Providers and payers have become more collaborative as payment models evolve toward value. Still, the payer-provider relationships can seem one-sided—decidedly in favor of the payer—as hospitals continue to face declining reimbursement and rising costs. Health leaders looking to optimize revenue cycle management, help reduce denials and improve financial performance should arm their teams with the right tools to better predict payer behavior.
Discussions around predictive analytics in healthcare are often centered on tech-enabled modeling to predict patient needs and behaviors with the aim of better care management, quality improvements and population health initiatives. In the revenue cycle, modeling is also used to improve the patient’s understanding of their portion of the cost of care to help improve collections at the time of service.
“Predictive analytics, coupled with business and artificial intelligence, allows you to deploy resources more effectively because you have a good idea of what is going to happen.”
Hospitals have used predictive data models to help improve patient collections however, the benefits that predictive analytics can bring to the revenue cycle don’t end with patient access. In recent years, hospitals have learned the value of using similar data models to assess a payer’s propensity to pay or deny claims.
“In the last six years, we have leveraged predictive analytics to model the behavior of health plans,” Dawn Castro, vice president of Revenue Cycle with Conifer Health Solutions, recently told Becker’s. “We apply what we’ve learned from the models to predict, with regular accuracy, whether a payer will dispute, deny or pay a claim and in what timeframe.”
Understanding the reimbursement patterns that payers have adopted around a variety of claim types allows an organization to leverage its revenue cycle workforce more strategically. For example, if it is known that a particular payer isn’t likely to remit payment for care delivery until day 14 (regardless of the claim type or when it was submitted), the team won’t be prompted to work an account until day 15 or 16. Organizations can use these insights to redirect the workforce to resolving more complicated accounts or performing tasks that require more critical thinking.
Scott Rowe, vice president and CIO with Conifer, said these capabilities can be especially useful for provider organizations with limited resources. “Having this insight is like being able to see around corners,” he said. “Predictive analytics, coupled with business and artificial intelligence, allows you to deploy resources more effectively because you have a good idea of what is going to happen.”
The right team, the right data, the right price
As hospitals reboot their approaches to revenue cycle management, leaders should consider predictive analytics capabilities when looking to integrate new revenue cycle technology solutions. However, simply having the capacity to execute predictive analytics isn’t enough if you haven’t considered all the implications that come with new-found intelligence about the hospital’s financial performance.
Conifer has provided tech-enabled revenue cycle management for more than two decades and has compiled a substantial amount of data around payer behavior in that time. The company also has a sizeable workforce dedicated to making sure this data is regularly assessed in a manner that is as operationally relevant as possible.
“The value of the data is only as good as the talent and the technology tools that you’ve invested in to turn it into business intelligence that can be used to improve your revenue cycle’s performance,” Ms. Castro said. “We have a dedicated Optimization Team that interprets the data and keeps our predictive models refreshed in near real-time.”
Smaller provider organizations with limited resources will find it difficult to create the infrastructure to capture and analyze significant amounts of data without the aid of a technology partner. And while large health systems may be able to tap into their own data to form predictive models, this undertaking still requires considerable investments of time and funds.
“One of the biggest barriers to the adoption of predictive analytics in the revenue cycle is the sheer capability and resources needed to build and maintain the necessary infrastructure,” Mr. Rowe said. “Not every organization has the money to hire an expensive, full-time team of data scientists to build data models that are standardized across the enterprise.”
Trust the people, trust the technology
While the use of predictive analytics can be a game-changer for revenue cycle management, it is no replacement for human ingenuity. And, contrary to popular belief, the application of emerging technologies like artificial intelligence and robotic process automation does not necessarily coincide with workforce reductions. In fact, the uses of tech-enabled tools in the hospital revenue cycle are simply meant to elevate the importance and capacity of the staff. Predictive analytics, AI and automation free staff from doing tedious tasks that require little skill.
“Not every organization has the money to hire an expensive, full-time team of data scientists to build data models that are standardized across the enterprise.”
Communicating these realities to the staff is essential to generate buy-in. The effective application of predictive models often highlights the need to reimagine traditional revenue cycle processes, so the staff needs to believe that the new technology and updated protocols actually work. To achieve this, front-line users should be included in the implementation process to ensure all parties are properly informed of the technology’s capabilities and allowed to voice any concerns.
“Trust is the number one thing,” Ms. Castro said. “We get staff involved in workflow development … and we give them a seat at the table in case they find something they’d like to change down the road.”
Reimagining the revenue cycle to drive innovation takes time. Involving staff in the rollout of new technologies is key to building trust. Without the support of the humans who do the work, the application of even the most advanced technology will likely fall short of expectations.