Much has been written about the advent of Artificial Intelligence (AI) in the healthcare payment lifecycle, but has it been oversimplified? The straightforward answer is yes.
Many articles have characterized AI as the end-means antidote that will solve all revenue cycle management (RCM) issues and easily streamline the reimbursement process from beginning to end. In fact, many of the products in today’s marketplace promise a complete solution based almost entirely on AI. But, looking at this from a fresh perspective, the reality is something more nuanced and collaborative.
The AMA uses the term “augmented intelligence” rather than artificial intelligence to “[focus] on AI’s assistive role, emphasizing that its design enhances human intelligence rather than replaces it.” This perspective captures both the current limitations and potential of AI as supportive technology, as well as the indispensability of human decision-making for the best patient outcomes.
Healthcare is Swimming in Data
Today, the healthcare sector is quite literally overwhelmed with data. In a third-party billing and reimbursement environment that is unique to the healthcare sector, insurance payers and governmental agencies dictate the processes required while also deciding the governing rules and regulations. From how to code each specific diagnosis to how much annual fee schedules will pay, many outside factors go into the reimbursement for any patient/provider encounter.
This leaves providers and healthcare systems at a significant disadvantage. They are responsible for not only delivering care but also obtaining their contractually discounted reimbursement. Staying up to date can be overwhelming.
With each payer issuing their own predetermined processes, forms, and fee schedules, AI’s true value proposition is to make sense of the voluminous data and to continuously learn and adapt to the constantly changing payer parameters. With real-time access to insurance payer information and clearinghouses, AI-driven software can create and maintain a current and up-to-date library of information for processing claims with the latest available data.
AI as a Digital Labor Support Force
In the past few years, the potential for AI to replace human workers in the healthcare RCM process has been overstated. While AI certainly has its strengths and can make quick work of the often redundant and burdensome administrative workflow associated with patient access and RCM, human intelligence (HI) will always be needed to provide context, continuity, and empathy.
An Example Affecting Reimbursement: Prior Authorizations
Perhaps a better way to look at AI’s current potential is to view it as helping humans make better decisions rather than automating 100% of RCM processes. Take prior authorizations for example — they are the least automated patient access function with less than 21% of providers using a fully automated prior authorization process.
This means that providers and staff performing prior authorizations manually must spend an average of 16 hours per week to process 40+ individual pre-auths per provider. And while many insurance payers publicly agreed to relax their prior authorization requirements during the COVID-19 public health emergency, 52% of providers reported that this was never fully (or even partially) implemented.
The Synergy of AI + HI in the Healthcare Payment Lifecycle
Continuing with the case example of prior authorization, applying AI-driven advanced automation to the process can allow repetitive tasks to be completed in real-time with little human intervention, including:
– Determining if a prior authorization is required
– Continuous learning regarding prior authorization rules specific to each insurance payer
– Real-time processing of prior authorizations based on information gleaned from the patient’s EHR/EMR
– Electronic follow up on unanswered requests 24/7
However, HI is required when dealing with requests for clinical documentation, emergency or semi-emergent care, or outliers that bring complexity and demand nuance. Empathic decisions are often required which means HI needs to be augmented and adapted with AI, rather than replaced. The most valuable systems are those that help people make better decisions and actions while learning continuously. In other words, AI helps automate tasks by incorporating intake data, making sense of it, and presenting it in a way that allows human workers to get the best possible outcomes for their patients.
In other words, AI, at present not being sufficiently advanced to replace humans as a technology- eg, is only able to work in a set of predictable use cases at the current moment so humans need to keep track of the changing rules and regulations, to provide domain-specific intelligence in making the AI, and make complex higher-level decisions.
What’s to be Gained By a Streamlined Prior Authorization and RCM Process?
Whether looking at prior authorizations and the patient access process or focus on back-end solutions geared towards accounts receivables and denials management, Augmented Intelligence — the reciprocal and collaborative effort of AI and HI, is poised to take administrative processes off the plate of humans who have more important things to do. This will free up providers and staff members to do the following:
– Improve patient care and increase the time available to spend with patients
– Reorganize workflow to more adequately meet or exceed best clinical and operational practices
– Increase revenue through reduced mistakes/errors, fewer claim denials, and increased patient capacity
Improving the Patient Experience and RCM Outcomes
From a business perspective looking past the current operational constraints, AI can remove many administrative processes from the plates of humans who can better spend their time on patient care using AI-synthesized information for better outcomes. If an organization can rethink business/operational systems beyond the people-centric systems in place today, the opportunities to improve the patient experience are boundless when the deluge of information is tamed and human energy refocused.
As healthcare costs continue to shift toward patient consumers and they assume a more financially relevant role in healthcare, provider groups, and healthcare organizations are finding it important to reallocate capital and resources toward business models that support the patient experience and improve outcomes.
By incorporating a holistic and comprehensive AI + HI solution, organizations are reorienting from a delivery-centric model to a customer-centric focus, thereby increasing their competitiveness while positively impacting their bottom line. When weighing future best practices, consider that decision-making in healthcare shouldn’t be completely automated because it still requires human intelligence and human empathy. Thus, AI is best used in an “augmented intelligence” capacity, where it assists human decisions. Importantly, it also allows us to reallocate human resources to essential patient care tasks.
About Navaneeth Nair
Navaneeth Nair is Vice President of Products at Infinx Healthcare which provides leading-edge AI-assisted end-to-end solutions across the payment life cycle, including patient access, prior authorization, and revenue optimization. Navaneeth has over 20 years of experience in healthcare, where he has specialized in leading large-scale technology product and solutions development.