Rural Residence and One-Person Households Are Associated with Diagnostic Delay in Pulmonary Tuberculosis in a Low-Incidence European Setting

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

General Comments

This manuscript investigates factors associated with diagnostic delay among adults with pulmonary tuberculosis in Croatia, with emphasis on geographic and social determinants. The topic is relevant for tuberculosis control strategies in low-incidence settings and addresses an important public health issue.

The study presents potentially valuable findings, particularly regarding rural residence and partnership status as predictors of diagnostic delay. However, several methodological aspects require clarification, and some conceptual inconsistencies - especially regarding the use of the term social isolation - should be addressed before the manuscript can be considered for publication.

In addition, improvements are needed in the definition of outcomes, description of data collection procedures, interpretation of findings, and contextualization of results within the broader epidemiological literature.

 

Abstract

Objective: The study aims to evaluate causes of diagnostic delay in adults with confirmed tuberculosis in Croatia.

Study design: Retrospective observational study.

Outcome: Time between symptom onset and initiation of treatment.

Results: The study reports diagnostic delay among rural residents and retired individuals living without a partner.

Conclusion: The manuscript refers to social isolation, but the variable evaluated corresponds only to the absence of a partner. Therefore, the use of the term social isolation is not appropriate.

 

Methodology

Adult patients diagnosed with tuberculosis and treated between 2019 and 2022 in Croatia were included. However, ethical approval for the study was granted only in 2021. It is unclear whether the questionnaires mentioned in the manuscript were applied after treatment or retrospectively. It is also unclear when healthcare professionals were interviewed. Since healthcare professionals manage many patients, this represents a potential source of recall bias and should be clarified. Consider removing this information if it cannot be properly justified.

Regarding patient-reported data, it is unclear whether the information about symptom onset was obtained from medical records or from questionnaires applied directly to patients. The manuscript should clearly distinguish which variables were collected from medical records and which were obtained through questionnaires.

There are inconsistencies in formatting within the methodology section (e.g., spacing before paragraphs and between section titles and initial paragraphs).

The authors evaluated partnership status but did not collect additional information about household composition or patients’ social interactions with relatives or friends. Therefore, the use of the term social isolation, including in the title, is not appropriate, since this variable was neither assessed nor explored in the study.

The manuscript defines Total Diagnostic Delay (TDD) as the time between symptom onset and initiation of tuberculosis treatment. This definition requires clarification, as in several tuberculosis delay frameworks, this interval is typically referred to as total delay. The authors should clarify which conceptual framework was adopted and ensure consistency with established terminology.

The authors define extreme delay as the 75th percentile of the maximum length of delay and cite reference (8). However, it is unclear whether this definition refers specifically to TDD or to another outcome. Reference (8) describes two delay components: patient delay (PD), defined as the interval between symptom onset and first consultation, and health system delay (HSD), defined as the interval between first consultation and initiation of tuberculosis treatment. The cited reference also states that delays longer than the median were classified as prolonged delays, and delays exceeding the 75th percentile were classified as extreme delays. The manuscript should clarify how this classification was applied in the present study.

 

Results

The titles of the tables could be improved. Since this is an epidemiological study, it would be useful to specify the study population more clearly (e.g., n = 116 adults diagnosed between 2019 and 2022).

In Table 2, the formatting of the percentage symbol (%) is incorrect in the last row.

Overall, the description of the results is adequate; however, the text could be made more concise and objective by reducing the repetition of terms already defined by abbreviations.

 

Discussion

Formatting should be revised throughout the section, particularly spacing and punctuation.

The manuscript does not sufficiently discuss the magnitude and public health implications of diagnostic delay, such as increased transmission, disease severity, or mortality risk. The discussion would benefit from better contextualization of the findings and clearer emphasis on their epidemiological relevance.

The type of diagnostic approach is important when evaluating factors associated with treatment delay in tuberculosis. It would be useful to clarify whether the diagnosis was based on screening or confirmatory testing, since diagnostic pathways influence treatment initiation time. In some countries, empirical treatment may begin before microbiological confirmation, and this should be discussed.

The manuscript uses the term social isolation incorrectly, since only partnership status was evaluated. The discussion should instead address the role of living with a partner in tuberculosis diagnosis and treatment (e.g., adherence and persistence during treatment). Social context was not fully assessed, as there is no information about the number of people living in the same household or family structure. Considering that tuberculosis is a communicable disease, discussion of these aspects would be relevant.

The suggested protective role of partnership status should be interpreted cautiously given the observational design of the study.

The discussion regarding the association between the absence of chest pain and shorter TDD, and the presence of chest pain with prolonged delay, is limited. The authors suggest referral to investigations for malignancy or cardiovascular disease, but do not explore similarities and differences between these diagnostic pathways. A deeper discussion of symptom presentation could contribute to improving clinical recognition of tuberculosis. The unexpected association between chest pain and prolonged diagnostic delay would benefit from further discussion, including possible referral bias or alternative diagnostic pathway explanations.

Although other evaluated variables (e.g., sex, age) were not statistically significant, their relevance in tuberculosis epidemiology should still be discussed. The authors should consider whether associations with these variables have been reported in previous studies and why they may not have been observed in the present population.

It would also be valuable to propose strategies for improving early tuberculosis diagnosis, including public health interventions (screening campaigns), improved notification systems, primary care diagnostic algorithms, digital triage tools, or mobile tuberculosis units. These approaches are already implemented in high-incidence settings and could enrich the discussion.

 

References

Of the 32 references cited, 16 were published more than five years ago, corresponding to approximately 50% of the reference list. The authors should evaluate whether more recent literature could be incorporated, particularly in the Discussion section.

 

Author Response

Reviewer 1.

General Comments

This manuscript investigates factors associated with diagnostic delay among adults with pulmonary tuberculosis in Croatia, with emphasis on geographic and social determinants. The topic is relevant for tuberculosis control strategies in low-incidence settings and addresses an important public health issue.

The study presents potentially valuable findings, particularly regarding rural residence and partnership status as predictors of diagnostic delay. However, several methodological aspects require clarification, and some conceptual inconsistencies - especially regarding the use of the term social isolation - should be addressed before the manuscript can be considered for publication.

In addition, improvements are needed in the definition of outcomes, description of data collection procedures, interpretation of findings, and contextualization of results within the broader epidemiological literature.

RESPONSE: We thank the reviewer for this thoughtful and constructive overall assessment of our manuscript. We appreciate the recognition of the relevance of the topic and the potential value of our findings for tuberculosis control in low-incidence settings.

We also thank the reviewer for highlighting the areas requiring improvement. In response, we have carefully revised the manuscript throughout. Several methodological sections have been substantially clarified, including the definitions of all delay-related outcomes, study procedures, data sources, and analytical approach. The Methods section now more clearly distinguishes information obtained from structured patient questionnaires, medical records, and registry documentation.

Regarding the conceptual concern about the use of the term “social isolation,” we fully agree that the original wording was imprecise. After re-evaluating the dataset and source questionnaire, we clarified that the measured variable referred to household composition (one-person household vs. multi-person household), rather than social isolation or partnership status. Accordingly, this terminology has been corrected consistently throughout the title, Abstract, Methods, Results, Tables, and Discussion.

We have also expanded the Discussion section to improve interpretation of the findings and to better contextualize the results within the broader epidemiological literature, particularly in relation to tuberculosis diagnostic delay in low-incidence European settings, rural healthcare access, ageing populations, and changing household structures.

For the reviewer’s convenience, all revisions made in response to these comments have been highlighted in red throughout the revised manuscript to facilitate review.

We believe these revisions have substantially strengthened the manuscript and addressed the reviewer’s important concerns.

Abstract

Objective: The study aims to evaluate causes of diagnostic delay in adults with confirmed tuberculosis in Croatia.

Study design: Retrospective observational study.

Outcome: Time between symptom onset and initiation of treatment.

Results: The study reports diagnostic delay among rural residents and retired individuals living without a partner.

Conclusion: The manuscript refers to social isolation, but the variable evaluated corresponds only to the absence of a partner. Therefore, the use of the term social isolation is not appropriate.

RESPONSE: We thank the reviewer for this important comment regarding the terminology used to describe the household-related variable. Upon re-evaluation of the original dataset and source questionnaire, we identified that the wording used in the manuscript (“living without partner”) did not accurately reflect the variable collected. The intended meaning referred to individuals living alone, i.e., residing in a one-person household, rather than marital or partnership status. This distinction was unfortunately lost during translation and manuscript preparation. To correct this, we have revised the manuscript throughout and replaced the previous terminology with the more accurate expression household composition (one-person household vs. multi-person household) or living in a one-person household, as appropriate. The title and Abstract and keywords (page 1, lines 2,16,25,30,31,34) has been updated accordingly. We agree that this revised terminology more precisely reflects the measured variable and avoids overinterpretation related to partnership status or social isolation.

 

Methodology

Adult patients diagnosed with tuberculosis and treated between 2019 and 2022 in Croatia were included. However, ethical approval for the study was granted only in 2021. It is unclear whether the questionnaires mentioned in the manuscript were applied after treatment or retrospectively. It is also unclear when healthcare professionals were interviewed. Since healthcare professionals manage many patients, this represents a potential source of recall bias and should be clarified. Consider removing this information if it cannot be properly justified.

Regarding patient-reported data, it is unclear whether the information about symptom onset was obtained from medical records or from questionnaires applied directly to patients. The manuscript should clearly distinguish which variables were collected from medical records and which were obtained through questionnaires.

RESPONSE: We thank the reviewer for these important comments and agree that the original manuscript did not sufficiently distinguish the sources of collected data. The Methods section has been substantially revised to clarify that all data were collected retrospectively from routinely available sources during the study period (December 2019 to December 2022), during which eligible consecutive patients were included until the predefined sample size required for statistical analysis was reached.

Specifically, a standardized structured questionnaire routinely completed by patients at hospital admission provided information on age, sex, place of residence, education level, employment status, household composition, place of birth, comorbidities, smoking and alcohol consumption, and the onset and type of presenting tuberculosis symptoms. Age, sex, comorbidities, and presenting symptoms were subsequently cross-checked using medical records and registry documentation. Data on microbiological findings and dates of first healthcare contact, tuberculosis diagnosis, and treatment initiation were obtained from medical records and national registry documentation. We also want to clarify that healthcare professionals were not interviewed. The previous wording was imprecise and has been removed from the revised manuscript. Finally, because symptom onset was based partly on patient self-report at admission, we have explicitly acknowledged the possibility of recall bias in the revised Methods and Limitations sections. Regarding ethical approval, approval was obtained in 2021 for the ongoing retrospective observational study using routinely collected data from the defined study period. Patients enrolled after approval provided written informed consent in accordance with institutional requirements. The Methods section has been revised to clearly distinguish the data sources for each variable and to improve transparency of the retrospective design.

Changes are evident in the page 3, paragraph 2.4.Data collection, lines 114-128

There are inconsistencies in formatting within the methodology section (e.g., spacing before paragraphs and between section titles and initial paragraphs).

RESPONSE: We thank the reviewer for noting these formatting inconsistencies. Spacing and paragraph formatting throughout the Methodology section have been carefully revised to ensure consistent presentation between section titles and text, as well as between paragraphs.

The authors evaluated partnership status but did not collect additional information about household composition or patients’ social interactions with relatives or friends. Therefore, the use of the term social isolation, including in the title, is not appropriate, since this variable was neither assessed nor explored in the study.

RESPONSE: We thank the reviewer for this valuable observation and agree that the term “social isolation” was not sufficiently precise based on the variable as originally described. After re-examining the original dataset and questionnaire, we identified that the variable did not assess partnership status or broader social relationships, but rather household composition, specifically whether the participant lived alone or in a multi-person household. The original wording in the manuscript did not adequately reflect this distinction. In response to the reviewer’s comment, we have revised the terminology throughout the manuscript. The term “social isolation” has been removed from the title, Abstract, Results, and Discussion, and replaced with the more accurate expressions one-person household, living alone, or household composition (one-person vs. multi-person household), depending on context. We fully agree that broader social isolation would require dedicated measures of social contact, support networks, or interpersonal relationships, which were not collected in this study. The revised manuscript now reflects only the measured household-level variable and avoids overinterpretation.

Changes are visible in page 3, paragraph 2.2. Predictors of delay , lines 90-92, and page 4, paragraph 2.5. Sample size and study power, lines 136.

The manuscript defines Total Diagnostic Delay (TDD) as the time between symptom onset and initiation of tuberculosis treatment. This definition requires clarification, as in several tuberculosis delay frameworks, this interval is typically referred to as total delay. The authors should clarify which conceptual framework was adopted and ensure consistency with established terminology.

The authors define extreme delay as the 75th percentile of the maximum length of delay and cite reference (8). However, it is unclear whether this definition refers specifically to TDD or to another outcome. Reference (8) describes two delay components: patient delay (PD), defined as the interval between symptom onset and first consultation, and health system delay (HSD), defined as the interval between first consultation and initiation of tuberculosis treatment. The cited reference also states that delays longer than the median were classified as prolonged delays, and delays exceeding the 75th percentile were classified as extreme delays. The manuscript should clarify how this classification was applied in the present study.

RESPONSE; We thank the reviewer for this valuable comment regarding the classification of delay outcomes. In response, we have revised the manuscript to align more closely with established tuberculosis delay frameworks. The primary outcome is now termed total delay (TotD), defined as the interval between symptom onset and treatment initiation. In addition, we now added and  define prolonged delay ( PrD) as delays longer than the median total delay of the study population, and extreme delay  (ED) as delays equal to or exceeding the 75th percentile of the distribution, consistent with previously published approaches. These revised definitions are now clearly presented in the Methods section, and terminology has been standardized throughout the manuscript.

Changes are visible in page 3, paragraph 2.3. Outcomes , lines 100-105, and page 4, 2.2 Statistical analysis, lines 158

 

Results

The titles of the tables could be improved. Since this is an epidemiological study, it would be useful to specify the study population more clearly (e.g., n = 116 adults diagnosed between 2019 and 2022).

RESPONSE: We thank the reviewer for this helpful suggestion. The titles of all tables have been revised to provide clearer epidemiological context by explicitly specifying the study population, sample size, and study period where appropriate. This has improved clarity and interpretability of the presented data.

Changes are evident in the - page 5, paragraph 3.1. Sociodemographic and clinical variables,  lines 170-171, 176-177, 181-182, also page 7, paragraph 3.2.1. Total delay, lines 222-223, and page 8, paragraph 3.2.3. Extreme delay, line 245-246

In Table 2, the formatting of the percentage symbol (%) is incorrect in the last row.

RESPONSE: We thank the reviewer for noting this formatting error. The percentage symbol in the last row of Table 2 has been corrected in the revised manuscript.

Overall, the description of the results is adequate; however, the text could be made more concise and objective by reducing the repetition of terms already defined by abbreviations.

RESPONSE: We thank the reviewer for this helpful suggestion. The Results section has been carefully revised to improve conciseness and readability. Repetitive wording has been reduced, and terms already defined by abbreviations are now used more consistently throughout the text to provide a clearer and more objective presentation of the findings. Furthermore, as requested by the other reviewer, we have added Table 4 and removed redundant results from the text.

Changes are evident in the - page 6, paragraph 3.2.1 Total delay -lines 186, 192, 193,196, 205, 212, and page 8, paragraph 3.2.3. Extreme delay, lines 236, 239, 241, 243, and paragraph 3.3. Secondary outcomes,linesss 256, 266,262,265, 266

 

Discussion

Formatting should be revised throughout the section, particularly spacing and punctuation.

RESPONSE: We thank the reviewer for noting these formatting inconsistencies. Spacing and paragraph formatting throughout the Result section have been carefully revised to ensure consistent presentation between section titles and text, as well as between paragraphs.

The manuscript does not sufficiently discuss the magnitude and public health implications of diagnostic delay, such as increased transmission, disease severity, or mortality risk. The discussion would benefit from better contextualization of the findings and clearer emphasis on their epidemiological relevance.

RESPONSE: We thank the reviewer for this valuable comment. The Discussion section has been expanded to better emphasize the magnitude and epidemiological relevance of diagnostic delay. We now explicitly discuss the potential consequences of prolonged delay, including extended infectiousness, increased community transmission, more advanced disease at presentation, and poorer clinical outcomes.

Changes are visibile in page 9, lines 279-283

The type of diagnostic approach is important when evaluating factors associated with treatment delay in tuberculosis. It would be useful to clarify whether the diagnosis was based on screening or confirmatory testing, since diagnostic pathways influence treatment initiation time. In some countries, empirical treatment may begin before microbiological confirmation, and this should be discussed.

RESPONSE We thank the reviewer for this important comment. We have clarified the diagnostic pathway in the revised manuscript. Patients were not detected through screening programs, but through routine clinical evaluation of symptomatic individuals. In our cohort, 85 patients had rapid direct microbiological confirmation, whereas 31 were confirmed subsequently by culture. We did not perform formal subgroup comparisons according to diagnostic modality, as this was beyond the primary scope of the present study and sample size was limited. However, we agree that diagnostic certainty and confirmation pathway may influence treatment initiation timing, and this issue has now been acknowledged in the Discussion section.

Changes are evident in Disscusion paragraph - page 9, lines 283-286, we have laso mentioned that in Methods part, paragraph : 2.1. Study design, setting, and population, lines 76-79

The manuscript uses the term social isolation incorrectly, since only partnership status was evaluated. The discussion should instead address the role of living with a partner in tuberculosis diagnosis and treatment (e.g., adherence and persistence during treatment). Social context was not fully assessed, as there is no information about the number of people living in the same household or family structure. Considering that tuberculosis is a communicable disease, discussion of these aspects would be relevant.

The suggested protective role of partnership status should be interpreted cautiously given the observational design of the study.

RESPONSE: We thank the reviewer for this thoughtful and important comment. We agree that the original terminology was imprecise. Following re-evaluation of the dataset, we clarified that the analysed variable reflected household composition (one-person vs. multi-person household) rather than partnership status or broader social isolation. Accordingly, the term “social isolation” has been removed throughout the manuscript. The Discussion section has been substantially revised to focus on the potential role of household composition in healthcare-seeking behaviour, timely symptom recognition, and practical support during the diagnostic process. We also incorporated the reviewer’s important point that, because tuberculosis is a communicable disease, household context may influence both early recognition and opportunities for household exposure when diagnosis is delayed. Finally, we agree that the association should be interpreted cautiously. The revised manuscript now explicitly states that detailed information on family structure, household crowding, and social support was not available, and that the observational design precludes causal inference or interpretation of a direct protective effect.

Changes can be detected in pages 9 and 10, lines 313-327

The discussion regarding the association between the absence of chest pain and shorter TDD, and the presence of chest pain with prolonged delay, is limited. The authors suggest referral to investigations for malignancy or cardiovascular disease, but do not explore similarities and differences between these diagnostic pathways. A deeper discussion of symptom presentation could contribute to improving clinical recognition of tuberculosis. The unexpected association between chest pain and prolonged diagnostic delay would benefit from further discussion, including possible referral bias or alternative diagnostic pathway explanations.

RESPONSE:

We thank the reviewers for these insightful comments regarding the association between chest pain and prolonged delay. We agree that the original discussion was too brief and that this finding requires cautious interpretation.

In the revised manuscript, we substantially expanded the Discussion section. We now emphasize that chest pain is a non-specific symptom that may initially prompt investigation for more common alternative diagnoses such as cardiovascular disease, pleural disorders, pulmonary embolism, or malignancy, potentially delaying consideration of tuberculosis. We also discuss possible referral bias and differences in diagnostic pathways in low-incidence settings.

Furthermore, we acknowledge that residual confounding cannot be excluded, including factors such as smoking history, immunosuppression, pain severity, and socioeconomic characteristics that were not fully captured in the model. We also added reverse causality as a possible explanation, whereby prolonged undiagnosed tuberculosis may itself lead to chest pain. Accordingly, the revised manuscript now interprets chest pain not as a causal determinant, but as a possible marker of more complex or prolonged diagnostic trajectories.

Changes are visible in page 10, lines 339-349

 

Although other evaluated variables (e.g., sex, age) were not statistically significant, their relevance in tuberculosis epidemiology should still be discussed. The authors should consider whether associations with these variables have been reported in previous studies and why they may not have been observed in the present population.

RESPONSE: We thank the reviewer for this important comment. The Discussion section has been revised to better contextualize sex-related differences in diagnostic delay. We now acknowledge that longer delays among women have been reported in some settings and are often attributed to social and structural barriers rather than biological factors. At the same time, we note that such patterns are not universal. In our cohort, no significant sex-related differences were observed, which is consistent with earlier Croatian findings. We added that this may reflect the context of a high-income European setting with comparatively good healthcare accessibility and greater gender equity, where differences in healthcare-seeking barriers between women and men may be less pronounced. Changes are visible in page 9, lines 288-294

 

It would also be valuable to propose strategies for improving early tuberculosis diagnosis, including public health interventions (screening campaigns), improved notification systems, primary care diagnostic algorithms, digital triage tools, or mobile tuberculosis units. These approaches are already implemented in high-incidence settings and could enrich the discussion.

RESPONSE: We thank the reviewer for this valuable suggestion. The Discussion section has been expanded to include practical strategies for improving early tuberculosis diagnosis, including standardized primary care diagnostic pathways, targeted awareness campaigns, digital triage tools, mobile outreach services for rural populations, and strengthened notification systems. We also note that selected approaches used in higher-incidence settings may be adaptable to low-incidence countries where diagnostic delay persists. Changes are visble in page 10, lines 383-391

 

 

References

Of the 32 references cited, 16 were published more than five years ago, corresponding to approximately 50% of the reference list. The authors should evaluate whether more recent literature could be incorporated, particularly in the Discussion section.

 

RESPONSE: We thank the reviewer for this valuable comment. The reference list has been carefully updated  ( now 46 ) and expanded in the revised manuscript. Several more recent studies and reviews have been incorporated, particularly in the Discussion section, to better reflect current evidence on tuberculosis diagnostic delay, low-incidence settings, vulnerable populations, and contemporary public health approaches. Older references were retained where they remain foundational, nationally relevant, or necessary for comparison with earlier studies.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for this insightful paper on an important subject.

I have minor comments

1. Lines 55-56. References to the definitions of individual delay components will be useful here, or a brief text of the definitions.  In particular, add some indication of what extreme delay refers to.
2. Remove lines 58-66- these are guidance instructions and should not be part of the manuscript.
3. Lines 71-72. Add that there are people who had self-presented 
4. In the tables, N should be n
5. Table 3: Heading should be Symptom, not Comorbidity

Additional comment :

This topic, while not original, is very relevant for low TB incidence settings and also for high TB settings where TB incidence is declining- there is a real risk of missing TB  cases where there is no active attention to maintain a high index of suspicion or to consider TB early when patients present with TB-related symptoms.

Diagnostic delay is one of the factors hindering TB elimination and is generally frequently reported and seen as more relevant to high TB burden settings. This manuscript addresses a key gap in the field by raising awareness regarding this topic in low-incidence settings. This is critical to address what the authors termed the “low-incidence paradox.”  It raises critical awareness for low TB incidence settings, since delayed TB diagnosis can have poor outcomes that include mortality and transmission to other people, thus sustaining the pool of community transmission, which in turn hinders TB elimination efforts.

 

Author Response

Reviewer 2.

1. Do you consider the topic original or relevant to the field . This topic, while not original, is very relevant for low TB incidence

settings and also for high TB settings where TB incidence is declining-

there is a real risk of missing TB  cases where there is no active

attention to maintain a high index of suspicion or to consider TB early

when patients present with TB-related symptoms.

2. Does it address a specific gap in the field? Please also explain why

this is/ is not the case. Diagnostic delay is one of the factors hindering TB elimination and i generally frequently reported and seen as more relevant to high Tbburden settings. This manuscript addresses a key gap in the field by raising awareness regarding this topic in low-incidence settings. This is critical to address what the authors termed the “low-incidence

paradox.”  It raises critical awareness for low TB incidence settings,

since delayed TB diagnosis can have poor outcomes that include mortality

and transmission to other people, thus sustaining the pool of community

transmission, which in turn hinders TB elimination efforts.

 

RESPONSE: We thank the reviewer for the careful evaluation of our manuscript and for these clear and constructive comments. We appreciate the reviewer’s helpful suggestions, which have improved the clarity, accuracy, and presentation of the manuscript.

In response, we have revised the Introduction to better define the multidimensional components of tuberculosis-related delay and to clarify the meaning of prolonged and extreme delay. Editorial/template guidance text identified by the reviewer has been removed. We have also clarified in the Methods section that the study population consisted of self-presenting adult patients who entered routine clinical care because of symptoms and were subsequently diagnosed with microbiologically confirmed pulmonary tuberculosis.

In addition, all table formatting issues noted by the reviewer have been corrected, including standardization of lowercase “n” notation and revision of the Table 3 heading from “Comorbidity” to “Symptom.”

For the reviewer’s convenience, all changes made in response to these comments have been highlighted in blue throughout the revised manuscript.

We sincerely thank the reviewer for these valuable comments, which helped improve the quality and readability of the manuscript.

 

1. Lines 55-56. References to the definitions of individual delay components will be useful here, or a brief text of the definitions.  In particular, add some indication of what extreme delay refers to.

RESPONSE: We thank the reviewer for this helpful suggestion. A brief explanatory sentence has now been added to the Introduction section to provide early conceptual context regarding the multidimensional components of tuberculosis-related delay. We also clarified that the study additionally examined prolonged delay, defined as total delay above the median, and extreme delay, referring to patients in the highest quartile of total delay. Detailed operational definitions remain provided in the Methods section.

Changes are visible in page 2, lines 57-60

1. Remove lines 58-66- these are guidance instructions and should not be part of the manuscript.

RESPONSE: We have removed guidence instructions

1. Lines 71-72. Add that there are people who had self-presented 

RESPONSE: We thank the reviewer for this helpful suggestion. The Study Design and Population section has been revised to clarify that the cohort consisted of self-presenting adult patients who entered routine clinical care because of symptoms and were subsequently diagnosed with microbiologically confirmed pulmonary tuberculosis.

Changes are visible in page 2, 2.1. Study design, setting, and population, line 73

 

1. In the tables, N should be n

RESPONSE: We thank the reviewer for noting this formatting issue. The table notation has been revised throughout the manuscript, and uppercase “N” has been replaced with lowercase “n” where appropriate.

1. Table 3: Heading should be Symptom, not Comorbidity

RESPONSE: We thank the reviewer for identifying this error. The heading of Table 3 has been corrected from “Comorbidity” to “Symptom” in the revised manuscript.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

I would like to congratulate the authors on this manuscript. The evaluation of this topic provides valuable support for the management of a significant public health problem.

I have the following comments:

 The Introduction section contains typographical errors: 40.  It also appears that a portion of the text is template material from the journal's submission guidelines. 53-66 Furthermore, the rationale presented in the first paragraph requires, at a minimum, revision. The manuscript lacks references to studies that would justify the importance of this topic in the national and European context, despite such literature being available for citation.

The authors should include in the Methodology section precise and unambiguous operational definitions for each component of the delay. Ideally, these definitions should be supported by citing an international standard, if one was applied. Additionally, the manuscript does not state how missing data, if any, were handled.

The narrative text in this section is difficult to read due to the inclusion of extensive statistical data. I suggest that the authors move these detailed statistics to tables and simply refer to the tables in the text. This would significantly improve readability.

The phrasing chosen by the authors regarding the link between chest pain and diagnostic delay could be confusing to the reader. Presenting it as an "unexpected finding" is certainly a possibility, but the explanation is difficult to fully grasp as currently written:

270 -273 “An unexpected finding was the association between chest pain and prolonged TDD. Chest pain may initially prompt evaluation for alternative conditions such as malignancy or cardiovascular disease, diverting patients into non-tuberculosis pathways.”

 The Conclusions could be rephrased to more strongly emphasize the central paradox highlighted by the manuscript: that in developed countries, better living conditions can paradoxically lead to longer diagnostic delays for tuberculosis.

Comments on the Quality of English Language

The manuscript presents important research, but its overall clarity is hampered by non-native English phrasing. While the content is strong, the text lacks fluency, which may cause confusion and require the reader to re-read sections to grasp the intended meaning. I strongly recommend that the manuscript be professionally edited by a native English speaker to improve its readability and impact.

Author Response

Reweiver 3.

We sincerely thank the reviewer for the kind and encouraging comments. We greatly appreciate the recognition of the relevance of our study and its potential contribution to the management of this important public health issue. Your positive assessment is highly valued and all revisions made in response to your comments have been highlighted in green throughout the revised manuscript for ease of review.

Open Review

(Comments and Suggestions for Authors

I would like to congratulate the authors on this manuscript. The evaluation of this topic provides valuable support for the management of a significant public health problem.

I have the following comments:

 The Introduction section contains typographical errors: 40.  It also appears that a portion of the text is template material from the journal's submission guidelines. 53-66 Furthermore, the rationale presented in the first paragraph requires, at a minimum, revision. The manuscript lacks references to studies that would justify the importance of this topic in the national and European context, despite such literature being available for citation.

RESPONSE: We thank the reviewer for these valuable comments. The Introduction section has been carefully revised and proofread to correct typographical and formatting errors. Any residual template text from the journal submission format has been removed.

In addition, the rationale for the study has been substantially strengthened. We have expanded the background section by incorporating references to studies from European low-incidence countries reporting persistent tuberculosis diagnostic delays, and by improving the national context for Croatia. The revised Introduction now emphasizes the relevance of demographic ageing, increasing numbers of one-person households, migration, and potential healthcare access inequalities as factors that may hinder timely tuberculosis recognition in a low-incidence setting.

We believe these revisions have improved both the clarity and justification of the study objectives. Changes are visible in page 2, lines 47-53.

 

The authors should include in the Methodology section precise and unambiguous operational definitions for each component of the delay. Ideally, these definitions should be supported by citing an international standard, if one was applied. Additionally, the manuscript does not state how missing data, if any, were handled.

RESPONSE: We thank the reviewer for this important methodological comment. The Methodology section has been revised to provide clear and explicit operational definitions for all delay components, including total delay, patient delay, health system delay, diagnostic delay, treatment delay, prolonged delay, and extreme delay. These definitions are now presented in the Outcomes section and supported by previously published tuberculosis delay frameworks cited in the manuscript.

We have also clarified data completeness. No missing data were present for the variables included in the final analyses; therefore, no imputation procedures or case exclusions due to missing values were required.

Changes are visible in page 3, lines 101-105, and page 4 ; lines 152-153.

 

The narrative text in this section is difficult to read due to the inclusion of extensive statistical data. I suggest that the authors move these detailed statistics to tables and simply refer to the tables in the text. This would significantly improve readability.

 

RESPONSE: We thank the reviewer for this helpful suggestion. The Results section has been revised to improve readability by substantially shortening the narrative text and removing extensive statistical detail from the main body of the manuscript. A new table ( Table 4) has been added to present the detailed univariable statistical results, while the text now focuses on the principal findings and refers readers to the relevant tables for complete data.

 

The phrasing chosen by the authors regarding the link between chest pain and diagnostic delay could be confusing to the reader. Presenting it as an "unexpected finding" is certainly a possibility, but the explanation is difficult to fully grasp as currently written:

270 -273 “An unexpected finding was the association between chest pain and prolonged TDD. Chest pain may initially prompt evaluation for alternative conditions such as malignancy or cardiovascular disease, diverting patients into non-tuberculosis pathways.”

 The Conclusions could be rephrased to more strongly emphasize the central paradox highlighted by the manuscript: that in developed countries, better living conditions can paradoxically lead to longer diagnostic delays for tuberculosis.

 

RESPONSE We thank the reviewer for these thoughtful comments. The Discussion section has been revised to provide a clearer and more cautious interpretation of the association between chest pain and prolonged delay, emphasizing that chest pain may lead to evaluation for more common alternative diagnoses and may represent a marker of more complex diagnostic pathways rather than a direct causal factor.

We have also revised the Conclusions to more clearly emphasize the paradox of tuberculosis control in developed low-incidence settings: that successful reductions in disease burden and improved living conditions may inadvertently lower clinical suspicion and contribute to delayed diagnosis.

Changes  are visible in page 10, lines 339-349, and page 11 lines 393-397

Comments on the Quality of English Language

The manuscript presents important research, but its overall clarity is hampered by non-native English phrasing. While the content is strong, the text lacks fluency, which may cause confusion and require the reader to re-read sections to grasp the intended meaning. I strongly recommend that the manuscript be professionally edited by a native English speaker to improve its readability and impact.

RESPONSE: We thank the reviewer for this constructive comment. The manuscript has been carefully revised throughout to improve grammar, clarity, and overall readability. In addition, the text underwent comprehensive language editing using professional English-language assistance, and all sections were further reviewed by the authors to ensure accuracy and fluency. We believe these revisions have substantially improved the quality and clarity of the manuscript.

Author Response File: Author Response.docx

Reviewer 4 Report

Comments and Suggestions for Authors

This manuscript addresses an important and timely topic, namely diagnostic delay in pulmonary tuberculosis within a low-incidence European setting, with particular emphasis on social determinants such as rural residence and social isolation. However, several issues need to be addressed before the manuscript can be considered for publication.

• The definition of "extreme delay" as ≥75th percentile of total diagnostic delay appears arbitrary and lacks generalizability, as the 75th percentile may vary significantly across different populations or sample groups. A stronger justification for this threshold is needed, ideally supported by existing literature or clinical relevance. Alternatively, it would be beneficial to consider using clinically meaningful cut-offs (e.g., >90 or >120 days) or to discuss how the chosen threshold might influence the interpretation of the results.
• The timing of symptom onset is based on patient self-report. This introduces a substantial risk of recall bias, which may affect all delay-related outcomes. The limitations section should be expanded accordingly.
• The finding that chest pain is associated with a longer diagnostic delay is counterintuitive and requires more careful interpretation. This association may reflect several complex factors that merit critical analysis. The following explanations are possible:
  • Chest pain is a symptom that often raises concerns about serious conditions such as cancer or cardiovascular diseases. As a result, there may be a diagnostic diversion toward these conditions, delaying the diagnostic process for tuberculosis.
  • There may be unaccounted-for variables in the model, such as the severity of chest pain, the patient's clinical history (e.g., smoking, immunosuppression), or other socio-demographic factors that influence both the likelihood of delayed diagnosis and the perception of pain. If these factors are not adequately controlled for, they could distort the observed association.
  • Another possible explanation is reverse causality. It is possible that a delayed diagnosis of an underlying condition, such as tuberculosis, might cause symptoms like chest pain, rather than chest pain being the cause of diagnostic delay. In other words, chest pain could be a sign of an already advanced disease that has not been diagnosed in time, leading to delays in treatment.
• Editorial instructions are still present in the Introduction (e.g., “The introduction should briefly place the study…”). These should be removed.
• Duplicate wording (e.g., “Correspondence: Correspondence:”) should be corrected.
• In Table 2, the denominator (N=44) should be clearly explained (patients with comorbidities only?).
• Table 3 appears mislabeled (“Comorbidity” instead of “Symptoms”). The denominator (N=44) is inconsistent with rows (e.g., patients reporting cough were 111).
• Row n. 174 is unclear: “The analysis was initially using log-transformed values…” (missing verb?).

Author Response

Reviewer 4.

This manuscript addresses an important and timely topic, namely diagnostic delay in pulmonary tuberculosis within a low-incidence European setting, with particular emphasis on social determinants such as rural residence and social isolation. However, several issues need to be addressed before the manuscript can be considered for publication.

We thank the reviewer for the careful evaluation of our manuscript and for recognizing the importance and timeliness of the topic. We appreciate the reviewer’s constructive comments and valuable suggestions for improvement.

In response, we have carefully revised the manuscript to address all concerns raised. Methodological aspects have been clarified, definitions of delay-related outcomes have been refined, interpretation of key findings has been strengthened, and editorial inconsistencies have been corrected throughout the text. We also revised the terminology related to household and social determinants to ensure that it accurately reflects the variables measured in the study and avoids conceptual overinterpretation.

These revisions have substantially improved the clarity, rigor, and overall quality of the manuscript. For the reviewer’s convenience, all changes made in response to these comments have been highlighted in orange throughout the revised manuscript.

We sincerely thank the reviewer for the thoughtful feedback, which has helped strengthen the manuscript considerably.

 

• The definition of "extreme delay" as ≥75th percentile of total diagnostic delay appears arbitrary and lacks generalizability, as the 75th percentile may vary significantly across different populations or sample groups. A stronger justification for this threshold is needed, ideally supported by existing literature or clinical relevance. Alternatively, it would be beneficial to consider using clinically meaningful cut-offs (e.g., >90 or >120 days) or to discuss how the chosen threshold might influence the interpretation of the results.

RESPONSE: We thank the reviewer for this important comment. In the revised manuscript, we clarified that extreme delay was defined a priori as total delay equal to or exceeding the 75th percentile of the study distribution in order to identify patients experiencing substantially prolonged delays, consistent with approaches used in previous tuberculosis delay studies 

In response to reviewer feedback, we additionally introduced prolonged delay, defined as total delay above the median, to improve interpretability and to identify a broader subgroup with delays exceeding the typical experience of the cohort. We agree that percentile-based thresholds may vary across populations and are not directly interchangeable between settings. Therefore, these measures were used as relative within-cohort indicators rather than universal clinical cut-offs. This limitation and its implications for interpretation have now been acknowledged in the manuscript.

Changes are visible in page 3, paragraph 2.3. Outcomes, lines 101-105

 

• The timing of symptom onset is based on patient self-report. This introduces a substantial risk of recall bias, which may affect all delay-related outcomes. The limitations section should be expanded accordingly.

RESPONSE:

We thank the reviewer for this important comment. The Limitations section has been expanded to more explicitly acknowledge the possibility of recall bias related to self-reported symptom onset. We now state that inaccurate recall may have influenced estimates of total delay and its individual components.

Changes can be detected in page 11, paragraph Disscusion, lines 376-382

 

• The finding that chest pain is associated with a longer diagnostic delay is counterintuitive and requires more careful interpretation. This association may reflect several complex factors that merit critical analysis. The following explanations are possible:
  • Chest pain is a symptom that often raises concerns about serious conditions such as cancer or cardiovascular diseases. As a result, there may be a diagnostic diversion toward these conditions, delaying the diagnostic process for tuberculosis.
  • There may be unaccounted-for variables in the model, such as the severity of chest pain, the patient's clinical history (e.g., smoking, immunosuppression), or other socio-demographic factors that influence both the likelihood of delayed diagnosis and the perception of pain. If these factors are not adequately controlled for, they could distort the observed association.
  • Another possible explanation is reverse causality. It is possible that a delayed diagnosis of an underlying condition, such as tuberculosis, might cause symptoms like chest pain, rather than chest pain being the cause of diagnostic delay. In other words, chest pain could be a sign of an already advanced disease that has not been diagnosed in time, leading to delays in treatment.

RESPONSE: We thank the reviewers for these insightful comments regarding the association between chest pain and prolonged delay. We agree that the original discussion was too brief and that this finding requires cautious interpretation.

In the revised manuscript, we substantially expanded the Discussion section. We now emphasize that chest pain is a non-specific symptom that may initially prompt investigation for more common alternative diagnoses such as cardiovascular disease, pleural disorders, pulmonary embolism, or malignancy, potentially delaying consideration of tuberculosis. We also discuss possible referral bias and differences in diagnostic pathways in low-incidence settings. Furthermore, we acknowledge that residual confounding cannot be excluded, including factors such as smoking history, immunosuppression, pain severity, and socioeconomic characteristics that were not fully captured in the model. We also added reverse causality as a possible explanation, whereby prolonged undiagnosed tuberculosis may itself lead to chest pain. Accordingly, the revised manuscript now interprets chest pain not as a causal determinant, but as a possible marker of more complex or prolonged diagnostic trajectories.

Changes are visible in page 10, paragraph Disscusion, lines 339-349

• Editorial instructions are still present in the Introduction (e.g., “The introduction should briefly place the study…”). These should be removed.

RESPONSE; We thank the reviewer for identifying this oversight. Residual editorial/template instructions in the Introduction section have been removed in the revised manuscript.

• Duplicate wording (e.g., “Correspondence: Correspondence:”) should be corrected.

RESPONSE: We thank the reviewer for noting this formatting error. The duplicated wording in the correspondence section has been corrected in the revised manuscript.

Changes are visible in page1, lines 8

• In Table 2, the denominator (N=44) should be clearly explained (patients with comorbidities only?).

RESPONSE: We thank the reviewer for this helpful comment. Clarifying text has now been added below Table 2 stating that n = 44 refers to participants with at least one recorded comorbidity, and that percentages were calculated within this subgroup..

• Table 3 appears mislabeled (“Comorbidity” instead of “Symptoms”). The denominator (N=44) is inconsistent with rows (e.g., patients reporting cough were 111).

RESPONSE: We thank the reviewer for identifying this error. Table 3 was mislabeled in the previous version, and the denominator was incorrect. This has now been corrected in the revised manuscript: the heading has been changed to “Symptoms,” and the denominator has been revised to n = 116, corresponding to the total number of study participants.

• Row n. 174 is unclear: “The analysis was initially using log-transformed values…” (missing verb?).

RESPONSE: We thank the reviewer for noting this unclear wording. The sentence in the Results section has been revised for grammatical accuracy and clarity. It now states that regression analyses were performed using log-transformed TotD values to account for skewness. Changes are visible in page 7, line 210.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have addressed the suggested revisions, resulting in a manuscript that is clearer and more coherent. Specifically, (a) the term “social isolation” was appropriately replaced with “one-person household”, which better reflects the context presented; and (b) Total Diagnostic Delay (TDD) was replaced with total delay (TotD), which more accurately represents the analysis performed.

Furthermore, the ethical issues previously raised have been clarified and are now clearly described in the manuscript.

Reviewer 4 Report

Comments and Suggestions for Authors

My comments have been addressed