Patient Characteristics and Clinical Indicators Influencing Oncologists’ Perceptions of Prognosis in Ovarian Cancer: A Discrete Choice Experiment

Objective To quantify how patient characteristics and treatment outcomes influence oncologists' perceptions of prognosis in ovarian cancer. Methods A cross-sectional survey incorporating a discrete choice experiment was conducted among 150 US-based oncologists (medical, hematologic, and gynecologic). Participants completed 12 choice tasks comparing hypothetical patient profiles defined by eight attributes (age, stage, BRCA/HRD status, cytoreduction outcome, ascites, anti-VEGF use, progression-free survival, and CA-125 status). Hierarchical Bayesian models estimated preference weights and relative attribute importance. Exploratory subgroup analyses examined differences by practice setting, specialty, and case volume. Results Perceived prognosis was primarily driven by longer progression-free survival (2 verses 10 years; utility range=3.43), earlier stage (II verses IV; range=2.89), and CA-125 status (rising verses low/normal; range=2.82). These attributes were more than twice as influential as BRCA/HRD status, cytoreduction outcomes, or ascites. Anti-VEGF use had minimal impact. Exploratory subgroup analyses suggested variability in preferences may exist: academic oncologists placed greater emphasis on age (12.4% verses 7.6%), while community oncologists weighted progression-free survival and CA-125 more heavily . Medical/hematologic oncologists emphasized stage more than gynecologic oncologists (21.8% verses 16.6%). Lower-volume clinicians assigned greater importance to CA-125. Conclusions Oncologists' perceptions of prognosis in advanced ovarian cancer are predominantly influenced by dynamic disease indicators, progression-free survival, stage, and CA-125, rather than baseline biomarkers or surgical outcomes. Variability across practice settings and specialties highlights heterogeneity in prognostic assessment. As therapeutic advances extend progression-free survival, understanding how clinicians interpret these signals may improve prognostic communication and support more aligned clinical decision-making.