Risk Prediction Models for Contrast Induced Nephropathy: A Systematic Review

Background: Acute kidney injury following exposure to radiocontrast, often referred to as contrast-induced nephropathy (CIN), has been associated with significant morbidity and mortality. Accurate models are needed to predict the risk of CIN in order to better counsel patients and evaluate potential prophylactic therapies. Objective: To systematically review prediction models for CIN. Design: We searched MEDLINE, EMBASE, and CINAHL databases from inception through 2015, and performed hand searches of the retrieved reference lists. We conducted dual reviews to identify studies published in the English language of prediction models tested with patients that included derivation and validation cohorts. We extracted data on baseline patient characteristics, procedural characteristics, modeling methods, metrics of model performance, risk of bias and clinical usefulness. Results: We identified 16 studies describing 12 prediction models, including seven studies with external validation of these models. There was significant heterogeneity among the included studies. Ten of 12 models were validated internally and six were validated externally (some models had more than one external validation study). Discrimination was variable in studies that were validated internally (c-statistic 0.610.95) and externally (c-statistic 0.57-0.86). Only one study presented reclassification indices. Page 2 of 39 https://mc.manuscriptcentral.com/bmj BMJ 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Condential: or Rview O ly 3 Conclusions: Risk prediction studies for CIN have many inherent limitations. Most of the models in clinical use have modest predictive ability, and are only relevant to individuals receiving contrast for coronary angiography. Further research is necessary to develop a clinically useful model which can accurately determine the risk of CIN across a series of clinical settings and prevent associated adverse consequences. Page 3 of 39 https://mc.manuscriptcentral.com/bmj BMJ 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Condential: or Rview O ly 4