Effect of Providing Stratification of Low Risk Penicillin Allergies on Penicillin Allergy Label Removal in ICU Setting: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Effect of Providing Stratification of Low Risk Penicillin Allergies on Penicillin Allergy Label Removal in ICU Setting (NCT03702283), a na na study. The primary outcome is Penicillin Allergy Label Removal.

2. Clinical Question

The study addresses whether an ICU-based risk-stratification approach can be used in the process of removing low-risk penicillin-allergy labels. Its intervention was not simply a test dose: it was an alert with a risk-stratification tool and recommendations about whether to use an oral amoxicillin challenge order set. The primary outcome concerned documentation in the medical chart, rather than a direct measure of whether a patient was truly allergic.

The registry describes the work as a single-arm intervention pilot intended to inform a future stepped-wedge randomized trial. The current study therefore provides no randomized estimate of how the intervention compares with another approach. Its primary question is descriptive in the posted record: whether labels were removed among the low-risk patients targeted by the intervention. The number of such patients and the number whose labels were removed are not reported in the registry record, so the primary outcome cannot be quantified from that record.

3. Trial Design

This completed study was registered as a single-group intervention pilot in an ICU setting. The registry lists Vanderbilt University Medical Center as sponsor, an enrollment of 240, and the condition as penicillin allergy. The recorded start date is 2019-03-31, and the recorded primary completion date is 2021-03-31. The study phase is listed as not applicable. Masking is listed as none, and allocation is listed as not applicable.

The intervention was a Penicillin Allergic Risk Stratification Best Practice Alert. Its description says the alert provided access to a risk-stratification tool and recommendations on whether to use an oral amoxicillin test-dose challenge order set for patients classified as low risk. The study lists one intervention group, described as penicillin-allergic ICU patients. No concurrent control group is described, so the design cannot separate changes associated with the intervention from changes that might occur over time or through other aspects of care.

The primary outcome was penicillin-allergy label removal. It was defined as the proportion of low-risk patients whose labels were removed from the allergy section of the medical chart. The assessment time was hospital discharge, approximately 7 days after ICU transfer. The registry also lists secondary outcomes concerning adverse events, communication about penicillin allergy in the discharge summary, and antibiotic utilization by patients.

The registry presents the study as a pilot intended to support future research, not as a randomized trial. It does not report protocol details such as how risk categories were assigned, whether clinicians consistently acted on the alert, or how participants were selected for analysis. Those details should not be inferred from the intervention description alone. Most importantly for interpretation, the registry record does not report the primary outcome result or an inferential analysis of that outcome.

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: Of those who were eligible, with a low-risk PAL, all 240 patients were offered an oral challenge with amoxicillin via informed consent, and 205/240 (85.4%) patients agreed to participate.

Overall Study

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 Penicillin Allergy Label Removal, measured over Hospital discharge at approximately 7 days after ICU transfer.

The primary endpoint concerns whether a low-risk penicillin-allergy label was removed from the allergy section of the medical chart by hospital discharge, approximately 7 days after ICU transfer. The outcome is expressed as a proportion: the number of eligible low-risk patients whose labels were removed relative to the relevant low-risk patient group. The registry describes the endpoint as a count of participants, but its stated measure is the proportion with label removal.

This endpoint captures a documentation change that is directly connected to the intervention’s purpose. A risk-stratification alert may support consideration of an oral amoxicillin test-dose challenge in patients judged to be low risk; label removal records whether the chart was changed. It does not, by itself, establish the accuracy of the allergy classification, the clinical reason for each chart change, or what happened during later treatment.

For a proportion to be interpreted, both the numerator and denominator matter. The numerator would be the low-risk patients whose labels were removed, and the denominator would be the low-risk patients assessed for this outcome. Neither count, nor the resulting proportion, is reported in the registry record. Consequently, the primary endpoint’s observed size and precision cannot be assessed from the posted record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies a binary primary outcome: a patient’s penicillin-allergy label was either removed from the medical chart or it was not. The stated summary measure is a proportion among patients with low-risk penicillin allergy. A proportion conveys how common an outcome was in a defined group; its interpretation depends on a clear numerator and denominator. Neither is reported for the primary outcome in the registry record, and no primary estimate or analysis is listed.

What does a confidence interval mean?

In general, a confidence interval describes the uncertainty around an estimate under a specified statistical method. For a proportion, it reflects both the observed proportion and the amount of information available, which is related to the number of participants assessed. A narrow interval indicates greater statistical precision than a wide interval, but an interval does not remove bias from selection, measurement, or study design. No confidence interval for the primary outcome is reported in the registry record, so its precision cannot be evaluated.

What can a single-arm analysis estimate?

In general, a single-arm study can describe the outcome frequency in the group observed. It cannot directly estimate the difference between intervention and control groups when no concurrent control group exists. Comparing the observed result with a historical or external benchmark would require a justified comparator and careful attention to differences in populations, definitions, and care settings. No such comparative analysis is reported. The registry lists allocation as not applicable and describes one intervention group, so the study should not be read as a randomized comparison.

For a binary endpoint, one common general approach is to estimate the observed proportion and its uncertainty using a method appropriate to the sample size and data structure. This is a general description, not a claim about the method used in this trial: the registry does not report an analysis method for the primary outcome. It also does not report a model, adjustment variables, hypothesis test, or primary-outcome estimate. An analyst should not infer that a particular method was used merely because the endpoint is binary.

How should the planning calculations be understood?

The registry’s sample-size assumptions describe planning illustrations based on an observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The effect to detect is described as a stated fraction of the observed effect. Such calculations are hypothetical planning aids, not observed outcomes and not evidence that a study had a particular probability of finding a statistically significant result. They are not “observed power.” The assumptions describe a two-arm planning framework, whereas the registered pilot is single arm. No sample-size justification for the pilot is reported in the registry record, and the primary observed effect needed to interpret the planning basis is not reported there.

More generally, sample-size calculations depend on the target effect, outcome variability, chosen error rate, allocation, and statistical method. A calculated sample size indicates what a design would plan to recruit under those assumptions; it does not guarantee precision or determine the result. For a binary outcome, the anticipated event proportion is particularly relevant. When the observed outcome and its denominator are not available, readers cannot judge how closely a planning assumption matched the study’s actual experience. Any sample-size values should be understood as planning illustrations rather than a description of the evidence observed in this pilot.

Why does the absence of a comparator matter?

In general, random allocation helps make groups comparable on average, supporting a causal interpretation of differences between groups. This pilot does not report randomized allocation or a comparator group. As a result, any change in labeling documented among participants cannot be attributed to the alert alone from this design. Other influences on clinical decisions and charting remain possible. The record’s description of a future randomized study indicates that the pilot was intended to inform later research, but it does not substitute for a comparative analysis in the current study.

Finally, event counts and proportions answer different descriptive questions. A count tells how many participants experienced an outcome; a proportion relates that count to a defined group. The adverse-event record includes counts and at-risk denominators, while the primary label-removal result is absent. The counts in the safety record should not be repurposed as a denominator for the primary endpoint: they refer to a different outcome and the record does not state that the safety population is identical to the low-risk primary-outcome population.

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("Intervention Group- Single Arm")
n      <- c(205)
events <- c(201)
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  Intervention Group- Single Arm  205  201  98.0  95.1  99.5  95.1  99.2  95.1  99.2

11. Multiplicity, Missing Data and Analysis Populations

The registry lists five outcome measures and identifies penicillin-allergy label removal as the primary outcome. Secondary outcomes include adverse events, communication about penicillin allergy in the discharge summary, and antibiotic utilization by patients. The registry record reports no analyses, so it does not show which outcomes were analyzed, how any analyses were prioritized, or whether results were adjusted for making multiple statistical comparisons.

In general, examining several outcomes can increase the chance of at least one apparently noteworthy result arising by chance when each is tested separately. A statistical plan may address this through prespecification, adjustment, or a clear distinction between primary and exploratory outcomes. The registry does not report a multiplicity strategy. This absence should not be taken as evidence that no such strategy existed outside the registry, only that it is not reported in the registry record.

The record does not report missing-data counts, reasons for missingness, or a method for handling missing outcomes. It also does not specify the analysis population for the primary endpoint, such as whether it included all enrolled patients, only those classified as low risk, or only those with documented discharge status. The endpoint description implies a low-risk population, but the operational rules for inclusion in the denominator are not given. Without those details, readers cannot assess whether missing chart information or exclusions affected the reported endpoint; in any case, the endpoint result itself is not reported.

No imputation method, sensitivity analysis, or analysis set is reported. These omissions matter because missing outcomes can affect both the estimated proportion and its precision, and different rules can lead to different interpretations. The registry provides no basis for determining whether missingness occurred or how it was handled.

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 serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: morbilliform exanthem following amoxicillin challenge (0.5%), morbilliform exanthem during subsequent penicillin treatment at vumc (0.5%).

The intervention group safety record lists 205 participants at risk for serious adverse events and one participant affected. For other adverse events, it lists 205 participants at risk and two affected. The record lists zero deaths. These are descriptive counts from the single intervention group, not a comparison with an untreated or alternative-care group.

The registry lists two serious-event terms: “Morbilliform Exanthem following Amoxicillin Challenge” and “Morbilliform Exanthem During Subsequent Penicillin Treatment at VUMC.” Each term is accompanied by a reported frequency of 0.5%. The listed terms describe the events recorded in the registry, but they do not provide detailed timing, clinical course, management, or adjudication. The record does not establish from these entries alone whether an event was caused by the alert or a challenge.

At-risk denominators are useful context for counts, but these safety figures should not be treated as the primary endpoint denominator. The primary outcome concerns label removal among low-risk patients, and its denominator is not reported. Because there is no comparison group, the safety entries also cannot establish how event frequency differed from what might have occurred under another strategy.

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

The registry record establishes that a completed, single-arm ICU pilot evaluated a risk-stratification alert and named label removal among low-risk patients as its primary outcome. It also describes the outcome’s intended timing and lists adverse-event counts for the intervention group. These details clarify what the study set out to assess and what safety events were recorded.

The record does not report the primary outcome result. It therefore does not show how many low-risk patients had labels removed, what proportion that represented, or how precisely that proportion was estimated. Nor does it report a statistical analysis comparing outcomes across groups. The study’s effect on label removal cannot be quantified from the registry record.

Even if a primary proportion were available, the single-arm design would describe the outcome in the intervention group rather than establish what the result would have been without the alert. No causal conclusion about the intervention follows from a single group alone. A chart label changing is also not equivalent to a complete assessment of allergy status, and the registry does not report whether labels remained removed or how later care was affected.

The adverse-event entries are likewise descriptive. They record events and at-risk participants, but there is no comparator to estimate relative safety, and the registry does not provide enough detail to attribute events to a particular part of care. The absence of a reported primary result should not be interpreted as evidence of no label removals or of no effect. It means that the registry record does not report the information needed to evaluate that outcome.

The stated pilot purpose was to inform future research. The findings available in the registry are therefore best understood as a record of study design and limited safety reporting, not as a comparative estimate of effectiveness. The record does not establish whether risk stratification increased label removal, whether another strategy would perform differently, or whether the observed approach would have the same outcomes in other settings.

14. Why This Trial Matters Statistically

This trial addresses a practical statistical question: how to evaluate a clinical workflow change when the intended outcome is a chart action and the intervention is delivered in a complex care setting. Defining label removal as the primary endpoint makes the target measurable, but interpretation still depends on a clear eligible population, a defined denominator, and complete reporting of the observed result. The registry states the endpoint and timing but does not report its numerator, denominator, or estimate.

The single-arm pilot design is relevant to feasibility and preparation for later research, but it has important limits for causal inference. Without a concurrent comparison group, an observed change in documentation could reflect influences other than the alert. A future randomized design, as described in the study rationale, could address a comparative question more directly, provided its methods and outcomes are fully reported.

The record also illustrates why safety counts and efficacy outcomes must be read separately. The reported adverse-event counts do not supply a result for label removal, and the safety population cannot automatically be treated as the primary-outcome population. Clear definitions and denominators are essential when outcomes concern different events and potentially different groups.

Finally, registry completeness affects what readers can learn. The record names secondary outcomes and records safety information, but lists no analyses and does not report the primary outcome result, missing-data approach, multiplicity handling, or statistical methods. These gaps prevent an assessment of the size and precision of the main outcome. They also reinforce the distinction between a trial’s stated objective and evidence that can be evaluated from its posted results.

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 main limitation for interpreting this trial is that the registry record does not report the primary endpoint result. The number of low-risk patients whose labels were removed, the denominator used, and an estimate of the proportion are absent. No confidence interval or statistical analysis of the primary outcome is reported, so neither outcome frequency nor precision can be assessed from the record.

The single-arm design provides no concurrent comparison group. It cannot distinguish changes associated with the alert from changes in clinical practice, patient mix, documentation, or other influences. Allocation is listed as not applicable and masking as none. The record also does not report whether the risk-stratification tool was applied consistently, how many patients received a challenge, or how recommendations were followed.

Details needed to interpret the target population are limited. Although the primary endpoint concerns low-risk patients, the record does not report further eligibility criteria, the method for classifying risk, or the analysis population. It does not report missing-data counts or methods for handling missing primary outcomes. The relationship between the enrollment listed in the record and the population assessed for the primary endpoint is not specified.

The registry lists secondary outcomes but reports no analyses and no multiplicity approach. It does not report a statistical analysis plan, sample-size justification for this pilot, or endpoint-specific estimates. Safety reporting includes event counts and listed terms, but not detailed event-level information or a comparator. These limitations constrain what can be concluded from the posted record; they do not establish that information absent from the registry was never collected.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Effect of Providing Stratification of Low Risk Penicillin Allergies on Penicillin Allergy Label Removal in ICU Setting — NCT03702283.
• Registry record last updated 2022-05-13; status last verified 2022-01.
• Results first posted 2022-05-13.
• 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, single-arm ICU pilot sponsored by Vanderbilt University Medical Center, with an enrollment of 240. The intervention was a best-practices alert containing penicillin-allergy risk stratification and recommendations about an oral amoxicillin test-dose challenge order set for patients classified as low risk. The primary outcome was removal of low-risk penicillin-allergy labels by hospital discharge, approximately 7 days after ICU transfer.

The primary outcome result and analyses are not reported in the registry record. The intervention group safety record lists 205 participants at risk for serious adverse events, one affected, and two participants with other adverse events; zero deaths are listed. The study has no concurrent comparator, so the record does not provide a randomized estimate of the alert’s effect or comparative safety.