Contraceptive Discontinuation, Switching and Reproductive Consequences: Analysis of 1,539,071 Episodes Contributed from 61 Countries That Participated in DHS: Population Base-Analysis

Mohamed Ali, World Health Organization (WHO)
John G. Cleland, London School of Hygiene and Tropical Medicine (LSHTM)

Data from DHS contraceptive calendars are used to assess discontinuation of method-specific use, switching within three months following method-related discontinuation, and reproductive consequences at 12 months. We add to the extensive existing evidence in four main ways: (1) trends in discontinuation and switching for countries with 3+ surveys; (2) the addition of country-level factors (human development index (HDI), method skewness, and unmet need) in the analysis of predictors; (3) for switchers, comparison of the 12-month discontinuation for origin and destination methods; (4) assessment of reproductive consequences at 12 months for four main self-reported reasons for discontinuation (method-related, desire for pregnancy, no further need, other/DK). Preliminary results for topics 1 and 2 are available. Among 104 countries, statistically significant (p<0.05) declines in discontinuation were observed in 40 cases versus significant increases in 24 cases. Among 60 countries, significant declines in switching were found in 19 and significant increases in 21. Predictors of method-related discontinuation were assessed by multi-level competing risk hazard models. A pronounced relationship of discontinuation with HDI was found but not with method skewness and unmet need. Switching was much higher in countries with high than low HDIs, with low method skewness, and with low unmet need.

Keywords: Family Planning and Contraception, Multi-level modeling

See extended abstract.