A vortex of trends threatening health system finances have made using advanced analytics including artificial intelligence (AI) and machine learning (ML), to streamline the revenue management cycle, both necessary and inevitable.
For nonprofit hospitals as the well-known saying goes “no margin, no mission” and there is no question that margin has come under pressure like never before. According to Moody’s, nonprofit hospitals’ profits declined significantly amid the pandemic. Specifically, nonprofit hospitals had a median operating margin of 0.5% in fiscal year (FY) 2020, down from 2.4% in FY 2019, and an operating flow margin of 6.7%, down from 8.4% in FY 2019. Kaufman Hall research showed that 2021 hospital revenue would likely be down between $53B and $122B due to the lingering effects of COVID-19.
At the same time, Americans who have health insurance through their employers are increasingly being pushed into high-deductible plans, making them less confident that they can afford care compared with traditional plans, according to the National Center for Health Statistics.
Today, more than 30 percent of a large health system’s revenues are collected directly from patients, up from less than 10 percent just a decade ago. We have indeed entered a new “patient-as-payer” paradigm, which has now forced hospitals to thereby assume more financial risk of patient nonpayment, while also taking on more financial risk from payers in value-based contracting arrangements.
And let’s not forget that the U.S. adult uninsured rate has been continuously rising over the last 4 years.
Clearly, health systems are in dire need of some financial relief, and naturally, many are increasingly looking to automate portions of the revenue cycle through enhanced AI & ML platforms and methods to increase payment yield and velocity while optimizing FTE productivity. In fact, according to Fortune Business Insights, the worldwide market for revenue cycle management (RCM) solutions is expected to reach $268 billion by 2027, a jump of 250% from 2017.
Why AI and ML?
By automating important – but also routine – RCM tasks, health systems can free staff to work on more productive, highly-complex endeavors that require human intervention, while reducing labor costs, mitigating preventable revenue leakage and improving the overall patient experience.
AI algorithms “learn” by performing repetitive, high-volume tasks and building knowledge as they complete those tasks over time. Intelligent analytics and workflow powered by AI & ML can handle larger and more complex activities as they gain more experience – with the ultimate goal of approximating human intelligence to quickly solve complex problems. AI and ML are at their best when automating tedious, high-volume tasks that are repetitive and prone to human error. While AI offers numerous benefits to health systems’ RCM processes, below we’ll examine a few in more detail.
Increasing staff productivity: Hospitals employ armies of workers focused on patient and payer follow-up, devoting countless hours to resolving a variety of complex claim problems including denials, underpayments and slow payers. These laborious tasks often involve one-off phone calls to inquire why a claim was denied, then leaving the staff member to make various attempts to research information in different hospital systems, request medical records and the like, all with the goal of getting the claim paid at expected reimbursement and as quickly as possible.
AI and ML can upend this process, predicting denials and other claim problems based on results from millions of other claims historically submitted to the same payer. In this respect, the AI system is effectively pre-empting the payer’s denial, by alerting staff to the specific missing or incomplete information that is likely to cause the denial in the first place.
Improving patient communication and loyalty: As patient financial responsibility rises, insured rates fall and U.S. income inequality soars to levels not seen since the Great Depression, it’s time for health system executives to face an uncomfortable reality: It’s virtually certain that patient collections will become increasingly difficult in the future unless hospitals change their current approach to RCM.
Fortunately, this is where Applied Analytics, AI and ML come into play. AI can change billing from a more provider-centric process to one that is more patient-focused, personalizing the process with whatever combination of variables have been shown to yield results from that individual patient — or patients who embody similar characteristics.
Those variables may include factors such as preferred method of payment (Apple Pay, Venmo, debit card, interest-free payment plan), preferred channel of communication (text, email, phone call) as well as different times of the day or month to deliver communication and different types of content within the communication itself. As the consumerism of healthcare paradigm continues to take center stage, the experience delivered to patients must, too. Healthcare providers must now forge long-term relationships with patients, similarly to retail-centric experiences. Consider your preferred airline carrier where you are signed up for miles and rewards. They know your preferences and patterns. They communicate with you in the manner you’ve designated and can offer targeted value offerings to enrich the consumer experience and ultimately loyalty.
This is where AI and ML can be leveraged to discover the types of communications individual patients are most responsive to. Delivering said retail-like experiences can ensure more predictable and even an uptick in net revenue as the industry shifts from episodic healthcare to building more long-term healthcare consumer loyalty. We must pivot, putting AI and ML to work, gathering longitudinal consumer intelligence and thereby creating a continuous learning curve focused on value to the most important constituent – the healthcare consumer.
Collecting small balances: Low-dollar account balances sometimes confound health system financial departments because they lack the resources to chase these low-value amounts, and staff time is typically devoted toward touching higher-value accounts. The result is that these low-dollar accounts, which add up over time, are sent to collections, leading to unrealized or delayed revenue.
By simply automating the billing process though AI and catering it to individual patient needs, hospitals stay on top of low-value accounts, increasing net revenue performance and enabling staff to focus on more complex tasks.
It’s no doubt that increased margin pressure and lack of business automation has resulted in tremendous financial angst for most hospital providers and if recent trends persist without new solutions, the situation for many health systems is likely to grow worse before it gets better. Injecting applied analytics, AI and ML into the RCM process should be the inevitable next step for health systems to achieve much needed financial relief.