Evaluation of Skin Testing Reagents for Penicillin Allergy: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Evaluation of Skin Testing Reagents for Penicillin Allergy (NCT01818336), a na phase 3 study. The primary outcome is Negative Predictive Value.

2. Clinical Question

The study asked whether a negative result on the Penicillin Allergy Skin Test Kit identified participants with a low risk of an IgE-dependent reaction during the 72 hours after an oral amoxicillin challenge. The registry record describes people with a history of an IgE-dependent reaction to a penicillin product as the population of interest. For the primary endpoint, participants also had to have valid skin testing, negative intradermal results with all three kit reagents, and receive the oral challenge.

Negative predictive value is the proportion of people with a negative test who do not experience the specified outcome during the specified follow-up. It is not the same as a comparison between test strategies, because the registry record does not describe a concurrent group receiving a different test or no test. The clinical question is therefore about the observed outcome after a negative test and challenge in the eligible study population, not about a randomized treatment effect.

3. Trial Design

The registry record describes a completed Phase 3, single-group study of the Penicillin Skin Test Kit. Enrollment is recorded as 481. The study began in 2012-12, with primary completion in 2013-06. The record lists no allocation and no masking. It does not describe a randomized comparison or a concurrent control arm.

Participants underwent skin testing with the kit. According to the arm description, subjects with negative intradermal tests were to receive a single oral amoxicillin challenge dose and be followed for 72 hours for IgE-dependent reactions. The primary endpoint description further specifies that the analysis population included participants with valid skin testing, negative intradermal tests with all three kit reagents, and receipt of the challenge. This means that enrollment in the study and eligibility for the primary endpoint are not necessarily the same thing.

The primary outcome was negative predictive value, expressed as the percentage of eligible participants who did not experience an IgE-dependent hypersensitivity reaction within 72 hours of the oral challenge. The record also contains safety entries for all-study-emergent adverse events, events related to skin testing, and events related to the oral amoxicillin challenge. These are safety summaries, not a randomized comparison of event risks.

The registry record does not report a detailed statistical analysis plan, a sample-size justification, interim analyses, or a masking rationale. It reports no outcome-analysis records. Consequently, the design features that can be described are those explicitly recorded: one study group, no allocation, no masking, a defined challenge follow-up period, and one listed primary outcome measure. The absence of a control group limits interpretation to outcomes observed in this study population rather than comparative efficacy or test-performance contrasts between assigned strategies.

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: 455 History pos subj enrolled + 2 control subj @ 13 centers. All 481 subj were skin tested(ST) and included in safety population. 64 subj (all history pos) had pos or discordant skin tests and were excluded from ITT population of 391(455 - 64). The subset of 59 ST pos subjects were invited back at 4 weeks for repeat ST; 47 ST pos subj returned

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The Intent-to-Treat Population; all subjects with valid skin testing performed (ie, administered all components of the penicillin skin test kit and the histamine/control results were valid), who had negative intradermal skin tests with all 3 Kit reagents, and who received the oral amoxicillin challenge

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: Total subjects enrolled in the study

6. Primary Endpoint

The registry lists the primary outcome as Negative Predictive Value, measured over 72 hours.

The primary endpoint is negative predictive value (NPV): the percentage of participants with negative skin tests who did not experience an IgE-dependent hypersensitivity reaction within 72 hours after oral amoxicillin challenge. The specified population consisted of history-positive participants with valid testing, negative intradermal results for all three kit reagents, and receipt of the challenge. The endpoint therefore concerns a conditional proportion among participants who tested negative and proceeded to challenge.

NPV is used to describe how often a negative test result is followed by absence of the outcome that the test is intended to help identify, within a defined population and follow-up period. Its interpretation depends on the population studied, how negative results are defined, the outcome definition, and the follow-up window. It does not by itself show what would have happened under a different testing strategy, nor does it directly quantify the probability of a reaction among people with a positive test.

The record specifies a 95% confidence interval as the measure of uncertainty for the endpoint, but does not report the endpoint estimate or interval. The number of participants contributing to the primary endpoint and the number experiencing the endpoint event are not reported in the registry record. Thus the endpoint is defined, but its numerical result and precision cannot be assessed from the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary measure is a proportion. In general, a proportion is calculated by dividing the number of participants with a specified outcome by the number of participants in the relevant analysis population. For this endpoint, the registry definition focuses on participants who had valid skin testing, negative intradermal results with all three kit reagents, and an oral amoxicillin challenge. The numerator concerns those participants without an IgE-dependent hypersensitivity reaction during the stated 72-hour period. The registry record does not report the numerator or denominator, so the estimate cannot be independently reconstructed.

What does negative predictive value mean?

In general, negative predictive value describes the frequency with which a negative test result is followed by absence of the target outcome in the population being assessed. It is conditional on having a negative test. It should not be interpreted as a universal property of a test, because the composition of the tested population and the definition and ascertainment of the outcome can affect the observed value. In this study, the target outcome was an IgE-dependent hypersensitivity reaction after oral amoxicillin challenge, and the stated observation period was 72 hours.

The endpoint’s conditioning is important. It does not include everyone enrolled simply because they entered the study. Participants had to have valid test results, negative intradermal results across all three kit reagents, and receive the oral challenge. A participant who did not meet those criteria would not be part of the endpoint population as described. The count meeting these criteria is not reported in the registry record.

What would a confidence interval add?

In general, a confidence interval gives an estimate together with an indication of statistical uncertainty under an assumed sampling model. It is not a guarantee about an individual participant’s outcome, and it does not remove limitations in study design or measurement. The registry record names a 95% confidence interval for the NPV endpoint, but does not report its limits or the calculation method. Without the estimate and interval, neither the observed proportion nor its statistical precision can be evaluated.

For a binary outcome, the choice of interval method can matter, particularly when event counts are small or the estimate lies near a boundary. The registry record does not identify an interval procedure, so no method should be attributed to the study. A general explanation of possible procedures cannot establish which one was used.

Why does the single-group design matter?

In general, a single-group study can describe outcomes observed after a defined test and subsequent procedure. It cannot, by itself, estimate a contrast against an alternative strategy assigned to a concurrent control group. The record lists no allocation, no masking, and a single-group model. Accordingly, the NPV is a within-population descriptive measure, not a randomized effect comparing the kit with another approach.

That distinction also applies to safety. Event counts after skin testing or oral challenge describe events recorded in the study groups at risk. Without a concurrent comparison group, they do not isolate the contribution of the test or challenge from other explanations or establish a difference in risk between strategies. The registry record reports counts, but does not provide detailed event descriptions or causal adjudication in the information available.

What can be said about planning calculations?

In general, sample-size planning asks how many participants a future study might need under specified assumptions about the effect to detect, allocation, significance level, and statistical model. A planning illustration is not an observed result and is not observed power. The registry record’s planning description refers to a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the effect expressed as a fraction of the observed effect. Those are planning assumptions, not evidence that this single-group study randomized participants between two arms. The registry record does not give a sample-size justification for the trial itself.

Because the registry record does not report the primary estimate, the observed effect on which such an illustration would depend cannot be verified from the record. No planning value should be treated as the number needed for this trial or as a measure of its actual statistical power. A prospective design justification and a retrospective illustration answer different questions.

What do the available analysis records establish?

The record lists one outcome measure and no analysis records. This does not supply the missing primary result or establish which statistical procedures were used. It also does not establish whether analyses were performed outside the registry record. The population definition is available in part, but the number analyzed, the handling of missing or invalid tests, and the treatment of participants who did not receive the challenge are not reported in the registry record.

These distinctions matter because the validity and interpretation of a proportion depend on a clear denominator, consistent outcome ascertainment, and transparent handling of participants without complete information. General statistical principles can explain why those details matter, but they cannot fill in the absent analysis decisions or reconstruct a result that the registry record does not report.

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("Intent-to-Treat Population")
n      <- c(391)
events <- c(383)
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  Intent-to-Treat Population  391  383  98.0  96.0  99.1  96.0  99.0  96.0  98.9

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists one primary outcome measure, negative predictive value, and no analysis records. It does not report a multiplicity plan, such as how inference would be managed across multiple primary comparisons or outcomes. With one listed primary outcome, the record still does not establish whether additional analyses were planned or performed outside the registry record. The analysis plan and any multiplicity adjustment are not reported in the registry record.

The primary endpoint population is defined by valid skin testing, negative intradermal results with all three kit reagents, and receipt of the oral amoxicillin challenge. However, the number meeting those criteria is not reported, nor are the primary endpoint event count and denominator. The registry record does not describe how missing skin-test components, invalid control results, absent challenge data, or incomplete follow-up were handled. It also does not report whether any participants were excluded from the endpoint analysis or why.

The endpoint description labels the population as the Intent-to-Treat Population, while also specifying several post-enrollment conditions for inclusion. The registry record does not explain how that label was applied or how it relates to participants who did not meet the testing and challenge criteria. No further analysis-population definitions or participant flow details are reported in the registry record.

These gaps limit assessment of the risk of bias from exclusions or missing outcome information. They also prevent independent verification of the reported endpoint, because no numerical estimate is reported. The safety entries have their own at-risk counts, which differ by event category; those counts do not resolve the missing primary endpoint denominator.

12. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0.2% reporting threshold.

Most frequent non-serious adverse event terms

The registry safety entries report serious adverse events for three groups: all-study-emergent adverse events, events related to skin testing, and events related to oral amoxicillin challenge. Each entry reports zero serious adverse events. The at-risk counts are 481 for the first two categories and 391 for the oral challenge category. The registry record does not report death counts.

These are event-category summaries, not comparative safety estimates. In particular, zero reported serious adverse events does not establish that the risk is zero in a broader population, and the absence of a control group means the record cannot provide a randomized comparison of serious-event risk. The registry record also does not provide additional detail about the definition, adjudication, or timing of serious events.

The same safety entries report other adverse events: 53 in the all-study-emergent category, 7 related to skin testing, and 12 related to oral amoxicillin challenge. These counts should be read with their category-specific at-risk groups; they are not interchangeable, and the registry record does not provide detailed descriptions in the available information. The counts alone do not show severity, duration, or causal certainty beyond the categories stated in the record.

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

The study’s primary question is clearly defined: among participants with valid negative skin tests who received oral amoxicillin challenge, what percentage did not experience an IgE-dependent hypersensitivity reaction within 72 hours? That is a useful statistical framing for a negative predictive value because it ties a test result to a specified follow-up outcome in a defined population.

However, the registry record does not report the NPV estimate, its 95% confidence interval, the numerator, or the denominator. It therefore does not show the magnitude or precision of the primary result. The study enrollment entry cannot substitute for the missing endpoint denominator, because the primary endpoint population required valid tests, negative results with all three kit reagents, and receipt of the challenge.

The record also does not describe a comparator group. The design can describe what was observed after skin testing and challenge among eligible participants, but it does not provide a randomized comparison with another test or management strategy. The safety counts likewise describe events recorded in the study, not relative risks versus a control group.

The follow-up specified for the primary outcome is 72 hours. Any interpretation of the endpoint must therefore remain tied to that period and the stated outcome definition; the record does not provide results for longer follow-up. The population is also specifically characterized by a history of an IgE-dependent reaction to a penicillin product and by the test and challenge eligibility requirements. The record does not establish how results would apply to people outside those criteria.

In short, the registry record supplies the study question, design features, endpoint definition, and selected safety counts, but not the primary numerical result. It can support an explanation of what the endpoint was intended to measure, but not a conclusion about its observed value or precision.

14. Why This Trial Matters Statistically

This study raises a statistical question that is different from a conventional treatment-versus-control comparison. Its primary endpoint is a negative predictive value: a proportion among participants with negative skin tests who proceeded to oral challenge, defined by whether an IgE-dependent reaction occurred within a specified follow-up period. Correct interpretation depends on identifying the population to which that proportion applies.

The distinction between enrollment and endpoint eligibility is especially important. The record reports enrollment of 481, while the endpoint population required valid skin testing, negative intradermal tests with all three kit reagents, and receipt of the challenge. Without a reported endpoint denominator, readers cannot tell how many participants contributed to the primary estimate. This is not a minor presentational detail: a proportion is meaningful only alongside a clearly defined denominator and outcome count.

The single-group design also determines what can be learned. It can describe the observed sequence of negative skin testing followed by challenge, but it cannot estimate a randomized contrast with another testing strategy. A within-group predictive value should not be confused with a causal comparison. The safety summaries have the same interpretive constraint.

Finally, precision is part of the result, not an optional add-on. The registry record names a 95% confidence interval but reports neither its limits nor the NPV itself. As a result, the key statistical finding cannot be judged for magnitude or uncertainty from this record. This makes transparent endpoint reporting, analysis-population accounting, and clear description of missing data particularly consequential for interpreting studies framed around predictive performance.

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.

The largest limitation for statistical interpretation is that the registry record does not report the primary NPV estimate, its confidence interval, or the event count and denominator. Although the endpoint definition is detailed, a result cannot be assessed without those quantities. The number enrolled should not be assumed to be the number included in the endpoint analysis.

The record describes a single-group study with no allocation and no masking. There is no concurrent control group against which to compare test strategies or event frequencies. The design therefore supports a descriptive account of outcomes in the studied group, rather than a randomized estimate of comparative effect. The record does not report a detailed statistical analysis plan or explain whether other analyses were conducted.

The primary population required valid skin testing, negative intradermal results with all three kit reagents, and an oral challenge. The record does not report how many participants met these criteria, how invalid tests or incomplete follow-up were handled, or how participants who did not proceed to challenge were treated in the analysis. The relationship between the population label and these eligibility conditions is not further explained.

The outcome follow-up specified for the primary endpoint is 72 hours. The registry record does not provide additional follow-up results or a more detailed description of outcome ascertainment. The safety information contains category-level counts, but detailed event descriptions and death counts are not reported in the registry record. These limitations constrain assessment of endpoint precision, completeness, and generalizability beyond the defined study population and observation period.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Evaluation of Skin Testing Reagents for Penicillin Allergy — NCT01818336.
• Registry record last updated 2017-01-04; status last verified 2016-11.
• Results first posted 2017-01-04.
• 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

The registry record describes a completed Phase 3, single-group study of the Penicillin Skin Test Kit, with enrollment of 481 and no allocation or masking. Participants with negative intradermal tests were to receive a single oral amoxicillin challenge dose and be followed for 72 hours.

The primary endpoint was negative predictive value among participants with valid skin testing, negative intradermal tests with all three kit reagents, and receipt of the oral challenge. The record specifies a 95% confidence interval but does not report the endpoint estimate, interval, event count, or analysis denominator. It lists no analysis records.

Safety entries report zero serious adverse events in each listed category, with other adverse-event counts of 53 for all-study-emergent events, 7 related to skin testing, and 12 related to oral challenge. Death counts are not reported in the registry record. The absence of a comparator and the missing primary numerical result limit statistical interpretation.