An Analysis of the Risk Factors and Outcomes of Patients with COVID-19 Admitted to a Non-Acute Hospital

Round 1

Reviewer 1 Report

This is a study of patients admitted with COVID in 2019 and 2020. As the data is old, the relevance of this study is very limited. It is strongly recommended that more recent data is obtained, making this study more informative when comparing two different time periods. Note case fatality rates of COVID has been declining.

Comments and recommendations

• Should describe and focus on main topic early in the first two paragraphs, and throughout the paper.
• Research setting needs to be described in detail to understand the context. What is a non-acute hospital, what type of patients are admitted there and what treatment is given?
• Description of targeted included patient cohort should be specified in the methodology, not in the results section. Are these patients admitted to hospital due to COVID, admitted with COVID (admitted for another reason and have COVID), or developed COVID in hospital? These groups are different.
• Mortality rate is very high and would be consistent with the COVID strains during the study period of 2019 to 2020. The relevance of this study is therefore very limited. It would only be useful if there is more recent data to compare.
• Should describe abbreviations on first use, in body of paper and table.
• Table 1 results. Presume this is univariate analysis. Should do a multivariate analysis to define independent risk factors. This is important, if possible.
• The unexpected results from the WIMD and mortality needs more examination and discussion – is it s sampling error or is social deprivation only relevant if it is associated with severe disease, poor health and frailty, which require more research.
• The discussion needs to be updated. There are many risk factors for COVID mortality that were not examined, and if not examined, should be discussed. There was no mention of the different COVID strains, antiviral treatment or vaccination.

As above.

Author Response

Major comments

This is a study of patients admitted with COVID in 2019 and 2020. As the data is old, the relevance of this study is very limited. It is strongly recommended that more recent data is obtained, making this study more informative when comparing two different time periods. Note case fatality rates of COVID has been declining.

Response/changes:

Clarified that this is a first wave observational study, noting that this data provides early insight into the risk factors associated with COVID-19 infection, which has since evolved with the evolution of management and of the disease itself, thus it may be relevant for future pandemics before this stage of the disease management is reached. 

 

Comments and recommendations

• Should describe and focus on main topic early in the first two paragraphs, and throughout the paper.

Response/changes:

Some changes made to the Background section, to highlight the focus of this article. I have also removed some of the previous Background, removing some of the general information regarding COVID-19 and the pandemic.

 

 

• Research setting needs to be described in detail to understand the context. What is a non-acute hospital, what type of patients are admitted there and what treatment is given?

Response/changes:

Some explanation given for this in the results section - essentially the hospital does not have an A&E department or an ICU, so the majority of the acute intake was through a different hospital in South Wales, with some patients who were suitable for ICU escalation and sick enough to require it were transferred to that hospital. There was some general medical intake / care, as well as the specialties noted. Some patients were occasionally transferred from the other hospital to this one (e.g. rehabilitation).

 

• Description of targeted included patient cohort should be specified in the methodology, not in the results section. Are these patients admitted to hospital due to COVID, admitted with COVID (admitted for another reason and have COVID), or developed COVID in hospital? These groups are different.

Response/changes:

I have moved this information to the Methods section, and clarified the question regarding the patient numbers and COVID diagnosis.

 

• Mortality rate is very high and would be consistent with the COVID strains during the study period of 2019 to 2020. The relevance of this study is therefore very limited. It would only be useful if there is more recent data to compare.

Response/changes:

I have now highlighted this in the discussion and the limitations - the stage of the pandemic, but also the potentially more frail population may contribute to the higher mortality seen here.

 

• Should describe abbreviations on first use, in body of paper and table.

Response/changes:

Abbreviations expanded.

 

• Table 1 results. Presume this is univariate analysis. Should do a multivariate analysis to define independent risk factors. This is important, if possible.

Response/changes:

I have acknowledged the lack of multivariate analysis in the limitations, and mentioned this in the methods (stats paragraph).

 

• The unexpected results from the WIMD and mortality needs more examination and discussion – is it a sampling error or is social deprivation only relevant if it is associated with severe disease, poor health and frailty, which require more research.

Response/changes:

We agree this is an interesting/unexpected result. Whether this is true or if there is bias/confounding contributing to the result is difficult yo say. I have added an additional paragraph in this section in the discussion.

 

• The discussion needs to be updated. There are many risk factors for COVID mortality that were not examined, and if not examined, should be discussed. There was no mention of the different COVID strains, antiviral treatment or vaccination.

Response/changes:

Additional paragraph added to the discussion and the limitations. This study reflects outcomes during the first wave of the COVID-19 pandemic, prior to the emergence of later viral variants, vaccination programmes, and routine antiviral treatment.

Reviewer 2 Report

The manuscript analyses risk factors and outcomes of patients with COVID-19 admitted to a non-acute hospital. The issue is important from a public health perspective. Some suggestions could be indicated to improve the manuscript.

 

1. An objective of the study was:” This study was carried out to gain more understanding of which factors contribute to an increased risk of admission to hospital and death due to COVID-19”. However, risk factors of admission are not reported, and it is only mentioned: “Increased age is associated with increased risk of admission, the majority of patients between 81-90 years of age (Table 1)(Page 4, paragraph 1, line 4). Could you explain this situation?

 

2. The statistical analysis is not adjusted for potential confounding factors, and confounding bias could occur.

 

3. For the control of potential confounding factors, Poisson regression could be used.

 

4. In Table 2, risk ratios are reported in the title. However, hazard ratios are reported in Table 2. In the Materials and Methods section, relative risk is mentioned. In general, hazard ratios are used when the time of exposure is considered. Could you clarify this issue?

 

5. Treatments of the patients are not mentioned, and it may be important and probably has some effect on the probability of death. Reporting patients’ treatments could be convenient to better understand the manuscript.

 

 

 

1. Participation rate of the study. During the study period from December 2019 to June 2020, how many COVID-19 patients were admitted to the hospital, and how many were included in the study.

 

2. A mention of the statistical program employed in the study may be useful for the readers.

 

4. Few patient risk factors were reported. Other risk factors could be the COVID-19 exposure type (family, work, social, nursing homes, and others), ethnicity, and socio-economic levels.
5. In Tables 1 and 2, there are several acronyms: CVA, IHD, CKD, COPD, ICULOS, (IQR), NEWS2, and HLOS. Could you explain its meaning?

 

5. The study has some limitations that the authors do not mention, including the retrospective design, based on registers, and no control of confounding factors.

 

6. The authors should follow the guidelines of the references in the journal.

 

 

 

Author Response

 

1. An objective of the study was:” This study was carried out to gain more understanding of which factors contribute to an increased risk of admission to hospital and death due to COVID-19”. However, risk factors of admission are not reported, and it is only mentioned: “Increased age is associated with increased risk of admission, the majority of patients between 81-90 years of age (Table 1)(Page 4, paragraph 1, line 4). Could you explain this situation?

Response/changes:

We have made changes to the background section and results to better fit with this comment.

 

1. The statistical analysis is not adjusted for potential confounding factors, and confounding bias could occur. 
2. For the control of potential confounding factors, Poisson regression could be used.

Response/changes:

This is mentioned in the results and acknowledgement if limitations in the analysis is made in the limitations paragraph.

 

1. In Table 2, risk ratios are reported in the title. However, hazard ratios are reported in Table 2. In the Materials and Methods section, relative risk is mentioned. In general, hazard ratios are used when the time of exposure is considered. Could you clarify this issue?

Response/changes:

Hazard ratio is changes to Relative Risk in Table 2.

 

1. Treatments of the patients are not mentioned, and it may be important and probably has some effect on the probability of death. Reporting patients’ treatments could be convenient to better understand the manuscript.

 Response/changes:

Treatments is not discussed as part of this study. However during the study period the treatment for COVID-19 was largely supportive care and oxygen, with no routine use of antivirals or vaccination at this time.

 

Detailed comments

1. Participation rate of the study. During the study period from December 2019 to June 2020, how many COVID-19 patients were admitted to the hospital, and how many were included in the study.

 Response/changes:

I appreciate the lack of clarity here - I have altered the methods/results section to clarify this.

 

1. A mention of the statistical program employed in the study may be useful for the readers.

Response/changes:

Sentence included in the results.

 

1. Few patient risk factors were reported. Other risk factors could be the COVID-19 exposure type (family, work, social, nursing homes, and others), ethnicity, and socio-economic levels.

Response/changes:

Data on ethnicity, exposure setting, and occupation were not available and could not be analysed.

 

1. In Tables 1 and 2, there are several acronyms: CVA, IHD, CKD, COPD, ICULOS, (IQR), NEWS2, and HLOS. Could you explain its meaning?

Response/changes:

I have included an abbreviations note with the table.

 

1. The study has some limitations that the authors do not mention, including the retrospective design, based on registers, and no control of confounding factors.

Response/changes:

I have expanded the limitations paragraph.

 

1. The authors should follow the guidelines of the references in the journal.

Response/changes:

I have adjusted the references.

Round 2

Reviewer 1 Report

This new version is an improvement from the original manuscript. Limitations of the study are appropriately expressed.

As the data examined in this study is from a period some time back, the relevance and focus of the study would need to be on the examination of risk factors and outcomes in the first wave of COVID 19 with review and discussion of others viral respiratory pandemics within a similar period where appropriate vaccines and/or treatments were not available. More discussion of similar literature on other pandemics is preferred to identify and discuss common patterns, where available.

The authors have addressed most of my main concerns.

Additional comments are noted as above.

Author Response

Thank you for your review and your feedback. We appreciate that this study reflects the first wave of the pandemic and that things have moved forward significantly since then. We have included some additional multivariate analysis in this updated draft, after some discussion with our statistician.

Reviewer 2 Report

The authors have addressed some suggestions in our review. However, insufficiencies in the manuscript remain. Without a multivariable analysis, the results of this study are not completed.

 

Without a multivariable analysis, the results of this study are not completed.

Author Response

Comments: Without a multivariable analysis, the results of this study are not completed.

Response: 

Thank you for your review and your feedback. We have included some additional multivariate analysis in this updated draft, after some discussion with our statistician, as suggested following your review. There is mention of this in the statistics section of Methods, with an additional table (Table 3) included in the results with some explanation. There is also a brief addition referring to this additional analysis in the discussion.