Diagnostic Errors in Japanese Community Hospitals and Related Factors: A Retrospective Cohort Study

Round 1

Reviewer 1 Report

This manuscript is a very interesting study and an important effort to improve clinical outcomes related to diagnostic errors.

Please clarify the statement of research aim in the Abstract and Introduction sections.

Clear research aims can provide clear research conclusions.

Please take the following points into consideration and revise them.

Unnecessary capitalization and abbreviations were used.

In addition, in the letters in the table, please unify the capitalization so that all letters begin with a capital letter.

Also, when using abbreviations, please provide an explanation of the abbreviation.

 

Line102

There are already several papers on the definition of diagnostic errors (please refer to the following). Please improve the text of the criteria for diagnostic errors used in this study.

 

https://apps.who.int/iris/bitstream/handle/10665/252410/9789241511636-eng.pdf

 

https://www.ncbi.nlm.nih.gov/books/NBK338594/

 

https://journals.lww.com/journalpatientsafety/Fulltext/2022/12000/Defining_Diagnostic_Error__A_Scoping_Review_to.7.aspx

 

There are two identical expressions in the figure. Please delete "Figure. Flow chart of patient selection" above the figure, as it is redundant.

 

Line 144

The authors mentioned that "Univariate and logistic regression analyses were performed."

In fact, it is assumed that an independent t-test (Welch's t-test is appropriate here) was performed for the quantitative variables that were divided into two groups. Please refer to the following reference.

 

https://www.voxco.com/blog/welchs-t-test/

 

https://libguides.library.kent.edu/spss/independentttest

 

https://rips-irsp.com/articles/10.5334/irsp.82

 

How did you analyze the values expressed as percentages in Table 1?

Perhaps you performed a chi-square test, but please describe the exact statistical treatment method.

Also, how did you analyze the values indicated as percentages in Tables 1 and 2?

I assume you probably performed a chi-square test, but please describe exactly how you performed the statistical processing.

Also, please describe the statistical treatment method under Tables 1 and 2.

 

The authors state that "The total number does not always equal the number of cases "n," owing to missing 200 data in some cases. However, an exact "n" should be given for each case so that other researchers can analyze it.

 

Table 2.

I believe that the logistic regression analysis is interpreted differently.

Logistic regression analysis does not analyze significant differences.

Female patients (reference: Male patients) OR=0.566, 95% CI= 0.34-0.94, p=0.029

This value indicates that "the effect (association) for misdiagnosis is lower for female patients compared to male patients." This is also the case for oxygen intake.

 

On the other hand, Table 2 shows that "no variable was found to be associated with the risk of diagnostic errors."

 

Age, level of consciousness, systolic blood pressure, BMI, and oxygen support," which were found to be significantly different in the bivariate significance tests, were not applicable as risk factors in the logistic regression analysis. I believe that, for level of consciousness, systolic blood pressure, and BMI, these risk factors for diagnosis have been identified that you can assume that physicians are using this information for an accurate diagnosis.

 

Additionally, it is worth noting that the nighttime hours were not a risk factor for misdiagnosis. Can we not assume that your hospital has a good nighttime working system support (team care system)?

 

Please improve your statistical processing methods, your results, and further improve your discussion. This would make a good research report.

We would be happy to help. Please describe.

 

The authors state that "The total number does not always equal the number of cases "n," owing to missing 200 data in some cases. However, an exact "n" should be given for each case so that other researchers can analyze it.

 

Table 2.

I believe that the logistic regression analysis is interpreted differently. Logistic regression analysis does not analyze significant differences.

 

Female patients (reference: Male patients) OR=0.566, 95% CI= 0.34-0.94, p=0.029

This value indicates that "the effect (association) for misdiagnosis is lower for female patients compared to male patients." This is also the case for oxygen intake.

 

On the other hand, Table 2 shows that "no variable was found to be associated with the risk of diagnostic errors."

 

Age, level of consciousness, systolic blood pressure, BMI, and oxygen support," which were found to be significantly different in the bivariate significance tests, were not applicable as risk factors in the logistic regression analysis. I believe that, for blood pressure and BMI, those risk factors for diagnosis have been identified so that you could discuss that physicians are using this information for an accurate diagnosis.

 

Additionally, it is worth noting that the nighttime hours were not a risk factor for diagnostic errors. What do you think your hospital has a good nighttime working system support (interdisciplinary team care system)?

 

Please improve your statistical processing methods, your results, and further improve your discussion. This procedure would make a good research report.

I hope you will find it useful.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Thank you for giving me the opportunity to review this manuscript. I have a few comments for the authors to consider

 My primary concern is the methods section. it is unclear to me what it is the survey exactly? how was it administered ? and how is this study a retrospective one?

 

other comments:

 

- in the introduction section (line 63-73) please add references.

 

- please add a clear research question or objectives to the study.

 

- Why did the authors pick 6 medications to be a cut point ? it is very uncommon. 

 

- please explain CCI to the readers.

 

- There are few terminologies that require revising such as time zone.

 

-the discussion section should be revised. Some subheadings need to be expanded on (e.g., CCI)

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

I am pleased that your manuscript has been improved.

However, I believe the following points need further improvement.

 

Line 12-17

You can revise the sentences as below:

This study aimed to clarify the reality of diagnostic errors in regional hospitals in Japan. A 10-month retrospective cohort study was conducted from January to October 2021 at the emergency room of Oda Municipal Hospital in central Shimane Prefecture, Japan. Participants were divided into groups with or without diagnostic errors, and independent variables of patient, physician, and environmental factors were analyzed using Fisher’s exact test, univariate (Welch’s t-test), and logistic regression analyses.

 

Line 142-143

You can revise the sentences as below:

Univariate and logistic regression analyses were performed. In univariate analysis, continuous variables were analyzed using Welch's t-test and nominal variables using Fisher's exact test.

 

Line 175

What is the meaning of the NOT alert?

 

Line 185

The authors mentioned that "Univariate analysis revealed significant differences."

What statistical analysis methods were used in the univariate analysis?

 

First group:

　Sample size = 121

　    Mean = 74.0200

　Standard deviation = 18.9500

 

Second group:

　Sample size = 803

　    Mean = 79.3000

　Standard deviation = 16.0300

 

Test of Equal Variance for Two Groups

　  F value = 1.39750

　Degrees of freedom = (120, 802)

　  P-value = 0.010744 (two-tailed probability)

 

Normal t-test (when equal variances can be assumed)

　  t-value = 3.29353

　Degrees of freedom = 922

　  P-value = 0.00103

 

When equal variances cannot be assumed (Welch's method)

　  t-value = 2.91194

　Degrees of freedom = 147.01733

　  P-value = 0.00415 (exact value corresponding to fractional degrees of freedom)

 

Based on a two-tailed F test, σ1 (n=121) is considered as unequal to σ2  (n=803) (p-value is 0.0107). Since p-value < α, H0 is rejected.

The average of Group-1's population is considered to be not equal to the average of Group-2's population.

 

Thus, you have selected the Normal t-test (that is Student t-test), but you have to select the analysis result of Welch's method.

 

Line 196

There are still some unclear areas in the explanatory text for Table 1.

What is NOT alert?

Are there any data in Table 1 that show differences in the level of awareness?

 

There are other minor errors, so please check the comments in the pdf and correct them carefully.

Thank you so much.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Thank you for addressing my comments.

Polypharmacy cutoffs often are often set at 5 medications. Only handful of studies that used 6 as a cutoff. 

please review this article 

https://bmcgeriatr.biomedcentral.com/articles/10.1186/s12877-017-0621-2

That being said, I highly recommend the authors to make it clear for the readers that they chose 6 medications as a cutoff and provide a reference.

Author Response

Please see the attachment.

Author Response File: Author Response.docx