Response to comments: 14.11.2014 Journal: BMC Family Practice Manuscript ID: 1534478001341961 Date submitted by the author: Authors: Al-ani, Salwan; Spigt, Mark; Laue, Johanna; Melbye, Hasse Title: Predictors for treatment with antibiotics and systemic corticosteroids in acute exacerbation in a cohort of primary care patients diagnosed with asthma or chronic obstructive pulmonary disease Response: Thank you Ms Eloisa Nolasco for your feedback. Here is our response to your comments. ………………………………………………………………………………………………….. Editor´s Comments to Author Comment 1: Again, please refer in the discussion to asthma and how asthma exacerbation should be managed differently than COPD exacerbation. Asthma should be mentioned in the main findings and it should be discussed in the following section separately or you make a strong argument that asthma and COPD are basically the same. Response: The new analyses based on the GP’s diagnosis have made it easier to compare the “COPD patients” with “asthma patients”. This is now mentioned in the main findings (but not in tables). The patients with both diagnoses are more frequently treated with both kinds of medication than those registered with only one diagnosis. Chest findings are the strongest predictors of systemic corticosteroids among the asthma patients. That corticosteroids treatment is in accordance with the GINA guidelines is now included in the discussion. Comment 2: The low number (< 5) of patients in some categories is violating the assumptions of Chi2_Test. Please use Fisher’s exact test as suggested by Micheal Mehring. Response: This has now been done. Comment 3: Did you attempt a multivariate analysis? I am aware that the number of subjects is low for multivariate analysis, but given that the clinical decision to treat with Abx or steroids in practice is generally is generally based on a combination of findings, your univariate analysis of the data is of limited interest. Response: Yes, multivariate analysis has now been included (table 5 and table 6). Lumping the chest findings together made the number of variables to some degree acceptable. Comment 4: Rephrase line 263 The GPs emphasized chest findings together with CRP rapid test and pulse oximetry more strongly than symptoms when deciding type of treatment. Some would argue that chest findings are symptoms. Response: We have changed it into: The GPs relied on the findings by physical examination, together with CRP rapid test and pulse oximetry more strongly than respiratory symptoms when deciding type of treatment. (the last paragraph of the discussion). …………………………………………………………………………………………………... Reviewers' Comments to Author Reviewer 1: Michael Mehring Major comments: Comment 1: The presented prescription rates of antibiotics and systemic corticosteroids from Figure 2 are not comprehensible with the mentioned numbers in the text (Line 157-161). Additionally, it would be easier to understand if these results are also presented in table 2. Response: The text in the second paragraph of the results has been changed and the legend to figure 2 has also been improved. Comment 2: The percentages in Table 3 and 4 are not understandable, please check the provided information or explain the percentages in the text. Response: The figure text has been changed to make the dominator in the percentages more clear. However, the denominators cannot be presented, but they can be found in table 2. Comment 3: The assumption that hospitalized patients received both antibiotics and systemic corticosteroids has to be addressed as a limitation in the discussion/limitation section. Response: We have added hospitalization in the previous year as a separate variable in table 2; the hospitalized patients are no longer added to the categories of treatment in the previous year. Comment 4: Due to the small numbers in the groups might be a different test than the Chi2 appropriated (for example Fishers exact). Response: We now use the Fisher’s Exact Test. Comment 5: There is no mention why the inclusion criterion over 40 years was chosen, especially as this trial investigates in asthma which could lead hereby to an underrepresentation of this population. Please mention the cause for this age. (Homogeneity of the groups?) Response: We chose the age limit to have a considerable proportion of COPD patients in the sample. Comment 6: Please provide a mean or median age of the groups (in text or table). Response: Mean age of the main subgroups has been added (first paragraph of the results). Comment 7: As it is correctly described in the limitation section (254-260) that the used classification with FEV1/FVC is not crystal clear, it is not correct to interpret a group with COPD or Asthma. Please formulate this interpretation with more caution throughout the paper. Response: The definition of the COPD group is based on spirometry, and we think this has been made clear. Patients with asthma as the primary disease are nevertheless probably represented among the patients in this group. We now use the expression “spirometry indicating COPD” to remind the reader that the “COPD” is not clear-cut. Comment 8: Please highlight in the table headings that the FEV1/FVC is “postbronchodilator test”. Response: The recommended changes have been made. Comment 9: The classification of two groups under use of FEV1/FVC seems very practical and reflects the realistic primary care setting. It would be interesting to know if the results are the same when classifying after the already registered diagnoses from GPs instead of the FEV1/FVC in COPD/Asthma or both. Response: This has now been included. (in Results). …………………………………………………………………………………………………... Reviewer 2: Markus Bleckwenn Major Compulsory Revisions: Comment 1: In 59% of patients had the asthma in this study. Of these, 70% were also performed by family physicians as asthma patients. Therefore, I would also expect the first results concerning the asthmatics: What are the most important predictors of prescription antibiotics / cortisone in asthmatics? The answer to these questions should be further clarified at the beginning of the respective sections. Response: We have made new analysis based on GP’s diagnosis that addresses this comment. Comment 2: The discussion is conducted mainly on the results of the COPD patients. Generally I'm missing in the discussion of the possible clinical consequences due to the acquired data, especially for asthmatics. The discussions on possible follow-up studies I could find at the end of the discussion interesting. Response: Treatment of asthma has now got more attention in the discussion. Comment 3: In Figure 2 it can be seen that in 41% of asthma patients in the exacerbation of antibiotic was added. Since antibiotics according to the guidelines cited in the rule are not necessary, I would want an answer to the question in the discussion: Give your dates clues as how to reduce the rate of antibiotic prescriptions? Response: Our study cannot give advice to treatment; it just shows how GPs assess the patients. Comment 4: With the introduction is the clinical benefit of new predictors not understands. In my opinion the search for predictors especially in asthma therapy would be useful, since there are no clear criteria especially for antibiotic therapy, such as in COPD (Anthonisen) there. Perhaps the idea of the study could be better formulated in the introduction. Response: As mentioned, antibiotics are not recommended in the current guidelines in patients suffering from pure exacerbation of their asthma (without concurrent pneumonia or any other bacterial infection). Anthonisen criteria is one the factors which can help the physicians decide the type of treatment in patients with COPD exacerbation and as you can see in the results of our study it was not even significant and other factors, such as CRP, oxygen saturation and chest findings can be used alone, or together with Anthonisen criteria to decide the type of treatment in this group of patients. Discretionary Revisions: Comment 1: There are guidelines for the treatment of the two diseases asthma and COPD. A reasonable question would be how these guidelines are implemented in general practice. A direct comparison between the treatment and the guidelines would be interesting. Response: We have tried to give more attention to this question, referring also to GINA guidelines in the discussion.