Table 2.

Modeling scenarios: HIV risk and contraceptive use

	Scenario 1	Scenario 2	Scenario 3
Underlying HIV risk for HC users
HR	1.0	1.2	1.0
Contraceptive method
Injectable hormonal contraception	10.8%	10.8%	10.8%
Oral hormonal contraception	4.0%	4.0%	4.0%
Condoms (consistent use)	91.1%	91.1%	50%
Other	None	None	None

Injectable and oral contraceptive use in the model is assumed to be the same as that reported at enrolment among uninfected women in Heffron et al. (Table 1 in [7]). Condom use in scenarios 1 and 2 is derived from all follow-up intervals in initially uninfected women (Table 2 in [7]); scenario 3 assumes a lower overall level of condom use. Condom use is randomly assigned in the model independently of other contraceptive use. With no misreporting, condoms are assumed to be used consistently by all couples assigned as condom users. Condom misreporting can be varied independently among HC users and nonusers. We assume that there is no contraceptive switching through the simulated follow-up period and no other contraceptive methods in use.