Well being insurers are starting to undertake AI to help prior authorization selections. However is that this an excellent factor? Specialists weighed in throughout a panel dialogue held by KFF on Thursday.
One panelist mentioned she has some questions on using AI in prior authorization, and added that there needs to be extra transparency on this subject and the way usually prior authorization requests utilizing AI are overturned.
“I believe that as we see using AI improve, one query we’ve is, what’s the information that’s going into these algorithms? What information are these algorithms based mostly on? Are they probably the most present information? Do these algorithms embrace outdated research that won’t mirror the most effective medical information that we’ve proper now? How usually are they being up to date? Are they being inspired to disclaim care, at the very least on the first degree?” mentioned Anna Schwamlein Howard, principal of coverage growth on the American Most cancers Society Most cancers Motion Community.
Nonetheless, there has just lately been extra scrutiny about using AI in healthcare, which is an efficient factor, in line with Kaye Pestaina, vp and director of the Program on Affected person and Client Safety at KFF. She famous that simply final week, there was a Senate listening to on AI in healthcare.
One other panelist echoed the necessity for transparency on the subject of AI in prior authorization. Nonetheless, he famous that AI and newer applied sciences even have the chance to enhance and velocity up the prior authorization course of.
“We speak about most cancers, it takes over 4 weeks to get in to see an oncologist or radiation oncologist right this moment, and I hate to suppose that a part of that delay is the results of individuals having to take care of prior authorization. So any lower within the latency interval of getting individuals handled is a vital factor. And I believe so long as you’ve received the transparency and you may perceive what these algorithms are doing, then I believe it’s probably an important enchancment general within the course of. I wouldn’t be afraid of it,” mentioned Dr. Troyen Brennan, adjunct professor of well being coverage and administration at Harvard T.H. Chan Faculty of Public Well being. Brennan can be a former government at CVS Caremark and Aetna.
Dr. Fumiko Chino, a radiation oncologist at Memorial Sloan Kettering Most cancers Middle, mentioned that she welcomes “our new pc overlords with some caveats.”
“We all know that datasets are very flawed and that for instance, marginalized populations are more likely to have undocumented stage or they might be lacking key parts from their EMR notes that might result in boundaries and subsequently could disproportionately face denials,” Chino mentioned. “You then’ve skilled a machine based mostly on a dataset that’s basically racist. I believe that’s in the end what we’ve to combat in opposition to.”
How will using AI in prior authorization have an effect on affected person belief? Schwamlein Howard famous that the typical affected person isn’t eager about this.
“They’re specializing in getting higher,” she mentioned.
Photograph: Piotrekswat, Getty Photos
