Optimization of early diagnosis of ovarian cancer using an original software ScOv

Abstract

The most common causes of treatment failure in ovarian cancer (ОС) include late diagnosis due to the absence of screening programs for its early detection and lack of vigilance by general practitioners and gynecologists. World experience suggests that screening of 100% of women has a minimal impact on mortality. Examination of women at high risk for OC is a sensible alternative to mass screening. This examination can be performed using universal computerized screening, which is a very promising and cost-effective method.To identify the symptoms significantly associated with early-stage OC, we questionnaired 100 patients with morphologically verified stage IA–IC OC, admitted to the Department of Gynecology, Tula Regional Oncology Dispensary between 2010 and 2017 (experimental group). The control group included 200 women without malignant diseases, who underwent preventive medical examination in outpatient units of Tula. We analyzed the significance of 22 clinical symptoms and developed original computer software ScOv, which was subsequently used for identifying patients at high risk of OC among women aged 40 years and older by evaluating their complaints. The sensitivity and specificity of ScOv were 73.7 % and 88.8 % respectively. A retrospective check of the developed program demonstrated its high efficiency and prognostic value.