Artificial intelligence (AI) has the potential to catch fraud, waste, and abuse in the claims submission process. The Dental AI Council (DAIC) recently spoke with Dr. Linda Vidone, chief clinical officer and vice president of clinical management at Delta Dental of Massachusetts, as well as a founding member of the DAIC, about the technology’s impact.
DAIC: How did you first become aware of AI and its applications to your business?
Dr. Vidone: It was interesting. It happened almost overnight. At the National Association of Dental Plans (NADP) meeting last year, all the AI companies were showing their demos, and it was amazing. We went from never hearing much about AI to having three or four companies ready to show a carrier how they can easily identify fraud, waste, and abuse. I’ve never seen anything happen so fast.
DAIC: Since this happened so quickly, are there areas where you feel these AI vendors may not understand all the nuances to the challenges carriers face?
Dr. Vidone: The companies that I’ve seen, I think they’re spot on in terms of what needs to be done. But I’d have to say not all of them understand all the nuances of identifying and addressing fraud, waste, and abuse.
I’ve been in the industry 20 years now, and it used to be you’d worry about the scaling and root planing claims. You’d worry about upcoding extractions from simple to surgical. You’d worry about too many buildups. Well, now the technology’s gotten more sophisticated, and I have to say that bad actors have kept up and gotten more sophisticated, too. So, it’s not just the top five codes anymore. It’s more complex.
Take fillings, for example. You ask, “What’s filling-related fraud, waste, and abuse?” Well, now these bad actors are upcoding. They’re adding more surfaces to the fillings. It’s the combination of codes on a claim. AI holds the potential to help identify those upcoded claims sooner rather than later. And actually I feel in some cases AI would prevent the fraud from even happening.
DAIC: We do hear from a lot of AI vendors that the better you get at combatting fraud, waste, and abuse, the less likely it becomes, because people who know they’re going to get caught won’t do it.
Dr. Vidone: Yes, and it takes time. Let’s consider scaling and root planing. Every company wants to be able to identify bone loss. Well, I think some tools are more sophisticated in the AI field than others. And you know what, let’s face it, it takes a while to develop an AI with that skill. You can’t just program an AI system overnight. You have to train it. It takes many, many claims. You need the data.
DAIC: Every insurance carrier has slightly different standards and approaches these things in different ways. Is that a problem for some of these AI providers?
Dr. Vidone: Not in my experience. AI companies are building flexible products to respond to varying industry needs. And what we need today will likely change in the future. Each carrier has to address its own compliance requirements and strategies to keep up with change.
DAIC: Carriers have already been using AI of different types to conduct statistical analysis or analyze natural language in claims. What we’re talking about now is the computer vision side of AI and these newer diagnostic systems that can read x-rays. Is that something that every carrier is already doing?
Dr. Vidone: Dental carriers leverage technology differently. I can’t speak for all carriers, but I do think many are investigating how to better utilize AI for things like identifying duplicate x-rays. So, for instance, if a bad actor submits the same exact x-ray for five different patients, an AI tool could quickly identify those kinds of duplications.
DAIC: That sort of computer-vision AI is very distinct from the one that’s looking at statistical analysis. Would you like to see investigation into the possibility of bringing those two things together, or at least creating processes internally on your side to allow you to integrate those analyses?
Dr. Vidone: Absolutely. Yes. I think there’s a lot of good research that you could do with claims data.
DAIC: Is there, do you think, a tolerance among carriers for sharing the data that could provide for interesting research into this sort of synergy? Or is this proprietary stuff where getting those solutions isn’t going to be easy because carriers won’t part with their data?
Dr. Vidone: It would be tough to do without a full review of the proprietary and privacy issues. But I think a wide range of oral health stakeholders are interested in finding ways to better understand data across populations to identify trends and inform public policy, too.
DAIC: One of the concerns in general, whenever AI is brought up, is, “Is it going to take away my job, or is it going to reduce the need for a human role in in a process?” Do you see that as a valid concern?
Dr. Vidone: While I understand that concern, AI is a tool. In order for these tools to work most effectively, we will need highly skilled individuals like dental consultants to engage in clinical review.
If we continue using the example of duplicate x-rays, AI is going to help identify all these duplicate x-rays, and our clinical experts will be able to also focus on more complicated clinical review cases. It’s really going to enhance and help identify some of the bad actors in dentistry.
DAIC: So AI just makes sure that their time is being used efficiently?
Dr. Vidone: Exactly. If I’m paying a dental consultant, I want them applying their expertise rather than reviewing claims that a computer or someone without clinical experience could be helping to review.
DAIC: AI is usually brought in as an enterprise solution for efficiency and cost savings. But one thing we’ve heard from various people is, “Bottom line, whatever we do with AI, it should be conceived with the intention that the patient care comes first” and another thing we’ve heard is that insurance carriers do care about the satisfaction of their network providers. Broadly speaking, do you think the use of AI by carriers can contribute to improvements in provider satisfaction or patient care?
Dr. Vidone: Absolutely. AI tools hold the potential to improve both the patient and provider experience. That future is what I find most intriguing about AI in dentistry. We’ve been talking primarily about the ways that insurance carriers are thinking about AI today but in the future, things like real-time review and approval for covered benefits will be a game changer in many ways.
DAIC: You still receive a certain number of claims in analog form, mailed in, faxed, or scanned. Is there a need on the carrier side for some sort of standardization of what information and what quality of information is required for submissions?
Dr. Vidone: Standardization translates to speed. If all dentists are submitting their claims electronically for example, it means we can adjudicate that claim, even if it requires additional review, much faster than if the information is submitted via snail mail, for example. Having an electronic submission standard also helps mitigate potential human error.
DAIC: Provided there’s a technology in place that allows every single claim to be reviewed, is there an argument to be made for the benefit of submitting imagery with all claims? With AI able to facilitate new levels of consistency, does an organization like the NADP need to bring all the carriers together to develop industry-wide claims submission protocols and standards?
Dr. Vidone: It is a good question to raise. Regardless of the level of standardization, I think consistency within individual organizations is possible, with the help of AI, and will benefit the entire clinical ecosystem.
Dr. Linda Vidone is the chief clinical officer and vice president of clinical management at Delta Dental of Massachusetts. She also is a founding member of the Dental AI Council (DAIC), a non-profit devoted to helping define the future of artificial intelligence in dentistry through research, education, and thought leadership. To join to the DAIC’s effort, visit dentalaicouncil.org/membership.