News Stay informed about the latest enterprise technology news and product updates.

Success of ACO model hinges on clinical data analytics

Getting physicians and hospitals to sign up for ACO adoption will require some heavy IT lifting. Clinical data analytics will support these potential payment game-changers.

BOSTON -- Whether the accountable care organization (ACO) model that the Centers for Medicare and Medicaid Services (CMS) envisions will happen or not, health care providers can rest assured that new, similar payment models -- based on primary care physicians managing a patient's health dossier, with the results graded against quality measures -- will come to pass. Health IT leaders need to prepare for this sea change by understanding and implementing clinical data analytics technology.

"We are at an inflection point in health care in the United States, and the old is about to leave us, as [President Barack] Obama always says, because it's not sustainable," said Dr. Allan Goroll, Harvard Medical School professor and Massachusetts General Hospital primary care physician. Goroll spoke at a recent Massachusetts Health Data Consortium event at Suffolk University law school.

Goroll said that it's not the government, patients or physicians who are forcing the change, but, instead, large employers. They're getting crushed by higher premiums for what amounts to poor health care for their employees, when comparing cost and quality in our country to that in others, he said. "We are moving basically from payment for volume to payment for … value."

Systems similar to the ACO model are already popping up in the private-payer space, including Blue Cross Blue Shield of Massachusetts' Alternative Quality Contract (AQC) program, which offers reimbursement bonuses to participating physicians according to how they score on quality indicators.

Dana Gelb Safran, senior vice president for performance measurement and improvement, gave a presentation on AQC results, which mirrored a recent CMS webinar on the potential of the ACO model. From her payer perspective, Safran said employers are not "buying health, but a lot of care" -- which the AQC was designed to help fix. Its goal was not necessarily to cut costs, but just to slow down their growth.

The analytics are going to change. Instead of productivity -- which is really just value -- it's going to be in terms of how much value was created.

Dr. Allan Goroll, Harvard Medical School professor and Massachusetts General Hospital primary care physician

She agreed with other conference speakers that clinical data analytics will be the engine driving the bonuses these new payment schemes promise for the health care providers who can somehow increase quality while lowering claims. But health IT leaders have a couple problems to solve before they can leverage the value of fledgling clinical data analytics technology: A lack of data that's been collected to date, and a lack of agreed-upon benchmarks to create from that data.

Collecting benchmarking data a challenge for ACO model

While some large data sets -- think hospital discharge databases or insurance claims databases -- could be mined to create quality benchmarks, Goroll said those data sets right now represent the old transactional, fee-for-service system U.S. health care is trying to leave behind. For these new quality-based payment schemes, the data collected will be "fundamentally different."

"Instead of me sitting in my office and spending 30 minutes a day filling out forms to justify the care that I have given -- that data's going to be irrelevant -- what we're going to need is data on how my patients are doing and how cost-effectively I've been able to get that," Goroll said. "The analytics are going to change. Instead of productivity -- which is really just value -- it's going to be in terms of how much value was created."

Goroll said it will take some time for medical leaders to determine what data should be collected, what analytics to run on it, and how to adjust for regional and economic health disparities. A difficult task, he added, will be determining how to reward physicians for improved patient health relative to their previous visits as well as for their performance relative to other physicians.

Clinical data analytics large part of Steward plan

Before BCBS launched the AQC program, Safran expected physicians to bristle at the concept. In fact, she said, none have so far, and many have received it with enthusiasm.

John Donlan, vice president of managed care at Steward Health Care System, said his group of seven community hospitals and numerous ambulatory offices, adding up to about 5,000 physicians in New England, is billing itself as an "accountable care organization." To that end, the system is aggressively pursuing payment schemes such as the CMS ACO model and BCBS's AQC, which he lumped together as "risk-based payments." Steward hopes to succeed in these programs by driving more health care into local, community physician offices and hospitals.

The challenges of clinical data analytics for quality-based systems include the ability to track the progress (and therefore assess the quality of care) of patients who seek care outside the Steward network, which, obviously, is off Steward's information systems. Because patients have a right to see any physician or go to any hospital, that's a data collection dilemma.

Another problem is time. While CMS plans to get the ACO model off the ground in the next few years, it will take forward-thinking organizations like Steward several more years to build the IT infrastructure needed to integrate data from labs, prescribing systems and medical records with historical data to perform the required analytics.

Donlan said that, eventually, he sees Steward's IT infrastructure supporting robust data systems that don't just crunch past data but perform predictive analytics that can help improve quality through preventive care. That, and using mapping data to determine what illnesses are occurring disproportionately in which regions, and working with local health officials to tackle their causes -- thereby improving health through analytics on a whole-community scale.

"The sophistication around the analytics and the data needs increases every day," Donlan said.

Let us know what you think about the story; email Don Fluckinger, Features Writer.

Dig Deeper on Clinical data analytics software and systems

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.