Demographic, behavioural and occupational risk factors associated with SARS-CoV-2 infection in UK healthcare workers: a retrospective observational study

Objective Healthcare workers (HCWs) are at higher risk of SARS-CoV-2 infection than the general population. This group is pivotal to healthcare system resilience during the COVID-19, and future, pandemics. We investigated demographic, social, behavioural and occupational risk factors for SARS-CoV-2 infection among HCWs. Design/setting/participants HCWs enrolled in a large-scale sero-epidemiological study at a UK university teaching hospital were sent questionnaires spanning a 5-month period from March to July 2020. In a retrospective observational cohort study, univariate logistic regression was used to assess factors associated with SARS-CoV-2 infection. A Least Absolute Shrinkage Selection Operator regression model was used to identify variables to include in a multivariate logistic regression model. Results Among 2258 HCWs, highest ORs associated with SARS-CoV-2 antibody seropositivity on multivariate analysis were having a household member previously testing positive for SARS-CoV-2 antibodies (OR 6.94 (95% CI 4.15 to 11.6); p<0.0001) and being of black ethnicity (6.21 (95% CI 2.69 to 14.3); p<0.0001). Occupational factors associated with a higher risk of seropositivity included working as a physiotherapist (OR 2.78 (95% CI 1.21 to 6.36); p=0.015) and working predominantly in acute medicine (OR 2.72 (95% CI 1.57 to 4.69); p<0.0001) or medical subspecialties (not including infectious diseases) (OR 2.33 (95% CI 1.4 to 3.88); p=0.001). Reporting that adequate personal protective equipment (PPE) was ‘rarely’ available had an OR of 2.83 (95% CI 1.29 to 6.25; p=0.01). Reporting attending a handover where social distancing was not possible had an OR of 1.39 (95% CI 1.02 to 1.9; p=0.038). Conclusions The emergence of SARS-CoV-2 variants and potential vaccine escape continue to threaten stability of healthcare systems worldwide, and sustained vigilance against HCW infection remains a priority. Enhanced risk assessments should be considered for HCWs of black ethnicity, physiotherapists and those working in acute medicine or medical subspecialties. Workplace risk reduction measures include ongoing access to high-quality PPE and effective social distancing measures.

(other), mixed ethnicity, or Black or Black British (African) ethnicity" Discussion: Overall, there is a lack of comparison with similar studies from other countries, whether the risk factors for COVID-19 infection among HCW are the same or different, and the reason of why is it so?
Discussion: first paragraph-authors mentioned that they have identified targetable risk factors for future pandemics. I believe that this is an over-assumption, because the study was based solely on this COVID-19 pandemic.
Discussion-3rd paragraph: I believe there is still a lot of room for discussion regarding ethnicity. Rather than unknown, authors may discuss on health equalities, access to healthcare, housing, discrimination and many other factors regarding why minorities has higher risk of getting COVID-19 infections. One such example could be found at CDC: https://www.cdc.gov/coronavirus/2019ncov/community/health-equity/race-ethnicity.html Discussion-4th paragraph: The last few sentences of the paragraph regarding social distancing comes abruptly after a long elaboration on PPE. The authors might consider expand the discussion regarding social distancing and putting it into a separate paragraph for better readability.
Discussion-5th paragraph: I am not sure why "hospitalization with COVID-19 increase with age" comes into picture, because it was not presented in the Results section. It also did not cohere with the subsequent description on physiotherapist's risk.
Discussion-6th paragraph: "We think that the risk factors discussed within this paper are unlikely to be greatly affected by a change in the risk of infection in new variants and remain broadly generalisable as risk factors for HCW infection." Authors may explain what is their reasoning for such assumption.

REVIEWER
Martineau, Adrian Queen Mary University of London REVIEW RETURNED 12-Jul-2022

GENERAL COMMENTS
This well-written manuscript reports findings of a retrospective observational study investigating risks for anti-S IgG/A/M seropositivity in a cohort of 2258 UK healthcare workers. Strengths include the large size and detailed characterization of risks inside and outside the workplace, allowing for adjustment for potential confounders. The findings demonstrate the combined importance of exposure at home and in the workplace, esp. outside of ICU / Infectious Diseases wards where use/quality of PPE may be less/lower. The study also highlights a massive ethnic disparity in risk, which is currently unexplained. The serology test employed utilized detected 3 classes of antibody and has been validated with high sensitivity and specificity. There are some limitations though. 1. Domestic staff (n=0) and porters (n=7) were seriously underrepresentedthis is a limitation and should be explicitly acknowledged. These groups were potentially at very high risk of infection, but their experience has not been captured in this study. 2. The study was conducted in the pre-vaccination era, which could constrain relevance of its results for the current situation where the majority of HCWs are vaccinated against SARS-CoV-2. Populationbased studies e.g. https://doi.org/10.1101/2022.03.11.22272276 have reported that vaccination has been effective in ablating / attenuating increased risk of SARS-CoV-2 infection associated with Black or Asian ethnic origin and increased BMI, for example. It was also conducted before emergence of delta / omicron variants, which could also constrain relevance for the current situation, but to a lesser extent. These limitations should be acknowledged in the Discussion. The Discussion currently states: "We think that the risk factors discussed within this paper are unlikely to be greatly affected by a change in the risk of infection in new variants and remain broadly generalisable as risk factors for HCW infection". I tend to agree re variantsbut the rollout of vaccination is likely a game changer that could dramatically affect risk factors reported here.
3. The authors acknowledge that the retrospective nature of the questionnaires could have introduced recall bias -especially likely to operate re subjective factors like PPE availability. Since respondents knew their serostatus at the time of questionnaire completion, this could have become a self-fulfilling prophecy (people who knew they got infected likely to blame lack of PPE). There is also a potential risk of imprecision in answers due to the time elapsed which should also be acknowledged. 4. Clarity re dates: I found the chronology difficult to understand. The authors write: "Questionnaire invites were sent between October and November 2020 and covered the period between March 2020 and June 2020 (the time of serological sampling). Questions relating to behavioural and demographic factors were separated by time periods covering March -May and June -July to account for differences in behaviour and exposures outside of occupational environments due to the instigation (March 2020) and easing (June 2021) of the first UK national "lockdown" measures'. Is there a typo here (June 2020?). 'Covered the period' is ambiguous, as it could refer to when the questionnaires were completed, or the time period referred to in the questionnaire. Suggest something along the lines of 'Questionnaire invites were sent between October and November 2020 and questions within them related to participants' recalled behaviour during two periods -March-May 2020 and June-July 2020'. Was the time of serological sampling in June 2020 only or was it done over 4 months (Mar 20 to Jun 20?). The wording is ambiguous.
Elsewhere it is written that ''HCWs were sent questionnaires spanning Mar-Aug 2021.' I am struggling to reconcile this with other dates reported.
A flow chart to illustrate the chronology of serology sampling, consent, questionnaire issue / completion etc would help to clarify the methods. 5. I was interested to see that BMI did not associate with risk of seropositivitya contrast to findings from others e.g. PMID 35189888. Could the authors speculate why this may have been the case? 6. Comments to the Author: This paper described the risk factors of COVID-19 infection among healthcare workers. While similar findings were reported in available literatures, this paper may add consolidate further on risk factors among HCW in the local context. Overall, the paper is methodologically sound, but the discussion needs further depth and support with reasoning.
Abstract-design: the study design was not clearly stated This has been updated.
Abstract-conclusion: I am not sure whether the term "novel" is still relevant.
The wording has been changed.
Methods-Population and setting: This sentence was not clear and need rephrasing. "The definition of COVID-19 working for the purpose of risk stratification included clinical areas designated as either "Red" (patients with PCR-confirmed SARS-CoV-2 infection) or "Amber" (patients for whom there is a high clinical suspicion of ." This been updated to improve clarity Methods-Questionnaire: there was a lack of description on the validation process of the questionnaire, is there any pre-testing and reliability testing performed? And whether the questionnaire was only available in one language?
This has been clarified.
Methods-there was a lack of description on the sample size calculation.
The was no indication for a sample size calculation, as this questionnaire was sent to all available respondents enrolled in the original longitudinal study. Additionally, the were no prior data available to assess differences between groups on the metrics assessed in the analysis to inform a sample size calculation.
This has been clarified in the text.
Methods-statistical analysis: It was not mentioned what software was used for statistical analysis This has been included.
Results: It is valuable to report the response rate, i.e. what is the total number of HCW invited?
This has been included in the results text.
Results: I am not sure why OR was not included in this sentence. "Other demographic factors that were positively associated with seropositivity include identifying as being Asian or Asian British (other), mixed ethnicity, or Black or Black British (African) ethnicity" This has been included.
Discussion: Overall, there is a lack of comparison with similar studies from other countries, whether the risk factors for COVID-19 infection among HCW are the same or different, and the reason of why is it so?
This has been updated.
Discussion: first paragraph-authors mentioned that they have identified targetable risk factors for future pandemics. I believe that this is an over-assumption, because the study was based solely on this COVID-19 pandemic.
We believe that these targetable risk factors could serve as a framework for targeting and lowering risk in HCWs during future pandemics, particularly those driven by respiratory pathogens. The wording of this sentence has been changed to clarify that this is not a firm conclusion.
Discussion-3rd paragraph: I believe there is still a lot of room for discussion regarding ethnicity.
Rather than unknown, authors may discuss on health equalities, access to healthcare, housing, discrimination and many other factors regarding why minorities has higher risk of getting COVID-19 infections. One such example could be found at CDC: https://www.cdc.gov/coronavirus/2019ncov/community/health-equity/race-ethnicity.html This has been included.
Discussion-4th paragraph: The last few sentences of the paragraph regarding social distancing comes abruptly after a long elaboration on PPE. The authors might consider expand the discussion regarding social distancing and putting it into a separate paragraph for better readability.
This has been done.
Discussion-5th paragraph: I am not sure why "hospitalization with COVID-19 increase with age" comes into picture, because it was not presented in the Results section. It also did not cohere with the subsequent description on physiotherapist's risk.
This is in relation to the proportion of workload of physiotherapists being undertaken with elderly hospitalised patients. This paragraph has been restructured and clarified to clarify this.
Discussion-6th paragraph: "We think that the risk factors discussed within this paper are unlikely to be greatly affected by a change in the risk of infection in new variants and remain broadly generalisable as risk factors for HCW infection." Authors may explain what is their reasoning for such assumption.
This sentence was qualified by the preceding sentence stating that subsequent variant risk factors are not identical. Further, we have expanded to add that vaccination is likely to have had more impact on risk than subsequent variants per reviewer 2's comments below.
This well-written manuscript reports findings of a retrospective observational study investigating risks for anti-S IgG/A/M seropositivity in a cohort of 2258 UK healthcare workers. Strengths include the large size and detailed characterization of risks inside and outside the workplace, allowing for adjustment for potential confounders. The findings demonstrate the combined importance of exposure at home and in the workplace, esp. outside of ICU / Infectious Diseases wards where use/quality of PPE may be less/lower. The study also highlights a massive ethnic disparity in risk, which is currently unexplained. The serology test employed utilized detected 3 classes of antibody and has been validated with high sensitivity and specificity.
There are some limitations though.
1. Domestic staff (n=0) and porters (n=7) were seriously under-representedthis is a limitation and should be explicitly acknowledged. These groups were potentially at very high risk of infection, but their experience has not been captured in this study.
This has been explicitly acknowledgedand referenced to our previous work, where porters and domestic staff were at higher risk of infection.
2. The study was conducted in the pre-vaccination era, which could constrain relevance of its results for the current situation where the majority of HCWs are vaccinated against SARS-CoV-2. Populationbased studies e.g. https://doi.org/10.1101/2022.03.11.22272276 have reported that vaccination has been effective in ablating / attenuating increased risk of SARS-CoV-2 infection associated with Black or Asian ethnic origin and increased BMI, for example. It was also conducted before emergence of delta / omicron variants, which could also constrain relevance for the current situation, but to a lesser extent. These limitations should be acknowledged in the Discussion. The Discussion currently states: "We think that the risk factors discussed within this paper are unlikely to be greatly affected by a change in the risk of infection in new variants and remain broadly generalisable as risk factors for HCW infection". I tend to agree re variantsbut the rollout of vaccination is likely a game changer that could dramatically affect risk factors reported here.
Discussions around these suggestions have been included and expanded.
3. The authors acknowledge that the retrospective nature of the questionnaires could have introduced recall bias -especially likely to operate re subjective factors like PPE availability. Since respondents knew their serostatus at the time of questionnaire completion, this could have become a self-fulfilling prophecy (people who knew they got infected likely to blame lack of PPE). There is also a potential risk of imprecision in answers due to the time elapsed which should also be acknowledged.
Both of these points are already acknowledged and discussed in the existing limitations paragraph, however have now been further expanded.
4. Clarity re dates: I found the chronology difficult to understand. The authors write: "Questionnaire invites were sent between October and November 2020 and covered the period between March 2020 and June 2020 (the time of serological sampling). Questions relating to behavioural and demographic factors were separated by time periods covering March -May and June -July to account for differences in behaviour and exposures outside of occupational environments due to the instigation (March 2020) and easing (June 2021) of the first UK national "lockdown" measures'. Is there a typo here (June 2020?). 'Covered the period' is ambiguous, as it could refer to when the questionnaires were completed, or the time period referred to in the questionnaire. Suggest something along the lines of 'Questionnaire invites were sent between October and November 2020 and questions within them related to participants' recalled behaviour during two periods -March-May 2020 and June-July 2020'. Was the time of serological sampling in June 2020 only or was it done over 4 months (Mar 20 to Jun 20?). The wording is ambiguous.
Elsewhere it is written that ''HCWs were sent questionnaires spanning Mar-Aug 2021.' I am struggling to reconcile this with other dates reported.
A flow chart to illustrate the chronology of serology sampling, consent, questionnaire issue / completion etc would help to clarify the methods.
These have been clarified and a flow chart has been added ( Figure 1).
5. I was interested to see that BMI did not associate with risk of seropositivitya contrast to findings from others e.g. PMID 35189888. Could the authors speculate why this may have been the case?
BMI was not assessed in this analysis. We assessed a univariate relationship with participants previously being told they were "overweight" or "obese" in a medical setting. This did not reveal any positive association, likely due to a combination of variability in measurements and reporting, plus lack of sensitivity of a binary variable when compared to discrete BMI calculations. The was no indication for a sample size calculation, as this questionnaire was sent to all available respondents enrolled in the original longitudinal study. Additionally, the were no prior data available to assess differences between groups on the metrics assessed in the analysis to inform a sample size calculation.
This has been clarified in the text. Yes, this has been changed. environments due to the instigation (March 2020) and easing (June 2021) of the first UK national "lockdown" measures'should this be June 2020?
Yes, this has been changed.
Elsewhere it is written that ''HCWs were sent questionnaires spanning Mar-Aug 2021.' Should this be March to July 2020?
Yes, this has been changed.
2. Abstract and elsewhereif odds ratios adjusted, then they should be reported as such This has been checked and updated where necessary.