Back

On the lag between deaths and infections in the first phase of the Covid-19 pandemic

Chrusciel, P.; Szybka, S.

2021-01-04 public and global health
10.1101/2021.01.01.21249115 medRxiv
Show abstract

One of the key issues in fighting the current pandemic, or the ones to come, is to obtain objective quantitative indicators of the effectiveness of the measures taken to contain the epidemic. The aim of this work is to point out that the lag between the daily number of infections and casualties provides one such indicator. For this we determined the lag during the first phase of the Covid-19 pandemic for a series of countries using the data available at the server of the John Hopkins University using three different methods. Somewhat surprisingly, we find a lag varying substantially between countries, taking negative values (thus the maximum daily number of casulties preceding the maximum daily namber of new infections) in countries where no steps to contain the epidemic have been taken at the outset, with an average lag of 7 {+/-} 0.3 days. Our results can be useful to health authorities in a search for the best strategy to fight the epidemic. Key MessagesO_LIThe lags between the maximum daily infections and casualties during the first phase of the Covid-19 pandemic differ widely between countries. C_LIO_LIThese lags are clear for some countries, but impossible to determine confidently for most. C_LIO_LIIn some countries the day at which the maximal number of daily deaths is attained precedes the day of the maximal number of casualties, indicating a failure to protect the most vulnerable part of the population. C_LIO_LIThe lags can serve as an objective quantitative measure of the effectiveness of the measures taken to contain the epidemic. C_LI

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.