Letters | With talk of a sixth Covid-19 wave in Hong Kong, can we trust data models?

Given that different assumptions made by researchers will generate very different projections, claims made by health experts should be accompanied by a disclosure of assumptions made in the research

Reading Time:2 minutes

Shoppers wearing masks flock to Causeway Bay on Easter Sunday on April 17. Photo: Nora Tam

Published: 11:30am, 19 May 2022

Feel strongly about this letter, or any other aspects of the news? Share your views by emailing us your Letter to the Editor at letters@scmp.com or filling in this Google form. Submissions should not exceed 400 words, and must include your full name and address, plus a phone number for verification.

A comment from Professor Gabriel Leung, dean of the faculty of medicine at the University of Hong Kong, about a potential sixth Covid-19 wave has sparked debate. Mathematical modelling is often used to project Covid-19 trends and guide policies. Professor Leung’s team at HKU used data models to predict the number of Covid-19 infections, deaths and hospitalisations in different scenarios during the fifth wave of Covid-19 in Hong Kong.

Can we trust these predictions? According to the projections published by the HKU team on February 21, the daily number of Covid deaths in Hong Kong would peak at nearly 100 by late March and the cumulative number of deaths by mid-May would be around 3,206. In a subsequent update on March 14, which incorporated data on outbreaks in elderly care homes, the estimated cumulative death toll by May 1 was revised to 5,102. On May 1, what actually happened was that a cumulative total of 9,100 Covid-related deaths were recorded since the beginning of the fifth wave, nearly 4,000 more than what the experts had anticipated.

Projections using mathematical modelling involve assumptions and estimates made by researchers. Previous reports published by the HKU team clearly stated the assumptions and estimates made, that is, the basis on which their predictions were made. For example, one of the assumptions influencing the predicted outcome was the estimated effectiveness of tightened social restrictions announced by the government.

Different assumptions made by researchers will generate very different projections. It is important that any claims made by health experts – like the one related to the sixth wave – should be accompanied by a disclosure of assumptions made in the research. The lack of such transparency will only preclude informed discussion of new findings.

Predictions can go wrong, especially when they involve changing human responses and environmental factors, coupled with a new, evolving virus. In the United Kingdom, several models by leading scientists have been criticised for overestimating the number of daily infections and deaths at different phases of the pandemic.