Radiological Explorations of Patients with Upper or Febrile Urinary Tract Infection

Round 1

Reviewer 1 Report

It is an interesting article, which highlights the excess of imaging tests that are performed.

In the title, it is not quite correct to write retrospective cohort study, since cohort study is by definition prospective.

In Figure 1, the first table indicates 103 patients and the second 120 patients after excluding 17 losses.

I miss a figure representing the ROC curve in Table 3, with the different factors analyzed.

Moderate editing of English language required

Author Response

Reviewer 1:

• In the title, it is not quite correct to write retrospective cohort study, since cohort study is by definition prospective.

Reply : The study is a retrospective analysis of a cohort, prospectively included for another study. Since it is confusing, we decided to delete this from the title. We also changed the word “study” for “analysis” in the sentence in line 68: “We performed a retrospective analysis on a cohort of patients with APN who had previously been enrolled prospectively [17].”

• In Figure 1, the first table indicates 103 patients and the second 120 patients after excluding 17 losses.

Reply : Thank you for this comment. This point was also raised by a second reviewer. The figure was not clear enough and we changed it to start with the “120 patients” bloc. We also added a comment next to figure title, to explain the origin of the 120 eligible patients.

• I miss a figure representing the ROC curve in Table 3, with the different factors analyzed.

Reply : We decided to add in the appendix the 16 ROC curves for single factors, and the 2 ROC curves of the Prediction rules (appendix Figure1a-c).

• Moderate editing of English language required

Reply : The previously submitted manuscript was already corrected by a professional editing service. We sent the revised manuscript for a second check.

Reviewer 2 Report

Minimally invasive imaging studies such as ultrasound are considered to be performed when a symptomatic urinary tract infection is suspected. Gender and age have an impact.

The number of patients in the top box in Figure 1 is not 103?

In the fourth column of Table 3, there is a wide gap between "under" and "immunosuppressive".

Author Response

Reviewer 2:

• Minimally invasive imaging studies such as ultrasound are considered to be performed when a symptomatic urinary tract infection is suspected. Gender and age have an impact.

Reply : Thank you for this interesting comment. We agree with you. POCUS are often performed by the ER team, with no risk and minimal/or even no expense. This is of utmost importance for men, with the increased risk of urinary retention.

• The number of patients in the top box in Figure 1 is not 103?

Reply : Thank you for this comment. This was also pointed by another reviewer. The figure wasn’t clear enough and we decided to change it. It start now with the “120 patients” that were eligible.

• In the fourth column of Table 3, there is a wide gap between "under" and "immunosuppressive"

Reply : Thank you for this. We changed for a left adjusted format (that looks better).

Reviewer 3 Report

This article is a very interesting study of the compliance rates of the two criteria for imaging for upper and febrile UTI, as well as a study of the justification of these criteria. I think your paper worthy of this journal.

There are a few things that are unclear, so please check

Minor revision

Table2: There is no asterisk on the p-value in EAU recommendation, is it not Fisher’s exact test?

Why is the number of cases imaged different between Van Nieukooop’s criteria and EAU criteria?

Appendix Table1: The number of clinically relevant findings is 15 cases, but urgent urological disorder is 8 cases and non-urgent urological disorder is 9 cases, so the numbers do not match. Similarly, the number of obstructive and non-obstructive stones is 9 cases, but urgent is 4 and non-urgent is 7 cases, so the numbers do not match. Please confirm.

Other than the above, I don't see any problem.

Author Response

Reviewer 3:

• Table2: There is no asterisk on the p-value in EAU recommendation, is it not Fisher’s exact test?

Reply : The number of cases per cells was large enough to perform a Chi-squared test. Nevertheless, in order to make it simpler in the table, we decided to use Fisher’s exact test for both rows (Fisher test is more conservative than Chi-squared test). The results were statistically significant in both tests.

• Table2: Why is the number of cases imaged different between Van Nieukooop’s criteria and EAU criteria?

Reply : The first takes into account the early imaging (within 24h) in order to apply the Van Nieukooop’s criteria, whereas the second considered all the images performed during the hospital stay (in order to account for not resolving fever). This is explained in the methods section, line 106-111, in the results section line 160-162, and in Table 2 headings of each rows. We decided to add the period in the heading (early = within 24h; delayed > 24h), to make it clearer.

• Appendix Table1: The number of clinically relevant findings is 15 cases, but urgent urological disorder is 8 cases and non-urgent urological disorder is 9 cases, so the numbers do not match. Similarly, the number of obstructive and non-obstructive stones is 9 cases, but urgent is 4 and non-urgent is 7 cases, so the numbers do not match. Please confirm.

Reply : This was explain in the headings of the Table S1: “Since patients can have several abnormal findings, anomaly totals are not cumulative.” We decided to make it clearer with a comment in the footnote.