• Request a demo of the AI-driven technology that's changing healthcare

    Ready to see how we're starting the healthcare revolution? Complete the form below and a member of our team will be in touch with you soon about a demo.

  • This field is for validation purposes and should be left unchanged.
Blog   |   4.29.19

Building a Sustainable Quality Measurement Process with AI

Building a Sustainable Quality Measurement Process with Artificial Intelligence (AI)

Plans and providers are increasingly evaluated and compensated based on the value of care they provide. 34% of U.S. healthcare dollars are tied to value in some way, and that number continues to increase year over year. In 2018 alone, 190 million members were covered by healthcare organizations implementing HEDIS quality measures, and CMS paid $6.3 billion in bonuses for Medicare Advantage (MA) plans who had strong quality performance under the Star Ratings program.

All of these stats reinforce a single theme: As healthcare organizations enter into more value-based arrangements, the need for accurate, scalable quality measurement becomes critical. Whether they’re taking on upside or downside risk, or simply receiving bonus payments for specific care activities, plans and providers need to understand how they’re performing against predefined quality measures. If they perform well, they have the potential to bid for new territories, acquire more members, receive higher rates from government programs like Medicare and Medicaid, and secure additional bonus or incentive payments. If they don’t, they may lose out on revenue, market share, and renewals.

While the industry shifts away from fee-for-service payments to fee-for-value has had many positive impacts on care delivery, cost containment, and overall alignment of incentives across government programs, private payers, and providers, these benefits come with a price: laborious and expensive administrative overhead required to perform quality measurement activities.

The Challenges of Quality Measurement

Currently, there is no universal set of measures used across all quality programs, markets, or payment arrangements. The most widely adopted measure set to date is HEDIS, which itself has over 90 measures and dozens of sub-measures. Medicare’s Shared Savings Program (MSSP) has over 30 measures. Other programs have varying numbers of quality measures with different reporting requirements. For payers and providers participating in multiple value-based arrangements, the time and resources required to do quality measurement work can add up fast.

The amount of effort required to measure and report on quality depends on the measure set. Some quality measures involve simple calculations based on information coded into claims. Others require provider staff members to perform chart abstractions in search of specific patient test values, screenings, or procedures. Finding evidence for these clinical measures is manual, time consuming, susceptible to human error, and expensive. It also pulls nurses and other provider staff away from high-value activities that impact patient experience and care.

How AI Can Improve the Quality Measurement Process

Manual quality reviews aren’t sustainable for provider groups—particularly smaller physician practices who don’t have the luxury of hiring additional administrative staff to scale their efforts. Fortunately, artificial intelligence (AI) has the potential to supplement people-driven processes and make quality measurement programs sustainable. AI algorithms can quickly and accurately locate relevant patient data in medical records for quality measurement, streamlining the review process and surfacing trends that can be used for provider education efforts to help physicians improve their clinical practice.

In the future, AI could also further the development of more meaningful clinical quality measures. Today, many quality measures are process-based measures that are easy to assess from structured data or manual reviews. Using AI-driven measurement processes, providers could evaluate clinical outcome and patient experience measures more easily, which could open the door to other kinds of performance metrics that paint a clearer picture of care quality.

Given the proliferation of value-based care, quality measurement activities aren’t likely to slow down or become any less complex. AI can help payers and providers keep up with the increasing demands of quality programs without putting undue burden on staff workloads and administrative budgets.

Want to streamline your quality measurement process? Apixio’s AI-powered Quality Identifier solution can help.

Ashley Taylor Anderson Profile Image
Author
Ashley Taylor Anderson