Artificial intelligence, machine learning, robotics — all terms we hear a lot in 21st-century healthcare, but often in a forward-looking context. And while it’s true that the application of many promising emerging technologies has yet to be seen, many are currently in deployment — and driving real results. In other words, the robots are here.
In the effort to address healthcare overspending and fight fraud, waste and abuse in the system, payment integrity is a key area of focus for both healthcare organizations and regulatory bodies. And with process improvement at the crux of efficiency, cost reduction and compliance, artificial intelligence and machine learning-driven technologies are enabling healthcare payers to identify, recover and even prevent payment inaccuracies with unprecedented precision.
At HMS, we’re deploying several new and emerging technologies to maximize cost containment and improve outcomes. Here are the four new tech applications that are currently driving the greatest results in payment integrity — and where we see the greatest long-term potential.
1. Artificial Intelligence (AI) & Machine Learning
There are so many groundbreaking applications of AI in healthcare, and the possibilities are, quite literally, limitless. From medical image analysis to precision medicine to administrative efficiency, AI is having a profound impact on the way we deliver and manage care.
With regard to health plan payment integrity, artificial intelligence is being used to automate the claims review process for more accurate billing and payment, significantly reducing the need for pay and chase. Leveraging data from a vast and expanding network of health plans and states — and the knowledge accumulated from processing billions of claim records annually — ML-driven algorithms allow us to select claims for review that are most likely to be paid improperly or that have the greatest likelihood of recovery.
While exact definitions vary and can get quite complex, the key component of ML systems is that they continuously learn from the data they’re fed, which aids in pattern detection and facilitates continuous improvement. Through our experience and data, we are progressively improving our methodologies for predicting, identifying and recovering improper payments to drive value for all healthcare stakeholders.
2. Autonomous Coding Review Using Natural Language Processing
Natural language processing (NLP) is a branch of AI that essentially analyzes human language. When we think about the potential for NLP in claims administration, we might picture things like smart text detection, predictive text and optical character recognition (OCR) that replace or augment manual processes. And though these technologies are important, language is about so much more than words — and medical coding language is no exception.
NLP systems that are able to analyze beyond written text alone are providing us with a deeper, more contextual understanding of the types of coding languages being used. By taking what we learn from our NLP analyses and applying ML algorithms, we are able to improve the accuracy of our findings and, again, continuously improve our detection systems.
At HMS, NLP is redefining the way we review medical records and claims to ensure services are accurately coded, billed and paid. Records are read autonomously, which eliminates the risk of human error while speeding up both the review and recovery process.
3. Robotic Process Automation
The concept of robotics in healthcare is probably more likely to conjure images of robot-assisted surgery than robot-assisted payment integrity processes, but in fact, robotics is playing an important role in healthcare cost containment.
Using software-based robotics process automation, we are able to streamline many of the complex functions required to identify and address improper claims payments. Rather than having to wait for a trigger before moving to the next task, robotics is allowing us to automate these paths and move seamlessly through the identification and recovery process for quicker, more accurate results.
4. Advanced Data Matching
Patient identity management continues to be a challenge that has a substantial impact on the care continuum. With the increasing volume and complexity of healthcare data, ensuring the long-term accuracy and security of patient health information is becoming more difficult — and more critical — than ever. Interoperability and mechanisms that facilitate cross-organizational data exchange are key to addressing this issue for payers, providers and patients.
Drawing on the industry’s most current and comprehensive eligibility data source, we are using AI-driven match logic to obtain third-party coverage data from external sources and deliver the information at multiple points throughout the member journey, including before services are rendered. This capability is playing an enormous role in healthcare cost avoidance — with full coverage information from the start, providers are able to secure prior authorizations from the appropriate payer and bill correctly the first time, improving payment accuracy and reducing administrative rework at both the payer and provider level.
Integrating Big Data, New Tech Into Your Operations
While many of these new and emerging technologies promise to optimize processes and reduce costs, getting started can be a bit daunting. Our advice is to start small — identify a specific use case that would benefit from process optimization and for which you can easily establish a desired outcome and measure the results. This will allow you to put the right team in place and establish an agile framework that can be scaled up as you take on more projects.
Are there any processes in your organization that could benefit from the implementation of AI, ML or another new technology?
Find this blog as originally published on FierceHealthcare, a resource for healthcare industry news and insights.