Women's contraceptive discontinuation and switching behavior in urban Senegal, 2010-2015

Background: With the focus of global and national family planning initiatives on reaching "additional user" targets, it is increasingly important for programs to assess contraceptive method discontinuation and switching. This analysis calculated the discontinuation rate and method-specific discontinuation rates, examined reasons given for contraceptive discontinuation, and assessed characteristics associated with subsequent contraceptive switching and abandonment among women living in urban areas of Senegal. Methods: Data came from the Measurement, Learning & Evaluation project's 2015 survey of 6927 women of reproductive age living in six urban sites (Dakar, Pikine, Guediawaye, Mbao, Kaolack and Mbour). Information on contraceptive use and discontinuation for the five years preceding the survey were recorded in a monthly calendar. Single decrement life tables were used to calculate discontinuation rates. Descriptive analyses were used to assess reasons for discontinuation and method switching after discontinuation. A multinomial logistic regression was used to estimate the likelihood of being a non-user in-need of contraception, a non-user not in-need of contraception, or a method switcher in the month after discontinuation, by sociodemographic and other characteristics. Results: The 12-month discontinuation rate for all methods was 34.7%. Implants had the lowest one-year discontinuation rates (6.3%) followed by the intrauterine device (IUD) (18.4%) while higher rates were seen for daily pills (38%), injectables (32.7%), and condoms (62.9%). The most common reasons for discontinuation were reduced need (45.6%), method problems (30.1%), and becoming pregnant while using (10.0%). Only 17% of discontinuations were followed by use of another method; most often daily pills (5.2%) or injectables (4. 2%). In the multivariate analysis, women with any formal education (primary, secondary or higher) were more than 50% more likely to switch methods than remain in need of contraception after discontinuation than women with no education or Koranic-only education (RRR = 1.59, p-value = 0.004; RRR = 1.55, p-value = 0.031). The likelihood of switching compared to being "in need" was also significantly higher for women who were married and who discontinued traditional methods. Conclusions: To support increased contraceptive method use, women with no education and unmarried women are priorities for counseling and information about side effects and method switching at the time of method adoption.