Background
Little is known about how sexual and reproductive (SRH) health can be made accessible and appropriate to adolescents. This study evaluates the impact and sustainability of a competitive voucher program on the quality of SRH care for poor and underserved female adolescents and the usefulness of the simulated patient (SP) method for such evaluation.

Methods
28,711 vouchers were distributed to adolescents in disadvantaged areas of Managua that gave free-of-charge access to SRH care in 4 public, 10 non-governmental and 5 private clinics. Providers received training and guidelines, treatment protocols, and financial incentives for each adolescent attended. All clinics were visited by female adolescent SPs requesting contraception. SPs were sent one week before, during (with voucher) and one month after the intervention. After each consultation they were interviewed with a standardized questionnaire. Twenty-one criteria were scored and grouped into four categories. Clinics' scores were compared using non-parametric statistical methods (paired design: before-during and before-after). Also the influence of doctors' characteristics was tested using non-parametric statistical methods.

Results
Some aspects of service quality improved during the voucher program. Before the program started 8 of the 16 SPs returned 'empty handed', although all were eligible contraceptive users. During the program 16/17 left with a contraceptive method (p = 0.01). Furthermore, more SPs were involved in the contraceptive method choice (13/17 vs.5/16, p = 0.02). Shared decision-making on contraceptive method as well as condom promotion had significantly increased after the program ended.

Female doctors had best scores before- during and after the intervention. The improvements were more pronounced among male doctors and doctors older than 40, though these improvements did not sustain after the program ended.

Conclusion
This study illustrates provider-related obstacles adolescents often face when requesting contraception. The care provided during the voucher program improved for some important outcomes. The improvements were more pronounced among providers with the weakest initial performance. Shared decision-making and condom promotion were improvements that sustained after the program ended. The SP method is suitable and relatively easy to apply in monitoring clinics' performance, yielding important and relevant information. Objective assessment of change through the SP method is much more complex and expensive.