The Impact of COVID-19 on Burns: A Brazilian Study

Round 1

Reviewer 1 Report

The authors of "The Impact of COVID-19 on burns: A Brazilian study" compared the demographics and burn characteristics of patients admitted to their burn center before and during the pandemic. Their hypothesis was that they would see more flame burns and more burns in the home, and see more burns related to hand sanitizers. They found a decreased number of admissions, more burns in the home, smaller TBSA burns and shorter lengths of stay for patients admitted during the pandemic. They compared their results with what has already been published.

This was an interesting question, but I have the following concerns.

The Introduction was short, and the objective was not clear. The introduction did not flow. There was speculation as to what the results would be. This would have been better stated as a hypothesis. 

Why not compare March 1, 2019 to December 31, 2019 to March 1, 2020 to December 31, 2020. Or March 18, 2019-December 31, 2019 to March 18, 2020 to December 31, 2020. The longer time period for the control group would lead to more patients anyway. How would this change your analysis?

Figure 1 is too small, it is difficult to see the differences in the figure

Table 1 the margins need to be fixed so that the table is easier to read (specifically the cause of burn)

Table 2 has no title or legend. It is also in a different format that Table 1. It would be easier to read if it was in a consistent format (also it is missing percentages).

Results mention "Figure 2-4" line 125 - either label the figures 2-4 or mention "figure 2a, 2b, 2c".

There is no breakdown for Peds burns like there is for the adult burns. Is there a reason?

The limitations mention that you were unable to delineate exactly how the burns occurred in the retrospective review. How certain are you of your conclusion about the have sanitizer and flame burns?

There is great potential for this work. 

Author Response

The authors of "The Impact of COVID-19 on burns: A Brazilian study" compared the demographics and burn characteristics of patients admitted to their burn center before and during the pandemic. Their hypothesis was that they would see more flame burns and more burns in the home, and see more burns related to hand sanitizers. They found a decreased number of admissions, more burns in the home, smaller TBSA burns and shorter lengths of stay for patients admitted during the pandemic. They compared their results with what has already been published.This was an interesting question, but I have the following concerns.

1.The Introduction was short, and the objective was not clear. The introduction did not flow. There was speculation as to what the results would be. This would have been better stated as a hypothesis.

Thank you for this comment. The Introduction was rewritten, as below:

“The rapid spread of Covid-19 virus led to exceptional strategies adopted worldwide to limit the outbreak; businesses considered as non-essential were canceled, outdoor activities limited, and social isolation recommended (1). The restrictions resulted in people staying mostly at home, implying the potential increase of their exposure to domestic accidents, such as burn injuries (2). In the state of Minas Gerais, Brazil, the first case of Covid-19 was confirmed on March 8, 2020, and the state of quarantine was declared on March 18, 2020. Furthermore, to reduce the spread of the virus, the authorities suggested the use of alcohol-based disinfectants, possibly raising the risk for flame injuries. (1; 2). However, while the rise of burn injuries during the outbreak was anecdotally reported by Brazilian physicians but not verified (2; 3), the increment of flame injuries caused by the external use of hand sanitizer, particularly affecting arms and trunk, was recently reported in Mina Gerais (4), suggesting that variations in epidemiological pattern were caused by changed behaviors. Understanding of the impact of exceptional events, such as a pandemic, on burn injuries trends is indispensable to develop appropriate preventive strategies and to respond to possible surge. However, it remains to be investigated whether occurrence, cause and severity of burns changed during the outbreak. For this reason, this study aimed to analyze possible differences in epidemiological trends in Brazil, comparing patients with burns admitted at the Burn Unit of Belo Horizonte, which is the main referral hospital of the state of Minas Gerais, before and during the pandemic.”

 

 

 

 

 

2. Why not compare March 1, 2019 to December 31, 2019 to March 1, 2020 to December 31, 2020. Or March 18, 2019-December 31, 2019 to March 18, 2020 to December 31, 2020. The longer time period for the control group would lead to more patients anyway. How would this change your analysis?

Dear reviewer, thank you so much for this comment. We did the data analysis again, from March to December 2019 (control group) and March to December 2020 (study group). As we can observe comparing both tables, including January and February 2020 did not change the results. We can see that place of injury, mechanism of burn, transferred from another city, TBSA% are the variables with P value < 0.05. This allows us to conclude that the statistical analysis was not affected. Here you can see another table, with the statistical analysis as you mentioned.

 

	All Patients  (n=781)	Control Group (n=402)	Study Group (n=379)	P
Sex   Female   Male	307 (40%) 474 (60%)	156 (39%) 246 (61%)	151 (40%) 228 (60%)	0.767^
Age (years)*	32 (9-48)	32.5 (13.75-48)	32 (6.25-48)	0.342#
Place of injury  Outside Work School Home Other	81 (11%) 69 (9%) 3 (0.4%) 478 (65%) 103 (14%)	38 (5%) 44 (12%) 3 (0.4%) 218 (29%) 62 (8%)	43 (12%) 25 (7%) 0 (0%) 260 (35%) 41 (5%)	0.002†
Cause of burn Scalds Flame Chemical Electrical Contact Other	258 (33%) 397 (51%) 23 (3%) 56 (7%) 31 (4%) 10 (2%)	134 (34%) 202 (50%) 7 (2%) 34 (8%) 18 (4%) 5 (2%)	124 (33%) 195 (52%) 16 (4%) 22 (6%) 13 (4%) 5 (1%)	0.245†
Mechanism of burn Accident  Caused by other Self-inflicted Other	603 (86%) 53 (6%) 58 (7%) 7 (1%)	341 (86%) 23 (5%) 31 (7%) 7 (1%)	322 (85%) 30 (8%) 27 (7%) 0	0.037†
Transferred from another city	371 (46%)	191 (50%)	147 (40%)	0.003^
Time to admission the at the João XXIII Hospital (days)*	1 (0-4)	1 (0-4)	1 (0-3.5)	0.828#
TBSA% *	15 (8-28)	15 (9-30)	13.5 (7-25)	0.038#
Location of the Burn Head and neck Core (Thorax, Abdomen and Back) Arms Hands Genitalia Legs Feet	470 (51%) 571 (62%)   577 (63%) 188 (21%)       86 (9%) 427 (47%) 107 (12%)	284 (53%) 345 (64%)   345 (64%) 96 (18%) 50 (9%) 257 (48%) 60 (11%)	186 (49%) 226 (60%)   232 (61%) 92 (24%) 36 (9%) 170 (45%) 47 (12%)	0.928^
Presence of FTB	357 (40%)	218 (41%)	139 (37%)	0.415^
In-patient status	770 (95%)	376 (95%)	370 (97%)	0.149^
Presence of inhalation injury	117 (15%)	59(15%)	58 (16%)	0.812^
Need for mechanical ventilation	142 (16%)	86 (16%)	56 (15%)	0.360^
In-hospital mortality	72 (8%)	39 (7%)	33 (9%)	0.696^
Length of stay (days)*	23 (15-41)	24 (15-45.75)	21 (14-35)	0.014#

 

3. Figure 1 is too small, it is difficult to see the differences in the figure

Thank you for this suggestion. The Figure was changed according to your advice.

4. Table 1 the margins need to be fixed so that the table is easier to read (specifically the cause of burn)

Thank you for this suggestion. We edited Table 1 according to your advice.

5. Table 2 has no title or legend. It is also in a different format that Table 1. It would be easier to read if it was in a consistent format (also it is missing percentages).

Thank you for this comment. We added a legend to Table 5 and harmonized the format to that of Table 1. We also added the missing percentages.

6. Results mention "Figure 2-4" line 125 - either label the figures 2-4 or mention "figure 2a, 2b, 2c".

Thank you for this comment. We changed the text as you suggested.

7.There is no breakdown for Peds burns like there is for the adult burns. Is there a reason?

Thank you for this comment. The main purpose of this study was to compare all patients with acute burns admitted before and during the pandemic, and pediatric patients are included in the study population and thus in Table 1.

During the outbreak, people stayed mostly at home potentially increasing their exposure to domestic accidents. Table 2 explored possible changes of circumstances of burns in children, since they are a recognized risk-group for domestic accidents. However, the pediatric population was not the main study-group.

8.The limitations mention that you were unable to delineate exactly how the burns occurred in the retrospective review. How certain are you of your conclusion about the have sanitizer and flame burns?

Thank you for your comment. We agree with your thoughts. Based on data retrieved by the medical records of the Burn Unit, less patients with burns were referred for specialized burn care during the pandemic, although patients admitted for specialized burn care had smaller TBSA% and shorter LOS. We omitted from the Conclusions the meaning of increased flame burns caused by the use of alcohol-based sanitizer since it was not specifically addressed in this study.

There is great potential for this work.

 

Reviewer 2 Report

 

This study compares two groups of patients retrospectively at the onset of pandemics from March to December and before pandemics from March to February next year and looks for possible differences in the epidemiology caused by pandemics.

 

There are some major issues I want to address:

 

1)    The major problem is that the data is collected in two different time periods. This I do not understand that since the data for the control group exists and, supposing that there are no differences in the referral system and organization of the burn care, it would be much better to have exactly same time for the data collection, e.g., from March to December, to be able to compare the true number of patients. This should be easy to change.

2)    Figures:

a.     Fig. 1 doesn´t have the control groups, why?

b.     Fig.2: The y-axis is about double of the highest value. This makes the columns very flat. Also, the columns are very wide. It might even be better to have columns for each variable before and after the pandemics to make it more understandable.

 

 

Minor issue:

line 78: What means “Data of interest was collected from medical records from the Burn Unit between January and June 78, 2021”?

The period of the data collection of the study group in the beginning of Materials and methods is mentioned but why not the period or the control group? Only in next section.

 

I think the authors should correct these things before it can be accepted for publication.

 

 

 

Author Response

Reviewer #2:

This study compares two groups of patients retrospectively at the onset of pandemics from March to December and before pandemics from March to February next year and looks for possible differences in the epidemiology caused by pandemics. There are some major issues I want to address:

1.The major problem is that the data is collected in two different time periods. This I do not understand that since the data for the control group exists and, supposing that there are no differences in the referral system and organization of the burn care, it would be much better to have exactly same time for the data collection, e.g., from March to December, to be able to compare the true number of patients. This should be easy to change.

Dear reviewer, thank you so much for this comment. We did the data analysis again, from March to December 2019 (control group) and March to December 2020 (study group). As we can observe comparing both tables, including January and February 2020 did not change the results. We can see that place of injury, mechanism of burn, transferred from another city, TBSA% are the variables with P value < 0.05. This allows us to conclude that the statistical analysis was not affected. Here you can see another table, with the statistical analysis as you mentioned.

 

	All Patients  (n=781)	Control Group (n=402)	Study Group (n=379)	P
Sex   Female   Male	307 (40%) 474 (60%)	156 (39%) 246 (61%)	151 (40%) 228 (60%)	0.767^
Age (years)*	32 (9-48)	32.5 (13.75-48)	32 (6.25-48)	0.342#
Place of injury  Outside Work School Home Other	81 (11%) 69 (9%) 3 (0.4%) 478 (65%) 103 (14%)	38 (5%) 44 (12%) 3 (0.4%) 218 (29%) 62 (8%)	43 (12%) 25 (7%) 0 (0%) 260 (35%) 41 (5%)	0.002†
Cause of burn Scalds Flame Chemical Electrical Contact Other	258 (33%) 397 (51%) 23 (3%) 56 (7%) 31 (4%) 10 (2%)	134 (34%) 202 (50%) 7 (2%) 34 (8%) 18 (4%) 5 (2%)	124 (33%) 195 (52%) 16 (4%) 22 (6%) 13 (4%) 5 (1%)	0.245†
Mechanism of burn Accident  Caused by other Self-inflicted Other	603 (86%) 53 (6%) 58 (7%) 7 (1%)	341 (86%) 23 (5%) 31 (7%) 7 (1%)	322 (85%) 30 (8%) 27 (7%) 0	0.037†
Transferred from another city	371 (46%)	191 (50%)	147 (40%)	0.003^
Time to admission the at the João XXIII Hospital (days)*	1 (0-4)	1 (0-4)	1 (0-3.5)	0.828#
TBSA% *	15 (8-28)	15 (9-30)	13.5 (7-25)	0.038#
Location of the Burn Head and neck Core (Thorax, Abdomen and Back) Arms Hands Genitalia Legs Feet	470 (51%) 571 (62%)   577 (63%) 188 (21%)       86 (9%) 427 (47%) 107 (12%)	284 (53%) 345 (64%)   345 (64%) 96 (18%) 50 (9%) 257 (48%) 60 (11%)	186 (49%) 226 (60%)   232 (61%) 92 (24%) 36 (9%) 170 (45%) 47 (12%)	0.928^
Presence of FTB	357 (40%)	218 (41%)	139 (37%)	0.415^
In-patient status	770 (95%)	376 (95%)	370 (97%)	0.149^
Presence of inhalation injury	117 (15%)	59(15%)	58 (16%)	0.812^
Need for mechanical ventilation	142 (16%)	86 (16%)	56 (15%)	0.360^
In-hospital mortality	72 (8%)	39 (7%)	33 (9%)	0.696^
Length of stay (days)*	23 (15-41)	24 (15-45.75)	21 (14-35)	0.014#

 

2a. Fig. 1 doesn´t have the control groups, why?

2b. Fig.2: The y-axis is about double of the highest value. This makes the columns very flat. Also, the columns are very wide. It might even be better to have columns for each variable before and after the pandemics to make it more understandable.

Thank you for these suggestions. We have added the control groups to the Figure 1 and changed the Figure 2 according to Your advice.

3. What means “Data of interest was collected from medical records from the Burn Unit between January and June 78, 2021”? The period of the data collection of the study group in the beginning of Materials and methods is mentioned but why not the period or the control group? Only in next section.

Thank You for these questions. The time-period January to June 2021 represents the time during which data was extracted by the co-authors from the medical records. This should not be confused with the actual study period. This has been clarified in the Materials and Methods, as following:

“The data source for this retrospective study was the medical records from the Burn Unit. Between January and June 2021, the study co-authors had access to the medical records and extracted information regarding the control group and the study group.”

I think the authors should correct these things before it can be accepted for publication.

 

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The revision of "The Impact of COVID-19 on burns: A Brazilian study" is a retrospective review of the epidemiological differences among burned admissions at the main referral hospital of the State of Minas Gerais, Brazil before and during the pandemic. The authors made the aim more clear and thus it was easier to follow. However, it may be better to evaluate admissions from 1 March 2019 to 31 December 2019 (similar time-frame from 1 March 2020 to 31 December 2020).   The longer amount of time in for admissions in one cohort may be confounding.

Author Response

Thank you for this comment. We changed the timeframe according to your suggestion. The control group consists in the new drafting of patients admitted from 1 March 2019 to 31 December 2019.  

Author Response File: Author Response.docx

Reviewer 2 Report

The authors have done the corrections but there are still some issues I like to address:

Although the statistics are the same even if the study periods are not similar, you should mention this in the discussion if you do not correct the periods. Otherwise, the readers will notice the same we both reviewers brought up.

 

-see line 142, an extra ”that”?

 

-Line 153, a dot missing between the sentences. Also, in line 151 at the end.

 

-Fig. 2 a and b. Why don´t you use the same model to depict the colums like in the c? Now there is a line going from scalds to flame which is confusing and is actually wrong. Better to have two colums next to each other, before and after covid.

 

Conclusions: The abstract conclusions differ from the article conclusions.

Author Response

Reviewer# 2

The authors have done the corrections but there are still some issues I like to address:

Although the statistics are the same even if the study periods are not similar, you should mention this in the discussion if you do not correct the periods. Otherwise, the readers will notice the same we both reviewers brought up.

 

Thank you for this comment. To avoid confusion, we changed the timeframe for the admission in the control group, which now consists of patients admitted from 1 March 2019 to 31 December 2019.

 

See line 142, an extra ”that”?

Line 153, a dot missing between the sentences.

Also, in line 151 at the end.

 

Thank you for noticing these inaccuracies. We corrected the text as you suggested

 

Fig. 2 a and b. Why don´t you use the same model to depict the colums like in the c? Now there is a line going from scalds to flame which is confusing and is actually wrong. Better to have two colums next to each other, before and after covid.

 

Thank you for this suggestion. We changed figures 2a and 2b, as following 

 

 

Conclusions: The abstract conclusions differ from the article conclusions.

Thank you for notice that. We changed the conclusion in the abstract as following:

“Less patients with burns were referred for specialized burn care during the pandemic, although patients admitted for specialized burn care had smaller TBSA% and shorter LOS. “