Data analytics helps Texas-based ACO find cost savings in MSSP
Accurate analysis based on easily verifiable data has been extremely helpful in identifying and reducing costs in the ACO, which up until the platform was adopted, were difficult to determine.
Texoma ACO is a 45-provider accountable care organization located in Wichita Falls, Texas, participating in the Medicare Shared Savings Program. Our ACO, comprised of leadership and providers from the 15-specialty Clinics of North Texas, was accepted in the MSSP in July 2012 and has 6,800 attributed patients.
Our leaders believed that clinical and financial data analytics would be essential for providers to achieve the clinical quality and cost metrics associated with the MSSP. Moreover, consistent performance reporting and benchmarking would support the organization with its long-range goal of transitioning to the era where compensation will be based on metrics associated with cost-effective, high-quality care.
To assist in this regard, we chose a population health management data analytics platform that unfortunately proved to be unusable. We eventually implemented a second system, this time from a different vendor, Lightbeam Health Solutions. Chosen after several months of due diligence, the second system delivered the analytics and performance reports our providers and administrators now rely on for insight into clinical quality and costs.
Our second system — which analyzes the ACO's clinical, claims, pharmacy, labs and other data — delivers a more accurate and timely view of patients' care history than just claims analysis alone, while also delivering insight into costs. This comprehensive view is fully transparent to the physicians' own records, helping earn buy-in and behavior adjustment among those whose performance or cost metrics are outside of program compliance. We believe achieving this provider buy-in has helped the ACO accelerate its progress in the MSSP, despite providers having little practical experience with pay-for-performance initiatives.
Negative first experience with analytics technology
Texoma ACO's founders recognized years ago that fee-for-service was being phased out of the industry in favor of value-based payments. In 2009, Clinics of North Texas' leaders began attending educational sessions by the Institute for Healthcare Improvement and participated in ACO Accelerated Development Learning Sessions sponsored by CMS. There they learned how population health management would be crucial for success under value-based payment contracts. Although our providers used EHRs, we lacked familiarity with the information technology tools necessary to extract and analyze data for monitoring patient populations.
Searching the market, we found a population health management system that the vendor claimed would suit our needs. After nine months, however, we struggled to generate even basic reports, such as a list of Medicare patients with one specific condition or a list of Medicare patients for a single physician.
The few reports we could create often contained errors, such as attributing conditions to patients that never received that diagnosis, for example. This lack of reliability posed a significant obstacle to performance-improvement collaboration. When physicians see errors on reports that they can easily confirm by reviewing their charts, they are unlikely to trust subsequent analysis generated by the technology.
After meeting several times face-to-face to resolve issues over several months, the vendor eventually stopped responding to our support requests. The accumulated difficulties we had with the platform and vendor finally prompted the ACO to make the difficult and expensive decision to uninstall the system.
Data transparency crucial for behavior change
Undeterred, we began investigating replacement systems. The second selection process was much more thorough for the vendors we considered. We demanded numerous demonstrations of how our data would be captured and normalized for consistency as well as how our providers and care coordinators could generate reports on-demand. Internally verifying the accuracy of reports was also crucial to educate providers and encourage their buy-in.
We immediately observed how much more user-friendly the second system was to use than the previous platform. The care coordinators, who would be the primary operators of the tool, lacked population health management experience using such technology and were particularly grateful for the intuitive interface. In addition, unlike others we considered, the platform captured and analyzed clinical, claims, pharmacy and lab results data to offer a comprehensive view of patients' care.
Our physicians, some of whom were concerned about having their performance evaluated on the population health management technology, became supportive of the insight delivered by the tool based on its data transparency. Demonstrating to the physicians that the data used to create these reports came from their own charts, claims and orders made a much greater impact in adjusting their attitudes and behaviors.
Informed by the analysis, some physicians were able to rapidly address care quality metrics, such as intervening with diabetic patients with uncontrolled HbA1c levels and scheduling mammograms for women who had not been screened within the recommended time frame. The ability to identify and address these care gaps proactively also positively influenced providers' perception of the technology.
Eventually, the data analysis became so embedded in physicians' workflows that some providers began requesting customized reports showing, for instance, the cost per each visit for patients with certain diagnoses. This indicated to administrators that while the patients' health was the physicians' top priority, they were beginning to understand how the ACO's financial performance affected their compensation as well.
Immediate cost savings
This newfound insight into patients' care made physicians much more aware of the costs. For example, early after the analytics platform implementation, an OB-GYN identified a significant prescription-cost anomaly for a patient.
The physician learned that the patient had been prescribed an injection by her primary care physician at the cost of $6,000 per month. After speaking with the patient, the OB-GYN determined the injection was not delivering a noticeable benefit, but the patient continued the treatment because she wanted to follow the primary care physician's recommendation. Administrators collaborated with both physicians to explore other evidence-based treatment options. Fortunately, an oral medication was found that costs only $40 a month and has also shown to be more clinically effective and tolerated by the patient.
In this one example, we improved our financial performance in the MSSP, but there are other benefits, too. With many patients now paying more out-of-pocket, such high-cost medications may either become a financial hardship or be avoided. Intervening with a lower-cost prescription reduced costs for the patient, provider and payer while improving the patient's condition and satisfaction.
Data-backed insight encourages collaboration
Apart from the insight into cost patterns for our providers, the ACO is also able to identify how caregivers outside the ACO, but within the care continuum, are impacting the organization's overall cost-per-patient. This is another key metric we can analyze through the population health management technology.
For example, Texoma learned that the home health agencies where we had referred patients were heavily contributing to our overall costs. After investigating, we determined that some patients referred to home health agencies continued to receive visits from caregivers, even though they acknowledged to no longer needing the assistance. This finding has led the ACO to explore less costly alternatives, such as having a clinician from our organization perform targeted home visits to patients driven by analysis from the population health management tool.
Accurate analysis based on easily verifiable data has been extremely helpful in identifying and reducing these costs, which up until the platform was adopted, were difficult to determine. In addition, having evidence to demonstrate to physicians why some care decisions may be unnecessary has helped identify more cost-effective protocols when appropriate.
This collaborative approach will be essential for our success in the MSSP in the years ahead. Beyond the MSSP, we hope to leverage this experience and our information technology tools to confidently pursue future Medicare- or commercial payer-sponsored value-based payment programs.
© Copyright ASC COMMUNICATIONS 2017. Interested in LINKING to or REPRINTING this content? View our policies by clicking here.