Required Research (part 2): Where & how can AI help the industry?
About the Author:
Ophir Tanz is an award-winning entrepreneur and CEO of Pearl, an AI company focused on solving problems in the dental industry.
In part one of this this three-part post, I wrote about the basic research required to validate why, on a fundamental level, the dental industry needs AI. Here, I’ll outline the second variety of research that will allow us to more efficiently capitalize on the technology’s potential.
To put AI to work, we need to understand the distinct needs and perspectives of each of the industry’s constituents. What matters to a practitioner is different than what matters to a DSO is different than what matters to an insurance carrier is different than what matters to an OEM, and so on. We need research to identify where, how and to what degree the infusion of AI will serve each of the various dental stakeholders’ priorities.
Let’s consider insurance carriers. A carrier’s emphasis is going to be on utilization management, which is its core cost center. So AI’s most obvious utility for a carrier would be in the adjudication of claims and related processes, where the technology can supplement the work of human examiners. But, before it’s going to adopt AI into its utilization management process, a carrier first needs to know what the potential cost savings will be. How many human examiners will it still need to employ? Will it increase denial rates––and what knock-on effects could that have for the network at large?
The answers to these questions aren’t just going to be useful to carriers. DSOs will want to hear them too, because, if denial rates do go up, then they need to be thinking proactively about how to get more claims approved––and they’ll probably want to bring an AI to bear on that problem. Attendant to that priority is consistency of care across its offices, talent retention and, of course, practice acquisition. These are all areas where AI could be brought to bear, but can, for example, AI allow a DSO to identify sufficient additional treatment opportunities to justify the cost of AI integration? Will inserting AI into the clinical process have any unforeseen impact on the relationship between a DSO and the practitioners it employs? We need research to answer these questions.
On the OEM side, there are also questions about how practitioners perceive AI-integrated products. Are they welcome? Is AI a threat? How does it affect patient care? All of those questions tie into a whole line of inquiry relating to the practitioner experience. Studies are needed to understand that perspective, because, after all, the practitioner will be a frequent end user of the technology.
I could go on listing from this roll of stakeholder-specific questions, but it suffices to say that all are important, everyone wants them answered and, until this point, any answers that may exist are not readily available. After all, where answers have been found, it is often because a company has conducted its own research––and there is little incentive among competing companies to share this kind of information. To provide answers that will make AI truly beneficial to every stakeholder, we need independent research – case studies, feasibility studies, cause-and-effect studies, pulse surveys, and financial and statistical analysis – that can be widely accessed and shared by all.
At the moment, everyone is climbing the same AI mountain, but by different routes and with different equipment. The DAIC gives us a base camp from which we can make the ascent together.
