Risk-prediction model of severe postoperative complications after primary debulking surgery for advanced ovarian cancer

Objectives To refine models to predict surgical morbidity and 90-day mortality after primary debulking surgery (PDS) for advanced epithelial ovarian cancer (EOC). Methods Women with stage IIIC/IV EOC who underwent PDS with curative intent between 1/2/2003 and 12/30/2011 were included. Patient characteristics, intraoperative and postoperative outcomes, and vital status were abstracted. Complications were graded using the Accordion classification. Nomograms were generated based on multivariate modeling. Results 138 (22.3%) of the 620 patients who underwent PDS experienced a grade ≥ 3 complication. Age (OR 1.21 per 10 years increase in age), BMI (OR 1.35 for BMI < 25 kg/m2 versus reference, OR 2.83 for BMI ≥ 40 kg/m2 versus reference), ASA score ≥ 3 (OR 1.49), stage (OR 1.69 stage IV) and surgical complexity (OR 2.32 high complexity versus intermediate) were predictive of an accordion grade ≥ 3 complication Within 90 days of surgery, 55 (8.9%) patients died. A multivariable model included age (OR 1.76 per 10 year increase in age), ASA score ≥ 3 (OR 3.28), preoperative albumin < 3.5 (OR 4.31), and BMI (OR 2.04 for BMI < 25 kg/m2 versus reference, OR 3.64 for BMI ≥ 40 kg/m2 versus reference) was predictive of 90-day mortality. Conclusion Using an independent cohort we report the importance of age, ASA score, preoperative albumin, FIGO stage, and surgical complexity, and BMI, to refine a prediction model for complications after PDS for advanced EOC. This information is useful in preoperative counseling and can be utilized to aid in patient-centered decision making and risk stratification.