Research Article

A Statistical Analysis of Positive Excess Mortality at Covid-19 in 2020-2021

Authors

  • Md Nurul Raihen Assistant Professor, Department of Mathematics and Computer Science, Fontbonne University, St. Louis, MO, 63105, USA https://orcid.org/0000-0003-2680-0658
  • Sultana Akter Teaching Assistant, MS in Statistics, Department of Statistics, Western Michigan University, Kalamazoo, MI, 49006, USA
  • Fariha Tabassum Teaching Assistant, MS in Sociology, Department of Sociology, Western Michigan University, Kalamazoo, MI, 49006, USA
  • Farjana Jahan Teaching Assistant, MS in Statistics, Department of Statistics, Western Michigan University, Kalamazoo, MI, 49006, USA
  • Md Nazmul Sardar Senior Officer, Product Development at Radiant Nutraceuticals Ltd (F. Hoffman-La Roche Ltd), Dhaka, 1000, Bangladesh

Abstract

When it comes to making assessments about public health, the mortality rate is a very important factor. The COVID-19 pandemic has exacerbated well-known biases that affect the measurement of mortality, which varies with time and place. The COVID-19 pandemic took the world off surveillance, and since the outbreak, it has caused damage that many would have thought unthinkable in the present era. By estimating excess mortality for 2020 and 2021, we provide a thorough and consistent evaluation of the COVID-19 pandemic's effects. Excess mortality is a term used in epidemiology and public health to describe the number of fatalities from all causes during a crisis that exceeds what would be expected under 'normal' circumstances. Excess mortality has been used for thousands of years to estimate health emergencies and pandemics like the 1918 "Spanish Flu"6. Positive excess mortality occurs when actual deaths exceed previous data or recognized patterns. It could demonstrate how a pandemic affects the mortality rate. The estimates of positive excess mortality presented in this research are generated using the procedure, data, and methods described in detail in the Methods section and briefly summarized in this study. We explored different regression models in order to find the most effective factor for our estimates. We predict the pandemic period all-cause deaths in locations lacking complete reported data using the Poisson, Negative Binomial count framework. By overdispersion test, we checked the assumption of the Poisson model, and then we chose the negative binomial as a  good fitting model for this analysis through Akaike Information Criteria (AIC) and Standardized residual plots, after that checking the P-value<0.05; we found some significant predictors from our choosing model Negative binomial model, and the coefficient of all predictors gave the information that some factors have a positive effect, and some has a negative effect at positive excess mortality at COVID-19 (2020-2021).

Article information

Journal

Journal of Mathematics and Statistics Studies

Volume (Issue)

4 (3)

Pages

07-17

Published

2023-08-03

How to Cite

Raihen, M. N., Akter, S., Tabassum, F., Jahan, F., & Sardar, M. N. (2023). A Statistical Analysis of Positive Excess Mortality at Covid-19 in 2020-2021. Journal of Mathematics and Statistics Studies, 4(3), 07–17. https://doi.org/10.32996/jmss.2023.4.3.2

Downloads

Keywords:

COVID-19, Excess Mortality, Pandemic, Poisson Regression, Negative Binomial regression, WHO (World Health Organization), Region