Www.dovepress.com



Supplemental AppendixThis appendix has been provided by the authors to give readers additional information about this work.Supplement to: Dingwei D, Alvarez PJ, Woods SD. A predictive model for progression of chronic kidney disease to kidney failure using a large administrative claims database.Table of ContentsSupplemental Table 1 Variables with significant univariate association to first identification of kidney failure in the prediction year 2Supplemental Figure 1 Comparison of logistic regression and other machine learning algorithms to predict progression of CKD stage 3 or 4 to kidney failure 3Supplemental Table 1. Variables with significant univariate association to first identification of kidney failure in the prediction year AgeDementiaLiver diseaseAllergyDepressionMalignant neoplasmsAnxietyDiabetes mellitusMetabolic syndromeARBERGMigraineAsthmaGenderMRA useAtrial fibrillationHepatitisObesityBeta-blocker useHuman immunodeficiency virus diseaseOsteoarthritisBusiness line (commercial vs Medicare Advantage)HyperkalemiaOsteoporosisCalcineurin inhibitor useHyperlipidemiaProportion of days covered ≥80%Cerebrovascular diseaseHypertensionPeripheral artery diseaseCongestive heart failureInflammatory bowel diseaseProspective ERG risk scoresCKD stageIron deficiency anemiaRAASi at optimal doseChronic thyroid disordersIschemic heart diseaseGeographic regionCongenital heart diseaseKidney stone historyRheumatoid arthritisChronic obstructive pulmonary diseaseCKD stage switchUrban vs rural residenceARB, angiotensin II receptor blocker; CKD, chronic kidney disease; ERG, episode risk group; HK, hyperkalemia; MRA, mineralocorticoid receptor antagonist; RAASi, renin-angiotensin-aldosterone system inhibitorSupplemental Figure 1: Comparison of logistic regression and other machine learning algorithms to predict progression of CKD stage 3 or 4 to kidney failure. The predictive performance of logistic regression, gradient boosting, neural network, and support vector machine are similar, ROC index all = 0.85; random forest works well, ROC index = 0.83; decision tree is the worst, ROC index = 0.73. ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download