Baked Egg or Egg Oral Immunotherapy for Children With Egg Allergy: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Baked Egg or Egg Oral Immunotherapy for Children With Egg Allergy (NCT01846208), a randomized phase 2 study. The primary outcome is Sustained Unresponsiveness to Egg Consumption at 2 Years..

2. Clinical Question

The trial’s stated purpose was to compare baked egg with egg oral immunotherapy for inducing sustained unresponsiveness to egg exposure in children. The primary endpoint focuses not only on consuming egg during treatment, but also on consuming it after a period without egg. That makes it a more demanding outcome than a measure of response while participants are receiving therapy.

The registry does not report enough detail to characterize every aspect of the intended clinical question. The specific eligibility criteria beyond the baseline challenge groups, study locations, and participant characteristics are not reported in the registry record. The result therefore needs to be read in the context of the defined study groups and endpoint rather than generalized to all children with egg allergy.

3. Trial Design

This was a phase 2, randomized, open-label, parallel-group study with an enrollment of 92. The registry lists the status as completed, a start date of 2013-07, and a primary completion date of 2017-09-28. Hugh A Sampson, MD, is listed as the sponsor. Additional details about the study setting and participating centers are not reported in the registry record.

Children who passed a baked egg oral food challenge at baseline were randomized to receive either egg oral immunotherapy or baked egg. The egg oral immunotherapy group received egg white solid. The baked egg group received home-baked goods and “safe” commercial products. Each group could undergo up to four oral food challenges as directed by the protocol. The registry does not report the randomization ratio or allocation concealment procedures.

A separate group of children who failed the baseline baked egg challenge was assigned to egg oral immunotherapy. Because the record describes this group as assigned rather than randomized, its outcomes should not be treated as another randomized comparison. Differences between this group and the randomized group could reflect their different baseline challenge results, among other factors not detailed in the record.

The primary endpoint was assessed at 2 years and included all treated participants in the analysis, according to the registry description. The record lists five outcome measures and eight analyses, but the full protocol and statistical analysis plan are not reported in the registry record. Specific details such as the intended analysis populations beyond the primary description, handling of missing outcomes, and planned interim analyses are not reported in the registry record.

The design permits a direct comparison between the two randomized groups for the primary question, while the assigned group provides a distinct, nonrandomized reference. The open-label design means masking was absent. Whether knowledge of group assignment affected participant behavior or outcome assessment cannot be determined from the registry information.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All treated participants were included in the analysis.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Population: Only treated participants are included in analysis

6. Primary Endpoint

The registry lists the primary outcome as Sustained Unresponsiveness to Egg Consumption at 2 Years., measured over 2 Years.

The primary endpoint is the count of participants with sustained unresponsiveness to egg consumption at 2 years. A participant met the registry definition by successfully consuming 7444 mg egg white protein in a desensitization oral food challenge and then, after an 8-10 week egg-free interval, successfully consuming 7444 mg egg white protein in another challenge after up to 2 years of therapy.

This endpoint is used to distinguish a response demonstrated during treatment from a response that remains after a period without egg. In general, a challenge-based endpoint provides a defined way to assess whether participants can consume a specified amount under the study’s test conditions. The egg-free interval is part of this particular endpoint definition; the registry does not establish how the result would translate to consumption outside the study setting.

The primary outcome is reported as participant counts in each group. The randomized baked egg group had 3 participants meeting the endpoint among 27, while the randomized egg oral immunotherapy group had 10 among 23. The assigned egg oral immunotherapy group had 7 among 39. These counts describe the posted results, but group counts alone do not capture the uncertainty in the comparison. The estimated risk differences and their confidence intervals provide additional information about the size and precision of the comparisons.

The endpoint’s timeframe and definition are specific to this trial. Other measures, such as unrestricted consumption of unbaked egg, address different outcomes and should not be treated as interchangeable with sustained unresponsiveness as defined for the primary analysis.

7. Results Posted in the Registry

• Egg OIT Randomized: 43.5% (10/23) versus 11.1% (3/27) on control; risk difference 32.4 percentage points (95% CI 8.9 to 55.8), risk ratio 3.91 (1.22–12.53); unadjusted two-proportion z-test p = 0.009.
• Egg OIT Assigned: 17.9% (7/39) versus 11.1% (3/27) on control; risk difference 6.8 percentage points (95% CI -10.1 to 23.7), risk ratio 1.62 (0.46–5.70); unadjusted two-proportion z-test p = 0.446.

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 Unrestricted Consumption of Unbaked Egg

• 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:
  • Egg OIT Randomized: 73.7% (14/19) versus 23.5% (4/17) on control; risk difference 50.2 percentage points (95% CI 21.9 to 78.4), risk ratio 3.13 (1.28–7.69); unadjusted two-proportion z-test p = 0.003.
  • Egg OIT Assigned: 40.0% (12/30) versus 23.5% (4/17) on control; risk difference 16.5 percentage points (95% CI -10.2 to 43.2), risk ratio 1.70 (0.65–4.45); unadjusted two-proportion z-test p = 0.252.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Egg OIT Randomized than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Egg OIT Assigned and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes cover several aspects of egg response and safety: desensitization to at least 4.444 Grams Egg White Solid, serious adverse events, and unrestricted consumption of unbaked egg. The record lists five outcome measures overall and reports three secondary outcomes. The secondary analyses therefore offer additional perspectives on the study, but they do not all measure the same underlying response as the primary sustained-unresponsiveness endpoint.

For unrestricted consumption of unbaked egg at 3 years after randomization, the posted counts are higher in the randomized egg oral immunotherapy group than in the baked egg group. The registry’s sponsor-reported comparison is statistically significant, and the unadjusted recalculation points in the same direction. In the comparison of randomized and assigned egg oral immunotherapy participants, the posted counts also indicate a higher observed frequency in the randomized group; the sponsor reports a significant difference, while the simple recalculation does not reproduce the sponsor estimate. The registry flags a discrepancy, and the sponsor’s analysis takes precedence.

Overall, the posted secondary pattern is compatible with the primary randomized contrast in that some egg-consumption outcomes favor the randomized egg oral immunotherapy group over the baked egg group. That pattern is supportive context, not an independent confirmation of the primary result: secondary measures differ in definition and timing, and some analyses compare groups that were not randomized against one another. The registry does not report enough information to establish whether these outcomes were prespecified for confirmatory testing.

Unadjusted secondary results carry less weight because multiple outcomes and comparisons create more opportunities for apparently notable findings. A multiplicity adjustment or a prespecified hierarchy can help address that issue, but whether either was used is not reported in the registry record. The secondary p-values should therefore be interpreted as individual, unadjusted results, not as evidence that the overall set of secondary findings has controlled error rates. The missing two listed outcome measures and their results are not reported in the registry record.

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

• Desensitization to >= 4.444 Grams Egg White Solid.
• Incidence of All Serious Adverse Events
• Number of Participants With Unrestricted Consumption of Unbaked Egg

9. Statistical Interpretation

10. Statistical Methods Explained

What does a risk difference describe?

A risk difference compares the proportion of participants who experienced an outcome in one group with the proportion in another. It expresses an absolute contrast, rather than a relative comparison. For this trial’s primary endpoint, the event is meeting the registry definition of sustained unresponsiveness. A positive or negative sign depends on which group is subtracted from which, so the comparison’s stated group order and the registry’s direction convention matter when reading the estimate.

The sponsor reported a risk difference of 32.4 for the randomized egg oral immunotherapy versus baked egg comparison. The registry also reports group counts, which indicate that the endpoint was observed in 10 of 23 randomized egg oral immunotherapy participants and 3 of 27 baked egg participants. The separate comparison involving assigned egg oral immunotherapy uses a different participant group and should not be interpreted as if it were randomized.

Why use an exact test?

The registry identifies Barnard’s Exact Test for its reported comparisons. In general, exact tests evaluate the observed group counts using a probability model for the data, rather than relying on a large-sample approximation for the test itself. Such methods can be useful when outcomes are binary and group sizes are limited. The test does not remove uncertainty or correct design limitations; it supplies a method for evaluating a particular contrast under its statistical assumptions.

The listed outcome is binary at the participant level: each participant either met the endpoint definition or did not. The group summaries are counts, not continuous measurements. The registry does not report a dispersion measure for this outcome, and a mean and standard deviation are not the natural summary for a binary event count.

What does the confidence interval add?

A confidence interval describes the uncertainty around an estimated effect using a method with a stated coverage level. In repeated use under the method’s assumptions, intervals constructed in this way would cover the target effect at the stated rate. For the primary sponsor-reported contrast, the interval is 8.9 to 55.8 at the 95% level. Its width reflects that the data do not pin down the effect to a narrow range. An interval should be read alongside the point estimate, group counts, and study design rather than as a guarantee that the true effect lies within its particular endpoints.

The interval also makes clear that an estimate is not a precise single value. The evidence remains compatible with a range of effect sizes. A confidence interval is not a probability distribution for the true effect, and its interpretation relies on the analysis method and assumptions. The registry does not describe additional modeling or covariate adjustment for the reported primary analysis.

How should p-values be understood?

For the primary comparison, the sponsor reports p=0.009. In general, a p-value is calculated under a null hypothesis and a specified statistical model. It describes how unusual data at least as discordant with that null model would be, if the null model and test assumptions held. It does not give the probability that the null hypothesis is true, the probability that the result occurred “by chance,” or the size of an effect.

The registry’s reported p-values are unadjusted. That means the values are presented without a reported correction for testing multiple outcomes or contrasts. Whether an adjustment was part of the original analysis plan is not reported in the registry record. Consequently, individual p-values should not be read as establishing that the full collection of trial findings has a controlled overall false-positive rate.

Why distinguish randomized from assigned comparisons?

Random allocation is intended to make the randomized groups comparable, on average, with respect to both measured and unmeasured characteristics at the start of a study. That supports interpreting differences between the two randomized groups as a treatment contrast, subject to issues such as adherence, missing outcomes, and the analysis method. The record labels the trial randomized and the primary comparison uses the two randomized groups.

The assigned egg oral immunotherapy group consisted of participants who failed the baseline baked egg challenge, unlike the randomized participants who passed it. Treatment assignment in that group was not described as randomized. A difference between that group and a randomized group can therefore reflect baseline eligibility differences as well as treatment or other factors. Its comparison has a different evidential role from the randomized contrast.

What is the role of recalculation?

Recalculated estimates based on posted summary counts are simple, unadjusted checks of the reported group summaries. They may help readers assess whether the reported direction and scale appear consistent with the counts, but they are not substitutes for the sponsor’s analysis. The registry flags a discrepancy, and the sponsor’s own analysis takes precedence. The record does not explain the discrepancy or provide enough information to resolve it independently.

In general, differences can arise when a recalculation and a sponsor analysis use different definitions, analysis populations, conventions, or methods. Those are possible explanations in statistical work generally, not established explanations for this trial. Which, if any, applies is not reported in the registry record. The defensible reading is to keep the sponsor-reported result distinct from the unadjusted re-calculation, rather than silently treating the two as identical.

11. Confidence Intervals and Precision

The primary sponsor-reported risk difference is 32.4, with a 95% confidence interval from 8.9 to 55.8. The interval is the main description of statistical precision: it shows that the estimate is compatible with a range of effect sizes, rather than identifying a single exact value. Its endpoints are separated enough to signal meaningful uncertainty about the size of the contrast, even though the reported interval excludes zero.

In general, a confidence interval is constructed by a statistical procedure that would cover the target quantity at its stated rate across repeated samples if the assumptions hold. It is not a statement that there is a 95% probability that this particular interval contains the true effect. The interpretation also depends on the analysis method and study design. The registry names Barnard’s Exact Test for the sponsor-reported comparison, but it does not provide further detail about interval construction beyond the reported 95% level.

The group counts help explain why precision is limited: the primary outcome was observed in 3 of 27 participants in the baked egg group and 10 of 23 in randomized egg oral immunotherapy. These are modest group sizes, so estimates based on the observed proportions remain subject to sampling variation. The registry’s overall enrollment was 92, but the two randomized groups in this endpoint analysis account for only part of that enrollment; an additional assigned group is reported separately.

The sponsor-reported interval for the comparison involving assigned participants is 2.0 to 49.1, with a 95% confidence level. That interval also reflects uncertainty, but its interpretation is further limited by the nonrandomized nature of the comparison. A confidence interval quantifies statistical uncertainty under the analysis, not uncertainty about whether the compared groups were exchangeable or whether results generalize to other populations.

Because the record flags a discrepancy between sponsor results and simple recalculations, the sponsor’s estimates should be treated as primary. The registry does not state why all recalculated and sponsor-reported figures differ. The intervals should therefore be read as the intervals attached to the sponsor’s reported analyses, not as a resolution of the discrepancy.

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 these binary outcomes, the figures are participants per arm. They are not total event counts. The planning assumptions are a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect to detect set to the stated fraction of the observed effect. The trial’s own original sample-size assumptions are not reported in the registry record.

The 80% power column gives the number of participants per arm that the planning calculation estimates would be needed to have an 80% chance of detecting the selected effect under those assumptions. The 90% power column uses a higher chance of detection, and therefore generally calls for more participants per arm. Power is a property of a planned design under specified assumptions; it is not a guarantee that a particular study will reach a result or a measure of the probability that its finding is true.

For the randomized primary comparison, the illustration based on 100% of the observed effect lists 29 participants per arm for 80% power and 38 for 90% power. Planning for 75% of the observed effect lists 47 and 62 participants per arm, respectively; planning for 50% lists 92 and 123. These examples show the general pattern: smaller effects are harder to detect, so the calculation requires more participants.

The illustrations do not establish what the trial was designed to detect, and they should not be interpreted as observed power. They are retrospective planning scenarios based on observed effects, which can themselves be uncertain. The assigned-group comparisons use different planning figures and are not randomized treatment comparisons. As with any such illustration, changing assumptions about the target effect, allocation, significance level, or calculation method would change the resulting participant counts.

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

Worked calculation for this trial

1. Control proportion p₀ = 0.111; treatment proportion p₁ = 0.435; average p̄ = 0.273.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6300 = 1.2348.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5870 = 0.4940.
4. n = (1.2348 + 0.4940)² / (0.3237)² = 28.5, rounded up to 29 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 <- 0.111111111111111                          # control proportion
p1 <- 0.434782608695652                          # 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(1846208)
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 = 29 (80% power), 38 (90% power)
 75% of observed effect: n per arm = 47 (80% power), 62 (90% power)
 50% of observed effect: n per arm = 92 (80% power), 123 (90% power)
power.prop.test: 28.5 
Simulated power at n = 29 per arm: 0.819 (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.

14. Analyses Reported by the Sponsor

The registry reports sponsor analyses using Barnard’s Exact Test. For the randomized primary comparison, the sponsor reports a risk difference of 32.4, a 95% confidence interval from 8.9 to 55.8, and p=0.009. The unadjusted recalculation from the posted group summaries is directionally consistent with the sponsor’s result, but the record flags a discrepancy. The sponsor’s own analysis takes precedence.

The discrepancy is more consequential for the comparison between randomized egg oral immunotherapy and assigned egg oral immunotherapy. The sponsor reports a risk difference of 25.5, with a 95% confidence interval from 2.0 to 49.1 and p=0.031. A simple unadjusted recalculation from the posted counts does not reproduce that estimate. The registry does not explain the difference, so there is no basis in the record to attribute it to a particular analysis choice or data handling decision.

These comparisons should remain distinct. The primary contrast is between two randomized groups; the comparison involving assigned participants includes a group that entered after failing the baseline challenge. The sponsor’s reported results describe the registry’s formal analyses, while the recalculations are checks based only on posted summary counts. Neither should be silently substituted for the other, and the discrepancy does not by itself reveal which underlying detail differs.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists five outcome measures, reports three secondary outcomes, and identifies eight analyses. The available record does not describe a hierarchy of primary and secondary tests, a multiplicity correction, or a prespecified approach for interpreting the full family of comparisons. Because several outcomes and contrasts are reported, unadjusted p-values should not be treated as if they account for all testing. Whether a formal multiplicity strategy was included in the protocol is not reported in the registry record.

The primary analysis description states that all treated participants were included. It does not fully explain how participants without an evaluable endpoint were handled, whether all treated participants had outcome data, or whether any imputation or sensitivity analyses were conducted. These missing-data methods are not reported in the registry record. The distinction matters because an analysis based on observed outcomes can be affected if outcome availability differs between groups or is related to participants’ underlying outcomes.

The record identifies randomized groups and a separate group assigned to treatment after failing the baseline challenge. It does not report further details about the analysis population for each sponsor comparison, such as whether the randomized analysis followed an intention-to-treat principle or how treatment discontinuations were handled. The primary endpoint says all treated participants were included, but that description alone does not resolve every population and missing-data question.

Interim analyses, stopping rules, and any associated statistical adjustments are not reported in the registry record. The full statistical analysis plan, original sample-size justification, and details of any protocol amendments relevant to analysis are also not reported in the registry record. These omissions prevent a complete assessment of whether the posted comparisons follow a prespecified confirmatory plan. They do not invalidate the posted results, but they limit how confidently readers can distinguish planned testing from additional analyses.

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: pneumonia (2.6%).

For serious adverse events, the registry reports 0 affected participants among 27 at risk in the baked egg randomized group, 0 among 23 in the randomized egg oral immunotherapy group, and 1 among 39 in the assigned egg oral immunotherapy group. The record reports no deaths in any of these groups. These are counts over the registry’s stated groups; the record does not provide event timing or a fuller account of clinical context.

The registry lists Pneumonia at 2.6% among the top serious adverse event terms. A term listing should be read separately from the group-level counts: the record does not provide enough detail in this summary to reconcile the term listing with the participant-level serious-event figures. The source of any apparent difference is not reported in the registry record.

Other adverse events are reported as affecting 27 of 27, 23 of 23, and 39 of 39 participants in the respective groups. The registry does not provide further detail about severity, duration, or how these other events were defined. Small group sizes and the limited information on event details restrict the conclusions that can be drawn from these counts. They describe what the registry reports and do not establish comparative safety across groups.

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

The registry’s primary result describes a difference in the observed frequency of sustained unresponsiveness between the two randomized groups. The sponsor reports a risk difference of 32.4, with a 95% confidence interval from 8.9 to 55.8 and p=0.009. The posted counts are 10 of 23 for randomized egg oral immunotherapy and 3 of 27 for baked egg. These results indicate that the observed endpoint frequency was higher in the randomized egg oral immunotherapy group in this study. They do not establish the size of the effect with precision beyond the interval, nor do they answer every question about outcomes outside the trial conditions.

The endpoint involved controlled oral food challenges and a defined egg-free interval. It is not the same as a broad measure of routine consumption or long-term outcome. The secondary measure of unrestricted unbaked egg consumption addresses a related but distinct question, at a different timeframe. Findings across these endpoints should be considered in light of their differing definitions and the registry’s reporting limitations.

The randomized comparison is more informative about the relative outcomes of the two treatment groups than the comparison that includes participants assigned after failing the baseline challenge. That assigned group followed a different entry pathway, so its results should not be interpreted as a randomized treatment contrast. The sponsor-reported and recalculated figures also do not fully agree for every comparison, and the reason is not reported in the registry record.

The statistical findings do not determine whether a particular treatment is appropriate for an individual, and they do not provide medical advice. They also do not establish results for children who would not meet the study’s entry conditions, for different treatment settings, or for outcomes not measured in the trial. The record does not report enough participant characteristics to assess how closely the enrolled groups represent other populations.

Finally, the primary p-value is unadjusted, and the record does not describe a multiplicity strategy or all missing-data procedures. Those limitations affect how the collection of results should be interpreted. The most defensible conclusion is limited to the posted randomized comparison, its defined endpoint, and the uncertainty represented by the sponsor’s confidence interval.

18. Why This Trial Matters Statistically

This trial illustrates why a clinical result should be read through its design, endpoint, and analysis together. The primary outcome is not simply egg consumption during therapy: it requires success in a challenge, an egg-free interval, and another successful challenge. That definition makes clear what the endpoint captures and why it cannot be treated as interchangeable with desensitization while treatment is ongoing or with unrestricted consumption of unbaked egg.

The study also distinguishes a randomized comparison from an assigned group formed by baseline challenge status. Keeping those comparisons separate is essential: randomized groups support a more direct treatment contrast, while a comparison involving the assigned group can reflect differences in who entered each group. The record’s structure makes this distinction statistically important even though both groups received egg oral immunotherapy.

The reported confidence interval adds information that a p-value alone cannot provide. It places the sponsor’s primary estimate in a range of compatible values and makes the uncertainty in the effect size visible. At the same time, the registry’s discrepancy flag shows why numerical checks based on posted counts should not be confused with the sponsor’s formal analysis. Transparent reporting of the analysis population and method would help explain differences that cannot be resolved from the registry record.

Finally, the trial shows why secondary outcomes and planning illustrations need careful interpretation. Related outcomes can provide context, but unadjusted secondary comparisons do not automatically confirm the primary result. Retrospective sample-size scenarios can illustrate how planning changes with a target effect, but they do not reveal the trial’s original design rationale or the chance that its observed finding is true.

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.

The registry record leaves several trial-specific questions unresolved. It reports the randomized and assigned groups but does not give detailed participant characteristics, study locations, the randomization ratio, or allocation concealment procedures. The baseline baked egg challenge separates the randomized participants from the assigned group, limiting direct comparison across those groups. The record does not provide enough information to judge how representative the enrolled children were of other children with egg allergy.

The open-label design means participants and study personnel were not masked, according to the record. The registry does not explain whether outcome assessors were masked or how oral food challenges were administered and adjudicated. Because the primary endpoint depends on challenge performance, further procedural detail would help readers understand how consistently the outcome was assessed.

The sponsor’s analysis and simple unadjusted recalculations do not fully agree for every reported contrast. The discrepancy is flagged, but its cause is not reported in the registry record. The formal statistical analysis plan, handling of missing outcomes, analysis populations for each comparison, and any sensitivity analyses are not reported in the registry record. The record also does not explain the multiplicity approach, if any, for the several outcomes and comparisons.

The primary outcome uses a specific challenge-based definition and timeframe. The results therefore do not directly describe every form of egg consumption or longer-term outcomes beyond those recorded. The secondary outcome of unrestricted unbaked egg consumption is a different measure, and some secondary comparisons involve the nonrandomized assigned group. These limits should remain in view when interpreting the overall pattern of results.

Finally, the posted group sizes are limited, and the confidence intervals show uncertainty around the estimates. The registry does not provide the original sample-size justification, so the planning illustrations based on observed effects cannot establish what effect the trial was originally designed to detect.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Baked Egg or Egg Oral Immunotherapy for Children With Egg Allergy — NCT01846208.
• Registry record last updated 2019-07-26; status last verified 2019-07.
• Results first posted 2018-10-30.
• 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

NCT01846208 is a completed phase 2, randomized, open-label, parallel-group study titled “Baked Egg or Egg Oral Immunotherapy for Children With Egg Allergy.” Enrollment was 92. Children who passed a baked egg oral food challenge at baseline were randomized to baked egg or egg oral immunotherapy; children who failed were assigned to egg oral immunotherapy.

The primary outcome was sustained unresponsiveness to egg consumption at 2 years, defined by successful consumption of 7444 mg egg white protein during a challenge and again after an 8-10 week egg-free interval. For the randomized comparison, the sponsor reported a risk difference of 32.4, a 95% confidence interval from 8.9 to 55.8, and p=0.009. The registry flags a discrepancy between sponsor analyses and simple recalculations; the sponsor’s own analysis takes precedence.

The record also reports secondary outcomes and adverse-event counts, but does not report several details needed for a full evaluation, including the original sample-size justification, full missing-data procedures, and multiplicity strategy. The assigned-group comparisons are not randomized contrasts.