Purpose: To identify prevalence, correlates and survival implications of non-surgically managed epithelial ovarian cancer (EOC).
Methods: The National Cancer Database (NCDB) was queried for EOC cases between 2003 and 2011. Type of treatment, survival data, reasons for non-surgical treatment, clinicopathologic and process-based factors were collected. Logistic regression identified independent predictors of surgical treatment; Cox proportional hazards regression modeled association between time to death and receipt of surgery.
Results: 172,687 of 210,667 patients (82%) received surgical treatment for EOC. 95% of patients treated non-surgically had stage III, stage IV or unknown stage disease. The reason for non-surgical treatment was unclear in 80% of cases. Black race and uninsurance were significantly associated with non-surgical treatment. Median survival time was 57.4months (95% CI: 56.8-57.9) for surgery with or without systemic treatment compared to 11.9months (95% CI: 11.6-12.2) for systemic treatment alone and 1.4months (95% CI: 1.3-1.4) for no treatment. Relative to surgical treatment, the adjusted hazard ratio for death associated with systemic treatment alone was 1.9 (p<0.001); hazard ratio for untreated patients was 4.7 (p<0.001). Among 29,921 patients older than 75 with Stage III/IV disease, 21.5% received only systemic treatment; 22.8% were entirely untreated.
Conclusion: 18% of EOC patients in the NCDB did not receive surgical treatment. These patients experienced significantly worsened survival. Prospective investigation is needed to determine how often apparent deviation from best-practices guidelines is clinically appropriate. Non-surgically treated patients may be at risk for poor access to gynecologic oncology care and deserve further study.
Keywords: Cancer care delivery research; Health disparities; Health policy; Health services research; Ovarian cancer.
Copyright © 2016 Elsevier Inc. All rights reserved.