External validation of predictive models for acute kidney injury following cardiac surgery

Overall, the lack of calibration of these models and the difficulty in using the models clinically because of the large number of variables limit their applicability.