Coverage, completion and outcomes of COVID-19 risk assessments in a multi-ethnic nationwide cohort of UK healthcare workers: a cross-sectional analysis from the UK-REACH Study

Abstract

Introduction: There are limited data on the outcomes of COVID-19 risk assessment in healthcare workers (HCWs) or the association of ethnicity, other sociodemographic and occupational factors with risk assessment outcomes.

Methods: We used questionnaire data from UK-REACH (UK Research study into Ethnicity And COVID-19 outcomes in Healthcare workers), an ethnically diverse, nationwide cohort of UK HCWs. We derived four binary outcomes: (1) offered a risk assessment; (2) completed a risk assessment; (3) working practices changed as a result of the risk assessment; (4) wanted changes to working practices after risk assessment but working practices did not change.We examined the association of ethnicity, other sociodemographic/occupational factors and actual/perceived COVID-19 risk variables on our outcomes using multivariable logistic regression.

Results: 8649 HCWs were included in total. HCWs from ethnic minority groups were more likely to report being offered a risk assessment than white HCWs, and those from Asian and black ethnic groups were more likely to report having completed an assessment if offered. Ethnic minority HCWs had lower odds of reporting having their work change as a result of risk assessment. Those from Asian and black ethnic groups were more likely to report no changes to their working practices despite wanting them.Previous SARS-CoV-2 infection was associated with lower odds of being offered a risk assessment and having adjustments made to working practices.

Discussion: We found differences in risk assessment outcomes by ethnicity, other sociodemographic/occupational factors and actual/perceived COVID-19 risk factors. These findings are concerning and warrant further research using actual (rather than reported) risk assessment outcomes in an unselected cohort.

Keywords: COVID-19; Ethnic Groups; Health Personnel; Risk assessment.

Figure 1

Formation of the analysed sample and derivation of outcome measures. *at time of questionnaire completion. NHS, National Health Service.

Figure 2

Adjusted ORs and 95% CIs are shown for the association of sociodemographic, occupational and perceived risk variables with four outcomes relating to NHS COVID-19 risk assessments derived from multivariable logistic regression. ORs are adjusted for age, sex, ethnicity, occupation and migration status. All outcome variables are binary and all analyses include only HCWs who were working for the NHS during UK national lockdown and at the time of questionnaire completion. Outcome 1 is whether or not an HCW was offered a COVID-19 risk assessment; the analysed sample includes all HCWs who provided information on their ethnicity and the outcome of interest. Certain risk perception variables were not included in this analysis (as shown by the lack of ORs) as it was felt that these were unlikely to influence being offered a risk assessment. Outcome 2 is whether or not an HCW chose to complete a risk assessment; the analysed sample includes all those in the previous analysis who were offered a risk assessment. Outcome 3 is whether or not changes were made to the working practices of HCWs after the risk assessment; the analysed sample includes HCWs who completed a risk assessment and provided information on the outcome with those indicating that work did not need to change excluded. Outcome 4 is whether or not an HCW wanted changes to be made at work; the analysed sample includes all HCWs who indicated their work did not change, excluding those who indicated that their work did not need to change. Derivation of all outcomes is described in detail in figure 1. In the occupation variable, nursing includes midwives, nursing associates and healthcare assistants; allied health professionals include pharmacists, healthcare scientists, ambulance workers and those in optical roles. BMI, body mass index; HCWs, healthcare workers; NHS, National Health Service.