Ethnic and social inequalities in COVID-19 outcomes in Scotland: protocol for early pandemic evaluation and enhanced surveillance of COVID-19 (EAVE II)

Abstract

Introduction: Evidence from previous pandemics, and the current COVID-19 pandemic, has found that risk of infection/severity of disease is disproportionately higher for ethnic minority groups, and those in lower socioeconomic positions. It is imperative that interventions to prevent the spread of COVID-19 are targeted towards high-risk populations. We will investigate the associations between social characteristics (such as ethnicity, occupation and socioeconomic position) and COVID-19 outcomes and the extent to which characteristics/risk factors might explain observed relationships in Scotland.The primary objective of this study is to describe the epidemiology of COVID-19 by social factors. Secondary objectives are to (1) examine receipt of treatment and prevention of COVID-19 by social factors; (2) quantify ethnic/social differences in adverse COVID-19 outcomes; (3) explore potential mediators of relationships between social factors and SARS-CoV-2 infection/COVID-19 prognosis; (4) examine whether occupational COVID-19 differences differ by other social factors and (5) assess quality of ethnicity coding within National Health Service datasets.

Methods and analysis: We will use a national cohort comprising the adult population of Scotland who completed the 2011 Census and were living in Scotland on 31 March 2020 (~4.3 million people). Census data will be linked to the Early Assessment of Vaccine and Anti-Viral Effectiveness II cohort consisting of primary/secondary care, laboratory data and death records. Sensitivity/specificity and positive/negative predictive values will be used to assess coding quality of ethnicity. Descriptive statistics will be used to examine differences in treatment and prevention of COVID-19. Poisson/Cox regression analyses and mediation techniques will examine ethnic and social differences, and drivers of inequalities in COVID-19. Effect modification (on additive and multiplicative scales) between key variables (such as ethnicity and occupation) will be assessed.

Ethics and dissemination: Ethical approval was obtained from the National Research Ethics Committee, South East Scotland 02. We will present findings of this study at international conferences, in peer-reviewed journals and to policy-makers.

Keywords: COVID-19; epidemiology; protocols & guidelines; public health.

Figure 1

Data flow diagram for linkage of census and administrative data comprising cohort blue box: census data yellow box: administrative data green box: linked data dark background: data processing. A&E, Accident and Emergency; CHI, Community Health Index; CO-CIN, COVID-19 Clinical Information Network; EAVE II, Early Assessment of Vaccine and Anti-Viral Effectiveness II; ECOSS, Electronic Communication of Surveillance in Scotland; eDRIS, electronic Data Research Innovation Service; GP, general practice; HEPMA, Hospital Electronic Prescribing and Medicines Administration; ISARIC, International Severe Acute Respiratory and Emerging Infection Consortium; NHS, National Health Service; NRS, National Records for Scotland; PIS, Prescribing Information System; RAPID, Rapid Preliminary Inpatient Data; SAS, Scottish Ambulance Service; SICSAG; Scottish Intensive Care Society Audit Group; SMR, Scottish Morbidity Records.

Figure 2

Directed Acyclic Graph of the causal pathways between ethnicity and COVID-19 yellow box: outcome blue box: exposure green box: confounder white box with green outline: confounder (unmeasured) white box with blue outline: mediator blue arrow: main effect green arrow: confounding light green arrow: unmeasured confounding light blue arrow: mediated effect.