Search

Objective: To determine if natural language processing (NLP) with machine learning of unstructured full text documents (a preoperative CT scan) improves the ability to predict postoperative complication and hospital readmission among women with ovarian cancer undergoing surgery when compared with discrete data predictors alone., Methods: Medical records from two institutions were queried to identify women with ovarian cancer and available preoperative CT scan reports who underwent debulking surgery. Machine learning methods using both discrete data predictors (age, comorbidities, preoperative laboratory values) and natural language processing of full text reports (preoperative CT scans) were used to predict postoperative complication and hospital readmission within 30 days of surgery. Discrimination was measured using the area under the receiver operating characteristic curve (AUC)., Results: We identified 291 women who underwent debulking surgery for ovarian cancer. Mean age was 59, mean preoperative CA125 value was 610 U/ml and albumin was 3.9 g/dl. There were 25 patients (8.6%) who were readmitted and 45 patients (15.5%) who developed postoperative complications within 30 days. Using discrete features alone, we were able to predict postoperative readmission with an AUC of 0.56 (0.54-0.58, 95% CI); this improved to 0.70 (0.68-0.73, 95% CI) (p < 0.001) with the addition of NLP of preoperative CT scans., Conclusions: Natural language processing with machine learning improved the ability to predict postoperative complication and hospital readmission among women with ovarian cancer undergoing surgery., Competing Interests: Declaration of Competing Interest The authors disclose no financial conflicts of interest with the presented work. Dr. Barber receives research grant payments to her institution from Eli Lilly, GOG Foundation, and the NICHD., (Copyright © 2020 Elsevier Inc. All rights reserved.)

Objective: Trillions of dollars pass to physicians from industry-related businesses annually, leading to many opportunities for financial conflicts of interest. The Open Payments Database (OPD) was created to ensure transparency. We describe the industry relationships as reported in the OPD for presenters at the 2019 Society of Gynecologic Oncology (SGO) Annual Meeting and evaluate concordance between author disclosures of their financial interests and information provided by the OPD., Methods: This is an observational, cross-sectional study. Disclosure data were collected from authors with oral and featured abstract presentations in the 2019 SGO annual conference. These disclosures were compared to data available for each author in the 2018 OPD, which included the amount and nature of industry payments., Results: We examined the disclosures of 301 authors who met inclusion criteria. Of 161 authors who had disclosure statements on their presentations,147 reported "no disclosures," and 14 disclosed industry relationships. The remaining 140 did not list any disclosure information. Sixty percent (184/301) of authors had industry relationships in the 2018 OPD, including 173 of 287 (60.3%) of authors who either reported no disclosures or did not have disclosure data available in their presentations. These transactions totaled over 43 million USD from 122 different companies, with most payments (46%) categorized as "Research or Associated Research." Accurate disclosure reporting was associated with receiving higher payments or research payments, and being a presenting author., Conclusions: Most authors at the SGO annual conference did not correctly disclose their industry relationships when compared with their entries in the OPD., Competing Interests: Declaration of competing interest ELB reports grant payments from the National Institutes of Health (K12 HD050121-12), GOG Foundation and Eli Lilly, and has received a food and beverage payment from Intuitive. EJT served as a consultant for Auris Health, Inc. and has received miscellaneous food and beverage payments from Intuitive, Covidien, Ethicon, Davol, Merck, and AstraZeneca. These relationships do not present a conflict of interest with the current work. The remaining authors have nothing to disclose., (Copyright © 2020. Published by Elsevier Inc.)