Population Health Management Impact on Healthcare Utilization and Costs in CKD: Post Hoc Analysis of the Kidney Coordinated HeAlth Management Partnership Trial.

Key Points: A population health management intervention for CKD reduced inpatient hospitalizations by 27% compared with usual care over a 1-year follow-up. Despite lower hospitalizations, total healthcare costs were not significantly different between population health management and usual care.

Background: CKD represents a substantial economic burden, particularly in the Medicare population in the latter stages of disease progression. There are potential opportunities to provide quality care through population health management (PHM) interventions in the hopes of improving downstream outcomes and costs. In Kidney Coordinated HeAlth Management Partnership, a pragmatic, cluster randomized trial, patients received a PHM, multidisciplinary team approach to improve CKD care or usual care. The primary objective of this study was to conduct a comparative analysis of the 1-year healthcare utilization between patients who received the PHM intervention compared to usual care with a secondary objective of comparing standardized costs.

Methods: A subset of Kidney Coordinated HeAlth Management Partnership patients who had available health insurance claims with enrollment for the full 12 months in the year after trial enrollment were included. Inpatient, outpatient, and pharmacy standardized costs were estimated using diagnosis-related groups, current procedural terminology, and National Drug Codes, respectively. Resource utilization was analyzed using negative binomial models, and costs were analyzed using two-part models. All analyses were adjusted for demographic and clinical characteristics. Subgroups were analyzed by age, sex, CKD stage, and diabetes status.

Results: Of the 1596 trial participants, 614 patients met inclusion criteria (PHM: 300; usual care: 314). Patients in the PHM arm had 27% fewer inpatient hospitalizations than usual care (incident rate ratio=0.73; 95% confidence interval, 0.54 to 0.99), but outpatient visits did not differ significantly. Total standardized costs were similar between the PHM and usual care treatment arms across inpatient, outpatient, and pharmacy categories.

Conclusions: The PHM intervention reduced inpatient hospitalizations but did not significantly affect healthcare costs over 1 year. The reliance on standardized costs and the short follow-up may have obscured potential differences. Longer term data would help provide insight into the economic and resource utilization effect of the PHM intervention.