 This paper reviews the current state of risk prediction models for type 2 diabetes. It finds that many studies are lacking in terms of quality, as they do not report on key details such as the number of events per variable, the selection process for risk factors, and the handling of missing data. Furthermore, some studies use unreliable methods such as univariate screening and categorization of continuous risk factors. This can lead to unreliable predictions and potentially misleading results. This article was authored by Oma Oma, Mallet Susan, Colin Scarry S and others.