PLoS ONE · 2014 · 41 citations · 31 references
MedicineBiostatisticsNephrologyDialysisHemodialysisRenal Replacement TherapyChronic Kidney DiseaseDiabetic Kidney DiseaseDialysis AdequacyDialysis ComplicationsMortality ModelingRenal DiseaseClinical Prediction RulesLogistic RegressionMacrovascular ComplicationsMortality PredictionPredictionPrediction Model DevelopmentPrediction ModelsPrognostic Prediction Models
BACKGROUND: While some prediction models have been developed for diabetic populations, prediction rules for mortality in diabetic dialysis patients are still lacking. Therefore, the objective of this study was to identify predictors for 1-year mortality in diabetic dialysis patients and use these results to develop a prediction model. METHODS: Data were used from the Netherlands Cooperative Study on the Adequacy of Dialysis (NECOSAD), a multicenter, prospective cohort study in which incident patients with end stage renal disease (ESRD) were monitored until transplantation or death. For the present analysis, patients with DM at baseline were included. A prediction algorithm for 1-year all-cause mortality was developed through multivariate logistic regression. Candidate predictors were selected based on literature and clinical expertise. The final model was constructed through backward selection. The model's predictive performance, measured by calibration and discrimination, was assessed and internally validated through bootstrapping. RESULTS: A total of 394 patients were available for statistical analysis; 82 (21%) patients died within one year after baseline (3 months after starting dialysis therapy). The final prediction model contained seven predictors; age, smoking, history of macrovascular complications, duration of diabetes mellitus, Karnofsky scale, serum albumin and hemoglobin level. Predictive performance was good, as shown by the c-statistic of 0.810. Internal validation showed a slightly lower, but still adequate performance. Sensitivity analyses showed stability of results. CONCLUSIONS: A prediction model containing seven predictors has been identified in order to predict 1-year mortality for diabetic incident dialysis patients. Predictive performance of the model was good. Before implementing the model in clinical practice, for example for counseling patients regarding their prognosis, external validation is necessary.
31
R: A Language and Environment for Statistical Computing
R Core Team · 2000 · 353K citations · Full text
Natural Sciences, Computer Science, Statistical Computing +10
An Introduction to the Bootstrap.
S. T. Buckland, B. Efron, Ryan J. Tibshirani · Biometrics · 1994 · 3.2K citations
Internal validation of predictive models
Ewout W. Steyerberg, Frank E. Harrell, Gerard Borsboom et al. · Journal of Clinical Epidemiology · 2001 · 2.6K citations
Prognosis and prognostic research: what, why, and how?
Karel G.M. Moons, Patrick Royston, Yvonne Vergouwe et al. · BMJ · 2009 · 1.4K citations