the safety, reactogenicity, and immunogenicity of Ad26.COV2.S across two dose levels and different primary and booster schedules. Participants included adults aged 18 to 55 years and adults aged 65 years and older who were in good health, with or without stable underlying conditions. Cohorts varied in age eligibility and vaccination schedule; some evaluated a single vaccination, some a two-dose primary regimen, and some included later booster vaccination.

The primary outcome reported in the registry is narrower than the overall study purpose. It counts participants in Cohorts 1a and 1b with solicited local adverse events during the 7 days after vaccination 1 in the primary regimen. Cohort 1a includes groups receiving different vaccine dose and schedule combinations and a placebo group. Cohort 1b also includes vaccine and placebo groups, but the listed group sizes are 5 participants each. Results for the primary outcome should therefore be understood in the context of the particular cohorts and outcome time window, rather than as a summary of all 1085 enrolled participants.

The registry lists 70 outcome measures and 24 posted secondary outcomes. These include antibody binding and neutralizing antibody outcomes. The original protocol’s detailed allocation procedures, sample-size justification, and statistical analysis plan are 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.

Before assignment: A total of 1085 participants were enrolled in the study, out of which 1076 participants received treatment. Remaining 9 participants did not receive any treatment and are excluded from the analyses.

Only the first 6 of 23 groups are shown.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Full analysis set (FAS) included all participants with at least one vaccine administration documented.

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.

Only the first 6 of 24 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Cohorts 1a and 1b: Number of Participants With Solicited Local (Injection Site) Adverse Events (AEs) for 7 Days After Vaccination 1 in the Primary Regimen, measured over 7 days post-vaccination 1 on Day 1 (Day 8).

The primary endpoint is the number of participants with solicited local adverse events for 7 days after vaccination 1 in the primary regimen in Cohorts 1a and 1b. The events were injection-site pain or tenderness, redness (erythema), and swelling. These are predefined local reactions at the injection site, used to characterize reactogenicity. The endpoint is a participant count: a participant is counted if they meet the event definition during the stated window. It does not describe the number of separate symptoms or episodes experienced by each participant.

Reactogenicity is distinct from broader safety outcomes and from immunogenicity. The endpoint records local events after vaccination; it does not measure whether vaccination prevents COVID-19, and it does not quantify antibody response. The registry describes an adverse event as an untoward medical occurrence that does not necessarily have a causal relationship with the study agent. Thus, a reported event is not by itself proof that vaccination caused it.

The analysis population is the full analysis set, defined in the registry as participants with at least one vaccine administration documented. The endpoint’s short observation window makes it a specific early post-vaccination outcome. It cannot characterize events beyond 7 days after the first vaccination or substitute for the study’s other safety and immunogenicity outcomes.

7. Results Posted in the Registry

• COHORT 1A: Ad26 5e10, Ad26 5e10(AHBV: Ad26 5e10): 64.9% (50/77) versus 10.4% (8/77) on control; risk difference 54.5 percentage points (95% CI 41.9 to 67.2), risk ratio 6.25 (3.18–12.29); unadjusted two-proportion z-test p = <0.001.
• COHORT 1A: Ad26 5e10, PL(AHBV: Ad26 5e10): 66.7% (50/75) versus 10.4% (8/77) on control; risk difference 56.3 percentage points (95% CI 43.6 to 68.9), risk ratio 6.42 (3.27–12.61); unadjusted two-proportion z-test p = <0.001.
• COHORT 1A: Ad26 1e11, Ad26 1e11(AHBV: Ad26 5e10): 76.0% (57/75) versus 10.4% (8/77) on control; risk difference 65.6 percentage points (95% CI 53.8 to 77.4), risk ratio 7.31 (3.75–14.27); unadjusted two-proportion z-test p = <0.001.
• COHORT 1A: Ad26 1e11, PL(AHBV: Ad26 5e10): 80.8% (59/73) versus 10.4% (8/77) on control; risk difference 70.4 percentage points (95% CI 59.1 to 81.7), risk ratio 7.78 (4.00–15.13); unadjusted two-proportion z-test p = <0.001.

8. Statistical Interpretation

9. Statistical Methods Explained

What kind of outcome is this?

This is a binary participant-level outcome: for the specified 7-day window, each participant either met the definition of having at least one solicited local adverse event or did not. The registry reports counts of participants with an event, along with the group sizes. A count is not a continuous measurement, and the result does not indicate how many different local symptoms an affected participant experienced. The endpoint description identifies pain or tenderness, erythema, and swelling as the solicited local events.

General statistical explanation: for a binary outcome, an event proportion is the number of participants with an event divided by the number assessed in that group. A risk difference compares two such proportions by subtraction. It is expressed on the absolute probability scale, so it describes the change in event frequency between groups rather than a relative multiplier. The recalculated estimates in the registry analysis are unadjusted contrasts based on the posted summary counts.

What do the risk differences and p-values convey?

The listed Cohort 1a risk differences compare each vaccine group with placebo. Their positive direction reflects the larger observed event proportion in each vaccine group than in placebo. The accompanying p-values are unadjusted and are all reported as <0.001. A p-value describes how incompatible the observed data are with a specified null model under its assumptions; it is not the probability that the null hypothesis is true, and it does not measure clinical importance.

General statistical explanation: an unadjusted comparison treats a given contrast separately and does not account for the possibility of multiple comparisons. When many contrasts or outcomes are examined, the chance of at least one small p-value under null conditions can increase. Whether and how to adjust depends on the prespecified testing strategy. The registry record does not report the trial’s multiplicity strategy or a sponsor analysis plan, so these recalculated p-values should not be interpreted as multiplicity-controlled findings.

Why do confidence intervals matter?

A confidence interval gives a range of effect estimates compatible with the observed data under the method used. It communicates uncertainty as well as direction. The recalculated intervals for the Cohort 1a differences exclude zero, consistent with the positive contrasts in the posted estimates. Their endpoints also show that the exact size of the difference remains uncertain. Confidence intervals do not account for every possible source of bias or design limitation, and they do not guarantee that the interval contains the true effect.

General statistical explanation: interval width depends on factors such as the number of participants and the distribution of outcomes. With limited group sizes, estimates can be imprecise even when a point estimate appears large. Conversely, a narrow interval indicates greater statistical precision for the particular comparison and model, not necessarily freedom from bias or broad applicability.

What is and is not specified about the analysis?

The registry defines the full analysis set as participants with at least one vaccine administration documented. That identifies the population used for the stated endpoint, but it does not give the details of how missing endpoint observations were handled, whether any participants were excluded after vaccination, or how protocol deviations were treated. Those details are not reported in the registry record.

The registry lists the primary outcome as a count and records no sponsor analysis count in the summary information. The comparisons presented are recalculated, unadjusted estimates from the posted summary statistics. They are not a substitute for a detailed prespecified analysis. The registry record does not report the original statistical model, covariate adjustment, planned estimand, sample-size justification, or interim-analysis plan. No assumption about those elements should be inferred from the recalculated estimates.

General statistical explanation: randomization is intended to create comparable groups on average, while masking can reduce some forms of expectation and assessment bias. Those design features do not eliminate chance imbalance, incomplete outcome information, measurement error, or limitations in generalizing to people outside the studied groups. The trial is recorded as randomized and double-blind, but the specific implementation details and any deviations from the intended design are not reported in the registry record.

Finally, these results concern solicited local events during a defined early period. The safety outcomes elsewhere in the registry include serious and other adverse events, while secondary outcomes include antibody measures. Those outcomes answer different questions and should not be combined conceptually with the primary local-event count. The registry record lists many outcomes, but the endpoint-specific counts alone cannot establish an overall safety profile or an immunogenicity conclusion.

10. Confidence Intervals and Precision

The Cohort 1a group sizes range from 73 to 77 participants, with 77 participants in the placebo group. For the four vaccine comparisons, the recalculated confidence intervals for the risk differences all lie above zero. This supports a positive observed contrast in these data, while the interval endpoints indicate that the precise size of each contrast is not known exactly. The posted intervals are not identical across groups, so precision differs somewhat by comparison.

Confidence intervals are especially useful alongside the event counts because a point estimate alone can give a false impression of certainty. General statistical explanation: a confidence interval reflects sampling uncertainty under a particular calculation method. It does not include uncertainty from unrecorded outcomes, departures from the analysis plan, outcome misclassification, or limits in how well the study participants represent other populations. The registry record does not report enough detail to assess all of those sources.

The small Cohort 1b groups, each with 5 participants, provide much less information for comparative inference than the Cohort 1a groups. Counts such as 4 of 5 or 5 of 5 are descriptive observations; they do not provide a stable basis for estimating a population event frequency with high precision. The registry record does not list recalculated confidence intervals or p-values for Cohort 1b comparisons.

The reported precision is specific to solicited local events after the first vaccination and to the analysis population stated for this endpoint. It should not be generalized to later doses, booster visits, serious adverse events, antibody outcomes, or COVID-19 prevention. Nor should statistical precision be confused with clinical importance: the interval quantifies uncertainty in an estimate, not the value or consequence of an event.

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 express participants per arm for 80% power and participants per arm for 90% power. They are based on a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each illustration uses the observed effect or a stated fraction of that observed effect as the effect to detect. For example, planning for 75% of the observed effect uses a smaller target effect than planning for 100% of the observed effect.

The 80% power column gives the illustrative number of participants per arm needed for an 80% chance of detecting the specified effect under those assumptions, if that effect were the true one. The 90% power column gives the corresponding illustrative number per arm for a 90% chance. They are planning quantities, not outcomes measured in the trial. In the listed comparisons, the illustrations require more participants per arm when planning for a smaller fraction of the observed effect.

These calculations are based on observed summary data, not on the trial’s original design rationale. The trial’s original sample-size justification and assumptions are not reported in the registry record. This is not observed power: it does not quantify the probability that the trial detected an effect after observing its results. The illustrations also do not account for features not listed in the stated assumptions, such as adjustment for multiple comparisons or missing outcome data.

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 = 0.104; treatment proportion p₁ = 0.649; average p̄ = 0.377.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6852 = 1.3430.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5664 = 0.4767.
4. n = (1.3430 + 0.4767)² / (0.5455)² = 11.1, rounded up to 12 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
p0 <- 0.103896103896104                          # control proportion
p1 <- 0.649350649350649                          # 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(4436276)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 12 (80% power), 15 (90% power)
 75% of observed effect: n per arm = 19 (80% power), 25 (90% power)
 50% of observed effect: n per arm = 38 (80% power), 50 (90% power)
power.prop.test: 11.1 
Simulated power at n = 12 per arm: 0.860 (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 reports several Cohort 1a comparisons for the primary local-event outcome, along with multiple cohorts and many other outcomes. The posted p-values for the Cohort 1a comparisons are unadjusted. How the sponsor planned to control the overall type I error across comparisons or outcomes is not reported in the registry record. Therefore, the individual p-values should not be read as if they necessarily account for all the testing conducted in the study.

The primary analysis population is described as the full analysis set, including participants with at least one vaccine administration documented. The endpoint counts and group sizes are posted, but the registry record does not report a detailed missing-data procedure. It does not state how a participant with an incomplete event diary or unavailable observation during the 7-day window would be classified, or whether any sensitivity analyses were conducted. No assumptions about these issues can be made from the summary counts alone.

The sponsor’s complete statistical analysis plan, planned covariates, interim-analysis procedures, and sample-size justification are not reported in the registry record. The registry summary does not report a sponsor inferential analysis for this outcome; the estimates discussed for Cohort 1a are simple unadjusted recalculations from posted counts. These do not establish which analysis was primary in the original protocol or whether additional analyses were conducted.

Randomization and double masking are recorded design features, but the registry record does not give the detailed methods used to implement them or information about deviations. It also does not describe how protocol deviations, withdrawals, or post-randomization exclusions were handled for this endpoint. These omissions limit assessment of the analysis population and of possible sources of bias. They do not change the descriptive fact that the listed numbers are counts among the stated group sizes, but they constrain broader statistical conclusions.

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: nephrolithiasis (20.0%), anaphylactic shock (3.3%), uterine prolapse (3.3%), asphyxia (3.1%), wrist fracture (1.7%).

The registry reports serious adverse events and other adverse events by group, with the number at risk and the number affected. These are separate summaries: serious adverse events are not interchangeable with the broader other-adverse-event counts. In Cohort 1a, the serious-event counts were 0 of 77, 1 of 75, 1 of 75, 1 of 73, and 2 of 77 across the four vaccine groups and placebo group, respectively. The listed death counts for these Cohort 1a groups are 0.

Across all listed cohorts, the registry includes some groups with recorded deaths and other groups with none. The counts should be read within their group and cohort, not combined into a single trial-wide rate without an appropriate denominator and analysis. A reported serious event or death in a group does not establish that the study vaccination caused it. The registry lists several serious-event terms with percentages, but the denominator and calculation basis for those percentages are not reported in the registry record. These summaries also do not convey event timing, duration, severity grading, or causal assessment for each participant.

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

The posted primary outcome answers a focused question about solicited local events during the 7 days after the first vaccination. In Cohort 1a, each listed vaccine group had more participants with such an event than the placebo group, and the unadjusted recalculated comparisons have p-values <0.001. The results therefore describe a clear difference in the observed counts for this particular endpoint and window.

That finding does not establish whether any individual event was caused by vaccination. The registry’s definition of an adverse event explicitly allows events without a necessary causal relationship to the study agent. Nor does the outcome describe the severity, duration, or impact of each local event. It is a count of participants meeting a solicited-event definition, not a complete account of experience or safety.

The endpoint does not measure protection against COVID-19. The study’s broader aims also included immunogenicity and safety, and the registry lists antibody outcomes and adverse-event summaries, but the primary local-event counts cannot answer those separate questions. In particular, these counts do not show whether vaccination prevents infection, hospitalization, or any other COVID-19 outcome.

The Cohort 1a findings should not be generalized automatically to Cohort 1b, the older-adult cohort, other schedules, later booster vaccinations, or populations not represented in these groups. Cohort 1b has only 5 participants per listed group, and the primary outcome comparison estimates are reported for Cohort 1a. The registry record also lacks detail on multiplicity, missing data, and the original analysis plan. The appropriate interpretation is therefore bounded: the posted Cohort 1a counts show a higher observed frequency of solicited local events in vaccine groups than in placebo during the stated early period, while the wider clinical and causal implications remain outside what this endpoint establishes.

16. Why This Trial Matters Statistically

This study is statistically informative because it separates a specific, measurable reactogenicity outcome from the larger questions of immunogenicity and prevention. Defining the event type and observation window makes the comparison interpretable: the outcome concerns participants with prespecified local injection-site events during 7 days after the first vaccination, rather than all adverse experiences over the full study.

The placebo-controlled randomized design provides a basis for comparing observed event counts between groups, while the double-blind design is intended to limit some influences on reporting and assessment. The group-level counts also make clear that the primary outcome draws on different amounts of information across cohorts. Cohort 1a has groups in the 73-to-77 participant range, whereas Cohort 1b has 5 participants per group. Those differences matter when judging how much confidence to place in a comparison.

The study also illustrates why estimates, uncertainty, and analysis scope should be considered together. Unadjusted comparisons can describe individual contrasts, but many comparisons raise questions about multiplicity. Confidence intervals communicate uncertainty in effect size; small groups generally give less stable estimates. Sample-size illustrations based on observed effects can help show how planning depends on the target effect, but they do not replace the trial’s original design justification.

Finally, the registry’s breadth of outcomes should not obscure the question answered by any one endpoint. Local reactogenicity counts cannot stand in for antibody response, broader safety, or disease prevention. Keeping those distinctions explicit is essential for sound statistical interpretation.

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.

The primary endpoint is limited to solicited local events after vaccination 1 and a 7-day observation period. It does not summarize local events after later doses or boosters, events occurring outside that window, or all adverse events. The outcome is reported as a participant count, so the posted result does not indicate the number of distinct symptoms, their severity, or their duration.

The displayed primary outcome covers Cohorts 1a and 1b, but recalculated comparative estimates are listed for Cohort 1a. Cohort 1b groups contain 5 participants each, which limits the information available for comparing those groups. The older-adult cohort is part of the overall study but is not part of the primary endpoint named in the registry record, so these primary findings do not directly describe that cohort.

The record does not report the original sample-size rationale, formal statistical analysis plan, multiplicity procedure, interim-analysis plan, or detailed missing-data handling. It also does not provide enough information to assess how incomplete diaries, withdrawals, protocol deviations, or other post-randomization issues were treated. The recalculated estimates are unadjusted summaries and may not represent the sponsor’s intended analysis.

These limitations constrain causal and general conclusions. Randomization and masking are recorded, but detailed implementation information is not reported in the registry record. The primary endpoint by itself cannot establish causation, overall safety, immunogenicity, clinical importance, or prevention of COVID-19. The posted counts should be interpreted as evidence about the specified event definition, groups, and time window only.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: A Study of Ad26.COV2.S in Adults (COVID-19) — NCT04436276.
• Registry record last updated 2025-02-04; status last verified 2025-01.
• Results first posted 2024-05-24.
• 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

NCT04436276 is a completed randomized, double-blind, placebo-controlled Phase 1/2a study of Ad26.COV2.S. The registry records 1085 participants enrolled across adult cohorts and vaccination schedules. The primary outcome counts participants in Cohorts 1a and 1b with solicited local adverse events during the 7 days after the first vaccination.

For Cohort 1a, the posted group counts show more participants with local events in each listed vaccine group than in placebo. The recalculated unadjusted risk-difference comparisons have p-values <0.001, with confidence intervals excluding zero. The results are limited to this short-term local-event endpoint; they do not establish causation or address COVID-19 prevention. The original analysis plan, multiplicity strategy, missing-data procedures, and sample-size justification are not reported in the registry record.