The description of menstrual bleeding patterns: towards fewer measures

Area of Science:

• Reproductive Health
• Contraception Research
• Statistical Analysis in Medicine

Background:

• Menstrual bleeding disturbances are a key factor in contraceptive method acceptability.
• Clinical trials often require women to maintain menstrual diary records.
• The World Health Organization (WHO) recommends summarizing bleeding patterns using ten indices.

Purpose of the Study:

• To determine if the ten WHO-recommended menstrual bleeding indices can be reduced to a smaller set of measures.
• To identify the most informative indices for describing a woman's bleeding pattern.
• To simplify the analysis and comparison of bleeding patterns across different contraceptive methods.

Main Methods:

• Utilized principal components analysis (PCA) on ten menstrual bleeding indices derived from women's calendar records.
• Evaluated the contribution of each index to capturing the essential dimensions of bleeding patterns.
• Identified indices that were less effective in measuring their intended aspects of the bleeding pattern.

Main Results:

• Three indices (range/maximum bleeding/spotting episode lengths, minimum bleeding-free interval) were found to be less informative.
• Four indices were identified as containing most of the essential information: number of bleeding/spotting episodes, mean episode/interval lengths, and range of bleeding-free interval lengths.
• These four indices effectively describe the amount, frequency, and variability of menstrual bleeding.

Conclusions:

• Principal components analysis can effectively reduce the complexity of menstrual bleeding pattern data.
• A subset of four indices provides a comprehensive summary of bleeding patterns, simplifying analysis and interpretation.
• This reduction facilitates better comparison of bleeding patterns in clinical trials and contraceptive acceptability studies.