Assessing changes in vaginal bleeding patterns in contracepting women

In clinical trials comparing alternative contraceptive methods, women often discontinue for disturbances in menstrual bleeding patterns. Vaginal bleeding diaries have been used to monitor these patterns, and this paper suggests two new methods of analysis of these data. One analysis measures trends over time in the occurrence of specific patterns and, a second analysis, the relation between a woman's current experience and the probability of continuing method use. Contrasts between these approaches and the reference period method of analysis are made. A randomised clinical trial of the relative efficacy of 100 and 150 mg doses of depot-medroxyprogesterone acetate (DMPA) is used for illustration.