AI’s role in reshaping revenue cycle management

Revenue cycle management (RCM) is the intricate sum of many steps. From front-end registration to back-end payment collection, every part of the RCM process is an opportunity to progress billing calculations and payments along to reduce the risk of errors that could negatively impact revenue.

This is especially true for independent practices, where just a few associates, or even one person, are responsible for the end-to-end workflow. With eligibility checks, claims creation and tracking, denials management, and payment collection, there is no shortage of work to be done — and very little room for mistakes or delays. In fact, the American Medical Association cites challenges in managing payment as a top reason that practices are forced to sell or go out of business, in addition to staff shortages and experiencing burnout from heightened administrative burden.

The promise of artificial intelligence (AI) to help automate and alleviate administrative burden for independent practices, which are more resource constrained than ever, cannot be overstated. However, the effectiveness of applying AI in RCM relies on first improving our understanding of the technology and then thoughtfully identifying the most impactful opportunities to leverage AI throughout the revenue cycle.

Addressing common pain points

According to a recent Inovalon survey, eligibility, prior authorization, denials and patient payments are among the most significant pain points in RCM where AI could provide improvements — and it’s not a coincidence that all these steps are closely related and usually labor intensive.

Replacing manual processes with AI-driven workflows can help reduce common mistakes. For example, AI can search various data sets to identify patients’ active coverage much faster than a manual approach, accelerating access to care and increasing patients’ satisfaction with their health care experience.

Additionally, using AI to monitor staff workflows can help physicians and other clinicians understand the common mistakes within their revenue cycle. AI can identify the most frequent mistakes their team is making, which may range from errors at the beginning of the registration process to back-end claims creation mistakes. Armed with better visibility into frequent errors unique to their team, RCM leaders can better address their most costly challenges.

AI also creates opportunities for practices to become more productive with the same number of staff by handling routine, data-intensive and administrative tasks, ultimately augmenting staff workload without replacing human expertise. Streamlining work and reducing burnout are the key opportunities that health care professionals have noted in assessing where AI can offer the most impact, with denials management following closely behind.

Reducing claims denials

AI can flag potential claims denials prior to submission and offer guidance to correct errors, which helps to prevent denials while improving efficiency and increasing reimbursement rates.

In the long term, AI can help billing teams shift from reactively managing denials to proactively predicting likely denials — a perfect opportunity to combine human expertise with analytics trained on RCM processes and robust claims data to effectively address denials before they even occur.

The importance of bridging education gaps around AI

Though the potential for AI to help address common challenges and costly mistakes is clear, many physicians remain hesitant to weave AI into their revenue cycle processes, making education a good first step. If revenue cycle leaders dive into learning about AI — how it works, what the analytics are trained on, how other organizations are succeeding with AI in RCM, and understanding inherent risks such as bias in models and of users — their teams are likely to follow.

At a user level, those who gain a wider understanding of AI will be more likely to adapt well when it’s implemented and be more prepared to identify and address risks. Staff will realize the value of using AI to manage tedious steps, with confidence that their work is being done correctly, and embrace the time gained by greater efficiencies to better focus on patients and exceptions to standard workflows rather than being buried in their computers.

Building user confidence is crucial

Each team member will have a varying level of optimism in using AI, so building user confidence across RCM teams is vital for success. According to the same Inovalon survey, the top concerns around AI in RCM include accuracy and reliability (31%), lack of familiarity/understanding (17%) and AI being too new/untested (15%). Given health care’s high-risk environment, these concerns are unsurprising as staff are focused on minimizing risk while learning to maximize the value of their time and the tools available to them.

It’s also important for practices to keep in mind that AI is not static; this isn’t a solution that is implemented once and the learning stops. Practices should look for models built on representative health care data sets, as well as RCM and clinical expertise. These models continuously learn as users work with the AI, increasing their ability over time to anticipate and address challenges throughout the revenue cycle. AI benefits from the input of those who work with the models every day and who are managing RCM data. The models continue to “train” as staff complete their tasks, getting more and more accurate and further helping practices operate efficiently to boost clinical and financial outcomes.

Additionally, models built specifically for health care can help ease concerns around accuracy and unfamiliarity, as the insights not only help optimize the revenue cycle through automation but also help practices navigate evolving payer rules.

Embracing AI in the revenue cycle now will dictate its long-term success

As independent practices and large organizations continue to expand their understanding of AI, the technology is also learning from them. AI needs the input of those with feet on the ground: the staff managing RCM data, performing front- and back-end operations and overseeing the end-to-end RCM workflow. These inputs improve the potential value of AI in RCM as the technology increases in speed and accuracy, all while building greater trust from users. Staff can see their AI-powered tools become more valuable over time as the model learns their data and unique workflows.

Ultimately, practices that take advantage of AI now, trusting the technology to improve efficiencies and guiding the models with the appropriate input, will experience the most success in the long run. Although AI is still novel to many physicians and other clinicians, the time is quickly approaching when it will be the norm for practices to realize greater efficiency, experience fewer errors and increase time to focus on patients, thanks to the impact of AI in their revenue cycle. This will undoubtedly set the savvy practices apart from those still working claims manually or in basic technology, positioning the AI adopters to better face industry challenges, such as rising operational costs and ever-evolving patient expectations.

Julie Lambert serves as the president and general manager of Inovalon’s provider business unit, where she leads the organization that delivers cloud-based software as a service solutions to tens of thousands of provider organizations across the country, empowering them to achieve better outcomes and economics. She leads and is responsible for the management, product portfolio, sales, customer success, operations and the overall performance of the provider business. Before Inovalon, she served in various roles at Optum Rx, UnitedHealth Group and Prime Therapeutics.