Evaluating the effects of tubal sterilization on menstrual function: selected issues in data analysis

Area of Science:

• Reproductive Health
• Biostatistics
• Epidemiology

Background:

• The Collaborative Review of Sterilization (CREST) is a large, prospective observational study.
• Tubal sterilization is a common method of contraception for women.
• Understanding long-term health effects, including menstrual changes, is crucial.

Purpose of the Study:

• To analyze menstrual function changes after tubal sterilization.
• To address data analysis challenges in longitudinal studies, specifically dependency and missing data.
• To investigate the association between tubal sterilization and menstrual dysfunction.

Main Methods:

• Longitudinal data analysis using generalized estimating equations (GEE) on over 5000 women followed for 5 years.
• Employed marginal and transitional modelling to assess menstrual dysfunction.
• Utilized data imputation techniques to evaluate potential bias from missing data.

Main Results:

• Marginal modelling indicated a statistically significant increase in menstrual dysfunction odds at 5 years post-sterilization.
• Transitional modelling provided adjusted rates of dysfunction based on baseline menstrual function.
• Imputed data models confirmed the trends observed in the original analyses, suggesting robustness.

Conclusions:

• Tubal sterilization is associated with a statistically significant increase in menstrual dysfunction.
• The GEE approach effectively handled longitudinal data, accounting for repeated measures.
• Data imputation did not alter the study's main findings regarding menstrual dysfunction trends.