A polynomial regression model for excess mortality in Mexico 2020-2022 due to the COVID-19 pandemic

TL;DR

This paper develops a polynomial regression model to estimate excess mortality in Mexico during 2020-2022 due to COVID-19, providing detailed statistical estimates based on national death registry data.

Contribution

It introduces a novel polynomial regression approach for precise excess mortality estimation using comprehensive death registry data from Mexico.

Findings

Estimated excess mortality of around 788,000 people

Male/female mortality ratio approximately 1.7

Estimated excess mortality rate of 626 per 100,000 inhabitants

Abstract

Based on the comprehensive national death registry of Mexico spanning from 1998 to 2022 a point and interval estimation method for the excess mortality in Mexico during the years 2020-2022 is proposed based on illness-induced deaths only, using a polynomial regression model. The results obtained estimate that the excess mortality is around 788,000 people (39.3%) equivalently to a rate of 626 per 100,000 inhabitants. The male/female ratio is estimated to be 1.7 times. As a reference for comparison, for the whole period 2020-2020 Mexico's INEGI estimated an excess of mortality between 673,000 with a quasi-Poisson model and 808,000 using endemic channels estimation.