Professor Daisuke Sano of Tohoku University, who developed a model for predicting the spread of the coronavirus by monitoring sewage water, is seen in Sendai on Nov. 19, 2021. (Mainichi/Makoto Fukazu)

       SENDAI -- A project to predict the coronavirus's spread by examining sewage is being carried out by a group at Tohoku University's School of Engineering.

       Amid concerns of a potential sixth infection wave emerging, the team estimates new case numbers for the coming week based on virus amounts in sewage. It hopes local governments will use the information to secure adequate hospital bed numbers.

       The research group, led by professor Daisuke Sano, has been investigating virus concentrations in waste water since August 2020. Samples are taken twice weekly for analysis at a treatment plant that about 70% of the city of Sendai's sewage flows to.

       From the third to fifth waves, a strong correlation between detected virus amounts and new infection cases emerged. A computational model based on the data was built with machine learning to produce predictions for the next week's new case numbers.

       The virus is detectable in sewage when one in 10,000 people is infected. Its latest forecast released Nov. 22 showed 25 new cases were expected between Nov. 22 and Nov. 28. Although the system is still in the testing stage, comparing the results with past projections reportedly allows for "increasing or decreasing trends to be seen."

       Sano and other researchers noticed links between norovirus concentrations in sewage water and norovirus patient numbers. Since 2017, they have been working with the Sendai Municipal Government to release data on norovirus concentrations in waste water via a dedicated website. Since Nov. 8, coronavirus prediction results have also been emailed to a mailing list with about 2,000 registered users so far.

       Nowadays, actual new infection totals tend to come in lower than the projections due to the progression of vaccinations and the spread of correct mask usage. In future, the actual data will be reflected in the calculation model to improve its accuracy.

       Sano said that in countries including South Korea and Germany, infections among older people have risen as the passage of time has reduced the effectiveness of their second vaccine shots.

       "In Sendai, elderly people vaccinated in May and June may see their shots' efficacy start to fade around December. If third shots aren't administered, new case numbers may approach predicted levels again."

       (Japanese original by Makoto Fukazu, Sendai Bureau)

       Font Size S M L Print Timeline 0