Cost-effectiveness of an HPV self-collection campaign in Uganda: comparing models for delivery of cervical cancer screening in a low-income setting.

With the availability of a low-cost HPV DNA test that can be administered by either a healthcare provider or a woman herself, programme planners require information on the costs and cost-effectiveness of implementing cervical cancer screening programmes in low-resource settings under different models of healthcare delivery. Using data from the START-UP demonstration project and a micro-costing approach, we estimated the health and economic impact of once-in-a-lifetime HPV self-collection campaign relative to clinic-based provider-collection of HPV specimens in Uganda. We used an individual-based Monte Carlo simulation model of the natural history of HPV and cervical cancer to estimate lifetime health and economic outcomes associated with screening with HPV DNA testing once in a lifetime (clinic-based provider-collection vs a self-collection campaign). Test performance and cost data were obtained from the START-UP demonstration project using a micro-costing approach. Model outcomes included lifetime risk of cervical cancer, total lifetime costs (in 2011 international dollars [I$]), and life expectancy. Cost-effectiveness ratios were expressed using incremental cost-effectiveness ratios (ICERs). When both strategies achieved 75% population coverage, ICERs were below Uganda's per capita GDP (self-collection: I$80 per year of life saved [YLS]; provider-collection: I$120 per YLS). When the self-collection campaign achieved coverage gains of 15-20%, it was more effective than provider-collection, and had a lower ICER unless coverage with both strategies was 50% or less. Findings were sensitive to cryotherapy compliance among screen-positive women and relative HPV test performance. The primary limitation of this analysis is that self-collection costs are based on a hypothetical campaign but are based on unit costs from Uganda. Once-in-a-lifetime screening with HPV self-collection may be very cost-effective and reduce cervical cancer risk by > 20% if coverage is high. Demonstration projects will be needed to confirm the validity of our logistical, costing and compliance assumptions.