Mahoney, Douglas

INTRODUCTION: Colorectal cancer remains a leading cause of cancer associated death in the United States and colonoscopy the primary screening strategy for prevention. Rates of adenomatous and serrated neoplasia detection are inversely associated with postcolonoscopy colorectal cancer. This crucial quality metric depends on accurate ascertainment of colorectal neoplasia findings from both endoscopy and histopathology records. We aimed to assess the feasibility of a random forest machine learning model to rapidly and accurately categorize colorectal neoplasia from electronic health record data. METHODS: A retrospective cohort study compared neoplasia detection rates among individuals undergoing colonoscopy at a large academic institution to develop a rule-based algorithm to categorize colorectal neoplasia from endoscopy reports and pathology systematized nomenclature of medicine – clinical terms. This cohort provided a large training set to develop a natural language processing system using a random forest approach to automatically classify unstructured pathology findings into adenoma, serrated, or advanced neoplasms. This system was manually validated through an independent holdout set. RESULTS: The training set comprised 35,953 unstructured pathology reports with matched systematized nomenclature of medicine – clinical terms from 95,188 unstructured colonoscopy reports. The final model was assessed on an independent holdout set of 337 manually annotated procedures obtaining an area under the receiver operating characteristic curve of 0.997 (confidence interval [CI] 0.994–1), 0.99 (CI 0.98–1), and 0.99 (CI 0.98–0.99) for prediction of adenoma, serrated, and advanced lesions, respectively. DISCUSSION: The random forest-based hybrid natural language processing system for classification of colonoscopy results was both accurate and explainable. NLP combined with effective machine learning algorithms can provide a scalable strategy for colonoscopy quality monitoring.

Background Sex‐related differences in morbidity and survival in bicuspid aortic valve (BAV) adults are fundamentally unknown. Contemporary studies portend excellent survival for BAV patients identified at early echocardiographic‐clinical stages. Whether BAV adults incur a survival disadvantage throughout subsequent echocardiographic‐clinical stages remains undetermined. Methods and Results Analysis was done of 3 different cohorts of consecutive patients with echocardiographic diagnosis of BAV identified retrospectively: (1) a community cohort of 416 patients with first BAV diagnosis (age 35±21 years, follow‐up 16±7 years), (2) a tertiary clinical referral cohort of 2824 BAV adults (age 51±16 years, follow‐up 9±6 years), and (3) a surgical referral cohort of 2242 BAV adults referred for aortic valve replacement (AVR) (age 62±14 years, follow‐up 6±5 years). For the community cohort, 20‐year risks of aortic regurgitation (AR), AVR, and infective endocarditis were higher in men (all P ≤0.04); for a total BAV‐related morbidity risk of 52±4% vs 35±6% in women ( P =0.01). The cohort's 25‐year survival was identical to that in the general population ( P =0.98). AR independently predicted mortality in women ( P =0.001). Baseline AR was more common in men ( P ≤0.02) in the tertiary cohort, with 20‐year survival lower than that in the general population ( P <0.0001); age‐adjusted relative death risk was 1.16 (95% confidence interval [CI] 1.05‐1.29) for men versus 1.67 (95% CI 1.38‐2.03) for women ( P =0.001). AR independently predicted mortality in women ( P =0.01). Baseline AR and infective endocarditis were higher in men (both ≤0.001) for the surgical referral cohort, with 15‐year survival lower than that in the general population ( P <0.0001); age‐adjusted relative death risk was 1.34 (95% CI 1.22‐1.47) for men versus 1.63 (95% CI 1.40‐1.89) for women ( P =0.026). AR and NYHA class independently predicted mortality in women (both P ≤0.04). Conclusions Within evolving echocardiographic‐clinical stages, the long‐term survival of adults with BAV is not benign, as both men and women incur excess mortality. Although BAV‐related morbidity is higher in men in the community, and AR and infective endocarditis are more prevalent in men, women exhibit a significantly higher relative risk of death in tertiary and surgical referral cohorts, which is independently associated with AR.