| Summary: | The aim of this thesis was to determine the diagnostic accuracy of outpatient endometrial evaluation using endometrial biopsy (EB), ultrasound scan (USS) and hysteroscopy (OPH) by conducting systematic quantitative reviews of the published literature. The optimum diagnostic strategy in terms of cost-effectiveness (cost per life year gained), was then established for the investigation of women with post-menopausal bleeding (PMB) for endometrial cancer, using the review data in a decision analysis designed to reflect current service provision. Meta-analyses showed that a positive test result following EB or OPH was more useful for predicting endometrial disease than USS, whereas a negative test result following USS was more useful for excluding endometrial disease than EB or OPH. The economic model included 12 diagnostic strategies and indicated that a strategy based on initial diagnosis with USS, using a 5mm double layer endometrial thickness cut-off, was the most cost-effective. Sensitivity analyses showed that initial investigation with EB or USS using a 4mm cut-off were also potentially cost-effective (incremental cost-effectiveness ratios under £30,000 per life year gained) at their most favorable estimates of diagnostic performance, in women under 65 years and at disease prevalence of 10% or more. The choice between initial testing with EB or USS will therefore depend upon patient age and preference, disease prevalence and the availability of high quality USS. In most circumstances women presenting for the first time with PMB should undergo initial evaluation with pelvic ultrasound using a threshold of 4mm or 5mm to define abnormal results.
|
|---|
