Characterizing a heterogeneous chronic patient population for redesigning person-centred bundled payment models using risk-mitigating measures.

Since 2010, most Dutch patients with diabetes mellitus type 2 (DM2), an increased risk of cardiovascular diseases (CVR), and chronic obstructive pulmonary disease (COPD), have been treated in single-disease management programs (SDMPs) provided by primary care cooperatives (PCCs). These SDMPs are funded through bundled payments. However, given the prevalence of multimorbidity among patients, there is a growing need for care that is more person-centred. We have previously published an alternative payment model that stimulates the integration of care required for a person-centred approach and in this paper, we demonstrate an operationalization of this model. We performed cluster analysis on claims data to distinguish between different subgroups of patients, predicted cluster probabilities with data available to general practitioners, designed different care packages and investigated the expected financial risk for PCCs of different sizes. We showed that mainly the size of the PCC and the content of the care package influenced the predicted losses or profits for the PCC. Two risk-mitigating measures-risk-adjustment and cost-capping-resulted generally in predicted losses or profits closer to 0, and therefore a reduced risk for the PCC.