Accountable care movement depends on clinical analytics

U.S. healthcare is in the very early stages of a transformation from a focus on volume to value-based purchasing. In other words, instead of getting paid for how many procedures and office visits they do, health systems are launching "accountable care organizations" (ACOs) that will get rewarded by payers for keeping groups of patients healthy, regardless of how many procedures, lab tests or visits it involves. They may also face financial risk if they fail to meet clinical quality measures.

Those recently formed ACOs present a new type of knowledge management challenge for healthcare organizations because they will require a sophisticated deployment of clinical analytics and collaboration between a larger pool of providers and across organizational boundaries in order to identify and close gaps in care.

So what is clinical analytics? By most definitions, it is a system that puts up-to-date clinical information in the hands of providers to help them improve both patient outcomes and compliance with evidence-based guidelines.

The technology roadmap for an ACO is coming into focus: Digitize the information in electronic health records, overlay actionable analytics and provide point-of-care clinical decision support, allowing caregivers to focus their valuable time on patient care and improving outcomes, says Mansoor Khan, CEO of Alere Analytics, a company that specializes in analytics and clinical decision support technology.

Claims and clinical data

Two pools of knowledge are in play: one to standardize care for all patients, and the other to meet the advanced needs of chronically ill patients. "You need to stratify those patients who  are  at higher risk," Khan explains. With diabetic patients, for instance, several measures are the most important to follow, including hemoglobin A1c, cholesterol levels and regular foot exams. Using data warehouses and clinical dashboards, health systems can remind physicians to focus on those treatments during office visits. Those tools also give the organization the capability to compare how providers are performing with chronically ill patients and how they can improve.

Khan says it is now possible to combine clinical and claims data in one comprehensive view to take risk stratification to the next level. "Previously it was only done with claims data," he explains. "Now that we can create evidence-based care gap scores for patients with both clinical and claims data, the fidelity goes way up. We can benchmark who your best providers are and look at their practices, and try to roll those practices out to your whole network."

Cynthia Burghard, who tracks ACO trends for IDC Health Insights, has given several presentations on the IT tools ACOs need. She says that beyond the analytics, ACOs also need tools to manage patients across settings of care, such as patient registries and consumer engagement tools to work on medication adherence and behavioral changes with patients. "That is why you need the analytics first: to help identify people most likely to make changes and devote resources to them," she says.

In a 2012 survey of healthcare leaders about accountable care, IDC Health Insights found that the top four reported capabilities for which healthcare organizations intend to use analytics are:

• the ability to identify patients/members in need of care management,
• clinical outcomes analysis,
• performance measurement and management, and
• clinical decision-making at the point of care.

Actionable data for physicians

Brown & Toland Physicians, an independent practice association of more than 1,500 physicians in the San Francisco Bay Area, is a "pioneer ACO" and is using a cloud-based clinical analytics tool called MinedShare from Humedica, a company that recently became a part of OptumInsight.

"Humedica's tool is relatively easy to use, has attractive features such as graphical views and offers both pre-delivered and ad hoc reports," explains Adrian Rawlinson, M.D., director of medical informatics for Brown & Toland. After working through data governance issues, Brown & Toland will begin pushing out reports with actionable data to physicians. "We have groups of measures on clinical standards involving pneumonia vaccine, obesity, tobacco use, diabetes and hypertension measures," Rawlinson adds.

Those reports will be published and sent to physicians as an e-mail link. "Eventually we want it embedded in the electronic medical record," Rawlinson says. "A real deficiency in the current generation of EMRs is the lack of point-of-care population management tools."

Ronnie Brownsworth, M.D., CEO of the Piedmont Clinic, an independent practice association with 380 employed physicians and 520 community-based physicians in the Atlanta area, recently said that the important thing about setting up a clinical analytics platform is that it allowed his organization to start tracking progressive improvement as it gradually moves toward an accountable care environment. In a recent Healthcare Informatics magazine webinar presentation, Brownsworth said, "It gave us a track record verifying that the work we do is making a difference, which helps get physicians' buy-in to move forward."

Piedmont is working with a clinically integrated IT platform with a built-in physician portal (from Recombinant by Deloitte, recomdata.com) that serves physicians using a very wide variety of electronic medical records.

In the same presentation, Mark Golberg, general manager of the Provider/Payer/ACO Market Sector for Recombinant by Deloitte, added that healthcare system leaders just starting down the ACO path need to consider what questions they want to answer in order to change care, and then think backward from there about the technology and data sources needed. "None of the predictive analytics will do what they say if you don't have the right data coming into your data warehouse and workflow applications," he said.

From push to pull

Another group using Humedica's MinedShare tool is Anceta, the clinical analytics subsidiary of the American Medical Group Association (AMGA). John Cuddeback, M.D., Ph.D., chief medical informatics officer for Anceta, gave a presentation at the recent Population Health and Care Coordination Colloquium in Philadelphia. He explained that MinedShare extends the AMGA's model for shared learning based on comparative clinical analytics. Medical groups share data to identify opportunities for improvement and to recognize best practices.

Anceta's large medical groups contribute data from their electronic health record systems, and physicians receive performance reports with comparative data. MinedShare's predictive analytics helps them identify patients who may soon fall into one of the chronic disease categories.