The Safety of Administering a Second Dose of a COVID-19 mRNA Vaccine in Individuals Who Experienced a Systemic Allergic Reaction to an Initial Dose: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for The Safety of Administering a Second Dose of a COVID-19 mRNA Vaccine in Individuals Who Experienced a Systemic Allergic Reaction to an Initial Dose (NCT04977479), a randomized phase 2 study. The primary outcome is Proportion of Participants Who Had Recurrent Systemic Allergic Reaction to Vaccine.

2. Clinical Question

3. Trial Design

This was a Phase 2, randomized, placebo-controlled crossover study. The masking designation was quadruple, while the registry record does not identify which specific roles were masked. Its central crossover design exposed participants to active vaccine and placebo on different days, rather than assigning separate groups to receive only one of these during the blinded phase. The arm descriptions state that a 24-hour washout period followed the first dose before participants crossed over. The record does not report the order-allocation procedure or details of how the crossover sequence was implemented.

The study population consisted of people who had experienced a systemic allergic reaction after their first full dose of Pfizer-BioNTech or Moderna COVID-19 vaccine. The brief summary gives an age range of 16-69. Participants were admitted to the NIH hospital and stayed for at least 4 days. The described assessments included urine samples, a nasal swab SARS-CoV-2 test, breathing tests, and intravenous lines in both arms. Clinical photography could be used if a rash developed. Four follow-up visits were planned, including phone and in-person visits, and allergy skin testing was also described.

The blinded active arm received active mRNA vaccine and then placebo; the placebo arm received placebo and then active vaccine. The primary endpoint analysis was defined using participants who received active vaccine in the crossover phase. A separate open-label booster was offered to participants who did not have a CoFAR Grade 3 or higher reaction after the blinded active dose, with the arm description placing it 5 months later. The registry lists a bivalent vaccine intervention, but its relationship to the primary endpoint is not further specified. The trial is marked completed, with enrollment of 20 and start and primary completion dates of 2021-09-08 and 2022-06-17, respectively.

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: Three participants withdrew prior to start of study. One participant was ineligible due to partial 2nd vaccine dose pre-enrollment outside of the study

First Period of Crossover

Washout Period

Second Period of Crossover

Open Label

Analysis population for the primary outcome, as defined in the registry: The analysis included participants who received the active study vaccine in the crossover phase as pre-specified in the protocol to only collect and report data from the active vaccination.

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 Proportion of Participants Who Had Recurrent Systemic Allergic Reaction to Vaccine, measured over Within the 3-hour post-vaccine observation period.

The primary endpoint is the proportion of participants who had a recurrent systemic allergic reaction after the active COVID-19 vaccine dose. It was assessed within the 3-hour post-vaccine observation period. The registry specifies a 95% confidence interval as the dispersion measure. It defines systemic allergic reaction as either a CoFAR Grade 1 reaction accompanied by elevated tryptase or a CoFAR Grade 2 or higher reaction regardless of tryptase level.

That definition combines clinical severity grading with a laboratory marker for the mildest grade included in the endpoint. CoFAR Grade 1 describes mild symptoms involving one organ system; Grade 2 describes mild symptoms involving at least two organ systems. Higher grades include more serious cardiovascular, lower respiratory, or gastrointestinal presentations, with the distinction between grades including responsiveness to treatment and end-organ dysfunction. Grade 5 is death. The definition therefore makes clear which events count, rather than treating every symptom after vaccination as a primary-endpoint event.

A proportion is useful for describing how often a defined event occurred in the outcome population. The confidence interval would communicate uncertainty around that estimated proportion. However, the registry record does not report the primary endpoint's event count, estimated proportion, or interval. The adverse-event figures are reported separately and should not be substituted for the primary endpoint result: the record does not establish that those categories use the same event definition or ascertainment as the primary outcome.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The endpoint is a binary event outcome: each participant either meets the registry definition of systemic allergic reaction during the observation period or does not. A proportion expresses the number of participants meeting that definition relative to the relevant outcome population. This is a direct way to describe an event frequency. It does not, by itself, establish why an event occurred or whether an observed rate differs from another condition.

What does the confidence interval convey?

The registry specifies a 95% confidence interval for the primary proportion. In general, a confidence interval describes uncertainty in an estimate arising from the observed data and the statistical procedure. It provides a range of values compatible with those data under the method's assumptions. It is not a guarantee that the true value lies in the particular interval, nor does it indicate the probability that an individual will experience an event. Without the primary proportion and interval, precision cannot be evaluated for this study.

The interval also helps prevent over-reading a point estimate. When event counts are limited, the observed proportion may be uncertain; a confidence interval makes that uncertainty visible. General statistical interpretation should consider the interval's width and the endpoint definition together. In this record, the interval is named as the dispersion measure, but no numerical interval is reported in the registry record.

Why does the endpoint definition matter?

Outcomes depend on rules for classifying observations. This endpoint uses the CoFAR grading scale and a tryptase condition: Grade 1 counts only when tryptase is elevated, while Grade 2 or higher counts regardless of tryptase. The grading description distinguishes mild symptoms limited to one organ system from mild symptoms across at least two systems, then describes severe and life-threatening grades using symptoms, treatment response, and end-organ dysfunction. This operational definition supports consistent classification, but interpretation still depends on how events were assessed and recorded. Those assessment procedures beyond the listed definition are not reported in the registry record.

The primary endpoint timeframe is within the 3-hour post-vaccine observation period. It therefore concerns reactions in that specified period, not all possible later events. The registry also lists secondary outcomes concerning severe reactions, mild or moderate reactions, anaphylaxis, and comparisons with reactions after the initial dose or placebo. The full outcome list contains repeated descriptions, and the record does not provide numerical results for those outcomes in the registry record for this analysis.

How does crossover affect interpretation?

In general, a crossover design gives participants more than one study condition at different times. Comparing conditions within participants can account for stable characteristics shared by a participant across periods. It also means that time, order, carryover, and the interval between conditions can matter. The registry describes active vaccine and placebo on different days, and the arm descriptions state a 24-hour washout. The record does not report the randomization sequence, whether order was balanced, or a statistical method for handling period or carryover effects. Those details cannot be assumed.

For this primary outcome, the record says that the analysis population was participants who received active vaccine during the crossover phase, with data collected and reported only from active vaccination. Thus, the endpoint is a proportion after active vaccine, not a reported within-person treatment contrast. Although the design includes placebo, a comparison is not established by the primary endpoint definition alone. The record lists zero analyses, and no adjusted or unadjusted inferential comparison is reported.

What can planning calculations establish?

General sample-size planning translates an assumed effect and design into a target amount of information. The planning assumptions described for illustrations use a two-sided significance level of 0.05, equal allocation, and a normal approximation, with the effect to detect expressed as a fraction of the observed effect. The assumptions are described as planning illustrations based on the observed effect, not as the trial's own design rationale. The registry record does not report a sample-size justification. Such planning calculations are not observed power and do not replace the trial's outcome estimate or confidence interval.

The description of planning assumptions alone does not establish that a particular effect was clinically important, that the study was designed to detect it, or that the resulting precision is adequate. Planning depends on assumptions, whereas an outcome analysis describes the data actually recorded. Since the primary numerical result is not reported in the registry record, the observed estimate and its uncertainty cannot be assessed from the primary outcome description.

Why are event categories not interchangeable?

Safety summaries commonly separate serious adverse events from other adverse events, while an efficacy or safety endpoint may use a specific clinical definition. General interpretation requires keeping the category, timeframe, population, and event definition aligned. Here, the primary outcome has a CoFAR-based definition and a 3-hour window. The adverse-event table reports serious and other event counts by group, but the record does not provide event-level details connecting those categories to the primary endpoint. Combining or substituting these counts would therefore change the question being answered.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Blinded Active Vaccine 2")
n      <- c(16)
events <- c(0)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Blinded Active Vaccine 2  16  0  0.0  0.0  20.6  0.0  19.4  0.1  19.5

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 13 outcome measures, including the primary endpoint and secondary outcomes concerning reaction severity, anaphylaxis, recurrence, and comparison with the initial vaccine reaction. Multiple outcomes create a general multiplicity issue: when many statistical questions are considered, the chance of at least one apparently notable result can increase if each is tested separately without accounting for the set of tests. Whether and how to address that issue depends on a prespecified analysis plan.

The registry record does not report a multiplicity adjustment, a hierarchy among secondary outcomes, or a statistical testing strategy. It also lists zero analyses. No p-values or outcome-specific inferential results are reported in the record, so there is no basis to assess how multiplicity was handled in practice. The existence of multiple registered outcomes does not by itself show that every outcome was tested or that any particular adjustment was used.

Missing-data handling is not reported in the registry record. The primary outcome population is described as participants who received the active vaccine in the crossover phase, but the record does not state how missing assessments, incomplete follow-up, or unavailable outcome classifications were handled. It also does not give a participant flow accounting for the primary endpoint. The enrollment figure and safety-table risk counts should not be used to infer a missing-data mechanism or explain the difference between populations.

The analysis population is described specifically for the primary endpoint, but other analysis populations, including any intention-to-treat or per-protocol populations, are not reported in the registry record. The record does not explain whether exclusions occurred or how participants who did not complete both crossover conditions were treated analytically. These omissions limit assessment of selection, missingness, and reproducibility. They do not establish that such issues affected the findings; they mean that their handling cannot be evaluated from the registry record.

12. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent non-serious adverse event terms

The safety figures list zero serious adverse events and zero deaths for Blinded Active Vaccine 2, Placebo, Open-Label Booster Vaccine, and Overall Study. Each displayed group lists 16 participants at risk for serious adverse events and 16 at risk for other adverse events. The other-adverse-event counts listed are 3 for Blinded Active Vaccine 2, zero for Placebo, 16 for Open-Label Booster Vaccine, and 16 for Overall Study. These are the categories and counts recorded in the registry.

These entries should be read as safety-table counts, not as the primary endpoint result. The registry does not report whether the other adverse events met the CoFAR definition of systemic allergic reaction, what those events were, or how their severity was distributed. The primary endpoint is defined for reactions after active vaccine during the 3-hour observation period; the safety counts are not explicitly tied to that definition and timeframe.

The table does not describe event-level attribution, timing, or overlap across groups. The open-label booster group is also a distinct study phase described as offered to participants meeting a reaction-severity condition after blinded active vaccination. The listed group counts should not be added or treated as independent participant totals. The registry record reports no serious adverse-event terms. These summaries are descriptive and, without event details or a comparative analysis, do not establish a causal interpretation.

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

The record establishes that investigators registered a randomized, placebo-controlled crossover study focused on a second mRNA COVID-19 vaccine dose among people with a systemic allergic reaction after their first dose. It defines a primary endpoint based on recurrent systemic allergic reaction following active vaccination, uses a CoFAR-based event definition, and specifies a 3-hour observation period and 95% confidence interval. These details clarify the intended question and how an event was to be classified.

The registry record does not report the numerical primary result. It therefore cannot establish the observed proportion, its precision, or how the event frequency compares with placebo or any other reference. No numerical comparison or primary outcome analysis is reported. The safety counts are a separate descriptive summary and cannot supply the missing primary endpoint estimate.

In general, a result from a study population is an estimate for the conditions and participants observed; it is not automatically a precise estimate for every population or setting. Interpretation would also depend on complete outcome reporting, the analysis population, and the handling of missing observations. The registry record does not provide enough statistical detail to evaluate these elements fully. The reported enrollment and safety group counts should not be mistaken for a complete account of endpoint ascertainment.

The study's design includes crossover exposure to active vaccine and placebo, but the primary question is framed as an active-vaccination proportion. A design that contains two conditions does not itself constitute evidence of a difference between them. Such a conclusion would require a reported comparative analysis suited to the design, and none is reported in the registry record.

Accordingly, the defensible interpretation is about the scope of the study and the limits of its registry reporting. The record documents the target population, intervention sequence, endpoint definition, and selected safety counts. It does not report the primary estimate or comparative findings needed to quantify the study's answer. This absence should not be interpreted as evidence for or against an effect; it leaves the primary numerical question unresolved in the registry record.

14. Why This Trial Matters Statistically

This trial is statistically notable because it combines a crossover design with a narrowly defined post-vaccination event endpoint. The crossover structure makes it important to distinguish a within-participant comparison from a simple proportion after active vaccination. The registry's primary outcome is the latter: it concerns participants receiving active vaccine and does not itself specify an active-versus-placebo contrast. That distinction helps prevent the design label from being mistaken for the estimand actually reported.

The endpoint also illustrates why event definitions and observation windows matter. The CoFAR grading criteria and tryptase rule determine which reactions count, and the 3-hour period limits the timeframe under consideration. A count from another safety category, or from a different timeframe, would answer a different question. Clear definitions improve interpretability, while missing outcome results prevent assessment of the estimated frequency and its uncertainty.

Another statistical issue is the separation between a registered study design and an analysis report. The record describes randomization, masking, crossover, and a confidence interval, but reports zero analyses and no numerical primary result. Without the estimate and interval, readers cannot assess precision. Without the analysis details, they also cannot examine how the crossover structure, missing data, or multiple outcomes were handled.

The safety summary further shows why descriptive counts require context. Serious and other adverse-event categories are not automatically interchangeable with the primary endpoint. Group counts can describe what was recorded in those categories, but they do not alone provide a comparative risk estimate or establish event causation. Statistical interpretation depends on matching the count to the population, definition, and timeframe of the question.

The trial matters as a case in reading registry evidence carefully: design features, outcome definitions, and posted results each contribute different information. Here, the design and endpoint are described, but the primary numerical result is not reported in the registry record. That reporting gap places a firm limit on statistical conclusions.

15. Limitations

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

A key trial-specific limitation for interpretation is that the registry record does not report the primary endpoint estimate, event count, or 95% confidence interval. It also lists no analyses. Thus, the primary question cannot be quantified from the registry record, and precision cannot be assessed. The separate adverse-event counts do not resolve this limitation because their event definitions are not shown to match the primary endpoint.

The registry does not describe the crossover sequence in sufficient detail to evaluate order or period effects. It gives a 24-hour washout period in the arm descriptions, but does not report whether order was balanced, how carryover was assessed, or what statistical method was used for the crossover. The primary endpoint is limited to active vaccination, further restricting what can be inferred about an active-versus-placebo comparison.

Missing-data handling, participant flow for the primary outcome, and the treatment of incomplete crossover participation are not reported in the registry record. Neither are the full statistical analysis plan, multiplicity strategy, or sample-size justification. These gaps limit evaluation of analytic choices and reproducibility, but do not establish that any particular bias occurred.

The registry lists a bivalent vaccine intervention, while the primary endpoint description and arm details do not clarify its role in the primary analysis. The safety record also provides no serious adverse-event terms and no event-level descriptions for other adverse events. As a result, the safety categories cannot be linked confidently to the primary CoFAR-defined outcome. Finally, this registry record alone does not supply outcome detail for broader interpretation beyond the specified study population and observation period.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: The Safety of Administering a Second Dose of a COVID-19 mRNA Vaccine in Individuals Who Experienced a Systemic Allergic Reaction to an Initial Dose — NCT04977479.
• Registry record last updated 2023-10-16; status last verified 2023-01.
• Results first posted 2023-10-16.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

NCT04977479 is a completed Phase 2 randomized, quadruple-masked, placebo-controlled crossover study sponsored by the National Institute of Allergy and Infectious Diseases. It enrolled 20 people who had experienced a systemic allergic reaction after an initial COVID-19 vaccine dose. Participants received active mRNA vaccine and placebo on different days; the primary endpoint concerned recurrent systemic allergic reaction after active vaccination within the 3-hour observation period.

The registry defines the endpoint using CoFAR grading and a tryptase condition and specifies a 95% confidence interval. It describes a later open-label booster for participants who did not have a CoFAR Grade 3 or higher reaction after the blinded active dose. The safety summary lists zero serious adverse events and zero deaths in each displayed group, alongside other adverse-event counts.

The registry record does not report a numerical primary outcome result, confidence interval, comparative analysis, missing-data approach, or detailed statistical plan. Those omissions limit quantitative interpretation of the trial's primary question.