Cost-Effectiveness of Introducing the SILCS Diaphragm in South Africa.

Background: Though South Africa has high contraceptive use, unintended pregnancies are still widespread. The SILCS diaphragm could reduce the number of women with unmet need by introducing a discreet, woman-initiated, non-hormonal barrier method to the contraceptive method mix.

Methods: A decision model was built to estimate the impact and cost-effectiveness of the introduction of the SILCS diaphragm in Gauteng among women with unmet need for contraception in terms of unintended and mistimed pregnancies averted, assuming that the available contraceptives on the market were not a satisfying option for those women. Full costs were estimated both from a provider's and user's perspective, which also accounts for women's travel and opportunity cost of time, assuming a 5% uptake among women with unmet contraceptive need. The incremental cost-effectiveness ratio is computed at five and 10 years after introduction to allow for a distribution of fixed costs over time. A probabilistic sensitivity analysis was conducted to incorporate decision uncertainty.

Results: The introduction of the SILCS diaphragm in Gauteng could prevent an estimated 8,365 unintended pregnancies and 2,117 abortions over five years, at an annual estimated cost of US$55 per woman. This comes to a cost per pregnancy averted of US$153 and US$171 from a user's and provider's perspectives, respectively, with slightly lower unit costs at 10 years. Major cost drivers will be the price of the SILCS diaphragm and the contraceptive gel, given their large contribution to total costs (around 60%).

Conclusions: The introduction of the SILCS diaphragm in the public sector is likely to provide protection for some women for whom current contraceptive technologies are not an option. However to realize its potential, targeting will be needed to reach women with unmet need and those with likely high adherence. Further analyses are needed among potential users to optimize the introduction strategy.