Volume 19, Issue 5 (12-2020)
TB 2020, 19(5): 55-67 |
Back to browse issues page
Ethics code: IR.SSU.SPH.REC.1398.130
Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
Askarishahi M, Mirjani Arjenen M, Vakili M. Comparison between Efficiency of Poisson Regression Model and Negative Binomial Regression in the Analysis of Factors Affecting Mortality from Cardiovascular Diseases in Yazd Province in 2017. TB 2020; 19 (5) :55-67
URL: http://tbj.ssu.ac.ir/article-1-3022-en.html
URL: http://tbj.ssu.ac.ir/article-1-3022-en.html
Shahid Sadoughi University of Medical Sciences , moasbio@gmail.com
Abstract: (1532 Views)
Introduction: Despite the advances in cardiovascular diseases, death caused by these diseases is still considered as the leading cause of mortality. In this study, some of the effective factors on the deaths caused by cardiovascular diseases were investigated
Methods: This cross-sectional analytical study investigated the efficacy of Poisson regression models and negative binomial regression models on factors affecting mortality from cardiovascular diseases. The death data were extracted from the death registration system for Yazd province in 2017.Gender, age, education, occupation, location, and city of death were also extracted for each deceased. The two regression models were then fitted to the data
Results: A total of 5,015 deaths were recorded, of which 1,642 were due to cardiovascular diseases.
Cardiovascular disease mortality rates were significant using negative binomial regression in terms of the educational variables, place of residence, type of residence, and age. Death rates caused by cardiovascular diseases were not significant for age and occupational, educational, and residential variables.
Conclusion: If the time of death is considered as an offset variable, the regression model of two negative sentences is more effective in showing the factors affecting death due to cardiovascular diseases according to AIC and BIC criteria. In the case that the total number of deaths is considered as the offset variable, the Poisson regression model is more efficient.
Methods: This cross-sectional analytical study investigated the efficacy of Poisson regression models and negative binomial regression models on factors affecting mortality from cardiovascular diseases. The death data were extracted from the death registration system for Yazd province in 2017.Gender, age, education, occupation, location, and city of death were also extracted for each deceased. The two regression models were then fitted to the data
Results: A total of 5,015 deaths were recorded, of which 1,642 were due to cardiovascular diseases.
Cardiovascular disease mortality rates were significant using negative binomial regression in terms of the educational variables, place of residence, type of residence, and age. Death rates caused by cardiovascular diseases were not significant for age and occupational, educational, and residential variables.
Conclusion: If the time of death is considered as an offset variable, the regression model of two negative sentences is more effective in showing the factors affecting death due to cardiovascular diseases according to AIC and BIC criteria. In the case that the total number of deaths is considered as the offset variable, the Poisson regression model is more efficient.
Keywords: Cardiovascular Disease, Negative Binomial Regression, Poisson Regression, Performance, Akaike Score, Bayesian Score
Type of Study: Research |
Subject:
General
Received: 2020/02/15 | Accepted: 2020/10/4 | Published: 2021/01/20
Received: 2020/02/15 | Accepted: 2020/10/4 | Published: 2021/01/20
Send email to the article author
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |
