Using national electronic health records for pandemic preparedness: validation of a parsimonious model for predicting excess deaths among those with COVID-19 – a data-driven retrospective cohort study
There have been over 5 million deaths due to COVID-19 globally, with more than 144,000 in the UK. Estimates of mortality have been the most reported and used statistics since the start of the pandemic, influencing policy and planning. Use of EHR informed early identification of risk factors for COVID-19 mortality, leading to social distancing and lockdown in the UK.
One of the crucial predictors of mortality in infectious disease is the existence of underlying chronic diseases. However, baseline mortality risk in people with underlying conditions is not used in common models for estimating the mortality of infection diseases. Most of the predictive models for mortality of infectious diseases, including COVID-19, are based on transmissibility, the severity of the infection, case fatality ratio, infection fatality ratio and R-0, many of which are unknown at the start of the pandemic.
In March 2020, we estimated 1-year COVID-19 mortality in England using the EHR of 3.8 million patients provided by Clinical Practice Research Datalink (CPRD). The main parameters in our predictive model were baseline all-cause mortality in people with underlying chronic conditions and combinations of scenario-based values of relative risk (RR) of mortality and infection rate (IR) of COVID-19.
This project aims to validate our previous model in 44.9 million patients aged 30 and older in NHS Digital's Trusted Research Environment (TRE) for England. In our development and validation models, we are interested in the direct effects of the pandemic on 1-year all-cause mortality in people with underlying chronic conditions listed at high-risk factors for COVID-19.
Mizani MA, Dashtban A, Pasea L, et al. Using national electronic health records for pandemic preparedness: validation of a parsimonious model for predicting excess deaths among those with COVID-19–a data-driven retrospective cohort study. Journal of the Royal Society of Medicine. 2022;0(0). doi: https://doi.org/10.1177/01410768221131897
- View the analysis code used in NHS Digital's TRE for England
- View the phenotyping algorithms and codelists used in NHS Digital's TRE for England
This is a sub-project of project CCU003 approved by the CVD-COVID-UK / COVID-IMPACT Approvals & Oversight Board (sub-project: CCU003_03).
Licensed under the Apache License, Version 2.0 (the "License"); you may not use this software except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0. Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.