New tool predicts risks of hospital admission and death from COVID-19

Photo credit: CC0 Public Domain

The BMJ today released a new risk tool developed by UK researchers to predict a person’s risk of being hospitalized and dying from COVID-19.

With the rise in cases in the UK and elsewhere and with winter approaching, there is an urgent need for reliable models that predict the likely course of COVID-19 to aid decisions about screening, hospitalization, treatment and vaccination.

The Risk Prediction Tool (known as QCOVID) uses readily available information about people, such as age, ethnicity, and whether they have certain pre-existing medical conditions (comorbidities), to identify those at highest risk for developing a serious illness. It is intended for the general adult population in the UK.

The tool offers differentiated information on the risk of a serious illness due to COVID-19 and was developed for doctors with patients in order to achieve a common understanding of the risk.

The tool needs regular updates as the pandemic develops and performance is closely monitored.

Some previous risk forecasting models have been developed. They were found to have a high risk of bias, raising concerns that these models may be unreliable in practice.

The UK research group aims to develop and validate a population-based predictive model to estimate the overall risk of infection with and subsequent hospitalization or death from COVID-19. Steps have been taken to alleviate known bias factors.

Their findings are based on data from more than 8 million patients, ages 19 to 100, from 1,205 general practitioners in England linked to COVID-19 test results and data from hospital and death records.

Data from 6 million patients were used to develop the model over a period of 97 days (January 24 to April 30, 2020) and an additional 2.2 million patients to evaluate its performance over two separate periods (January 24 to April 30, 2020) April 30, 2020 and May 1) to be validated by June 30, 2020) during the first wave of the pandemic.

Known factors such as age, ethnicity, deprivation, body mass index, and a range of comorbidities were used to develop the model to estimate the likelihood and timing of hospitalization or death from COVID-19.

During the study period, 4,384 deaths from COVID-19 occurred in the development group, 1,782 in the first validation period and 621 in the second validation period.

The model performed well, predicting 73% and 74% of the variation in time to death from COVID-19 in men and women, respectively.

People in the top 5% of projected risk of death accounted for 76% of COVID-19 deaths within the 97-day study period. 94% of COVID-19 deaths were from people in the top 20% of the projected risk of death.

The researchers point out that the model is intended to predict risk – it is not intended to explain which individual factors causally affect risk, and the results should not be interpreted in that way.

The absolute risks resulting from the model will change over time according to the prevailing COVID-19 infection rate and the extent of social distancing measures in place. Therefore, these should also be interpreted with caution. However, it is expected that the order of people in relation to their risk will remain relatively stable over time so that those at highest risk can be identified.

The researchers say QCOVID represents a robust risk prediction model that has the potential to support public health policies, from collaborative decisions to mitigate health and workplace risks to targeted recruitment for clinical trials and prioritization for vaccinations.

The model can also be recalibrated for different time periods and can be updated regularly as the pandemic develops.

Although QCOVID is specifically designed to inform UK health policy and interventions to manage COVID-19 related risks, it has international potential, subject to local validation.

In a linked editorial, researchers from the University of Manchester agree that QCOVID and the International Severe Acute Respiratory and Emerging Infection Consortium (ISARIC) 4C (Coronavirus Clinical Characterization Consortium) mortality score show advances in the quality of COVID-19 predictive models, however Care should be taken in interpreting the predictions generated by these models.

Given the rapidly changing nature of the disease and its management, they also emphasize the need to regularly update these models and closely monitor their performance across time and space.

They acknowledge that improved data on COVID-19 incidents “allow more accurate prediction” and with these qualifications they say, “We support the continued validation and impact assessment of these models.”

Follow the latest news on the Coronavirus (COVID-19) outbreak

More information:
Life Risk Prediction Algorithm (QCOVID) for the risk of hospitalization and mortality from COVID-19 in adults: National Cohort Study on Derivation and Validation, BMJ (2020). DOI: 10.1136 / bmj.m3731,

Editorial: Predictive models for the results of the coronavirus 2019, BMJ (2020).

Supplied by the British Medical Journal

Quote: New tool predicts risks of hospitalization and death from COVID-19 (2020, October 20), accessed October 20, 2020 from covid-. html

This document is subject to copyright. Except for fair trade for the purpose of private study or research, no part may be reproduced without written permission. The content is provided for informational purposes only.