Predicting long-term kidney allograft outcomes: pitfalls and progress

Abstract

Early identification of kidney transplant recipients at risk of progressive allograft dysfunction may allow clinicians to provide closer monitoring and more aggressive risk factor modification. In this issue, Raynaud et al. presented a latent class model that clustered kidney transplant recipients into 8 risk categories of post-transplant kidney function loss. This commentary discusses some of the advantages, but also challenges, of the use of latent class analyses, including the clinical applicability of models that are often derived from such approaches.