2.2. Outcome and covariates

The outcomes were COVID-19 death, and deaths from causes other than COVID-19 (hereafter “non-COVID death”). Cause of death was assigned using the underlying cause of death field (main/primary cause of death, coded in ICD-10) in the death registration. COVID-19 death was defined as any death with the underlying cause coded as U07.1 (“COVID-19, virus identified”) or U07.2 (“COVID-19, virus not identified”) ^[11]. Non-COVID deaths comprised all other deaths; these were also further sub-divided into categories covering the most common causes of death, namely cancer (ICD-10 chapter C), cardiovascular disease (chapter I), respiratory (chapter J), dementia/Alzheimer's disease (F00-03 or G30), and other (all other ICD-10 codes). Two sensitivity analyses were done to check that our findings were robust to the way COVID-19 deaths were defined: (i) only using the U07.1 (“virus identified”) code which would likely have higher specificity; (ii) counting a U07.1/U07.2 code anywhere on the death certificate as a COVID-19 death, in case of variation in how underlying causes were assigned. For the secondary analysis of deaths prior to the pandemic, the outcome was all-cause mortality; this was based on a record for death in primary care, because ONS death registration linkage for 2019 was not available.

Covariates considered in the analysis included health conditions listed in UK guidance on higher risk groups; ^[12] other common conditions that may cause immunodeficiency inherently or through medication, and other postulated risk factors for severe outcomes among COVID-19 cases. We included age (grouped as 18-39, 40-49, 50-59, 60-69, 70-79 and ≥80 years for descriptive analysis), sex, ethnicity (White, Mixed, South Asian, Black, Other, categories from the UK census), obesity (categorised as class I [body mass index 30-34.9kg/m²], II [35-39.9kg/m²], III [≥40kg/m²]), smoking status (never, former, current), index of multiple deprivation quintile (derived from the patient's postcode at lower super output area level). We also considered the following comorbidities: diagnosed hypertension, chronic respiratory diseases other than asthma, asthma (categorised as with or without recent use of oral steroids), chronic heart disease, diabetes (categorised according to the most recent glycated haemoglobin (HbA1c) recorded in the 15 months prior to 1^st February 2020), non-haematological and haematological cancer (both categorised by recency of diagnosis, <1, 1-4.9, ≥5 years), reduced kidney function (categorised by estimated glomerular filtration rate derived from the most recent serum creatinine measure (30-<60, 15-<30, <15 mL/min/1.73m² or a record of dialysis), chronic liver disease, stroke, dementia, other neurological disease (motor neurone disease, myasthenia gravis, multiple sclerosis, Parkinson's disease, cerebral palsy, quadriplegia or hemiplegia, and progressive cerebellar disease), organ transplant, asplenia (splenectomy or a spleen dysfunction, including sickle cell disease), rheumatoid arthritis/lupus/psoriasis, and other immunosuppressive conditions (permanent immunodeficiency ever diagnosed, or aplastic anaemia or temporary immunodeficiency recorded within the last year). The Sustainability and Transformation Partnership (STP, an NHS administrative region) of the patient's general practice was included as an additional adjustment for geographical variation in infection rates across the country.

Information on all clinical covariates was obtained by searching TPP SystmOne records prior to 1^st February 2020 (or prior to 1^st February 2019, for the 2019 secondary analysis cohort) for specific coded data, based on a subset of SNOMED-CT mapped to Read version 3 codes. All codelists, along with detailed information on their compilation are available at https://codelists.opensafely.org for inspection and re-use by the wider research community.