Common protocol for validation of the QCOVID algorithm across the four UK nations

Area of Science:

• Epidemiology
• Biostatistics
• Public Health

Background:

• The QCOVID algorithm is a vital tool for predicting SARS-CoV-2 infection and severe outcomes.
• It informs UK government policy on COVID-19, including shielding and vaccine prioritization.
• Statistical validation is essential to ensure the algorithm's reliability across different UK nations.

Purpose of the Study:

• To present a unified methodology for conducting and reporting validation studies of the QCOVID algorithm.
• To assess the QCOVID algorithm's performance in England, Northern Ireland, Scotland, and Wales.
• To ensure consistent and rigorous evaluation of a key public health tool.

Main Methods:

• Utilizing open, retrospective cohort studies across the four UK nations.
• Assembling linked datasets from primary/secondary care, virology, and death registrations within trusted research environments.
• Calculating performance metrics including Harrell's C, Brier Score, R², and Royston's D.

Main Results:

• Validation exercises are planned for England, Northern Ireland, Scotland, and Wales.
• Population-level data coverage will be sought within each nation.
• Standardized performance metrics will be reported.

Conclusions:

• A common procedure will ensure robust validation of the QCOVID algorithm.
• Findings will support ongoing policy decisions related to the COVID-19 pandemic.
• Dissemination includes policy-makers, conferences, and peer-reviewed publications.