A simple preprocedural score for risk of contrast-induced acute kidney injury after percutaneous coronary intervention


  • Conflict of interest: Nothing to report.

  • The authors Yong-Li Chen and Nai-Kuan Fu have contributed equally to this work.



To develop a simple scoring system based on preprocedural clinical features that is capable of predicting contrast-induced acute kidney injury (CI-AKI) before percutaneous coronary intervention (PCI).


CI-AKI is associated with increased in-hospital morbidity and mortality, prolonged hospitalization, and long-term renal impairment. Although several scoring methods have been developed to determine risk of CI-AKI, no simple scoring method based on PCI preprocedural clinical features yet exists for Chinese patients.


A total of 2,500 Chinese patients were randomly and retrospectively assigned in a 3:2 manner to create a training and validation dataset, respectively. CI-AKI was defined as an increase of ≥25% or ≥0.5 mg/dL serum creatinine within 5 days after PCI. Preprocedural clinical variables showing independent correlation to CI-AKI were used to derive the risk score from the training dataset and then subsequently tested in the validation dataset. The odds ratios from multivariate logistic regression were used to assign a weighted integer to age ≥70 years = 4, history of myocardial infarction = 5, diabetes mellitus = 4, hypotension = 6, left ventricular ejection fraction ≤45% = 4, anemia = 3, creatinine clearance rate <60 mL/min = 7, decreased high-density lipoprotein <1 mmol/L= 3, and urgent PCI = 3. Summation of the integers represented the total risk score.


The overall incidence of CI-AKI in the training dataset was 16.4% [246/1500; 5.4% for low (≤7) and 61.3% for very high (≥17) risk scores]. The rates of CI-AKI, 1-year dialysis, and 1-year mortality increased significantly with each group (Cochran-Armitage test of trend, P < 0.001). The risk score facilitated appropriate classification of patients with low and high risk for CI-AKI after PCI in the validation dataset (c-statistic = 0.82).


Risk classification based on the most significantly correlated parameters is useful for predicting CI-AKI before contrast exposure. The simple preprocedural score showed excellent predictive ability for identifying patients at high risk of nephropathy and those with deteriorative prognosis after PCI. © 2013 Wiley Periodicals, Inc.