New findings show that communities where hospitals, other healthcare providers, and community services work together to coordinate evidence-based hospital discharges and provide better support in the community, can see a 6 percent drop in hospitalizations and rehospitalizations, per 1,000 beneficiaries, in just the first two years.

The project relied upon Medicare’s Quality Improvement Organizations (QIOs) to anchor and guide the work, and the average community netted about $3 million  in annual savings for Medicare.

The results were released Jan. 22 by the Journal of the American Medical Association (JAMA) in an article titled "Associations between quality improvement for care transitions in communities and rehospitalizations among Medicare beneficiaries."

For the project, 14 QIOs, led by the Colorado Foundation for Medical Care (CFMC) as a national coordinator, participated in a three-year project in which the QIOs convened medical, community and social service providers and facilitated community-wide quality improvement activities to implement evidence-based improvements in patient care transitions. The QIOs' efforts included community organizing, technical assistance in implementing best practices, and monitoring of participation, implementation, effectiveness and adverse effects.

[See also: Colorado ACO yields big savings, reduced readmissions.]

QIOs in each state and territory are funded by the Centers for Medicare & Medicaid Services (CMS). Their goal is to help achieve national quality goals through focused efforts at the community and provider level. The QIO Program focuses on three aims: better patient care, better population health and lower health care costs through improvement.

“This project took an unusual, yet ultimately effective, approach to improving care transitions,” Jane Brock, MD, chief medical officer at CFMC and lead author for the JAMA article, said in a news release. “Rather than focus on one hospital ward, or 100 patients, it engaged whole communities to improve care for large geographically-defined populations, and it worked.”