Volume 10, Issue 3 (10-2014)
HSR 2014, 10(3): 558-570 |
Back to browse issues page
Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
Zahra Mousavi Azarang, Soleiman Kheiri, Morteza Sedehi. Bayesian Analysis of Zero-Inflated Count Regression Models and their Application to Analyzing the Number of Return to Blood Donation. HSR 2014; 10 (3) :558-570
URL: http://hsr.mui.ac.ir/article-1-716-en.html
URL: http://hsr.mui.ac.ir/article-1-716-en.html
1- MSc of Biostatistics, Shahrekord University of Medical Sciences, Shahrekord, Iran
2- Associated Professor of Biostatistics, Shahrekord University of Medical Sciences, Shahrekord, Iran (Corresponding Author): Email: kheiri@hbi.ir.
3- Assistant Professor of Biostatistics, Shahrekord University of Medical Sciences, Shahrekord, Iran
2- Associated Professor of Biostatistics, Shahrekord University of Medical Sciences, Shahrekord, Iran (Corresponding Author): Email: kheiri@hbi.ir.
3- Assistant Professor of Biostatistics, Shahrekord University of Medical Sciences, Shahrekord, Iran
Abstract: (911 Views)
Background: To prepare safe and adequate blood supply to meet patients’ needs and ensure a sufficient number of regular blood donors, knowledge about factors encouraging people to donate blood regularly is essential. Considering its importance, we aimed to identifying the effective factors of the return to blood donation in based on zero-inflated count regression models using Bayesian approach.Methods: This is a descriptive-analytical study performed on data of 864 blood donors in blood centers scattered across Shahrekord. The response variable was the number of return to blood donation during up to five years. The analysis was done based on zero-inflated regression models including: zero-inflated Poisson, zero-inflated Negative Binomial and zero-inflated generalized Poisson. Analyses of models were carried out based on Bayesian technique using Marko Chain Monte Carlo methods by WinBUGS. Models comparison were done using deviance information criterion (DIC).Findings: The zero-inflated generalized Poisson regression model had the lowest DIC compare to other models. The body weight had a significant positive effect and the age of volunteers had significant negative effect on the number of return to donation.Conclusion: The body weight and age of volunteers have significant positive and negative relationship to the number of return to donation; so much effort should be done to encourage moderate-weight and higher-age to donate more frequently.Key Words: Blood donation- Zero-Inflated- Count Regression Model- Bayesian Analysis- Markov Chain Monte Carlo
Type of Study: Research |
Subject:
education health and promotion
Received: 2020/07/16 | Accepted: 2014/10/15 | Published: 2014/10/15
Received: 2020/07/16 | Accepted: 2014/10/15 | Published: 2014/10/15
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |
