Challenges posed by accountable care drive new data uses

Accountable care organizations are gaining prominence as one of the nation’s leading approaches for making the shift to value-based care. Providers in these ACOs are facing a variety of business challenges, including gathering and using data to improve care.

In many cases, ACOs are finding they need to incorporate data from a variety of sources, placing new importance on activities such as master data management, analytics and population health management applications.

ACO data requirements are driving IT vendors to tweak offerings to make them appropriate for their needs.

“To be successful, ACOs need to bring data together from different systems—payers and providers,” says Pete Hess, vice president and chief data architect at Health Catalyst, a data warehousing, analytics and outcomes improvement company.

According to Hess, one of the keys for ACO success is the ability to gather the data and combine it into a data warehouse that can direct overall efforts. “Master data management (MDM) tools are critical to the success of these efforts.”

“Some MDM tool providers focus on patient identity resolution—the classic (enterprise master patient identifier) problem space—while others focus exclusively on reference data, which is the space of terminology management tools,” Hess says.

Hess sees MDM evolving to meet the evolving needs of value-based care. “Value-based care requires organizations to be more collaborative than ever before, which means data must be shared,” Hess says. “Shared data is always going to have discrepancies—MDM tools can help manage those discrepancies.”


After data is gathered, the importance of sharing that information rises in order to coordinate patient care throughout an organization.

Farzad Mostashari, CEO and co-founder of Aledade and former National Coordinator for Health Information Technology, has been a long-time advocate of population health. He says that his emerging company has been making inroads in guiding the industry toward what it considers to be the “true north” of value-based care, using a cloud-based population health platform.

Mostashari says Aledade clients have been able to use its applications to improve delivery of care typically associated with value-based care, for example achieving a 251 percent increase in preventive care visits at physician practices in 2015, while boosting revenue 66 percent.

Mostashari emphasizes the power of “truly integrating” clinical and claims data—which is difficult to do. But, he adds, for ACOs to be successful at value-based care will ultimately take more than just payer and clinical data.

Aledade has also added admission, discharge, transfer (ADT) data to the clinical and claims data. This has helped provide primary care physicians with immediate notification when their patients are newly discharged from the hospital, when they need to pay particular attention to followup care so as to avoid readmissions.

For example: Aledade uses predictive analytics to determine which patients in an ACO are the most vulnerable patients, then uses ADT data to alert their doctors when they are hospitalized. This improved communication enables doctors to speak to the patients within 48 hours of discharge. Readmission rates among these high-risk patients fell to 20 percent, compared with 33 percent before the application of data. “This is a great example of putting data together and making it actionable,” Mostashari said.

Despite the progress being made with the data, ACOs still face challenges, typically not on the technical side, Mostashari says. There are still situations that exist where provider organizations retain exclusivity of information to encourage network participation and sometimes health IT vendors charge a lot for interfaces, he says.

“The problem is mostly in trying to get the agreements in place to make sure everyone is financially taken care of,” Mostashari says.

Innovation is occurring in other areas care delivery, spurred on by the financial challenges of value-based care.

The Reed Group has found a way to use data to help physicians ensure that patients return to activity within evidence-based timeframes. Employers are starting to turn to this kind of data to lower healthcare costs, says Joseph Guerriero, senior vice president of the Reed Group. This same data could make a big difference for ACOs, as well, he adds.

The company’s MDGuidelines help organizations reduce the cost, compliance risk and complexity of employee absence, and are based on the American College of Occupational and Environmental Medicine evidence-based recommendations and more than 7 million observed cases collected and reviewed by clinical professionals. In addition, the Reed Group uses predictive modeling based on particular illnesses, psychosocial factors, geographic location, gender and other data to create the guidelines.

For example, Kaiser Permanente incorporated MDGuidelines into its electronic health records in 2012. According to a study published by the Integrative Benefits Institute (IBI), in the first year of using the guidelines, Kaiser saved 350,000 lost days per year and $30 million in one region of the Kaiser health system.

“Doctors aren’t normally trained to look at duration of absence,” Guerriero says. Worker’s Compensation is often fraught with a public image of patient resistance to wellness, but the opposite is true, says Guerriero says. Supplying physicians with these guidelines helps them direct patients toward recovery within expected timeframes, he adds.

In another approach, Evolent Health is focused on meeting the needs of population health through the use of data management.

According to Chad Pomeroy, Evolent’s chief technology officer, the company’s product combines administrative data, eligibility files, attribution lists, pharmacy claims—“everything a payer has”-- and pulls it into a platform, Pomeroy says. The company also connects into clinical platforms to get ADT information, laboratory values and CCD notes and combines it all in a warehouse. Then, Evolent does clinical profiling based on all the various aggregated information. Evolent also has the ability to push information from the platform back into the EHRs.

Doctors are gaining interest in all this added data. “Some doctors are secretly data junkies,” Pomeroy says. “Some are curious, but have never had access to high fidelity data. The new level of transparency is pulling them in.”

Value-based care is creating a tipping point in physician interest, he says. “It’s the new way they do business.”

Pomeroy offers words of caution and wisdom when it comes to data. “There’s a mountain of data and equally important is how you put it into action,” he says. The best question to ask is: what is going to empower value-based care, then work backwards from there.

With that in mind, Evolent has been working on pulling information from the unstructured data in a CCD--using “the loose data from notes,” Pomeroy says. Evolent pulls the data, parses it, stores it and mines it for key words.

Evolent’s parsing engine uses natural language processing (NLP)—a relatively new technology—to find words and relationships between words. For example, in a note a doctor might write “shortness of breath,” but never use “asthma,” and so it never gets coded, billed or into the patient claims record. This added missing information can make a risk analysis of patients more accurate, he says.

“There’s only so much that structured data can do for you,” Pomeroy says. “It needs to make sense to care providers. You can’t aggregate data for data’s sake.”

But nonetheless, Pomeroy can see a day when “outside” data that affects healthcare will be just as important to value-based care as the clinical and claims data. This date might include information on the patients’ exercise patterns, diet and community service.

Timothy “Dutch” Dwight, vice president at Medullan, a digital healthcare consulting firm, definitely agrees that alternative data, specifically retail data, could be of use to ACOs in their quest for value-based care. Understanding a patient population is critical. Information provided from healthcare purchases made at retail stores is available and usable ‘to track the behavior of consumers in a proactive way.”

After an ACO finds and determines which patients could be targeted for improving their health,t then they need to find a way to provide financial incentives to those patients. “Health plans are trying to come up with incentives to do that,” he says. “The question is; are these incentives making a difference?”

Dwight says he’s had a lot of interaction with companies trying to put ACOs together. The IT hasn’t caught up yet with what value-based care requires. More innovation is needed, he says. There is still a lot of unused data in the patient record, for starters.

Aside from data helping ACOs to accomplish their mission, data can also be used to help providers get a vision of their potential future in an ACO. John Kelly, principal business advisor for Edifecs, says data can be used to see where their risks would lie, if a provider organization joined an ACO. Organizations are having to sign on to an ACO “with a blindfold on,” he says, and “that’s why there are a lot of horror stories going on.”

Since members of an ACO rise or fall together based on their success to control costs and improve outcomes across the continuum of care, it behooves them to see the data horizontally ahead of time. “Risk-takers need to be able to look at all the data,” he says.

This requires the ability to test and manage thousands of data-sharing relationships. The data is likened to a supply chain, Kelly says. “The irony is: every other industry has been doing this. It’s about supply chain integration--all those things that other industries have already nailed.”

For reprint and licensing requests for this article, click here.