Researchers have found that a common tool used by clinicians for predicting kidney failure could be underestimating the risk of failure for some people.