Differences in COVID-19 Mortality: the implications of imperfect and diverse data collection systems
Résumé
Daily counts of COVID-19 deaths have quickly become a key indicator to monitor the current unprecedented health crisis. The total number of deaths attributable to COVID-19 appeared to be one reliable indicator to carry out international comparisons international comparisons, provided statistical information is harmonized and standardized. The urgency of the situation has led to substantial variations in the data collection process across countries that have been scrambling to produce information in real time, leading to the collection of imperfect mortality statistics. In this context, we analyze COVID-19 death counts from the “Demography of COVID-19 deaths” database (https://dc-covid.site.ined.fr/en/) and discuss their limitations. We describe and illustrate various data-related issues that may create difficulties for international comparisons. Based on various data characteristics, we propose a completeness-related classification of data sources and analyze and compare COVID-19 death counts for 16 countries using the proposed classification. Finally, we discuss the importance of understanding the characteristics of data collection and formulate some recommendations to deal with imperfect statistics.