Contraceptive discontinuation, switching, abandonment and their reproductive consequences: An analysis of 1,539,071 episodes of reversible method use contributed from 61 countries that participated in DHS: Population base-analysis

As contraceptive uptake rises in a country, discontinuation of use of a contraceptive method because of dissatisfaction and the ability and willingness to switch promptly to an alternative following such discontinuation become increasingly important determinants of the prevention of unintended pregnancy. About one third of unintended births in low- and middle-income countries are the result of discontinuation or accidental pregnancy while using a method. Examination of the dynamics of contraceptive use continuation and switching can reveal how well contraceptive services are meeting the family planning needs of women and couples. We performed a population-based analysis of 1,539,071 episodes of use of reversible contraceptive methods from calendar data collected in 149 demographic and health surveys in 61 countries. Cause-specific discontinuation rates in the presence of competing events, switching to an alternative method within three months of discontinuation, and conception rates after cause-specific discontinuation, were computed using the cumulative incidence method. Competing risks marginal models were used to produce sub-hazard ratios of discontinuation, switching and abandonment. Meta-analysis was used to estimate the method-specific pooled cumulative incidence rates for discontinuation, switching and abandonment. We found that discontinuation because of dissatisfaction was highest in socio-economically developed countries with low unmet need for contraception. Conversely, switching within three months of discontinuation was much higher in these countries, with the net result that contraceptive abandonment, defined as discontinuation but no switching, was 31% lower in most developed societies than in the least developed. Women who switched were five times less likely to experience pregnancy within 12 months than abandoners. In conclusion, high discontinuation rates should be regarded as an unfortunate inevitability, which may be resistant to improvements in the quality of services. Increases in the ability and willingness to switch is a more plausible pathway to the reduction of unintended pregnancy.