Search

Background and objectivesAcute kidney injury after surgery is associated with high mortality and morbidity. The purpose of this study is to develop and validate a risk prediction tool for the occurrence of postoperative acute kidney injury requiring renal replacement therapy.Design, setting, participants, measurementsThis retrospective cohort study had 2,299,502 surgical patients over 2015-2017 from the American College of Surgeons National Surgical Quality Improvement Program Database (ACS-NSQIP). Eleven predictors were selected for the predictive model: age, history of congestive heart failure, diabetes, ascites, emergency surgery, preoperative serum creatinine, hematocrit, sodium, preoperative sepsis, preoperative acute renal failure and surgery type. The predictive model was trained using 2015-2016 data (n=1,487,724) and further tested using 2017 data (n=811,778). A risk model was developed using multivariate logistic regression and machine learning methods.Main outcomesThe primary outcome was postoperative 30-day acute kidney injury requiring renal replacement therapy(AKI-D)ResultsThe unadjusted 30-day postoperative mortality rate associated with AKI-D was 37.5%. The renal risk prediction model had high AUC (area under the receiver operating characteristic curve, training cohort: 0.89, test cohort: 0.90) for postoperative AKI-D.ConclusionsThis model provides a clinically useful bedside predictive tool for postoperative acute kidney injury requiring dialysis.

BackgroundVTE is a serious postoperative complication after surgery with resultant higher morbidity and mortality. Despite years of experience with current risk models, rates continue to be high and more information is needed on individual patient risk in the prophylaxis era.Research QuestionsCan we assess the individualized risk of postoperative venous thromboembolism (VTE) for broad categories of surgery?MethodsThis study was performed using data from the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) Database. Patient data (n=2,875,190) from 2015-2017 were used for study analysis. Eight predictors were selected for the model: age, preoperative platelet count≥450 (×109/L), disseminated cancer, corticosteroid use, serum albumin ≤2.5 g/dL, preoperative sepsis, hospital length of stay and surgery type. The second model included 7 predictors without hospital length of stay. A predictive model was trained using ACS-NSQIP data from 2015-2016 (n=1,859,227) and tested using data from 2017 (n= 1,015,963). Primary outcomes are postoperative 30-day VTE, including deep vein thrombosis (DVT) and/or pulmonary embolism (PE).ResultsVTE occurred in 23,249 patients (0.81%) and 49.9% of VTE occurred after discharge from index hospitalization. The risk prediction model had high AUC (area under the receiver operating characteristic curve) for postoperative VTE of 0.78 (training cohort) and 0.78 (test cohort).InterpretationThis clinical prediction model is a validated, practical and easy-to-use tool to identify surgical patients at the highest risk of postoperative VTE and provide an individualized assessment of risk based on clinical factors and type of surgery. This prediction model may be used as a tool to assess individualized risk of postoperative VTE and promote broader discussion and awareness of the VTE risk during the perioperative period.