Contraceptive method use, discontinuation and failure rates among women aged 15-49 years: evidence from selected low income settings in Kumasi, Ghana.

Background: This paper provides estimates of contraceptive discontinuation and failure rates in a poor urban setting in Ghana. Contraceptive use is for the purposes of preventing unintended or mistimed pregnancies. Unfortunately, evidence abounds in many parts of the world where there is considerable levels of contraceptive failure and high levels of discontinuation resulting in unintended pregnancies.

Methods: We estimated discontinuation rates during a 12-month period since starting use by applying single and multiple decrement life table methods to the contraceptive calendar data collected in a survey of women in reproductive age of 15-49 years.

Results: Modern contraceptive method use was estimated to be 13.7% at the time of the survey. The results show that contraceptive method discontinuation vary markedly by type of contraceptive method but are high for almost all methods, except for implants (23.7%). Discontinuation rate for emergency contraception was estimated at 88.5%, withdrawal 87.6%, and male condom use 80.9%. However, discontinuation rates were moderately high for rhythm (63.6%), pills (65.6%) and injectables (56%). In terms of failure rates, overall contraceptive failure for all methods was estimated at 7.9%. The factors significantly associated with method failure include being within age bracket 40-44 years (OR = 0.3, p < 0.05), having secondary/higher education (OR = 0.4, p < 0.01), belonging to the richest household wealth scale (OR = 3.3, p < 0.01), currently in union with a partner (OR = 2.2, p < 0.01), and using contraceptive methods such as rhythm (OR = 5.6, p < 0.01) and withdrawal (OR = 3.7, p < 0.01). On the flip side, the odds for method discontinuation were significantly higher for women in their 20s and mid 30s, formerly in union (OR = 1.9, p < 0.05) and use of withdrawal method (OR = 1.4, p < 0.05) and lower for women formerly in union (OR = 0.4, p < 0.01) and use of implants (OR = 0.2, p < 0.01) and injectables (OR = 0.6, p < 0.01).

Conclusion: While contraceptives use is low, both discontinuation and failure rates are high and variable among different methods. Failure and discontinuation rates are lowest for long-acting methods such as implants while higher failure rates are more prevalent among women who rely on withdrawal and the rhythm methods.