Risk prediction of covid-19 related death and hospital admission in adults after covid-19 vaccination: national prospective cohort study

Julia Hippisley-Cox, Carol A C Coupland, Nisha Mehta, Ruth H Keogh, Karla Diaz-Ordaz, Kamlesh Khunti, Ronan A Lyons, Frank Kee, Aziz Sheikh, Shamim Rahman, Jonathan Valabhji, Ewen M Harrison, Peter Sellen, Nazmus Haq, Malcolm G Semple, Peter WM Johnson, Andrew Hayward, Jonathan S. Nguyen-Van-Tam

Research output: Contribution to journalArticlepeer-review

Abstract / Description of output

Objectives: To derive and validate risk prediction algorithms to estimate risk of COVID-19 mortality and hospitalisation in UK adults following one or two doses of COVID-19 vaccination.

Design: Population-based cohort study using the QResearch database linked to COVID-19 vaccination, SARS-CoV-2 results, hospitalisation, cancer registry, systemic anticancer treatment, radiotherapy and the national death registry.

Settings: Adults aged 19-100 years with one or two doses of COVID-19 vaccination between December 8, 2020 and June 15, 2021.

Main outcome measures:Primary outcome was COVID-19 death. Secondary outcome COVID-19 hospital admission. Outcomes assessed from 14 days following each vaccination dose. Models fitted in the derivation cohort to derive risk equations using a range of predictor variables. Performance evaluated in a separate validation cohort of GP practices.

Results: In the derivation cohort of 6,952,440 vaccinated patients, 5,150,310 (74.08%) had two doses. There were 2,031 COVID-19 deaths and 1,929 COVID-19 hospital admissions, of which 81 deaths (4.0%) and 71 admissions (3.7%) occurred 14 or more days after the second vaccine dose.

The risk algorithms included age, sex, ethnicity, deprivation, BMI, and a range of comorbidities. COVID-19 mortality incidence increased with age and deprivation, male sex and Indian and Pakistani ethnicity. Cause-specific hazard ratios were highest for those with Down’s syndrome (12.7-fold increase); kidney transplant (8.1-fold); sickle cell disease (7.7-fold); care home residency (4.1-fold); chemotherapy (4.3-fold); HIV/AIDS (3.3-fold); liver cirrhosis (3-fold); neurological conditions (2.6-fold); bone marrow transplant or a solid organ transplant (2.5-fold); dementia (2.2-fold); Parkinson’s disease (2.2-fold). Other conditions with increased risk (ranging from 1.2 to 2-fold) included chronic kidney disease, blood cancer, epilepsy, chronic obstructive pulmonary disease, coronary heart disease, stroke, atrial fibrillation, heart failure, thromboembolism, peripheral vascular disease and type 2 diabetes. There was a similar pattern of associations for the hospital admission outcome. There was no evidence that associations differed after the second dose, though absolute risks were substantially reduced.

The risk algorithm explained 74.1% (95% CI 71.1 to 77.0) of the variation in time to COVID-19 death. Discrimination was high with a D statistic of 3.46 (95% CI 3.2 to 3.7) and C statistic of 92.5. Performance was similar after each vaccine dose. In the top 5% of patients with the highest predicted mortality risk, the sensitivity for identifying COVID-19 deaths within 90 days was 78.7%.

Conclusion: This population-based risk algorithm performed well showing very high levels of discrimination for identifying those at highest risk of post-vaccination break-through COVID-19 death and hospital admission.
Original languageEnglish
Article number2244
JournalBritish Medical Journal (BMJ)
Publication statusPublished - 17 Sept 2021


Dive into the research topics of 'Risk prediction of covid-19 related death and hospital admission in adults after covid-19 vaccination: national prospective cohort study'. Together they form a unique fingerprint.

Cite this