Artificial intelligence (AI) is a concept that has moved very quickly from the realm of science fiction into real and practical utility in a number of different industries—including healthcare.
One of the most effective applications for AI-powered tools is in the field of revenue cycle management. Sophisticated AI technologies can sift through vast amounts of data and parse nuances in ways that can make tasks such as claims and denials management vastly more efficient and effective. Understanding how those tools work—and best practices associated with their introduction and optimization—should be a priority for healthcare professionals.
When it comes to using AI to maximize workflow in revenue cycle management systems, one of the most important requirements understanding your core challenges. Use outcomes-based analytics that add context to the data and determine what is and isn’t working.
Payer denials of submitted claims, for example, involve a rich set of data that could indicate a wide range of potential problems. The issue could be in the claims process itself, the documentation, the training, or the execution. But if you don’t understand the root of the problem, you won’t be able to design an effective solution—or even determine if an AI tool could be a productive remedy. AI-driven support is likely to be most effective when implemented throughout high-touch processes, or those processes that are regularly problematic and require frequent and continued intervention.
Another best practice in applying AI solutions is to define success—and the metrics you will use to measure that success—early in the process. Perhaps it’s to reduce the number of days for the average claim submission lives in the system, or to reduce the number of denials. Define success, measure your baseline well, and be clear about your desired outcomes.
Finally, once a solution is in place, don’t just “set it and forget it.” Rules and circumstances can change rapidly, especially in the fast-evolving world of revenue cycle management. And because what works today might not be as effective in the long term, it is essential to maintain data and revenue cycle vigilance going forward.
AI and robotic process automation (RPA) capabilities have been most meaningful to revenue cycle context in follow-up and denials workflow. The future of the claims follow-up workflow will feature significant automation of repeatable tasks without any manual intervention and active prioritization of inventory at the account level based on payment or denial risk factors. However, when manual inputs are required, AI can identify the point at which intervention can be most impactful, efficient and productive.
Additionally, AI can employ workflow specialization to actively assign specific accounts to the right staff members to maximize the expertise and capabilities of both individuals and the team. This strategic talent allocation allows those with greater specialization and advanced skillsets to be utilized where and when they are most impactful.
Providers can start small, leveraging available capabilities in their patient accounting system to automate certain tasks, to increase the sophistication of rules that route accounts to work queues, and to create better insights into workflow bottlenecks. With that in mind, revenue cycle teams should not be afraid to get creative and prove out limited trials to better build a case for further investment in more technology and more operational focus applied toward developing these new capabilities.
The Missing Link
One underappreciated element of successful AI tool implementation is the value of deep-dive account-level investigation. A big-picture data review can help point to large issues, but it isn’t until you look in-depth at individual examples and understand why denials happened that you can start to create data and turn anecdotal observations into a meaningful data set. Until that process is complete, the potential upside for using AI solutions in denials management will be limited.
Many medical practices are rightly excited about the potential of having these new tools at their disposal. However, some providers that have made significant investments in new tools and technology have experienced limited success in translating those tools to implemented solutions that actually drive financial outcomes.
In many cases, the missing link is an expert-generated data set that is richer and more revealing. Ideally, revenue cycle experts and data experts will work collaboratively to conduct a targeted review. With dedicated resources that “speak the language” and can clearly articulate problems and solutions, true integration across both stakeholder groups can unlock the full potential of AI-driven solutions.