Healthcare spending in the United States is distributed unevenly, with approximately 50% of expenses incurred by 5% of the population.2,3 Managing the care of these high-cost patients is of key concern to health systems. However, few programs aimed at addressing the needs of high-cost patients have demonstrated net cost savings.4 Evidence suggests that the only successful programs are ones that closely align care management interventions with the specific needs of a given patient.6 Programs that target high-cost patients rely on several methods in order to identify patients needing intervention, including risk prediction algorithms, chronic disease criteria or utilization thresholds. Such approaches depend on the use of single variables, or combinations of variables, to identify high-cost patients, thereby assuming homogeneity among high-cost patients. By using growth mixture models (GMM), we will be able to characterize the heterogeneity among high cost patients in terms of a finite number of latent growth classes. Inclusion of auxiliary information (covariates and distal outcomes) in the GMM will be necessary in order to understand and evaluate the fidelity and utility of the resultant trajectory profiles. A result of this study will be the classification of high-cost patients into subgroups based on their cost and utilization trajectories and relevant medical, psychological, behavioral, and social factors. The identification of these subgroups represents an opportunity for care management programs to more effectively align services to the meet the needs of the many types of patients they serve, which may lead to better health outcomes and reductions in utilization of acute care services and costs.
The identification of subgroups of high-cost patients has important public health relevance because it provides actionable information to care management programs so that they may more effectively tailor services to meet patient needs. If programs are more effective in their approach, patients may experience better health outcomes and decreased use of acute care services as a result. Focusing on the small percentage of patients who incur a majority of healthcare costs represents an efficient way to drive down healthcare spending in the US, which continues to increase year to year.
