The practical applications of AI in dentistry go well beyond the operatory
My thinking about how artificial intelligence (AI) technologies will improve dentistry—even fundamentally revolutionize it—has broadened considerably over the past several years.
As Lou Shuman, DMD, CAGS, explained in his February 2019 Dental Economics op-ed article, “How artificial intelligence is shaping dentistry,” we are far from the Hollywood vision of AI-powered humanoid robots. However, a range of AI technologies is already in use today in the form of computer software. AI-powered software automates routine tasks that are typically completed by humans. Sophisticated AI technology platforms can learn to make insights and find patterns in complex systems and vast amounts of data. Computer vision AI in particular holds promise for health care, since it allows analysis of content in images with superhuman speed and precision.
As a dentist looking to the potential of advanced technology, my AI musings initially tended to pass through a narrow clinical lens: How will AI benefit me and dental clinicians like me? I’d think about how AI could enable hyperaccurate, liability-free diagnostics and superintelligent treatment plans. I’d imagine perfectly fitted prosthetics and increased patient trust, and I’d see oral medicine finally integrating with systemic health care.
But it did not occur to me that AI could have a substantial impact on dental insurance beyond its obvious application to the automation of insurance claims submission.
Today, I would argue that insurance is an area where AI will prove tremendously beneficial to our industry, and sooner rather than later. From what I see, it’s clear that AI applications for dental insurance could have as big an impact on the dental industry overall as they do on the administration of treatment.
In fact, while AI applications for diagnostics and practice management are still under careful regulatory scrutiny, insurers in other industries are already benefiting from AI. According to the Center for Insurance Policy and Research, artificial intelligence is currently being used in claims processing, underwriting, fraud detection, and customer service. Auto insurers are using AI-powered virtual assistant chatbots to respond to customer questions 24/7 with natural-sounding language, saving time and money. Machine learning and computer vision AI allows insurers to more quickly and accurately assess and predict costs by analyzing data from historical information, sensors, and claims images.
Dental insurance is the economic engine of our industry. Let’s face it—as much as we may be loath to admit it, when the dental insurance field suffers, we all suffer.
First, dental insurance companies can claim a good deal of the credit for the integration of scientific and technological advances into the practice of dentistry, thanks to the fact that they enable mass consumption of dental care. Certainly every practicing dentist can attest to the fact that most patients will turn down cutting-edge treatment options if their insurance won’t foot a significant portion of the bill.
In addition, most practicing dentists can attest to the fact that as financial gatekeepers, insurance companies’ claims processing systems often arbitrate the quality and speed of patient care, since claims processing determines patients’ willingness to accept treatment. In fact, it was that side of the insurance equation that I first considered when examining AI’s near-term benefits to dentistry.
Previously, my considerations (and, yes, frustrations) as a doctor dealing with insurance companies neglected the insurer’s perspective. As dentists, we look at claims processing and see a system that, at best, slows us down and, at worst, aims to cheat us. If we take a step back and swap the optics, the darkly bureaucratic hassle we see in claims processing starts to look like an unfortunate but essential industry safeguard.
As it is, fraud and overdiagnosis, as well as waste and human error in the claims process, create a drag on insurers. Unfortunately, we know all too well that a few dishonest dentists can really skew the system for the worse. The Atlantic magazine recently covered one story of egregious dental insurance fraud and drew the conclusion that our profession is inherently susceptible to fraud.3 However, I know that the bad apples are the exception in our field, not the rule. It’s time we change the story.
Consider that a large insurance company may receive some 40 million claims annually. No insurance company, however large, has the manpower to review that many claims. In truth, only a fraction of claims are manually adjusted (I’ve heard estimates in the range of 3%–5%), and the rest are decided by methods that feel downright arbitrary and randomized. In reality, most claims aren’t decided by a trained professional, but instead by informed math models that by their very nature will statistically mark some good claims as without merit and some bad claims as worthwhile.
It is on that last point that we—and our patients—get frustrated. But if insurance companies blindly accepted every claim, then fraud, waste, and abuse would bankrupt the financial backbone of our industry, and we’d lose the only official intermediary between doctor and patient that helps to ensure appropriate alignment between diagnosis and treatment. Let’s not forget that fraud, waste, and abuse don’t just cost insured patients and dentists. It costs taxpayers billions in Medicaid system costs. Everyone can agree that this is a problem worth tackling.
So, how do we reconcile the problems doctors face when dealing with insurance companies with the problems insurance companies face when dealing with doctors?
Oral imaging is both the primary diagnostic evidence on which we prescribe treatment and the primary evidence on which insurers assess claims. It’s also the area in dentistry where AI can deliver the most immediate results by way of computer vision. In fact, AI systems based on computer vision technology for oral imagery analysis are on their way to surpassing the diagnostic abilities of humans.
Oral imaging diagnostic systems with computer vision AI will soon deliver superaccurate results at superhuman speed and bring the clarity of truth to dental insurance. They will guarantee doctors fair treatment from insurance companies, protect insurance companies from dishonest doctors, and give patients peace of mind when dealing with doctors and insurers alike.
No claim left behind—Manpower will no longer limit how many claims receive some level of review, because AI will evaluate all oral imagery accompanying every claim. Evaluative results can be used to assess a claim’s likelihood of approval, allowing insurance companies to rapidly and intelligently determine which claims need to be manually reviewed and which should be fast-tracked for approval.
Prevent insurance fraud, waste, and abuse—When all claims-associated images undergo analysis, then fraud, waste, and abuse in dental insurance claims will be largely eliminated. AI can detect instances of resubmission fraud in which dentists submit a single image depicting a particular pathology multiple times or submit the same image to more than one insurer.
AI can also flag potential instances of overdiagnosis and diagnostic error. Cleaning up diagnoses will reduce costs for all, lower liability for insurers and doctors, and mitigate risks to patients. And, of course, AI can instantly notify doctors when they have submitted the wrong image by mistake.
Automatic preclaim assessment—AI will eventually allow for full integration between dental practices and insurance imaging systems, providing real-time claims assessment and pretreatment cost transparency.
AI that evaluates images on the spot and communicates findings to the insurer will allow insurers to immediately inform doctors whether insurance will cover the prescribed treatment and, if so, how much insurance will pay.
Automatic preclaim assessment is truly the holy grail of AI for dental insurance. It will bring up-front financial clarity and diagnostic validation to a field of medicine that has been frequently lambasted for lacking both. Clarity of costs and validation of diagnoses will deliver an ancillary benefit to patients as well, since they are more likely to receive the correct treatment when diagnostics and treatment financials are discussed openly, in real time.
For some fairly obvious (though not always fair) reasons, insurance companies tend to occupy a place of dishonor in the hearts of dentists and patients alike. But like it or not, insurance is the dental industry’s financial engine. Anything that makes that engine more efficient is going to benefit dentistry as a whole. That’s why AI’s insurance applications will revolutionize the dental industry potentially as much as its therapeutic applications.