key: cord-0430888-89ooe6ix
authors: Bhaskaran, K.; Rentsch, C. T.; Hickman, G.; Hulme, W. J.; Schultze, A.; Curtis, H. J.; Wing, K.; Warren-Gash, C.; Tomlinson, L.; Bates, C.; Mathur, R.; MacKenna, B.; Mahalingasivam, V.; Wong, A. Y.; Walker, A. J.; Morton, C. E.; Grint, D.; Mehrkar, A.; Eggo, R. M.; Inglesby, P.; Douglas, I. J.; McDonald, H. I.; Cockburn, J.; Williamson, E. J.; Evans, D.; Parry, J.; Hester, F.; Harper, S.; Evans, S. J.; Bacon, S. C.; Smeeth, L.; Goldacre, B.
title: Overall and cause-specific hospitalisation and death after COVID-19 hospitalisation in England: cohort study in OpenSAFELY using linked primary care, secondary care and death registration data
date: 2021-07-19
journal: nan
DOI: 10.1101/2021.07.16.21260628
sha: 55ffec4f0d449a2bc3644eabfc1d58fdf0116cdf
doc_id: 430888
cord_uid: 89ooe6ix

Background: There is concern about medium to long-term adverse outcomes following acute COVID-19, but little relevant evidence exists. We aimed to investigate whether risks of hospital admission and death, overall and by specific cause, are raised following discharge from a COVID-19 hospitalisation. Methods and Findings: Working on behalf of NHS-England, we used linked primary care and hospital data in OpenSAFELY to compare risks of hospital admission and death, overall and by specific cause, between people discharged from COVID-19 hospitalisation (February-December 2020), and (i) demographically-matched controls from the 2019 general population; (ii) people discharged from influenza hospitalisation in 2017-19. We used Cox regression adjusted for personal and clinical characteristics. 24,673 post-discharge COVID-19 patients, 123,362 general population controls, and 16,058 influenza controls were followed for [≤]315 days. Overall risk of hospitalisation or death (30968 events) was higher in the COVID-19 group than general population controls (adjusted-HR 2.23, 2.14-2.31) but similar to the influenza group (adjusted-HR 0.94, 0.91-0.98). All-cause mortality (7439 events) was highest in the COVID-19 group (adjusted-HR 4.97, 4.58-5.40 vs general population controls and 1.73, 1.60-1.87 vs influenza controls). Risks for cause-specific outcomes were higher in COVID-19 survivors than general population controls, and largely comparable between COVID-19 and influenza patients. However, COVID-19 patients were more likely than influenza patients to be readmitted/die due to their initial infection/other lower respiratory tract infection (adjusted-HR 1.37, 1.22-1.54), and to experience mental health or cognitive-related admission/death (adjusted-HR 1.36, 1.01-2.83); in particular, COVID-19 survivors with pre-existing dementia had higher risk of dementia death. One limitation of our study is that reasons for hospitalisation/death may have been misclassified in some cases due to inconsistent use of codes. Conclusions: People discharged from a COVID-19 hospital admission had markedly higher risks for rehospitalisation and death than the general population, suggesting a substantial extra burden on healthcare. Most risks were similar to those observed after influenza hospitalisations; but COVID-19 patients had higher risks of all-cause mortality, readmissions/death due to the initial infection, and dementia death, highlighting the importance of post-discharge monitoring.

Working on behalf of NHS-England, we used linked primary care and hospital 23 data in OpenSAFELY to compare risks of hospital admission and death, overall and by specific cause, 24 between people discharged from COVID-19 hospitalisation (February-December 2020), and (i) 25

demographically-matched controls from the 2019 general population; (ii) people discharged from 26 influenza hospitalisation in 2017-19. We used Cox regression adjusted for personal and clinical 27 characteristics. 28 24,673 post-discharge COVID-19 patients, 123,362 general population controls, and 16,058 influenza 29 controls were followed for ≤315 days. Overall risk of hospitalisation or death (30968 events) was 30 higher in the COVID-19 group than general population controls (adjusted-HR 2.23, 2.14-2.31) but 31 similar to the influenza group (adjusted-HR 0.94, 0.91-0.98). All-cause mortality (7439 events) was 32 highest in the COVID-19 group vs general population controls and 1.73, 33

1.60-1.87 vs influenza controls). Risks for cause-specific outcomes were higher in COVID-19 survivors 34 than general population controls, and largely comparable between COVID-19 and influenza patients. 35

However, COVID-19 patients were more likely than influenza patients to be readmitted/die due to 36 their initial infection/other lower respiratory tract infection (adjusted-HR 1.37, 1.22-1.54), and to 37 experience mental health or cognitive-related admission/death (adjusted-HR 1.36, 1.01-2.83); in 38 particular, COVID-19 survivors with pre-existing dementia had higher risk of dementia death. One 39 limitation of our study is that reasons for hospitalisation/death may have been misclassified in some 40 cases due to inconsistent use of codes. 41

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) emerged in early 2020 and rapidly 50 spread around the world, infecting >140 million people globally. 1 Acute infection can be 51 asymptomatic or mild, 2 but a substantial minority of infected people experience severe COVID-19 52 requiring hospitalisation, 3 with age being a major risk factor, along with male sex, non-White 53 ethnicity and certain comorbidities. [4] [5] [6] Early in the pandemic, the proportion surviving hospitalisation 54 was around 50-70%, 7 though improved treatment guidelines and the identification of effective 55 therapies such as dexamethasone helped to improve survival rates. 8, 9 There is now a large and 56 growing population of people who have survived a COVID-19 hospitalisation, but little is known 57 about their longer-term health outcomes. 58

Emerging evidence suggests that a subset of people infected with SARS-CoV-2 can experience health 59 problems for at least several months after the acute phase of their infection, with fatigue, pain, 60 respiratory and cardiovascular symptoms, and mental health and cognitive disturbances being 61 among the problems frequently described under the term "post-acute COVID-19 syndrome"; 10 62 however, epidemiological characterisation of such sequelae remains immature. Small descriptive 63 studies of COVID-19 survivors have been suggestive of high incidence of a range of outcomes 64

including respiratory, cardiovascular, and mental-health related, but firm conclusions are difficult 65 due to lack of comparison groups. 11, 12 There remains limited evidence about post-COVID sequelae 66

across the full range of health outcomes. One recent study of US Department of Veterans Affairs 67 (VA) data examined a wide range of diagnoses, prescriptions and laboratory abnormalities among 68 30-day survivors of COVID-19, showing excess risks of several health outcomes in the 6 months 69

following infection, compared with the general VA population. 13 Whether these findings generalise 70

to the entire US population or other settings remains unclear. Another US study limited to people 71 aged <65 years also found excess risks of a range of clinical outcomes ascertained from health 72 insurance data among people with a record of SARS-CoV-2 infection. 14 A UK study of routinely-73 collected primary care and hospitalisation data described raised rates of all-cause hospital admission 74 and death among patients discharged following a COVID-19 hospitalisation; the authors also noted 75 raised risks of adverse respiratory and cardiovascular sequalae among the selected outcomes 76

investigated. 15 Only a general population comparator was used, making it difficult to disentangle 77 risks specific to COVID-19 from those associated with hospitalisation more generally; furthermore, a 78 hospitalised cohort is likely to have been more prone to health problems at the outset than the 79 general population comparator group. 80

To strengthen the evidence base in this important emerging area, we therefore aimed to investigate 81 the incidence of subsequent hospital admission and death, both overall and from a wide range of 82 specific causes, following a COVID-19 hospitalisation in England. We aimed to compare post-COVID 83 risks with two separate comparison groups: (i) the general population, and (ii) people hospitalised 84

for influenza prior to the current pandemic. 85 86 87 88 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

Study design and study population 90 A cohort study was carried out within the OpenSAFELY platform, which has been described 91

previously. 6 We used routinely-collected electronic data from primary care practices using TPP 92

SystmOne software, covering approximately 40% of the population in England, linked at the 93 individual patient level to NHS Secondary Uses Service (SUS) data on hospitalisations, and Office of 94

National Statistics death registration data (from 2019 onwards). We selected all individuals 95 discharged between 1 st February and 30 th December 2020 from a hospitalisation that lasted >1 day 96 and where COVID-19 was coded as the primary diagnosis (based on the International Classification 97

of Diseases (ICD)-10 codes U07.1 "COVID-19 -virus identified" and U07.2 "COVID-19 -virus not 98

identified") and who were alive and under follow-up one week after discharge (to avoid a focus on 99 hospital transfers and immediate readmissions/deaths). We excluded a small number of people with 100 missing age, sex, or index of multiple deprivation, which are likely to indicate poor data quality. Two 101 comparison groups were also selected: (i) people under follow-up in the general population in 2019, 102

individually matched 5:1 to the COVID-19 group on age (within 3 years), sex, Sustainability and 103

Transformation Plans (STP, a geographical area used as in NHS administration, of which there were 104 32 in our data), and calendar month (e.g. a patient discharged from a COVID-19 hospitalisation in 105

April 2020 was matched to 5 individuals of the same age, sex and STP who were under follow-up in 106 general practice on 1 st April 2019); (ii) all individuals discharged from hospital in 2017-2019 where 107 influenza was coded as the primary reason for hospitalisation and who were alive and under follow-108 up one week after discharge. 109

110

The outcomes were (i) time to first hospitalisation or death (composite outcome); (ii) all-cause 112 mortality; and (iii) time to first cause-specific hospitalisation or death. Hospitalisations were 113 identified from linked SUS data. All-cause mortality was identified using date of death in the primary 114 care record so that deaths before 2019 (in the influenza group) could be included (as linked ONS 115 data were not available prior to 2019); concordance of death dates between primary care and linked 116 ONS data has been shown to be high. 16 Cause-specific outcomes were categorised based on ICD-10 117 codes into infections (ICD-10 codes beginning with "A"), cancers except non-melanoma skin cancer 118 (C, except C44), endocrine/nutritional/metabolic (E), mental health and cognitive (F, G30 and X60-119 84), nervous system (G, except G30), circulatory (I), COVID-19/influenza/pneumonia/other lower 120 respiratory tract infections (J09-22, U07.1/2), other respiratory (J23-99), digestive (K), 121 musculoskeletal (M), genitourinary (N), and external causes (S-Y, except X60-84). For each of these, 122 the outcome was time to the earliest of hospitalisation with the relevant outcome listed as primary 123 diagnosis, or death with the relevant outcome listed as the underlying cause on the death 124 certificate. 17 The influenza control group was restricted to those discharged in 2019 for analyses of 125 these cause-specific outcomes, because we did not have linked death registration data (and thus 126 cause of death) for earlier years. 127

Other covariates considered in the analysis were factors that might be associated with both risk of 128 severe COVID-19 and subsequent outcomes, namely age, sex, ethnicity, obesity, smoking status, 129

index of multiple deprivation quintile (derived from the patient's postcode at lower super output 130 area level), and comorbidities considered potential risk factors for severe COVID-19 outcomes (see 131  Table 1 and footnotes for full specification of covariate categories and comorbidities). 132

Information on all covariates was obtained by searching TPP SystmOne records for specific coded 133 data, based on a subset of SNOMED-CT mapped to Read version 3 codes. Covariates were identified 134 using data prior to the patient's hospital admission date (for the COVID-19 and influenza groups) or 135 the index date (for the matched control group, i.e. 1 st day of the matched calendar month in 2019). 136

For the COVID-19 and influenza hospitalised groups, primary care data on ethnicity was 137 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 19, 2021. ;  supplemented with information from the hospitalisation record, to improve completeness. All 138

codelists, along with detailed information on their compilation are available at 139 https://codelists.opensafely.org for inspection and re-use by the wider research community. 140 141

Follow-up began on the 8 th day after hospital discharge for the COVID-19 and influenza groups, and 143

on the 1 st of the same calendar month in 2019, for the general population control group. Follow-up 144 ended at the first occurrence of the analysis-specific outcome, or the earliest relevant censoring date 145 for data availability/coverage for the outcome being analysed; the control groups were additionally 146 censored after the maximum follow-up time of the COVID-19 group ( with influenza controls (models adjusted for age, sex, STP and calendar month). The additional 156

covariates noted above were then added to the models. People with missing data on ethnicity, body 157 mass index or smoking were excluded from models that used these variables ("complete case 158

analysis", which is valid under the assumption that missingness is conditionally independent of the 159 outcome); 18 imputation was not used because these variables were thought to be missing not at 160 random in primary care (e.g. smokers more likely to have smoking status recorded). Cumulative 161 incidence of cause-specific hospitalisation/death outcomes were calculated with deaths from other 162 causes treated as a competing risk. Hazard ratios for these outcomes were then estimated from a 163

Cox model targeting the cause-specific hazard, with deaths from competing risks censored. 164 CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Results 170 24,673 individuals discharged after a COVID-19 hospitalisation were included, alongside 123,362 171 matched controls from the 2019 general population, and 16,058 individuals discharged after 172 influenza hospitalisation in 2017-19 (Appendix Figure A1 -A2). 173

At entry, the COVID-19 group had similar age and sex distribution to the general population groups 174 due to matching, but had younger median age and were more likely to be male than the influenza 175 group (Table 1) . Body mass index, smoking and ethnicity data were 93-99% complete in all groups, 176 except that ethnicity was 25% missing in the matched control group (no hospital-based ethnicity 177 records were available for this group). The COVID-19 group were more likely to be obese, non-White 178

and less likely to be current smokers than both comparison groups. Pre-existing comorbidities were 179 more common in both COVID-19 and influenza-discharged patients than general population 180 controls. COVID-19 patients had longer median duration of hospital stay and were more likely to 181

have received critical care during their admission than influenza patients. 182

Numbers of outcome events are shown in Appendix Table A1 . Cumulative incidence of subsequent 183

hospital admission or death after study entry in the COVID-19 group was higher than in general 184 population controls, but similar to that in the influenza group (cumulative incidence at 6 To further explore this, causes of death were examined (appendix Table A2 ). A substantial 191

proportion of deaths in the COVID-19 group had COVID-19 listed as the underlying cause (500/2022, 192 24.7%), while in the influenza group ≤5 deaths were coded with influenza as the underlying cause. This was further explored in a post-hoc analysis of specific outcomes within the mental health and 203 cognitive category ( hospitalisations). Higher rates of hospitalisations/deaths due to mood disorders, and 208 neurotic/stress-related/somatoform disorders were also observed in COVID-19 patients, but 209 confidence intervals were too wide to be conclusive. 210

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 19, 2021. following non-hospitalised 19 this is in contrast to a recent study using US health insurance 232 data which found raised risks of a range of outcomes among a relatively young cohort with mostly 233 (92%) non-hospitalised COVID-19 disease, compared with both the general population and people 234 with a record of other viral lower respiratory tract infections. 14 235

Our data showed that COVID-19 hospitalised patients were more likely to have baseline 236 comorbidities than general population controls, reflecting known associations between 237 comorbidities and risks of severe COVID-19 outcomes. 6 Differences in outcomes between 238 hospitalised patients and general population controls might therefore reflect baseline differences 239 not fully captured in our adjustment models, and might also reflect a generic adverse effect of 240 hospitalisation. 20 This is supported by the more similar risks we observed when COVID-19 survivors 241

were compared with people who had experienced influenza hospitalisation. However, all-cause 242 mortality was substantially higher after COVID-19 compared with influenza. A quarter of deaths after 243 a COVID-19 hospitalisation had COVID-19 listed as the underlying cause, but it is not clear from our 244 data whether patients experienced specific complications after hospital discharge that were then 245 attributed to Our analysis of cause-specific outcomes also suggested a disproportionate rate of dementia deaths 247 post-COVID-19, particularly among those with pre-existing dementia. Cognitive decline after 248 hospitalisation and critical illness have been previously described; 21,22 acute COVID-19 and 249 associated hospital admission, social isolation, and medications may have accelerated progression of 250 patients' dementia; it is unclear whether post-discharge care was adequate for this vulnerable 251 group. However, it is possible that deaths where the underlying cause was recorded as dementia 252 may have been due to progression of underlying health problems following an acute illness as well 253

as difficulty in managing these due to dementia. Due to small numbers, we could not confirm 254 whether higher rates of mood disorders and neurotic/stress-related/somatoform disorders after 255 COVID-19 compared with influenza were due to chance, but a number of previous studies outside 256 the pandemic context have found that critical illness is associated with raised risks of depression, 257 anxiety and post-traumatic stress. 23-25 It will be important to continue to monitor these outcomes as 258 more follow-up accumulates. 259 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 19, 2021. ; https://doi.org/10.1101/2021.07.16.21260628 doi: medRxiv preprint

We identified COVID-19 hospitalisations from a base population covering around 40% of the 261 population of England, giving us high statistical power. We examined a broad range of 262 hospitalisation and mortality outcomes, and were able to describe and adjust for a wide range of 263 personal and clinical characteristics using rich primary care data. 264 265

However, our study has some limitations. We relied on ICD-10 codes entered as the primary reason 266

for hospitalisation or underlying cause of death to define our cause-specific outcomes, but these 267 fields may not have been used consistently. 26 In particular, there might have been a tendency for 268 clinicians aware of a recent COVID-19 hospitalisation to code COVID-19 for a range of clinical 269 complications, masking more specific sequelae. Outcomes were classified in broad categories to 270

obtain an overview of post-COVID-19 disease patterns; more granular disease categories would be 271 of future interest but will require more follow-up to maintain statistical power. 

Patients surviving a COVID-19 hospitalisation were at substantially higher risk than the general 290 population for a range of subsequent adverse outcomes over a period of up to 10 months' follow-up 291 included in this study. Risks for most outcomes were broadly comparable to those experienced by 292 influenza hospitalisation survivors prior to the pandemic, but in the period following hospital 293 discharge COVID-19 patients had higher risks of all-cause mortality, readmission or death attributed 294

to their initial infection, and adverse mental health and cognitive outcomes; in particular, among 295 people with pre-existing dementia, we observed an excess of deaths where dementia was recorded 296 as the underlying cause. These findings suggest a need for services to support and closely monitor 297 people following discharge from hospital with COVID-19. 298 299 300 301 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 19, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 19, 2021. Patients were not formally involved in developing this specific study design that was developed 378 rapidly in the context of a global health emergency. We have developed a publicly available website 379 https://opensafely.org/ through which we invite any patient or member of the public to contact us 380 regarding this study or the broader OpenSAFELY project. 381 382 383 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 19, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 19, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 6th August 2020 controllers-to-share-information (accessed 6th August 2020). 500 501 502 503 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 19, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 19, 2021. ; CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 19, 2021. ; https://doi.org/10.1101/2021.07.16.21260628 doi: medRxiv preprint Figure 2 : Hazard ratios comparing exposed and controls for risk of 522 subsequent hospital admission or death (composite outcome) and all-cause mortality 523 524 *among those with complete data available (see Table 1 ) 525 526 527 528 529 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 19, 2021. ; https://doi.org/10.1101/2021.07.16.21260628 doi: medRxiv preprint CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted July 19, 2021. ; https://doi.org/10.1101/2021.07.16.21260628 doi: medRxiv preprint Figure 4 : Hazard ratios comparing exposed and controls for cause-538 specific hospital admission/deaths, adjusted for age, sex and geography 539 540 *In the influenza group, only patients entering the study in 2019 were included in analysis of cause-specific outcomes, as linked cause of 541 death data were only available from 2019 onwards.

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted July 19, 2021. ; https://doi.org/10.1101/2021.07.16.21260628 doi: medRxiv preprint Hospital admissions with duration >1 day in OpenSAFELY population, with COVID-19 listed anywhere on the hospitalisation N = 108,308

Discharged before admin censoring date N 41,542

Discharged on or after administrative censoring date (= 60d before last date of data availability in SUS) N = 33,711

Missing/implausible age (n<=5), sex (n<=5), or index of multiple deprivation (n=583) N = 587

Final included COVID-19 hospitalisations N = 24,673

Comparison groups: 1) Control population from the 2019 general population, matched on age (±3 years), sex, STP, calendar month N = 79,502 (full 5 matches found for >99% of exposed patients; >99% matches were within 1 year of age)

2) Patients hospitalised with influenza as the primary reason, during the period 2017-2019 N = 16,058 (6, 689 were discharged in 2019 and had linked data registration available) Alive for >7d after discharge N = 75,840

Died during hospitalisation (n=28,652) or within 7 days after discharge (n=1573), or hospital/hospice transfer (n=2243) N = 30,070

Complete age/sex/deprivation data N = 75,253 COVID-19 not the primary listed reason for hospitalisation N = 16,869 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted July 19, 2021. ; https://doi.org/10.1101/2021.07.16.21260628 doi: medRxiv preprint Figure A2 : Distribution of entry dates for those in the hospitalised and matched general population comparison groups 549 550 551 CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted July 19, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 

World Health Organization. WHO Coronavirus (COVID-19) Dashboard. 2021

Prevalence of Asymptomatic SARS-CoV-2 Infection

Filling the gaps in the characterization 390 of the clinical management of COVID-19: 30-day hospital admission and fatality rates in a cohort of 391 118 150 cases diagnosed in outpatient settings in Spain

High-dimensional characterization of post-acute sequalae of COVID-439 19

Risk of clinical sequelae after the acute phase of SARS-CoV-2 infection: retrospective cohort 442 study

The accuracy of date 447 of death recording in the Clinical Practice Research Datalink GOLD database in England compared 448 with the Office for National Statistics death registrations

Bias and efficiency of multiple imputation compared with complete-case 454 analysis for missing covariate values

Post-acute effects of SARS-CoV-2 infection in individuals not requiring hospital 457 admission: a Danish population-based cohort study. The Lancet Infectious Diseases

Post-hospital syndrome--an acquired, transient condition of generalized risk

Long-Term Cognitive Impairment after Critical Illness

Anxiety 469 symptoms in survivors of critical illness: a systematic review and meta-analysis

Posttraumatic stress disorder in critical illness survivors: a metaanalysis

Needham 475 DM. Depressive Symptoms After Critical Illness: A Systematic Review and Meta-Analysis

COVID-19 as the underlying cause of death: disentangling facts 478 and values

489 information/looking-after-information/data-security-and-information-governance/data-security-490 and-protection-toolkit (accessed 6th

Note: counts for cause specific deaths in square brackets are deaths which were not preceded by a previous hospitalisation in the same 553 outcome category; these are the deaths that counted as outcome events in the cause-specific analysis, and not the total number of deaths 554 from the given cause. *In the influenza group, only patients entering the study in 2019 were included in analysis of cause-specific outcomes, 555 as linked cause of death data were only available from 2019 onwards

COVID-19/Influenza/LRTI outcome category had COVID-19 codes (ICD-10 U07.1/2) and a further 461/1122 (41%) had pneumonia codes (J18) 557 as the primary diagnosis