HJAR May/Jun 2024

HEALTHCARE JOURNAL OF ARKANSAS I  MAY / JUN 2024 45 Niki Carter, DMD, MPH Dental Director Delta Dental of Arkansas that once the algorithm has been imple- mented with certain data, the information of the data sets may not be available when it needs to be employed. Other concerns include: 1. AI will replace jobs that humans have. While this may be true of certain tasks, it makes the overall task more efficient, faster, and less error prone. 2. Due to the data sets, AI can reinforce bias and discriminatory methods, due to how it was engineered and trained. 19 3. Even though the long-term use of AI may result in a decrease of healthcare costs, the short-term costs in imple- menting AI in dental practices, dental carriers, and other types of insurance payers, is increased. 20 4. There exist legal ramifications for use of AI. As it is rapidly entering dental ed- ucation, it will be extremely important to obtain permission of use, depending on the laws of the country. 21 The World Health Organization outlined guidance that provides six principles to en- sure AI works for people’s benefits: 22 1. “Protect human autonomy: The use of AI or other computer systems should not diminish human autonomy, so humans should remain in control of the healthcare system and medical decisions. 2. “AI must not harm people: Regulato- ry requirements and governance are needed to ensure the safety, accuracy, and efficacy of AI for health. 3.“Ensuring transparency, explainability, and intelligibility: AI-based technolo- gies must be understandable to devel- opers, medical professionals, patients, users, and regulators. 4. “Fostering responsibility and accountability: Humans require clear, transparent specifications of it lessens claim appeals, offers more trans- parency of the system, and pays faster and more efficiently. This area of AI is growing and improving and will likely advance in the near future. FWA increases the overall costs of vital healthcare for everyone, including busi- nesses, suppliers, hospitals, and providers. Plus, it can also be harmful to patients. 12 Ac- cording to the U.S. Department of Justice, it is estimated that more than $100 billion is lost to FWA nationally. 13 A huge edge that AI technology has compared to humans is that an enormous amount of data can be assessed extremely fast — as in seconds. Humans are slower, forgetful, and, even if exceptionally smart, the human brain can only process a limited amount of information per second. 14 In the case of FWA, AI can detect if a film has been submitted for insurance claims previously on another patient. This allows the payer to catch FWA versus the regular automa- tion processes that exists. 15 It also saves time since the required documents that accom- pany a claim are identified as missing byAI technology. 16 This newer technology should result in protection of patient healthcare and result in cost savings. Due to numerous claim submissions, only a small percentage route to dental consultants to ensure the treatment is medically necessary. AI makes it possible that the claims most essential to be reviewed by human eyes are routed to the consultant team. 17 With any new innovation, there are con- cerns and drawbacks. First, it takes massive amounts of data for AI to classify and ac- curately predict the tasks assigned. Second, data is needed from patient records, which are not only regarded as confidential and protected information, but there also exists a reluctance between institutions to inter- change health data. 18 Another challenge is Significant headway has been made inAI technologies of treatment planning. By as- sessing and inspecting the patient’s records, evidence-based treatment recommenda- tions can be made to assist the dentist in important data collected that may have been overlooked previously. Algorithms flag de- viations and point to inconsistencies that may have been missed previously, which could potentially affect patient care. 7 AI caries detection can be an aid to the dentist in diagnosing lesions, but AI can- not make the definitive diagnosis, as this requires the expertise of the dentist. The traditional approach without an adjunct advanced resource can be inconsistent, depending on what the “office day” holds. Fatigue levels or variance of provider expe- rience can affect caries detection. 8 Available today and currently in use are AI software programs that dental practitioners utilize in office to identify carious lesions consistently. These act as tools to enhance the examina- tion and offer additional information. Bone loss can be marked on films in peri- odontal disease, signaling there is potential disease present, which aids dentists in eval- uating the disease stage. 9 This identification and automatization function has multiple innovations, as in patient education, engage- ment, and appointment scheduling. 10 Use of AI is ever growing and expanding in the claims payment industry and in detecting fraud, waste, and abuse (FWA). In AI-supported claim system reviews, films can be analyzed and identified for those claims that contain teeth in question necessitating further review. Those that warrant further review are routed to den- tal consultants. Claims that are not routed for human eyes are approved and paid au- tomatically, which offers consistency and efficiency for both scenarios. 11 This is an ap- plication of AI that is a win-win situation as