A cost consequence analysis of six diagnostic strategies for ovarian cancer: A model-based economic evaluation

Objective: To assess the costs and consequences of six diagnostic strategies for ovarian cancer in pre-menopausal and post-menopausal women with symptoms in secondary care. Design: Economic evaluation alongside a prospective single-arm diagnostic accuracy study. Setting: NHS secondary care outpatients (2-week referrals, clinics, GP referrals, cross-specialty referrals) and inpatients (emergency presentations to secondary care). Sample: Two cohorts of 857 pre-menopausal and 1,242 post-menopausal newly presenting to secondary care with symptoms of suspected ovarian cancer. Methods: A model-based cost-consequence analysis (CCA) was conducted using a decision tree simulating patient pathways over 12-months. Diagnostic accuracy data were sourced from the ROCkeTS study and supplemented by literature. Main outcome measures: Cancer deaths, correct diagnosis proportion, and diagnostic yield. Results: No diagnostic strategy was optimal across all outcomes. Across both cohorts, the Risk of Malignancy Index (RMI) 200 was least expensive but had poor cancer death and diagnostic yield outcomes. The ADNEX 3% strategy had the highest diagnostic yield and lowest cancer mortality but was the most expensive. For pre-menopausal women, the IOTA ADNEX 10% strategy outperformed ORADS, ROMA, and CA125 in cost and outcomes. For post-menopausal women, the high cancer prevalence required a trade-off. In sensitivity analysis a two-step IOTA ADNEX 10% strategy outperformed ORADS, ROMA, and CA125 across all three outcomes, making the strategy a more balanced choice in both cohorts. Conclusion: At 12 months, no single diagnostic strategy was superior. Early diagnosis requires balancing cancer mortality, diagnostic yield, and cost. The IOTA ADNEX two-step strategy at 10% threshold provided the best trade-off across these factors and is recommended for practice.