The basic guts of the model require that at least a dialysis provider, nephrologist and one other Medicare provider be involved. The "one other Medicare provider" bit is likely meant to capture the obvious need that dialysis patients have for a multitude of services beyond dialysis and renal specific care. The ESCO has to have at least 500 members all of whom need to be on dialysis - patients with functional kidney transplants are not eligible.
The hypothesis behind ESCOs is that an alignment of financial incentives with clinical care goals will produce superior outcomes and patient experience at lower per capita health spend than traditional fee for service Medicare where providers are rewarded for the volume of services they deliver. The 10,000 foot view of "alignment of financial incentives" is that ESCOs will be able to earn a fraction of savings when health care spending is lower than a pre-determined base line and will be at risk for a fraction of overruns when spending is higher.
In order to avoid rationing of high value services, the ESCO model requires that participating groups report and achieve a defined set of quality metrics in order to earn their fraction of savings. These cross a wide variety of areas including preventative health, chronic disease management, care coordination, patient safety, patient and caregiver experience, and patient quality of life.
The slide deck below from the CMS open door forum forum this past week digs a bit deeper into the details.