Search

Background This study aimed to describe the mortality excess during the three first epidemic periods of COVID-19 in all regions of France. Methods Two complementary approaches were implemented. First, we described the number of death of patients infected with or diagnosed with COVID-19 in health care (HC) and medico-social (MS) institutions. Then, we estimated general all-cause mortality excess (all ages) by comparing the mortality observed with the expected mortality. We used a daily number of death model according to a negative binomial distribution, as a function of the long-term trend in mortality (penalized spline function of time) and its seasonal variation (cyclic spline function). The model provided expected mortality during epidemic periods with a 95% credibility interval. Each region defined three epidemic periods, including the overseas territories. Results The two approaches were consistent in the most affected regions but there are major regional disparities that vary according to the epidemic period. There is an east–west gradient in the relative excess of deaths from all-causes during each epidemic period. The deaths observed in HC and MS institutions alone do not explain the excess (or deficit) of mortality in each region and epidemic period. Conclusion An analysis by age group according to the two approaches and a comparison of death specific causes could provide a better understanding of these differences. Electronic death registration system (mortality by medical causes) would allow a rapid mortality related estimation to an emerging pathology like Coronavirus Disease-2019 (COVID-19) but is still insufficient for real-time medical causes of death monitoring. [ABSTRACT FROM AUTHOR]