How AI Can Reduce Your Risk of an Audit

There has been enough talk among healthcare professionals about artificial intelligence (AI) to last a lifetime. And invariably, any conversation about AI turns to fears of stolen jobs or additional work for already overburdened staff. But what if AI in physical therapy could do the jobs no therapist or support staff wants to do all while making everyone’s lives in the clinic a little better? 

If there’s one piece of clinical workflow that most any staff member would like to avoid, it’s compliance and—more specifically—audits. Whether it's the routine chart audits done internally, the bigger, badder audits done by payers, or Medicare’s targeted probe and educate (TPE) audits, reviews are inevitable—and you need to be ready. Rehab therapists are in luck, though—rather than creating more work,  AI can reduce your risk of an audit. How exactly? For that answer, let’s look at the ways AI can serve as a sort of digital assistant in your practice.

AI helps improve documentation compliance.

Although it’s great to have a record of past treatments, physical therapists document because they have to, not because they want to. And with long days of treatment and who knows how many patients to document, it’s inevitable that there may be some mistakes. That’s why a PXM platform leveraging AI software builds compliance checks on the back end of workflows—so clinicians can spend less time worrying about those things. Unlike providers or staff, AI can quickly and easily scrub every word of every note and look for patterns that will either yield a compliant note or a non-compliant one. The AI-powered platform then reports its analysis back to the provider to improve the compliance of future notes and scrub current ones in the event of an audit.

Using AI technology—interconnected with a PXM platform—will boost your practice’s ability to weather any audit. When a Medicare Administrative Contractor (MAC) comes calling for one of those dreaded TPE audits, being able to walk them through your compliance process—inclusive of machine learning models—will inevitably expedite the process to a favorable result and serve as an “insurance policy” for your documentation.

AI can streamline chart audits.

The current private practice standard of internal manual chart audits suffers from two big failings: 

• It is being done by other providers who may—or may not—have an eye for compliance, and
• It takes time to audit one chart, let alone every single chart. 

Needless to say, the current standard for audits is terribly inefficient—and ripe for an audit of its own. AI can help you get ahead of the curve by auditing every single note in a fraction of the time, allowing your staff to get back to patient care with improved insight into their documentation practices.

AI boosts reimbursable coding.

Having rehab therapists coding treatments and acting in part as medical billers often raises the question of whether they’re the person best suited for that role within the practice. Too often clinic directors and practice owners lament that their new hires have limited knowledge of how best to assign CPT codes for their interventions and documentation. Conversely, those new hires may call into question the ethics or legality of certain billing practices. 

The truth is that this difference in thinking between manager and employee is rooted in a poor understanding of proper CPT code application and the rules around coding, which are often too complex to fully grasp for providers—and even medical billers, in some cases.

With AI technology that uses deep learning to recognize patterns of words and apply them to the letter of the law in the CPT code, the software can begin to recommend better coding practices to therapists with concrete evidence—scrubbed directly from a compliant note—to support a claim. With AI assistance, clinicians have to worry less about their coding and get to spend more time in patient care. Employers, meanwhile, are happy with the improved reimbursements seen downstream. 

AI empowers staff with actionable data analytics.

Aside from behind-the-scenes workflows, AI can also be used to provide insights and metrics that staff and management can use to improve patient care. There’s evidence that rehab therapy clinics aren’t doing enough to harvest data that could help their business: in the 2023 State of Rehab Therapy report, only about half of providers overall are collecting outcomes or clinical data, with the rate of collection for other data even lower. 

With the right tools, practices can not only collect data about their operations, they can use it to learn where they can improve. Take documentation, for example—looking at the State of Rehab Therapy report again, 91% of providers are working on documentation outside of working hours.   

And yet data from Prediction Health AI shows that the best notes in terms of compliance are signed-off between 8 AM and 3 PM,  with CPT coding risk increasing after hours or more than 24 hours after treatment has been completed. With that knowledge in hand—as well as concerns about risks to privacy and compliance that come with documenting at home—practices could use the AI tools available to them to better accommodate documentation during the workday.

AI can help you understand how to document.

When it comes to writing, finding the right words can be challenging. (Believe me, I know.) And while writing patient notes isn’t the same thing as writing a blog, it matters what words you choose and how you use them when you’re communicating with an audience—in this case, other providers and potential auditors. 

A feature like Prediction Health’s AskAI tool offers a HIPAA-friendly tool to help you build your notes in case you suffer from a bad case of writer’s block. Some clinicians have found this tool particularly useful in building bullet points for an evaluation and assessment summary, while others have used it to create SMART goals.

Furthermore, we often hear about how under-documentation yields audits and compliance issues, but over-documentation can also invite additional scrutiny. Machine learning models in a PXM platform can scan every single word in a note, sift out the fluff and filler, and assist providers in writing succinct and accurate notes that accurately reflect the codes they use for the day’s treatment.

AI can help with burnout. 

No one goes to PT school (or OT or SLP school) because they love documentation, and only a few get as excited about billing and compliance as John Wallace, PT, OCS, SVP of Member Value at WebPT does. So when something comes along that can streamline and practically eliminate some of the burdensome documentation and regulatory requirements that are often attributed to burnout—which happened to rank third in our report for the 36.1% of respondents stating they are burned out—most any rehab therapist should be jumping at the chance to implement AI in physical therapy. 

An audit is a scary process for anyone on the receiving end. Having someone scrutinize every word, note, and code you used for a given case can feel invasive—and it can be difficult to stomach the criticism. If you’re open to adopting it in your practice, AI can reduce your risk of an audit altogether or at the very least, make the process less painful when Medicare—or someone else—comes knocking. From removing the need to document at home to canceling the quarterly manual chart audits or improving clinic-wide compliance, AI in physical therapy is here to help you avoid an audit and improve your Practice Experience Management.

Get started with WebPT and Prediction Health today.

If you’re interested in adding Practice Intelligence with Prediction Health AI as part of your PXM platform, fill out the form above and someone will reach out to set up a discovery call.