Contributors of Mammography Screening with Zero-Inflated Count Regression Models by Bayesian Approach

Abstract

Background: Mammography is a valuable tool for early diagnosis of breast cancer in asymptomatic women. Considering the high prevalence of breast cancer in Iranian women and the low participation in mammography screening program, the purpose of this study was to investigate the factors affecting frequency of mammography screening in women over 40 years of age using zero-. Methods: In this study, the required information about number of performing mammography in women’s’ lifetime, demographic characteristics and behavioral risk factors were obtained through interview based on a researcher-made questionnaire. To investigate the factors affecting mammography, zero inflation Poisson regression models were performed using Win Bugs software. Results: The mean (SD) age of women participating in this study was 49.87 (6.76). 77% of the participants have never undergone mammography, 8.9% once, 6.9% twice, 6.7% three times, and 0.5% more than three times. Age had a positive effect on the number of mammograms in the women who have perform mammograms at least once. Having a family history of cancer and breast cancer, middle compared to low economic status, higher compared to low education and menopause were significantly associated with lower probability of never performing mammography. Conclusion: Given the relatively low participation of women in mammography, more facilities are needed for high risk women (aged 40-70) to facilitate the diagnosis process.

Keywords

• Mammography
• Breast cancer
• Zero-Inflated
• Poisson Regression
• Bayesian Approach