Guideline-concordant Arteriovenous Fistula Care With the PREDICT-AVF Web App

The aim of this study was to develop a tool for prediction arteriovenous fistula (AVF) outcomes and guide shared decision-making about access appropriateness futility in the context recent Kidney Disease Outcomes Quality Initiative (KDOQI) guidelines. KDOQI guidelines discourage ongoing hemodialysis salvage attempts after two interventions prior successful use or three per year overall. data source international multicenter PATENCY-1 PATENCY-2 randomized trials, which enrolled patients undergoing new radiocephalic AVF creation. outcome number surgical endovascular required at 1 year. Cox, random survival forest, pooled logistic, elastic net recurrent time-to-event machine learning models were built using combination baseline characteristics postoperative ultrasound measurements. Models trained with 10-fold cross-validation hyperparameter selection within randomly split training dataset. Discrimination calibration performance assessed on holdout testing An interactive web application created, generates patient-specific predictions contextualized cohort included 914 patients; mean age 57 years (standard deviation, 13 years); 22% female. Radiocephalic AVFs created snuffbox (2%), wrist (74%), forearm (24%). Patients underwent an estimated 1.04 (95% confidence interval [CI], 0.94-1.13) first Cox either 4- 6-week 12-week information alone had best discrimination performance, areas under receiver operating characteristic curve 0.75 CI, 0.68-0.82) 0.77 0.71-0.83) forest model performed when restricting only (Table I). is deployed https://predict-avf.com (Fig 1). PREDICT-AVF can patient counseling guideline-concordant as part patient-centered end-stage kidney disease life plan both preoperative settings. Future versions will include maturation patency variety configurations.Table 1Discrimination each feature set, measured by area (AUROC [ie, C-statistic]) postoperativelyFeaturesCoxRandom forestPooled logistic regressionElastic netBaseline0.704 (0.639-0.769)0.715 (0.647-0.783)a0.696 (0.631-0.762)0.703 (0.638-0.769)Baseline + US0.710 (0.635-0.785)0.747 (0.676-0.818)a0.707 (0.632-0.782)0.708 (0.633-0.783)Baseline US0.731 (0.664-0.798)0.776 (0.712-0.841)a0.725 (0.657-0.793)0.738 (0.671-0.804)4- US0.747 (0.678-0.816)a0.722 (0.647-0.797)0.743 (0.674-0.813)0.745 (0.675-0.814)12-week US0.770 (0.708-0.831)a0.736 (0.669-0.803)0.761 (0.699-0.824)0.763 (0.700-0.825)US, Ultrasound.Perfect corresponds AUROC 1.Feature sets (age, sex, smoking status, diabetes, heart failure, renal transplant history, anticoagulation use, antiplatelet statin status time creation, central venous catheter failed accesses, vein artery diameter, anastomotic suture technique, location, side), addition measurements (flow volume, luminal stenosis), alone.aBest set. Open table tab