The description of menstrual bleeding patterns: Towards fewer measures

Publisher Website
Google Scholar
Abstract
The disturbance in menstrual bleeding induced by many methods of contraception is an important factor in their acceptability, and women participating in clinical trials of new methods are usually asked to keep a calendar record of the occurrence of bleeding. A recent WHO report recommends that, for analysis and presentation, each menstrual diary should be divided into successive periods, and the woman's bleeding pattern in each period summarized in ten indices. These indices are simple to calculate and easily understood by clinicians, but comparison of several groups is problematic because of the large number of summary statistics generated. Principal components analysis was therefore used to determine whether the indices could be reduced to a smaller number of measures. The analyses showed that three of the indices - the range and maximum value of bleeding/spotting episode lengths and the minimum bleeding-free interval - rarely measure their intended dimension of the bleeding pattern. Most of the essential information about a woman's bleeding pattern is contained in only four of the ten indices: the number of bleeding/spotting episodes, the mean lengths of episodes and intervals, and the range of bleeding-free interval lengths. Together, these indices describe the most important dimensions of a pattern: the amount, frequency and variability of the bleeding.