The Annals of thoracic surgery · 2026
Predicting Persistent Renal Failure Necessitating Dialysis in Patients Undergoing Elective, Open Thoracoabdominal Aortic Aneurysm Repair.
СтатьяНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Blackburn Kyle W, Karla Anirudha R, Zambetti Benjamin R, Nowrouzi Ryan, Green Susan Y, Weldon Scott A, Nguyen Lynna H, Xue Anna H, Orozco-Sevilla Vicente, Chatterjee Subhasis, Moon Marc R, Coselli Joseph S
Аннотация
BACKGROUND: Acute renal failure remains a significant complication after open thoracoabdominal aortic aneurysm (TAAA) repair and is associated with high mortality. Rigorous risk estimation for this complication is vital for these patients.
METHODS: Data were analyzed from 2809 patients who underwent elective, open TAAA repair between 1986 and 2024 at a single practice. Persistent renal failure was defined as the need for dialysis at time of operative discharge (including any hospital transfer) or death. The effectiveness of 4 predictive models-multivariable logistic regression, random forest, support vector machine, and gradient boosting machine-was compared by using the test C statistic (C) from an 80:20, 1000-iteration cross-validation scheme. The regression model was converted into a nomogram for patient counseling.
RESULTS: Persistent renal failure necessitating dialysis developed in 158 (5.6%) patients, of whom 90 (57.0%) experienced operative death. Of the 4 models, the regression model was the most predictive (C = 0.75 [0.72-0.77]). The highest performing machine learning model, the gradient boosting machine, had similar efficacy (C = 0.73 [0.70-0.76]). The final regression model included 8 factors: Crawford extent II repair (odds ratio [OR], 1.86; P = .001), chronic aortic dissection (OR, 1.72; P = .01), symptomatic aneurysm (OR, 1.47; P = .03), maximum distal aortic diameter ≥6 cm (OR, 1.44; P = .06), age (OR, 1.04; P < .001), preoperative estimated glomerular filtration rate (OR, 0.97; P < .001), male sex (OR, 0.65; P = .01), and Crawford extent I repair (OR, 0.43; P = .004).
CONCLUSIONS: The regression model, built using preoperative factors, effectively predicted persistent renal failure necessitating dialysis in patients undergoing elective, open TAAA repair. We converted this model into a nomogram to aid in preoperative counseling.
METHODS: Data were analyzed from 2809 patients who underwent elective, open TAAA repair between 1986 and 2024 at a single practice. Persistent renal failure was defined as the need for dialysis at time of operative discharge (including any hospital transfer) or death. The effectiveness of 4 predictive models-multivariable logistic regression, random forest, support vector machine, and gradient boosting machine-was compared by using the test C statistic (C) from an 80:20, 1000-iteration cross-validation scheme. The regression model was converted into a nomogram for patient counseling.
RESULTS: Persistent renal failure necessitating dialysis developed in 158 (5.6%) patients, of whom 90 (57.0%) experienced operative death. Of the 4 models, the regression model was the most predictive (C = 0.75 [0.72-0.77]). The highest performing machine learning model, the gradient boosting machine, had similar efficacy (C = 0.73 [0.70-0.76]). The final regression model included 8 factors: Crawford extent II repair (odds ratio [OR], 1.86; P = .001), chronic aortic dissection (OR, 1.72; P = .01), symptomatic aneurysm (OR, 1.47; P = .03), maximum distal aortic diameter ≥6 cm (OR, 1.44; P = .06), age (OR, 1.04; P < .001), preoperative estimated glomerular filtration rate (OR, 0.97; P < .001), male sex (OR, 0.65; P = .01), and Crawford extent I repair (OR, 0.43; P = .004).
CONCLUSIONS: The regression model, built using preoperative factors, effectively predicted persistent renal failure necessitating dialysis in patients undergoing elective, open TAAA repair. We converted this model into a nomogram to aid in preoperative counseling.
Предметные рубрики MeSH
HumansRenal DialysisMaleAortic Aneurysm, ThoracoabdominalFemaleElective Surgical ProceduresAgedPredictive Learning ModelsMiddle AgedPostoperative ComplicationsBoosting Machine Learning AlgorithmsNomogramsRisk AssessmentRenal InsufficiencyRandom ForestPrediction AlgorithmsRisk FactorsLogistic ModelsAcute Kidney InjuryAortic Aneurysm, ThoracicRetrospective StudiesSupport Vector Machine