Test-driving an ACO

One year into an accountable care organization, Norton Healthcare and Humana officials can point to results—which is more than anyone expected at the outset.

As hospitals, medical groups and commercial insurers grapple with how to launch an accountable care network and Medicare officials debate final rules for ACO shared-incentive payments, Norton, which owns four Kentucky hospitals, and insurer Humana have a year's worth of results from a limited pilot.

So far, their test has produced minor savings and more pronounced gains for some quality measures.

“That was not our goal for the first year,” says Dr. Thomas James III, corporate medical director of national network operations for Humana, who says officials expected new strategies adopted during the first year would need time before producing results.

Cancer screening, routine diabetes tests and medication monitoring all showed improvement during the first year of the pilot, which included 10,000 patients (all employees of Norton and Humana) and more than 300 doctors, including roughly 200 primary-care, 29 medical specialists and 89 surgical specialists.

“We did better than what we expected,” says James, a practicing physician who works at Norton hospitals on weekends.

Norton and Humana got a jump-start on accountable care from policy experts at the Dartmouth Institute for Health Policy & Clinical Practice and the Brookings Institution's Engelberg Center for Health Care Reform.

The Kentucky ACO is one of five locations included in a closely watched Dartmouth-Brookings pilot of accountable care.

Ongoing quality efforts, heightened attention to accountable care quality measures and access to new data from Humana helped improve performance during the first year, says Dr. Steven Hester, Norton's senior vice president and chief medical officer.

For example, Humana's claims data presents a more comprehensive view of doctors' prescribing patterns and can help identify doctors who could switch to generics from brand-name drugs to lower costs, he says.

Norton also hired two care coordinators for primary-care patients and is expected to hire three more. The health system's employed doctors did not enter into accountable care contracts, but already receive incentive payments tied to quality performance and information technology use, Hester says. Bonuses include incentives that vary each year as quality priorities change; Norton officials included nationally recognized quality measures selected by Dartmouth and Brookings for the pilot last year, he says.

More diabetic patients were tested for blood sugar control and cholesterol management in the first year of the accountable care group. Blood sugar testing increased to 93.4% of diabetic patients from 87.7% for the baseline year, and cholesterol management tests climbed to 91.8% of diabetics from 83.9%.

Fewer Norton patients received an imaging study during the first 28 days after being diagnosed with lower back pain. Imaging is unnecessary for most back-pain patients, a common reason for doctor's visits, and was one target to improve quality and lower healthcare costs. Norton reported imaging use dropped to 56.3% for newly diagnosed back-pain patients during the first year compared with 65.2% in the baseline year.

Among other quality gains: Breast cancer screening increased, as did cervical cancer screening, though marginally, Norton's first-year figures show. Also improved was medication monitoring for patients with long-term prescriptions.

In addition to the Kentucky partners, Dartmouth and Brookings selected hospitals and physician groups at four other locations, in Arizona, California and Virginia, for the pilot of the health payment reform model. Other locations include the Carilion Clinic, Roanoke, Va.; Tucson (Ariz.) Medical Center; and two independent practice associations, Monarch HealthCare, Irvine, Calif., and HealthCare Partners, Torrance, Calif.

Dartmouth and Brookings selected the locations for their geographic and operational diversity.

The Carilion Clinic, the first of the five pilot organizations to announce the Dartmouth-Brookings experiment, has struggled to find a private insurer willing to attempt shared savings (July 27, 2009, p. 7) and remains the only location yet to begin operating an ACO.

Humana joined the pilot in 2009 after the insurer's former chief actuary, John Berko, publicly promoted the Dartmouth-Brookings experiment, James says. Berko, at the time a member of the Medicare Payment Advisory Commission, also served as a Brookings visiting scholar.

Norton was an obvious partner, James says, because Humana had worked with the health system to develop quality reporting before the partners' accountable care efforts.

Preparations began well before shared savings became public policy under the 2010 Patient Protection and Affordable Care Act.

Months before the Affordable Care Act became law, officials met for the first time to discuss shared savings between Humana and Norton. In February 2010, the insurer and health system began weekly meetings. By June 2010, the Kentucky accountable care group finalized the method to identify patients included in the experiment. The partners signed an agreement in August that year.

Dartmouth and Brookings supplied initial quality measures, the formula to calculate potential savings and a method for identifying patients included in the pilot. Such details have proved contentious in draft regulations for Medicare accountable care. Nonetheless, the templates deliberately left many elements for Norton and Humana officials to settle.

Each test location was allowed to decide how to award incentive payments and divvy up bonuses—or potential losses, known among policymakers as “financial risk.” For example, HealthCare Partners agreed to financial risk for some quality measures and Monarch agreed to some partial capitation, which requires doctors to absorb losses if treatment costs more than the negotiated capitated, or lump sum, payment.

Hester says such flexibility is critical because the marketplace and relationships between health systems, doctors and commercial insurers vary across the U.S.

Norton and Humana agreed to share savings beyond an initial 2% reduction in costs for patients included in the pilot. Hospitals and doctors would receive 40% of any savings under the arrangement and 60% would be returned to Norton and Humana for workers' healthcare costs. Norton won't be liable for losses should healthcare costs accelerate instead of slow.

Patients enrolled in the accountable care pilot can continue to seek care outside of Norton, and employee benefit plans do not include incentives for patients to remain within the ACO, James says. Instead, Norton providers will compete on cost, quality and patient satisfaction, which the accountable care effort seeks to improve, he says. “An ACO is not an HMO,” James says.

More quality measures will be added to the accountable care pilot during the second year, which began in July, Hester says, although the Dartmouth and Brookings policy experts won't select them. Norton and Humana will identify measures where the partners see an opportunity for improvement. Second-year efforts will also target health spending growth, he says.

James says the partners will share first-year results with area employers in an effort to recruit more patients into the pilot.

For Norton, work to develop and operate the private-market ACO helped as the health system completed the Center for Medicare and Medicaid Innovation application to test various ACO models, Hester says. The CMS declined to say how many applied for the Innovation Center experiment—the plans are known as Pioneer ACOs—which policy-makers proposed as an alternative to more-restrictive plans for accountable care under draft Medicare rules. Congress outlined criteria for Medicare accountable care in the 2010 health reform law, but allowed for further experiments under the Innovation Center.

Norton gained experience identifying ACO patients and transferring data during the Dartmouth-Brookings pilot that proved useful for its Innovation Center application, Hester says. But the Medicare pilot will also present new challenges, including patients who are older than commercially insured patients. For example, that could mean more congestive heart failure patients, a condition more common among the elderly. Providers would need to develop new interventions for heart failure patients. Meanwhile, diabetes is more common among commercially insured patients.

Hester says Norton officials won't rule out other proposed reform models and must settle unanswered questions before committing as an accountable care Pioneer. Ultimately, that did not prevent Norton from applying to participate. It's unclear where health reform is going, Hester says. “Being involved in the process is better than waiting.”