Cost-effectiveness of using the Cervex-Brush (broom) compared to the elongated spatula for collection of conventional cervical cytology samples within a high-burden HIV setting: a model-based analysis

BACKGROUND From 2010 to 2014, approximately 2 million Pap smears from HIV-infected women were submitted to the South African National Health Laboratory Services (NHLS) through the national cervical cancer screening programme. The objective of this analysis was to determine whether using the plastic Cervex brush ("broom") would be a cost-effective approach to improve cytology specimen quality as compared to the wooden spatula used currently. METHODS A decision analysis model was built using the expected adequacy rates for samples collected with the spatula (<$0.02) and broom ($0.23) and the probability of detecting cervical dysplasia. NHLS data was used for testing volumes and rates of HIV-positivity, suitability of specimens, and presence of endocervical cells. Expected positivity of Pap smears in HIV-infected women (73 %), odds ratios of the effectiveness of the broom (OR: 1.57), and improved sensitivity when endocervical cells present (OR: 1.89) are from literature. NHLS costs were used for the collection devices and conventional cytology ($4.89). Cost of clinic visit is from WHO CHOICE ($8.36). RESULTS In 2010, 80 % of specimens submitted to NHLS were adequate for evaluation; in 2014, only 54 % met the same criteria. For HIV-infected women, according to the guidelines model, using the wooden spatula costs $6.25 million per year, $16.79 per woman tested. Under intended practice, for each additional HSIL case detected among HIV-infected women, the South African cervical cancer screening programme could save $13.64 (95 % CI: $13.52 to $13.76) by using the broom as its standard of care collection device through increased collection of endocervical cells and consequent reduction in repeat Pap smears. CONCLUSION Under a wide range of parameters tested using a simulation model, the more expensive plastic broom could save the South African cervical cancer screening programme money and increase detection of high-grade cervical dysplasia in HIV-infected women compared to the current wooden spatula.