I am working on a study design for a study to determine behaviors and structural barriers to health seeking at baseline for HIV negative young women 15-19 years, conduct interventions for 2 years and later study their behavioral changes and remaining structural barriers after intervention. This will be carried out in regions based on the HIV prevalence. I have the HIV prevalence for young women 15-19 years for the overall country, but do not have the specific prevalence for these 3 clusters/regions. How can I determine the sample size for this cross sectional study with this?