Data Analytics Essential for Optimizing Accountable Care

The University of Texas Southwestern launched a comprehensive and successful Accountable Care Organization with significant savings using the help of data analytics and a series of customized dashboards and reports to elicit physician engagement and buy-in

The University of Texas (UT) Southwestern launched a comprehensive and successful Accountable Care Organization (ACO) with significant savings using the help of data analytics and a series of customized dashboards and reports to elicit physician engagement and buy-in, according to a presentation at the HIMSS17 meeting.

Overall, since being put into place, the UT Southwestern system has resulted in a 20% improvement in ACO/Healthcare Effectiveness Data and Information Set (HEDIS) measures. Moreover, it generated $6 million in savings during the first year of the program and $30 million in the second year.

"UT is really driving impressive results and health IT is really at the center of this transition," said Sam Stearns, MS, MBA, Vice President, Analytic Consulting, Verscend Technologies, during the presentation. Verscend is working with UT Southwestern on the solution, along with other vendors.

"This is all based on a foundation of data sharing with monthly reports and communications across a very broad physician network. Its all around driving results, whether that's better quality or savings," he noted.

UT Southwestern launched their Accountable Care Network (UTSCAN) in 2014. At this point, they had 17,000 lives under value-based contracts. In 2015, they were the largest participant in the ACO program, and by 2016 they had 177,000 lives covered.

To help implement the program, UTSCAN developed regional population management units, or pods, to help break up over 3000 physicians into more than 45 manageable groups. These pods meet monthly to share data, review reports, improve their referral process, and discuss best practices. Getting physician buy-in and commitment to accountable care was key to making the system work, said Christy Cawthon, during the presentation.

"With value-based initiatives going on, there are a lot of programs out there and it's important to keep trying new strategies," said Cawthon, Manager, Decision Support, University of Texas Southwestern Medical Center. "Define your strategy, and execute on it. Build a partnership with your physicians and stay focused, as accountable care is a marathon, not a sprint."

At a high level, data for the UTSCAN system are pulled from databases for paid claims, electronic medical records (EMR), and admission, discharge, and transfers (ADT). From these data sources, UTSCAN has developed predictive analytics applications and reports, to fully integrate data across the enterprise. These data are shared with all providers across pods, to engage providers in the metrics and costs.

Each provider has access to the dashboards and reports to track their performance against key quality measurements and, importantly, each other, fostering competition to get the best results. For this, UTSCAN extracted and standardized all EMR data across the ACO, and then drilled down on specific care gaps. This gave providers clinically-related measures and targets, and, from a system level, this allowed for resource allocation.

The dashboard also provides cost analyses, including trending costs and utilization efficiency calculations. Each team can view different sets of data and tools. This further allows the provider to focus in on high-risk patients, to ensure these individuals are addressed before an incidence, the presenters noted.

Home health utilization also became a focus of the program, which is using paid claims data and a risk adjustment to determine how best to use this approach. Using this calculation, a list of home health agencies (HHA) was narrowed down to 44 with ≥90% efficiency. This was cross checked with CMS STAR ratings to narrow down the list to 20 recommended HHAs, which UTSCAN uses preferentially.

"The CMS STAR rating is a practical useful approach to help bring quality information that compliments the cost information. This helps provide value," said Stearns.

To optimize the use of home health, UTSCAN first got physician buy-in for the program and created a team focused on home health utilization. This system was fed into the dashboard, to provide a report on individuals and the HHA costs. This helped each physician engage in the conversation. Using this system, there was a 15% reduction in home healthcare costs.

"We don't want to eliminate HHA, we just want to get people to the right care setting, so they can better use the tools that are out there," said Cawthon.

However, no system is perfect, as noted by a user of the UTSCAN pod system. Oftentimes, the system will assign a patient to a provider who may have met the individual once 3 years ago, he said anecdotally. When it comes to contacting and meeting gaps, this disconnection causes issues. Despite this, the commenter was optimistic that a solution was out there for this issue.

The next steps for the system is to continue to optimize the home healthcare component while also monitoring the quality and efficiency of each HHA, according to the presenters. Additionally, the reports and dashboards will continue to evolve, as new data sources become available and are leveraged.