Net reclassification improvement and integrated discrimination improvement require calibrated models: relevance from a marker and model perspective

\udFor the last three decades, clinical prediction models have mainly been evaluated on the basis of their\udability to discriminate between persons who develop the event of interest and persons who do not, as\udquantified by the c-statistic or area under the receiver operator characteristic curve (AUC). The\udAUC considers sensitivity and specificity of the model over all possible cut-points of predicted risk.\udHowever, prediction models are often used to classify patients into risk categories that correspond to\uddiagnostic or therapeutic decisions. This provoked the idea of comparing models according to their\udability to adequately assign clinical risk categories based on absolute risk estimates. Analyses of\udrisk reclassification have hit the ground running: uptake of measures such as net reclassification\udimprovement (NRI) has been enormous, and guidance documents on evaluations of markers and\udprediction models embraced it as a step prior to full-blown cost-effectiveness analysis.More recently,\udseveral researchers reviewed the current applications of reclassification analysis and expressed concerns\udabout inappropriate use