[Home ] [Archive]   [ فارسی ]  
:: Volume 6, Issue 1 (Spring 2014 2014) ::
JNKUMS 2014, 6(1): 117-123 Back to browse issues page
Factors associated to pelvic prolapse in women, weighted logistic regression model in complex sampling studies
B Gholami Chaboki , Y Mehrabi * , F Ramezani Tehrani , A Kavousi
Abstract:   (710 Views)

Abstract Background & Objectives: The use of complex sampling designs in data collection of some studies such as case-control studies is becoming common, yet this method has often been ignored in analysis. Ordinary logistic regression is used to estimate the coefficients of model, instead. The purpose of the present study is to evaluate the fitting of weighted logistic regression compared to that of ordinary (OLR) and post stratified (PSLR) logistic regression. Materials and Methods: The data used in this study were collected using complex sampling method among urban women in four provinces of Iran. Age, educational level, BMI, marital age, number of pregnancies and abortions, total and vaginal number of deliveries were considered in this research. Dependent variable was pelvic organ prolapse. Ordinary, post-stratified and weighted logistic regression was fitted to the data. To compare models’ goodness of fit, ROC curve analysis was used. Results: In post-stratified and weighted methods the standard errors of estimates were almost equal and both larger than that of ordinary method. In ordinary method, number of vaginal deliveries, age, employment and BMI showed statistically significant association with prolapsed pelvic. Area under ROC curve in ordinary, post-stratified and weighted methods obtained as 0.75, 0.72 and 0.73, respectively. Conclusion: Results showed that although weighting reduces the bias of estimates by adjusting sampling errors, it does not necessarily shrink their variance.

Keywords: Keywords: Complex Sampling, Weighted Logistic Regression, Post-Stratification, Pelvic Organs Prolapse, Reproductive Morbidity
Full-Text [PDF 508 kb]   (341 Downloads)    
Type of Study: Orginal Research | Subject: Basic Sciences
Received: 2015/03/16 | Accepted: 2015/03/16 | Published: 2015/03/16
Add your comments about this article
Your username or Email:

Write the security code in the box >

XML   Persian Abstract   Print

Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:

Gholami Chaboki B, Mehrabi Y, Ramezani Tehrani F, Kavousi A. Factors associated to pelvic prolapse in women, weighted logistic regression model in complex sampling studies. JNKUMS. 2014; 6 (1) :117-123
URL: http://journal.nkums.ac.ir/article-1-324-en.html

Volume 6, Issue 1 (Spring 2014 2014) Back to browse issues page
مجله دانشگاه علوم پزشکی خراسان شمالی Journal of North Khorasan University of Medical Sciences
Persian site map - English site map - Created in 0.56 seconds with 30 queries by YEKTAWEB 3647