Review Reports - Patient-Reported Outcomes Following Beta-Lactam Antibiotic Allergy Delabeling in Allergy Work-Up

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This manuscript addresses an important and clinically relevant topic, namely the patient perspective on antibiotic allergy delabelling.

The authors compare patients de-labeld from beta-lactam antibiotic allergy with patients with active label and describe that de-labeled patients perceive higher safety when getting BL-antibiotikca prescribed. Quality of live based on scors is not different between groups (exept from environment domain).
Among de-labeled patients at follow-up approx. half of patients required antibiotics , of which 65% received BL-abs .
With all de-labeled patienst with BL-ab indication at follow-up in 35 % of cases physicians refused to give them despite de-labeling.

Although the study cohort is relatively small, the findings have the potential to provide meaningful insights into patient experiences and perceptions surrounding antibiotic allergy delabelling.

However several major comments needs to address.

Major comments

Further details on the approach to antibiotic delabelling would be helpful. For example, was a standardized assessment performed to determine the appropriate diagnostic approach for antibiotic allergy de-labelling?
More details on how the delabelling outcome was communicated and documented should be provided, both for the patient and for other healthcare providers. For example, were patients informed during a follow-up consultation, was the antibiotic allergy label removed from the electronic medical record, and was the outcome communicated to treating physicians?

The interviews appear to have been semi-structured, which may introduce a considerable risk of bias. Furthermore, the final item in Supplementary Table 1, “Opinion about the possible prescription of an antibiotic by an allergist rather than another specialist,” seems to be an open-ended question. The manuscript should clearly describe how responses to this question were collected, categorized, and analyzed. Without a detailed description of the analytical approach, the validity and reproducibility of these findings are difficult to assess.

The results section is overly detailed. For instance, the demographic characteristics of the study population, including age and sex, could be presented concisely in the text using summary statistics (e.g., range and SD), and the table 1 could therefore be omitted.

The entire Results section should be substantially revised to improve conciseness and readability.

The reported numbers and percentages appear inconsistent. For example, in lines 91-93, 37/46 equals 80%, not 74%. Several other percentages in this sentence also seem incorrect and should be verified against the original data.

The Results section requires substantial editing to improve conciseness and readability. Many data are presented redundantly in the text, tables, and figures. The findings could be summarized more efficiently by reporting only the key results and avoiding duplication across different presentation formats. In particular, Figure 2 does not appear to add information beyond that already provided elsewhere and could be omitted.

Line 90–93: The Results state that 74% of patients indicated drug provocation testing as the primary diagnostic method leading to de-labelling. However, it is unclear how many patients actually underwent drug provocation testing, particularly as several patients underwent multiple procedures. Please clarify whether patients could select multiple methods and report how many patients underwent each diagnostic procedure. If multiple methods could be selected, the term “primary diagnostic method” may be misleading.

If drug provocation testing was performed in all participants, the statement in the Discussion section, sentence 218 “the majority of delabelled patients indicated drug provocation testing as the primary diagnostic method leading to delabelling” may require further consideration, as 26% of patients did not identify it as their primary diagnostic method. This finding may warrant further discussion, particularly regarding patients’ perceptions of the reliability of drug provocation testing.

The authors report the frequency and proportion of cases with a clinical indication for beta-lactam antibiotic use. However, a major limitation of the study is the lack of information on follow-up duration after delabelling. This may introduce substantial bias, as the likelihood of requiring antibiotic treatment increases with longer follow-up. Moreover, both patients and physicians may become more confident in using beta-lactam antibiotics as more time elapses after delabelling, potentially influencing the reported outcomes. Therefore, follow-up time should be presented and taken into account when interpreting the results.

Basic and important patients characteristic are missing as comorbidity (repeated infections ect ) ; time between last clinical reaction; type of reaction (I versus IV) and severity of reaction (Anaphylaxis or SCAR). This should be presented in a baseline table and included in analysis and interpretation.

Subsequently; the data presented in Figure 3 and described in the text are of limited value, as no meaningful association can be established with either the type or severity of the reactions. As a result, the clinical relevance of these findings remains unclear.

The Discussion would benefit from a more integrated interpretation of the results. Currently, the findings are discussed individually under separate headings, resulting in a fragmented and overly long discussion. Synthesizing the results into a coherent narrative would improve readability and enhance the scientific impact of the manuscript.

Line 148–150: three patients appear to have experienced mild cutaneous symptoms following delabelling. Further information regarding the subsequent clinical course would be valuable. For example, did these patients discontinue treatment, did they undergo reassessment by an allergist?
Line 178: report the number of patients who did not receive any antibiotic prescription during the relevant follow-up period. This is important for interpreting the extent of subsequent antibiotic use following delabelling.

Is figure 4 based on because the questionnaire item in Supplementary Table 1 phrased as “Antibiotic use between the initial diagnosis and the delabelling.”? In that case clarification is required regarding whether all antibiotics shown represent alternative antibiotics prescribed specifically because of the beta-lactam allergy label. Or if some of these antibiotics may also constitute first-line treatment options, in which case the beta-lactam allergy label may not have influenced antibiotic selection.
In the current version of the manuscript, Figure 4 adds limited value to the overall interpretation of the results. Consider removing the figure.

The conclusions regarding safe antibiotic reuse appear somewhat stronger than warranted by the presented data, since the subgroup assessed for subsequent antibiotic use is small (n=17) with symptom recurrence occurring in 3 patients. Similarly, the statement in Sentence 197 that de-labelling “is associated with a low rate of symptom recurrence” may overstate the findings. A more cautious, descriptive formulation would be appropriate, with explicit acknowledgement of the limited sample size and the exploratory nature of these findings. In addition, enabling safe antibiotic reuse is not specifically addressed in the stated study aim. The study aim should therefore be aligned more closely with the outcomes assessed.

Minor comments

Line 374: the statistical software used for the analysis should be specified.
The figure numbering in the text is not always consistent with the corresponding figures.
Why is methods section placed between discussion and conclusion?
It is unclear what the authors whant to say with the data presented in line 97-102. This is not explained and nor further discussed . Should be removed?
Consider to present figure 2 in supplements.

Author Response

Response to Reviewer 1 Comments

We sincerely thank the Reviewer for the time and effort devoted to the careful evaluation of our manuscript. We greatly appreciate the constructive and insightful comments and suggestions, which have helped us improve the clarity, scientific rigor, and overall quality of the manuscript. We have carefully considered all comments and have revised the manuscript accordingly. Below, we provide a point-by-point response to each comment.

Further details on the approach to antibiotic delabeling would be helpful.
For example, was a standardized assessment performed to determine the appropriate diagnostic approach for antibiotic allergy de-labelling?

Thank you for pointing this out. We have added an information regarding the standardized diagnostic approach in the materials and methods section (section 4.1), line 347-349.
More details on how the delabelling outcome was communicated and documented should be provided, both for the patient and for other healthcare providers. For example, were patients informed during a follow-up consultation, was the antibiotic allergy label removed from the electronic medical record, and was the outcome communicated to treating physicians?

Thank you for this comment, as it will be a valuable addition to our paper. We have added an explanation of the important information about delabeling outcomes for both the patients and other physicians in the materials and methods section (paragraph 4.1), line 349-354.
The interviews appear to have been semi-structured, which may introduce a considerable risk of bias. Furthermore, the final item in Supplementary Table 1, “Opinion about the possible prescription of an antibiotic by an allergist rather than another specialist,” seems to be an open-ended question. The manuscript should clearly describe how responses to this question were collected, categorized, and analyzed. Without a detailed description of the analytical approach, the validity and reproducibility of these findings are difficult to assess.

Thank you for pointing this out. After your comment we have decided not to include this data in this paper, as the results could have been inconclusive due to possible incomprehension between researchers and patients. The analysis of the results indicated that patients could have misunders tood the question due to the complexity of the Polish language. Analysis of answers to this question are no longer included in our work. Additionally, we have changed the numbering of supplementary materials and updated the supplementary table 2 with an explanation of the question you mentioned.
The results section is overly detailed. For instance, the demographic characteristics of the study population, including age and sex, could be presented concisely in the text using summary statistics (e.g., range and SD), and the table 1 could therefore be omitted.

Thank you, we agree with this comment. We have removed Table 1 and added its content to the Results section (paragraph 2.1 Study population and baseline characteristics) in lines 78-85. We have also shortened this section.
The entire Results section should be substantially revised to improve conciseness and readability.

Thank you for pointing this out – we improved the results section by removing table 1, omitting the data mentioned in comment 3 and editing the paragraph 2.2 (lines 90-95) both linguistically and methodologically.
The reported numbers and percentages appear inconsistent. For example, in lines 91-93, 37/46 equals 80%, not 74%. Several other percentages in this sentence also seem incorrect and should be verified against the original data.

Thank you. We have verified and corrected the data in the results section: paragraphs 2.2, 2.3 and 2.5.
The Results section requires substantial editing to improve conciseness and readability. Many data are presented redundantly in the text, tables, and figures. The findings could be summarized more efficiently by reporting only the key results and avoiding duplication across different presentation formats. In particular, Figure 2 does not appear to add information beyond that already provided elsewhere and could be omitted.

Thank you for this valuable comment, however, our intention was to present data graphically, to improve the comprehensibility, as the follow-up results and decision making post delabeling presented in figure 2 are an important part of our study. We have additionally omitted the less significant statistical data in the results section (e.g. size effects), with the exception of p values, as you mentioned in the further comment, our study has a limitation of sample size.
We have edited paragraph: 2.1 Study population and baseline characteristics and removed the table 1.
Line 90–93: The Results state that 74% of patients indicated drug provocation testing as the primary diagnostic method leading to de-labelling. However, it is unclear how many patients actually underwent drug provocation testing, particularly as several patients underwent multiple procedures. Please clarify whether patients could select multiple methods and report how many patients underwent each diagnostic procedure. If multiple methods could be selected, the term “primary diagnostic method” may be misleading.

We have edited the paragraph 2.2 (lines 90-95) both linguistically and methodologically. We changed the method of presenting data to highlight the patients perception on diagnostic methods used in the delabeling process.
If drug provocation testing was performed in all participants, the statement in the Discussion section, sentence 218 “the majority of delabelled patients indicated drug provocation testing as the primary diagnostic method leading to delabelling” may require further consideration, as 26% of patients did not identify it as their primary diagnostic method. This finding may warrant further discussion, particularly regarding patients’ perceptions of the reliability of drug provocation testing.

Thank you for this comment. We have changed the discussion section (paragraph 3.2) to correspond with the prior results in paragraph 2.2 (lines 90-95). We added a reference to the work of Popiolek et al., highlighting the importance of patient education when performing oral provocation testing.
We have emphasized in this section that drug provocation remains the gold standard for confirming or excluding beta-lactam allergy. We have added the information about the delabeling methods we use in our department (physician interview, skin testing + oral provocation testing) in the materials and methods section (section 4.1), line 347-349.
The authors report the frequency and proportion of cases with a clinical indication for beta-lactam antibiotic use. However, a major limitation of the study is the lack of information on follow-up duration after delabelling. This may introduce substantial bias, as the likelihood of requiring antibiotic treatment increases with longer follow-up. Moreover, both patients and physicians may become more confident in using beta-lactam antibiotics as more time elapses after delabelling, potentially influencing the reported outcomes. Therefore, follow-up time should be presented and taken into account when interpreting the results.

Thank you for this comment. We found and incorporated the registration data in the Allergy Clinic database considering mean and median of time from delabeling to fixed study endpoint on November 4, 2025. (lines 128-133). We calculated the beta-lactam antibiotic courses per person-year (lines 151-156).
Basic and important patients characteristic are missing as comorbidity (repeated infections ect ) ; time between last clinical reaction; type of reaction (I versus IV) and severity of reaction (Anaphylaxis or SCAR). This should be presented in a baseline table and included in analysis and interpretation.

Thank you for this valuable addition. Although the mentioned details would surely increase the scientific value of this paper, we have decided to base this study on the interview with patients, with simple dichotomous (yes/no) or with the use of scale ranging from 1 to 10. The main goal of this study was to present the patients opinions regarding being delabelled or labelled and the quality of life of both groups.
The Discussion would benefit from a more integrated interpretation of the results. Currently, the findings are discussed individually under separate headings, resulting in a fragmented and overly long discussion. Synthesizing the results into a coherent narrative would improve readability and enhance the scientific impact of the manuscript.

We improved the overall structure of the discussion section and eliminated subtitles.
Line 148–150: three patients appear to have experienced mild cutaneous symptoms following delabelling. Further information regarding the subsequent clinical course would be valuable. For example, did these patients discontinue treatment, did they undergo reassessment by an allergist?

Thank you for this comment. We have added the information about planned reassessment in our clinic in the Discussion section (lines 280-284).
Line 178: report the number of patients who did not receive any antibiotic prescription during the relevant follow-up period. This is important for interpreting the extent of subsequent antibiotic use following delabelling.

Thank you for this comment – we have decided to include this data indirectly in figure 2, in which we describe the follow-up of patients from the delabelled group.
The conclusions regarding safe antibiotic reuse appear somewhat stronger than warranted by the presented data, since the subgroup assessed for subsequent antibiotic use is small (n=17) with symptom recurrence occurring in 3 patients. Similarly, the statement in Sentence 197 that de-labelling “is associated with a low rate of symptom recurrence” may overstate the findings. A more cautious, descriptive formulation would be appropriate, with explicit acknowledgement of the limited sample size and the exploratory nature of these findings. In addition, enabling safe antibiotic reuse is not specifically addressed in the stated study aim. The study aim should therefore be aligned more closely with the outcomes assessed.

We changed the wording in the last sentence of the introduction (aim of the study, lines 73-74, as one of the goals of the study was to explore patients’ experiences with beta-lactam antibiotic use after delabeling. In line 197, we added information about self-reported aspect of the results and omitted the line “it supports the safe reintegration of first-line antibiotics”, as the main goal of this study was to present patients opinions regarding delabeling, and no to clinically assess the safety of the process.
Minor comments

Line 374: the statistical software used for the analysis should be specified.
The statistical software used for the analysis is mentioned in materials and metods section (4.5 Statistical Analysis). Responses were coded and analyzed using Statistica software.

The figure numbering in the text is not always consistent with the corresponding figures.
The figures and tables numbering in the main manuscript and in supplementary materials now correspond to references in the text.

Why is methods section placed between discussion and conclusion?

Thank you for this comment. We followed the official Antibiotics journal template.

It is unclear what the authors whant to say with the data presented in line 97-102. This is not explained and nor further discussed . Should be removed?

We added corresponding paragraph in the discussion, commenting on these data (lines 245-251).

Consider to present figure 2 in supplements.
Thank you for this valuable comment, however, our intention was to present data graphically, to improve the comprehensibility, as the follow-up results and decision making post delabeling presented in figure 2 are an important part of our study. In our opinion it is important to see how patients and physicians actually behave after delabeling in context of antibiotic choice.

Reviewer 2 Report

Comments and Suggestions for Authors

Dear authors,

congratulations on your valuable work. Please, find here below some suggestions in order to further improve the quality of your paper.

You describe the study as a cross-sectional questionnaire-based survey, but the two groups were recruited from different periods and represent different clinical trajectories. The delabelled group underwent specialist evaluation between 2017 and 2025, whereas the non-delabelled group consisted of patients registered for diagnostic evaluation in 2025–2026. Therefore, these groups should not necessarily be interpreted as two comparable cohorts differing only in delabeling status. I recommend to discuss more explicitly the potential for selection bias and confounding related to this aspect. Importantly, the manuscript currently uses between-group differences to support conclusions regarding the effect of delabeling. Given the observational and cross-sectional design, the language should be tempered throughout the manuscript. The data demonstrate an association between delabeling status and perceived safety, but they do not establish that delabeling itself caused the observed differences.
Three of the 17 patients who received beta-lactams after delabeling reported mild cutaneous symptoms. The manuscript states that "delabeling might have been unsuccessful" in these cases and later refers to a low recurrence rate. This interpretation needs to be interpreted more cautionally. A rash or urticaria occurring after subsequent beta-lactam exposure does not necessarily establish recurrence of the original hypersensitivity, particularly when the study relies entirely on patient-reported symptoms and there was apparently no formal allergological reassessment. The current wording risks overinterpreting these events. I suggest describing them as "reported cutaneous symptoms following re-exposure" unless objective clinical reassessment was performed.
You state that two separate "authorial questionnaires" were used. This raises important questions regarding their development, validation, comparability, and potential measurement bias. Please provide more information regarding how the questionnaires were developed, whether they were pilot-tested and how; whether response options were standardized; whether investigators administering the telephone interviews were blinded to group allocation and whether interviewer effects were considered.
The study includes only 96 participants, and several outcomes involve relatively small denominators. The authors appropriately use non-parametric testing for WHOQOL-BREF domains, but the statistical approach could be strengthened. Moreover, the manuscript reports p-values and some effect sizes, which is useful, but confidence intervals would substantially improve interpretation, particularly for the proportions related to antibiotic exposure and recurrent symptoms.
The relationship between allergy delabeling and antimicrobial stewardship should be developed more carefully. You correctly highlight that delabeling is not sufficient if the information is not effectively communicated and acted upon by subsequent prescribers. The finding that 9 patients were reportedly not prescribed beta-lactams despite documented delabeling is particularly relevant from an implementation and stewardship perspective. I would encourage you to expand this section and discuss delabeling as a health-system intervention, rather than solely as an allergological procedure. In this context, the manuscript could also discuss the importance of effective communication across healthcare settings, electronic health record updating, standardized documentation, education of non-allergist prescribers, and multidisciplinary antimicrobial stewardship programs. Please, refer to recent evidence that support these call-to-action (e.g. https://pubmed.ncbi.nlm.nih.gov/39762000/; https://pubmed.ncbi.nlm.nih.gov/35876226/)
The manuscript identifies two different implementation barriers: patients may remain reluctant to use beta-lactams, while physicians may continue to avoid them despite delabeling. These are conceptually distinct phenomena and may require different interventions. I suggest restructuring this part of the Discussion to explicitly distinguish these two barriers and discuss their respective implications for implementation.
The limitations section mentions sample size, recall bias, symptom misclassification, the generic quality-of-life instrument, and single-center design. However, the cross-sectional nature of the comparison deserves greater emphasis. The delabelled and non-delabelled groups are not randomized or prospectively followed from a common baseline. Therefore, observed differences in perceived safety or quality of life may reflect differences in patient characteristics rather than the causal effect of delabeling. This limitation should be explicitly reflected in the Abstract and Conclusions as well.

Author Response

Response to Reviewer 2 Comments

You describe the study as a cross-sectional questionnaire-based survey, but the two groups were recruited from different periods and represent different clinical trajectories. The delabelled group underwent specialist evaluation between 2017 and 2025, whereas the non-delabelled group consisted of patients registered for diagnostic evaluation in 2025–2026. Therefore, these groups should not necessarily be interpreted as two comparable cohorts differing only in delabeling status. I recommend to discuss more explicitly the potential for selection bias and confounding related to this aspect. Importantly, the manuscript currently uses between-group differences to support conclusions regarding the effect of delabeling. Given the observational and cross-sectional design, the language should be tempered throughout the manuscript. The data demonstrate an association between delabeling status and perceived safety, but they do not establish that delabeling itself caused the observed differences.
Thank you for this valuable comment. We changed the wording used throughout the manuscript, focusing on comparisons between delabelled and non-delabelled groups. (eg. lines 73-75).
We have expanded the limitations section, highlighting the observational and cross-sectional design of our study.
Three of the 17 patients who received beta-lactams after delabeling reported mild cutaneous symptoms. The manuscript states that "delabeling might have been unsuccessful" in these cases and later refers to a low recurrence rate. This interpretation needs to be interpreted more cautionally. A rash or urticaria occurring after subsequent beta-lactam exposure does not necessarily establish recurrence of the original hypersensitivity, particularly when the study relies entirely on patient-reported symptoms and there was apparently no formal allergological reassessment. The current wording risks overinterpreting these events. I suggest describing them as "reported cutaneous symptoms following re-exposure" unless objective clinical reassessment was performed.

Thank you for mentioning this aspect of our work. We have implemented the information about planned re-assessment in the future in those 3 cases to determine the nature of those symptoms in the discussion (lines 290-303). We added a statement to highlight, that rash or urticaria in this case does not necessarily establish recurrence of the original hypersensitivity.
You state that two separate "authorial questionnaires" were used. This raises important questions regarding their development, validation, comparability, and potential measurement bias. Please provide more information regarding how the questionnaires were developed, whether they were pilot-tested and how; whether response options were standardized; whether investigators administering the telephone interviews were blinded to group allocation and whether interviewer effects were considered.

Thank you for this comment. The structured telephone questionnaires were specifically designed for the purpose of this study under the direct supervision of senior clinicians with expertise in allergy and infectious diseases. To ensure clarity, options were standardized, primarily using binary (Yes/No) formats. Questions focused on objective clinical outcomes, such as self-reported re-exposure and subsequent adverse reactions, minimizing ambiguity during data collection.

Blinding of the investigators conducting the telephone interviews was structurally impossible due to the nature of the study design. Because the two cohorts were at fundamentally different stages of their clinical pathway (e.g., post-delabeling patients were questioned regarding drug re-exposure, whereas the pre-diagnostic group had not undergone testing), group-specific questions were inherently required. Blinding interviewer personnel during live, interactive clinical follow-up calls would therefore have been impossible.
To minimize potential interviewer bias and standardize data collection, all telephone interviews were conducted exclusively by trained medical personnel who had undergone standardized protocol training prior to initiating the study. Interviews were administered using a strictly scripted structure to ensure uniform delivery across all participants.
The study includes only 96 participants, and several outcomes involve relatively small denominators. The authors appropriately use non-parametric testing for WHOQOL-BREF domains, but the statistical approach could be strengthened. Moreover, the manuscript reports p-values and some effect sizes, which is useful, but confidence intervals would substantially improve interpretation, particularly for the proportions related to antibiotic exposure and recurrent symptoms.

Thank you for this comment. We have added the confidence intervals in paragraphs 2.1, 2.4 and 2.5 of the results section. Regarding the other review of our study, we omitted the effect sizes in the data presentation, to shorten and clarify the results section.
The relationship between allergy delabeling and antimicrobial stewardship should be developed more carefully. You correctly highlight that delabeling is not sufficient if the information is not effectively communicated and acted upon by subsequent prescribers. The finding that 9 patients were reportedly not prescribed beta-lactams despite documented delabeling is particularly relevant from an implementation and stewardship perspective. I would encourage you to expand this section and discuss delabeling as a health-system intervention, rather than solely as an allergological procedure. In this context, the manuscript could also discuss the importance of effective communication across healthcare settings, electronic health record updating, standardized documentation, education of non-allergist prescribers, and multidisciplinary antimicrobial stewardship programs. Please, refer to recent evidence that support these call-to-action (e.g. https://pubmed.ncbi.nlm.nih.gov/39762000/; https://pubmed.ncbi.nlm.nih.gov/35876226/)

Thank you for this valuable addition to our discussion section. We have implemented the references to recent evidence by Van De Sijpe et. al and Mengato et. al to the discussion section of our manuscript to add the stewardship perspective to the manuscript.
The manuscript identifies two different implementation barriers: patients may remain reluctant to use beta-lactams, while physicians may continue to avoid them despite delabeling. These are conceptually distinct phenomena and may require different interventions. I suggest restructuring this part of the Discussion to explicitly distinguish these two barriers and discuss their respective implications for implementation.

We have implemented the reference to a study of Alagoz et. al, which assess the barriers to penicillin allergy de-labeling by interviewing 20 clinicians from various fields of medicine to the discussion section of our manuscript.
The limitations section mentions sample size, recall bias, symptom misclassification, the generic quality-of-life instrument, and single-center design. However, the cross-sectional nature of the comparison deserves greater emphasis. The delabelled and non-delabelled groups are not randomized or prospectively followed from a common baseline. Therefore, observed differences in perceived safety or quality of life may reflect differences in patient characteristics rather than the causal effect of delabeling. This limitation should be explicitly reflected in the Abstract and Conclusions as well.

Thank you for this valuable addition. We have expanded the limitations section, highlighting the observational and cross-sectional design of our study. (lines 335-340)
We haven’t modified the abstract, due to the limited word count. We added one final sentence to the conclusions, highlighting the importance of prospective, longitudinal studies on larger patient cohorts to establish causal trajectories and long-term clinical impacts

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript addresses a relevant clinical question: whether patients accept beta-lactam allergy delabeling and subsequently use these antibiotics. The patient-reported perspective is valuable, particularly the distinction between diagnostic assessment and its practical consequences. However, several issues require clarification before the findings can be interpreted reliably. I recommend major revision.

The conclusions should reflect the cross-sectional design. The study compares different patients after delabeling and while awaiting assessment; it does not measure individual improvement. Please replace statements that delabeling “improves” perceived safety with wording describing the observed between-group difference. Selection bias and differences in previous reactions, comorbidities, and access to care should be acknowledged, since similarity in age and sex does not establish group comparability.
The clinical basis for delabeling needs clearer description. Please distinguish diagnostic procedures verified in medical records from those recalled by patients, specify the antibiotics tested, and explain how allergy was excluded in patients who did not undergo drug provocation. The scope of label removal should also be clear, as exclusion of allergy to a particular penicillin is not interchangeable with evaluation of all beta-lactams. Additionally, report the time between delabeling and interview: assessments performed during 2017–2025 imply substantially different opportunities for antibiotic re-exposure.
Several numerical inconsistencies require correction. For example, 37 of 46 patients corresponds to 80.4%, not 74%, and 24 of 50 corresponds to 48%, not 54%. Please check all percentages against their denominators. Figure 4 also needs clarification because its caption refers to prescriptions, its axis refers to patients, and the displayed counts exceed the stated total. Multiple responses should be identified explicitly.
The post-delabeling findings should remain clearly identified as patient-reported. Three of 17 re-exposed patients described cutaneous symptoms, but these do not establish recurrent allergy or unsuccessful delabeling without further assessment. Please provide brief clinical details of these events and moderate the broader safety conclusions. Similarly, clarify whether the nine instances attributed to physician reluctance were independently verified. The hypothetical refusal reported by patients awaiting assessment should not be interpreted as persistent refusal among patients already delabelled.
The quality-of-life findings warrant cautious interpretation. The isolated Environment-domain difference should be discussed in relation to multiple comparisons and possible socioeconomic differences between groups. Nonsignificant findings in the remaining domains do not establish absence of an effect. Please ensure that the complete questionnaires, WHOQOL-BREF scoring methods, and detailed results are available in the supplementary material.

Finally, correct the figure cross-references, align figure captions with their contents, remove remaining template placeholders, and undertake focused language editing.

The study could provide useful exploratory evidence, but its contribution would be stronger with accurate numerical reporting, clearer clinical definitions, and conclusions limited to what this design can demonstrate.

Author Response

Response to Reviewer 3 Comments

The conclusions should reflect the cross-sectional design. The study compares different patients after delabeling and while awaiting assessment; it does not measure individual improvement. Please replace statements that delabeling “improves” perceived safety with wording describing the observed between-group difference. Selection bias and differences in previous reactions, comorbidities, and access to care should be acknowledged, since similarity in age and sex does not establish group comparability.

Thank you for this valuable comment. We changed the wording used throughout the manuscript, focusing on comparisons between delabelled and non-delabelled groups. (eg. lines 73-75).
We have expanded the limitations section, highlighting the observational and cross-sectional design of our study. We added one final sentence to the conclusions, highlighting the importance of prospective, longitudinal studies on larger patient cohorts to establish causal trajectories and long-term clinical impacts.
The clinical basis for delabeling needs clearer description. Please distinguish diagnostic procedures verified in medical records from those recalled by patients, specify the antibiotics tested, and explain how allergy was excluded in patients who did not undergo drug provocation. The scope of label removal should also be clear, as exclusion of allergy to a particular penicillin is not interchangeable with evaluation of all beta-lactams. Additionally, report the time between delabeling and interview: assessments performed during 2017–2025 imply substantially different opportunities for antibiotic re-exposure.
Thank you for pointing this out. We have added the information about the delabeling methods we use in our department (physician interview, skin testing + oral provocation testing) in the materials and methods section (section 4.1), line 347-349.
We changed the wording in paragraph 2.2
In paragraph 2.5 we added the data about time from delabeling to our study and beta-lactam antibiotics courses per person-year.
Several numerical inconsistencies require correction. For example, 37 of 46 patients corresponds to 80.4%, not 74%, and 24 of 50 corresponds to 48%, not 54%. Please check all percentages against their denominators. Figure 4 also needs clarification because its caption refers to prescriptions, its axis refers to patients, and the displayed counts exceed the stated total. Multiple responses should be identified explicitly.

Thank you for this valuable comment. The numerical data was checked and corrected. We have changed the caption of figure four, adding the information about possible multiple antibiotics groups prescribed for one patient during the allergy label.
The post-delabeling findings should remain clearly identified as patient-reported. Three of 17 re-exposed patients described cutaneous symptoms, but these do not establish recurrent allergy or unsuccessful delabeling without further assessment. Please provide brief clinical details of these events and moderate the broader safety conclusions. Similarly, clarify whether the nine instances attributed to physician reluctance were independently verified. The hypothetical refusal reported by patients awaiting assessment should not be interpreted as persistent refusal among patients already delabelled.

Thank you for this valuable comment. We have implemented the information about planned re-assessment in the future in those 3 cases to determine the nature of those symptoms in the discussion (lines 290-303). We added a statement to highlight, that rash or urticaria in this case does not necessarily establish recurrence of the original hypersensitivity. We omitted the data about hypothetical refusal in the future, and highlighted the actual post-delabeling follow up (figure 2, paragraph 2.5).
The quality-of-life findings warrant cautious interpretation. The isolated Environment-domain difference should be discussed in relation to multiple comparisons and possible socioeconomic differences between groups. Nonsignificant findings in the remaining domains do not establish absence of an effect. Please ensure that the complete questionnaires, WHOQOL-BREF scoring methods, and detailed results are available in the supplementary material.

Thank you for your comment. Official Polish translation of WHOQOL-BREF questionnaire was used in our work. ( Wołowicka, L.; Jaracz, K. Polska Wersja WHOQOL-100 i WHOQOL-BREF. ) The reference to this version is implemented in the methods section.
Finally, correct the figure cross-references, align figure captions with their contents, remove remaining template placeholders, and undertake focused language editing.

We have edited the figures numbering and captions. We corrected the possible language mistakes. One of the authors (Julia Glińska, MD) reviewed the manuscript linguistically, as she is a native speaker of the English language.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The clarity and readability of the manuscript have improved significantly. However, the Results and Discussion sections remain overly detailed and long, and the key messages could be presented more concisely.

The data shown in Figure 1 are already described in detail in the text (line 107). Therefore, Figure 1 appears redundant and could be removed.

Line 126-133
The interval between diagnosis of BL allergy and study participation is reported using both mean and median values. The authors should assess the data distribution and report only the summary measure that is statistically appropriate (mean for normally distributed data, median for non-normally distributed data). Reporting both measures for the same variable is not methodologically justified.

Author Response

The clarity and readability of the manuscript have improved significantly. However, the Results and Discussion sections remain overly detailed and long, and the key messages could be presented more concisely.

Thank you for the valuable comments regarding our manuscript. We have established more transparency of the results and discussion sections, by changing order of the paragraphs. Now the discussion corresponds to the previous data in result section. We have also improved the citation order.
The data shown in Figure 1 are already described in detail in the text (line 107). Therefore, Figure 1 appears redundant and could be removed.

Thank you for this constructive observation. We understand the concern regarding potential redundancy between the text and Figure 1. However, we respectfully suggest retaining Figure 1 in the main manuscript. We believe that this visual representation allows readers to rapidly grasp the core clinical trajectory and key epidemiological characteristics of our cohort at a glance, which is a central message of our study. Visualizing these data significantly enhances readability and accessibility compared to text alone.
Line 126-133
The interval between diagnosis of BL allergy and study participation is reported using both mean and median values. The authors should assess the data distribution and report only the summary measure that is statistically appropriate (mean for normally distributed data, median for non-normally distributed data). Reporting both measures for the same variable is not methodologically justified.

Thank you for this valuable comment.
We have changed the previously wrong wording in this section by replacing the word “initial diagnosis” with word “the day of beta-lactam allergy delabeling ” and corrected the results, by calculating the exact time between the date of successful delabeling and the exact day of interview for each individual patient, instead of the date of the last interview (November 4, 2025). We also used the median values with IQR, considering the non-normal distribution. (paragraph 2.5, lines 154-162)

Reviewer 2 Report

Comments and Suggestions for Authors

Dear authors,

thank you for having addressed all my suggestions. I have no futher comment to add. Congratulations!

Author Response

Dear authors,

thank you for having addressed all my suggestions. I have no futher comment to add. Congratulations!

We sincerely thank the Reviewer for their positive assessment, constructive feedback, and valuable time spent reviewing our manuscript. We are delighted that all concerns have been satisfactorily addressed.

Reviewer 3 Report

Comments and Suggestions for Authors

Thank you for revising the manuscript and responding to my comments. The revised wording regarding the cross-sectional design, the additional description of the diagnostic assessment, and the clarification that reported cutaneous symptoms do not necessarily indicate recurrent allergy are welcome improvements. However, several concerns remain unresolved, and some changes described in the response letter have not been consistently implemented in the manuscript.

Interpretation of between-group differences

Diagnostic assessment and follow-up

The revised methods provide additional information, but the specific antibiotics tested and the scope of allergy exclusion remain insufficiently clear. Please distinguish removal of a label concerning a particular penicillin from exclusion of allergy to all beta-lactams.

On page 4, the follow-up interval is described as the time between the initial diagnosis of allergy and study assessment. Please clarify whether this actually represents time since delabeling. Additionally, explain why follow-up was calculated to November 4, 2025, when interviews occurred over several months, and whether antibiotic exposure was captured through that endpoint for every participant.

Numerical consistency and Figure 4

Reported reactions and reluctance to use antibiotics

Quality-of-life findings

Adding a reference for the validated Polish WHOQOL-BREF does not resolve the concerns about multiple comparisons and potential socioeconomic confounding. Please address these issues and interpret the isolated Environment-domain finding cautiously.

Editorial consistency

The manuscript has improved, but these reporting and interpretation issues should be resolved before it can be accepted.

Author Response

Interpretation of between-group differences

The acknowledgment that the study cannot establish causality is appropriate. However, the limitations should more explicitly address potential selection bias and differences in previous reaction characteristics, comorbidities, and access to care. Similarity in age and sex alone does not establish comparability between the groups. Interpretations throughout the manuscript should remain consistent with the cross-sectional design.

Thank you for this important comment. We fully agree that baseline comparability extends beyond demographic variables such as age and sex, and that potential selection bias must be explicitly addressed.

We have substantially expanded the limitations section of discussion (lines 372-382) to address potential sources of selection bias, unmeasured confounding, and access to care. Specifically, we have highlighted that even though both patient cohorts were recruited from the same regional center and exhibited similar demographic profiles, unmeasured baseline differences such as prior allergic reaction severity, clinical comorbidities, individual motivation to undergo allergy testing, and regional variations in diagnostic access may have influenced group allocation and patient perception.

We have added the information about cross sectional character of the study in the last paragraph of introduction section (lines 74-77) and first paragraph of discussion section (line 214.)

Diagnostic assessment and follow-up

Thank you for pointing out this important issue.

We have changed the previously wrong wording in this section by replacing the word “initial diagnosis” with word “the day of beta-lactam allergy delabeling ” and corrected the results, by calculating the exact time between the date of successful delabeling and the exact day of interview for each individual patient, instead of the date of the last interview (November 4, 2025). We also used the median values with IQR, considering the non-normal distribution. (paragraph 2.5, lines 154-162)

Numerical consistency and Figure 4

The correction of 24/50 to 48% and the explanation of multiple responses are appreciated. Nevertheless, page 6, line 188, reports 56 patients comprising 21 delabeled and 25 non-delabeled patients; these numbers sum to 46. Furthermore, Figure 4 refers to “n = 56 prescriptions,” while the plotted values represent patients. Please reconcile the totals and clearly define the unit of analysis and denominators. Please also provide the absolute numerator underlying the revised 74% statement.

Thank you for this important comment.
We have corrected the data, as 35 non-delabelled patients were included in the analysis. We have changed the description of the figure 4. Out of 56 patients included in the analysis, 43 patients were prescribed an antibiotic from only one group, 12 patients from two groups and 1 patient from three groups, resulting in 70 prescriptions in total. We have provided the absolute numerators throughout the text of the manuscript, where needed.

Reported reactions and reluctance to use antibiotics

Although the discussion now acknowledges uncertainty regarding the three reported cutaneous reactions, the Figure 2 caption still describes them as “hypersensitivity recurrence.” Please consistently describe these as patient-reported symptoms unless recurrence has been clinically confirmed. The absence of severe reported events in this small exposed subgroup should also be interpreted cautiously.

Thank you for this important addition.
We have changed the caption of Figure 2 omitting the overstatement of “hypersensitivity recurrence”, highlighting the self-reported character of the reports, and added more complex and cautious description of the mild skin reactions and absence of severe events. (lines 144-146)

The response letter states that hypothetical refusal data were removed; however, this finding remains on page 8, lines 244–247. Please remove it or clearly distinguish it from observed behavior after delabeling. Statements about persistent patient reluctance should be supported by corresponding results. Please also clarify whether the reasons for physicians withholding beta-lactams were independently verified or based exclusively on patient reports.

Thank you for this valuable comment.
We have changed the wording in this paragraph, and added information in the results section, that 15 out of 46 delabelled patients (32.6%) still rated perceived safety at 5 points or lower. We omitted the statement about sociological and economical reason for this result and the persistent patient reluctance, as we were not measuring those factors using standardized methods. (lines 249-254). We have also corrected the information about patient-reported character of the data considering physicians reluctance considering prescription after confirmed delabeling (lines 148-152).

Quality-of-life findings

The reported Environment-domain difference (p = 0.029) also conflicts with the unqualified statement in the conclusions that no significant quality-of-life differences were observed. Please ensure consistency across the abstract, results, discussion, and conclusions. The complete supplementary results and questionnaires should be available for review, with sufficient information about scoring and missing-data handling.

Thank you for this important addition to our discussion section.
We have added the information about substantial caution that needs to be considered interpreting this isolated result, to strengthen our interpretation that no significant quality-of-life differences were observed. (lines 270-284).

Editorial consistency

Please remove the remaining supplementary-material template text on page 14 and irrelevant template abbreviations. Reference numbers 31, 38, 39, and 40 are duplicated and require correction together with their corresponding in-text citations. Further language editing is also necessary.

Thank you for this valuable comment.
We removed the remaining supplementary material template text and abbreviations.
We have rearranged the discussion section to correspond in order to the results section. We have changed the references order and thoroughly corrected the numbers avoiding duplications. We have implemented new citation of a work of Nguyen et. al considering patient reported outcomes (PROMs) and their limitations. We have added a paragraph in discussion section about alternative use of antibiotics during allergy label period, corresponding to results section, including citation of Vasquez et al., to compare our results with another study describing antibiotic use during the label