Using Artificial Intelligence to Improve the Hospital Revenue Cycle

Using Artificial Intelligence to Improve the Hospital Revenue Cycle

Each year when they draft their budgets, hospital executives and other healthcare leaders are challenged to find ways to cut costs while maintaining high-quality services within the hospital revenue cycle.

Fewer dollars are flowing into hospitals because of changes to the care delivery landscape, and the cash hospitals have available is often spent on emergent necessities, such as data security, attracting and retaining top talent, or overdue facility upgrades.

Coupled with the growing number of patients enrolled in high-deductible health plans and lower commercial payer and Medicare reimbursement, healthcare organizations are forced to be creative with the resources they have so their organizations can accomplish the most with the least.

Healthcare’s high-cost problem is systemic, and the onus is on healthcare organizations to identify the root causes of their high-cost areas and think outside the box to drive some of those expenses down.

 

The financial drivers healthcare executives care about most

Throughout my years in healthcare, I have come to understand the challenges hospitals and other health systems face in budgeting. 

So, what has the biggest impact on operating margins? Labor costs typically eat up more than half of a hospital’s operating revenue, and that costly challenge is poised to increase as organizations face continued pressure to raise salaries and wages. Hospitals are now rethinking the roles of some employees.

The revenue cycle is one area of the hospital that is labor-intensive. Hospitals have to ensure their organizations are paid for services rendered or, in other words, that they collect payments from patients and payers. Financial leaders need to consider revamping their organization’s approach to revenue cycle management.

Common holes in the revenue cycle

Hospitals are constantly battling denials as payers require authorization for more types of procedures or adjust policies without any notification to providers. A denied claim triggers effects that spill over to disrupt other aspects of the hospital’s finances, too. For example, after a payer denies a claim, administrators are forced to chase the denial, which often requires multiple calls to the ordering physician and working closely with the payer. This is a time-consuming, and not always rewarding, process. The claim then sits in accounts receivable until it is adjudicated, which can take weeks – even months.

Reprocessing claims drives up labor costs as billing staff is forced to devote more time to unpaid claims. According to research from the American Medical Association, the industry could save $15.5 billion each year if companies processed claims correctly the first time.

The hospitals and health systems I have worked with have pointed toward lower dollar accounts as another key area where revenue is lost or delayed. That is because hospitals often lack sufficient resources — such as staff or budget — needed to touch each and every patient account. What ends up happening is organizations focus primarily on the high-dollar accounts, which means many low-dollar accounts slip through the cracks. Unfortunately, all those missed payments from low-dollar accounts add up. 

Hospitals often turn to an outsourced vendor to tackle these low-value, high-volume accounts. But similar to in-house teams, these outside partners aren’t always able to physically act on all accounts. 

On the bright side, there’s a potential solution for this persisting challenge: inserting AI into revenue cycle processes.

 

Improving the revenue cycle, starting with artificial intelligence and automation

As leaders devise revenue cycle improvement plans, they should consider how various solutions could affect bottom lines and operating margins. Leaders should also determine the resources needed to fulfill plans, such as an outsourced partner or a tool built in-house or purchased elsewhere. AI-enabled automation is one such tool that may help. 

This intelligence-based category of automation tools frees staff from necessary but redundant and time-consuming tasks, and it enables the organization to reposition skilled workers toward meaningful work like quality assurance programs within the revenue cycle. Automating some processes within the revenue cycle is an affordable way for organizations to reduce labor costs and recover leaked revenue because it enables organizations to touch both high- and low-dollar accounts. 

When hospitals automate portions of the revenue cycle, they can execute a higher volume of claims at a cost-effective rate and regain their edge. And as the machine learning components of this automation learns, it unleashes meaningful insights to the organization and allows for continued opportunities for efficiencies and increased revenue recognition.

Automate the mundane using AI to move employees into modern roles

AI in the revenue cycle is really about revenue integrity, a hospital’s ability to achieve operational efficiency, compliance, and optimal reimbursement. The technology allows hospitals to repurpose employees and mitigate revenue leakage in ways that require more critical thinking. 

Automation takes high-volume, repetitive, and mundane tasks off employees’ workloads, allowing hospitals to expand production at a much lower cost. Hospitals can then re-invest these freed up resources into those tasks that require a more human touch such as empathy, creativity, and complex thought. 

AI allows organizations to do a few things differently with their revenue cycle staffing and strategy.

First, with redundant tasks accounted for via automation, hospitals can strategically reassign skilled workers to duties they’ve historically struggled to fill or to non-urgent but essential projects that have sat too long on the backburner. 

Second, AI enables organizations to address workforce shortages. For example, rural markets often have trouble attracting top talent and may struggle to fill important positions. On the other side of the coin, hospitals in competitive markets may struggle to retain employees they have invested in. By automating portions of the revenue cycle, rural hospitals can move loyal and committed staff members into positions that have been difficult to fill due to talent shortages. 

Third, AI can help talent retention in more competitive markets, as hospitals can reward loyal employees with more consequential roles as they are no longer needed for tasks that can be completed with automation.

Moreover, AI enables healthcare organizations to be more strategic with the work they outsource. Hospitals can work the accounts likely to be paid in-house while outsourcing the others to a third party.

Overall, AI is really a movement toward an exception-based work environment in which employees only take on tasks automation cannot. In such a workflow, people will only be notified to monitor automated tasks on an exception basis. Automation can help reduce or eliminate the risk for error made by a person charged with typing day in and day out or copying information verbatim from an insurance card or EHR system.

In the same way, AI helps control where sensitive patient information is sent. It eliminates the need for at least one set of eyes on the data and enables information to be securely sent and accessed, meaning hospitals have greater control over who accesses patient data and where it goes.

Getting started

One question I hear often is, “Where should I start?” My advice? Start simple.

The best tasks to automate are tedious, high-volume, and repetitive. In the revenue cycle, some of these processes include checking the status of claims, working denials, or reviewing eligibility. These tasks are typically time-consuming and copy-and-paste intensive. By automating them, a hospital can address every patient account, regardless of the dollar amount attached to it. 

After tackling the easy stuff, organizations can progressively deploy AI for more advanced capabilities. 

Still, justifying spending thousands of dollars on new technology can be difficult. Organizations are inundated with high expenses. Consider the EHR — a single hospital stands to spend millions of dollars on a new system. But this is the wrong way to think about automation. Hospital leaders shouldn’t think about AI and automation as another price tag, but as a path to savings. What automation comes back to is being able to accomplish more with less. 

Since automation enables staff to accomplish more work with fewer resources, hospitals can put additional quality controls and checks in place to help speed the time required for processing claims, reduce days in accounts receivable and reduce denials. 

That’s because automation via an AI system helps staff in a couple of important ways. First off, it completes nearly 100 percent of eligibility checks on the front-end of the revenue cycle. Additionally, it helps highlight changes in payers’ reimbursement policies, which helps employees act sooner to reduce A/R days or notice variations in accounts much earlier.  

Here’s an example: Before automation, one hospital I worked with was spending nearly 100 hours per day checking claims status. With automation, the hospital was able to get through that same amount of volume in 90 minutes per day, freeing up staff to further the revenue cycle claims management process rather than waiting for payers to respond before proceeding. Automation helps get more answers with less manpower, and AI helps them turn those answers into new solutions.

 

Planning for the future

Nearly every hospital executive needs options to do more with less. AI helps hospitals operate smarter, not harder. Through automation, hospitals level up to achieve exceptions-based environments in which employees are empowered to focus solely on work that requires the human touch. Whether a hospital is operating efficiently now or struggling and in need of improvement, every organization is at its best when people are serving the best purpose.  

 

About Braden Lambros

As an Executive Director of AI Transformation Strategy at Olive, an AI as a service (AIaaS) company, I observe the promise and potential of automation in cost control firsthand. I have devoted my entire career to healthcare, working as a consultant focused on revenue cycle and other labor or non-labor areas. I have worked with more than 25 healthcare organizations, from small hospitals to some of the largest health systems and academic centers in the U.S.

Click here to learn more about how Olive can drive efficiency and increased revenue recognition.

Applying AI, Automation to the Healthcare Revenue Cycle

Applying AI, Automation to the Healthcare Revenue Cycle

Inefficiencies in the healthcare revenue cycle represent opportunities to apply Artificial intelligence and automation when approached thoughtfully.

 

Artificial Intelligence (AI) and intelligent automation, in general, are changing the face of modern business, and this is as true in the world of healthcare as anywhere else. You may have at least a general understanding of how intelligent automation can benefit your healthcare business, but identifying the best use cases and defining a pragmatic action plan can be difficult.

Here are steps to assist healthcare professionals in identifying and implementing the best use cases for artificial intelligence within the revenue cycle processes of their organization.

1. IDENTIFY USEFUL TYPES OF INTELLIGENT AUTOMATION

The first step in getting started with automation is understanding the technologies available to you at a high level. There are two core types of intelligent automation available: AI and Robotic Process Automation (RPA).

AI learns and iterates as it goes by completing repetitive, high-volume, error-prone tasks while collecting intelligence on this work along the way. This approach allows the technology to ultimately identify and address larger, more complex, opportunities for automation. The goal of AI is to mimic human intelligence using computers, enabling them to solve complex problems quickly and to scale.

On the other hand, RPA simply takes a set of inputs and produces an output based on a predefined set of rules. RPA doesn’t “learn”— it reacts the same way every time. RPA can complete the same types of repetitive, high-volume, error-prone tasks while collecting simple metrics for reporting.

Regardless of the technology, the objective of intelligent automation in healthcare, or any other industry, is the same. Take repetitive, high-volume tasks that are done in a similar way every time and offload them to software, freeing up human capital to focus on more important tasks.

Human capital is often reduced to the number of hours human employees spend completing their work, but that view has to be expanded to truly see the impact intelligent automation can have on your organization. At the surface, using automation to replace repetitive tasks humans complete can save money — bots don’t require benefits or vacation time — and reduce errors because typos don’t exist in their world. But at a deeper level, automation can free up your human employees to focus on more complex skills, like customer service, patient advocacy, and empathy.

There are several forms of intelligent automation that offer a multitude of capabilities, but humans possess infinite intelligent skills. When you’re able to apply your staff’s skills to the most important parts of your business rather than to the processes that automation can conduct, that is when you truly see the value intelligent automation can bring to your healthcare organization.

2. UNDERSTAND THE APPLICATION OF INTELLIGENT AUTOMATION

With an understanding of AI and RPA, you can start to see where their applications lie. Rule-based business processes (e.g., insurance verification, data recording) are prime candidates for RPA. More complex judgment-based processes (e.g., eligibility checks that require a review of electronic health records) can benefit from the application of AI.

Categorizing your revenue cycle processes can help set the tone as you continue to brainstorm precise processes to automate and what that implementation should look like. Keeping that constant throughout your brainstorming helps solidify your choices.

3. BRAINSTORM TO IDENTIFY PROCESSES TO AUTOMATE

The next step is brainstorming, identifying specific processes in your business that can benefit. As you begin to approach specific use cases for intelligent automation within your healthcare organization, it is important to consider the impact automation can have and the speed with which these tasks can be completed with automation versus how they are done now.

There are several common places impact and speed can be targeted within the revenue cycle, and account updates are usually a great place to start. These kinds of updates are usually a massive undertaking and can touch many or all of your organization’s patient accounts, and are tedious tasks that have to be done correctly.

Impact is immediately recognizable as automation is much faster than humans, reduces errors to negligible numbers, and can prevent costly errors down the line, saving your organization money. In this kind of automation, speed and impact are intertwined.

4. APPLY PROBLEM AND SOLUTION-ORIENTED APPROACHES

To identify specific processes to offload to intelligent automation, you can use either a problem or solution-oriented approach.

The problem-oriented approach looks to identify bottlenecks and areas where employees spend the bulk of their time performing repetitive tasks. The solution-oriented approach looks to optimize workflows and makes sense when there are no clear bottlenecks. This approach identifies where key performance indicators (KPIs) can be positively impacted by implementing intelligent automation. Whichever approach you take at this stage, remember that every moving piece in a healthcare business has multiple high-volume administrative processes. Intelligent automation does not have to be limited to one department or team — everyone can reap the benefits directly by improving their daily work, or indirectly by experiencing the positive impacts of these improvements in other tangent workflows and cycles.

A NOTE ABOUT DIY OR AI-AS-A-SERVICE

Once you have identified where AI should be applied, you have to consider how you’ll get it implemented. Building an automation solution requires a team of highly skilled developers, project managers, and automation engineers. The DIY approach can make sense, but for many, the upfront costs and lack of technical knowledge create barriers. Partnering with an AIaaS provider makes the benefits of intelligent automation more easily accessible to businesses of all sizes and allows you to focus on core business competencies, instead of development.

Want to get started?

To take the first steps toward automation at your organization, download this free guide. You’ll learn how to identify the right processes for automation, and start to build the case for AI and automation.