Effect of Attribution Length in a Pediatric Medicaid Accountable Care Organization

In the current accountable care organization (ACO) environment, pediatric health systems and hospitals are striving toward better population health management and cost control as they begin to assume the risk for the use of health care resources and spending. In this issue of JAMA Pediatrics, Christensen and Payne have elevated the conversation on the impact of pediatric ACOs to a higher level of sophistication by exploring the relationship among attribution, resource use, and cost within a pediatric Medicaid population. The authors demonstrated promising findings within a Medicaid ACO involving Children’s Hospitals and Clinics of Minnesota and the State of Minnesota Department of Human Services. Specifically, the study found that ACO attribution was correlated with decreases in the use of inpatient care resources and spending, particularly among children with multiple chronic conditions. The authors acknowledged, however, that the total ACO cohort had been reduced by 50% after 12 months and by 70% after 24 months. Given that the study did not control for social determinants, the authors asked the following question: If the proportion of patients who became unattributed represent those whose social circumstances have a negative influence on their health, would successful efforts to maintain engagement have resulted in a reduction in the use of resources and cost in this group? In the context of social determinants for the Medicaid population, a corollary issue arises with the assumption that continued attribution is a useful proxy for consistent primary care.

These observations raise 2 critical questions for hospitals and health systems involved in pediatric Medicaid ACOs. First, what level of responsibility should we assume in sustaining patient engagement in the primary care medical home? Second, what types of efforts to maintain ACO attribution would yield a solid return on investment in a shared risk and shared savings environment?


JAMA Pediatrics


Eisen MH, Rubin DM