Health plans say AI is pushing healthcare costs higher
By: Emily Olsen (Healthcare Dive)
Artificial intelligence is making things more expensive? Who didn't see this coming?
S E E D E D C O N T E N T
Dive Brief:
- Health plans are projecting commercial healthcare costs will rise 9% next year, driven in part by increased adoption of artificial intelligence billing tools by providers, according to a report released Thursday by professional services firm PwC.
- Nearly 70% of surveyed plans ranked providers' use of AI documentation and coding products as a top three inflator next year, while about 20% called AI the number one inflationary trend.
- Still, AI isn't a major driver of growing healthcare costs compared with labor and supply cost inflation or increased healthcare utilization, said Glenn Hunzinger, U.S. health industries leader at PwC. "The ability to use technology and AI to more appropriately code or code things that they were never able to, that's the trend we're seeing," he said. "It does have an impact on that 9%, albeit it's not the biggest piece."
Dive Insight:
The projected medical cost trend is the highest in nearly two decades, according to the PwC report, which surveyed and interviewed actuaries at 27 health plans.
The analysis identified five inflators increasing the cost of care next year, including the heightened use of AI — one of the healthcare sector's most anticipated and hyped technologies.
Providers have been rapidly adopting a range of AI products, including tools that can record and draft clinical notes and suggest codes for billing. More complete and detailed documentation means clinicians can record more diagnoses and comorbidities during visits, code for more complex care and receive higher reimbursement — even if the clinical work or treatment doesn't change, according to the analysis.
That doesn't mean the more intense coding is inappropriate. Many health systems were probably missing correct codes due to the volume of care and complexity of their own internal systems, Hunzinger said.
Plus, margins are thin for many hospitals, and they're likely facing increased reimbursement pressure amid federal healthcare spending cuts, particularly to the safety-net insurance program Medicaid. That could motivate them to adopt AI billing tools to stay afloat, Hunzinger said.
And there are other trends pushing increased healthcare spending. One major driver is elevated costs for providers, which continue to manage increased labor and supply expenses in the wake of the COVID-19 pandemic.
Consolidation is also fueling that trend, as providers merge and gain increased negotiating leverage over payers, according to the report. Nearly 65% of survey respondents said contracting pressure from providers was a top three inflator next year.
Another inflationary trend is the Independent Dispute Resolution process set up by the No Surprises Act, a law that went into effect in 2022 to protect patients from unexpected out-of-network medical bills, according to the report. Providers are winning the vast majority of the disputes and earning more than they typically would for providing care.
Additionally, rising pharmacy spending — like for expensive GLP-1 medications — and increased demand for behavioral healthcare could drive rising costs.
And trends that could slow healthcare spending — like biosimilars and generic drugs or pushing care towards cheaper facilities — are already baked into the current trajectory, so they likely won't drive further improvement, according to the report.
Still, technology like AI could eventually start to bend the cost curve, Hunzinger said. For example, administrative work accounts for a large chunk of healthcare spending, so automating some of those tasks could make a dent. Plus, AI tools could lessen burnout and allow in-demand providers to focus on clinical tasks.
But innovation takes time to bring down costs, especially in a highly regulated industry like healthcare.
"Its a tough environment to operate in, in a flawless fashion," he said. "Because at the end of the day, it's patients, it's people. So that's why it probably takes a little bit longer."
Be a good human. Grossly derogatory comments, blatantly inflammatory comments, excessive trolling, mean-spirited taunting and insults, scamming, spamming, and off topic comments may be removed in compliance with the site's Terms of Service and the Code of Conduct. The moderators have the final word.
Tags
Who is online
117 visitors
There is a reason rich people are excited about AI. The other 99% of us will be paying all the costs.
AI has the potential to be a 'common good' that would benefit humankind. Of course, that's not going to happen because the rich grifters have hijacked the 'common good' to make themselves richer.
I cannot tell from the article how AI is increasing healthcare costs.
My best guess is that they are looking at the costs of process and infrastructure changes to incorporate more AI. That would be an upfront cost to reduce costs over time. But again, I cannot tell from what is written here.
Yes, this is what I would expect.
Except it will not reduce consumer costs over time, it will reduce corporate costs thus increasing corporate profits.
I agree that if the healthcare companies are indeed passing upfront costs onto consumers, they are unlikely to lower costs over time once the cost savings kick in.
Healthcare, oil, every large corporation you can think of. Once prices go up, those prices never go back down (at least not all the way) but corporate profits always go up.
That does tend to be the case.
Health providers, apparently, are using AI to maximize billable coding for the medical care that has been delivered. AI is allowing billing for any request for information or complaint mentioned by a patient. A patient asking if a mole looks cancerous can now be a billable diagnosis. And AI can select the coding to maximize payment for that diagnosis.
AI is proving to be expensive labor. Using a PhD level intelligence for data entry tasks doesn't seem cost effective, does it? Reducing the number of workers won't necessarily lower the 'systemic' costs that will be passed on to consumers.
Are you suggesting that AI is making the rules?
Human beings misusing technology! And I take issue with viewing current AI as Ph.D. Level intelligence. Better to view it as Ph.D. level assimilation of knowledge based on statistical norms.
Where did that come from? AI used by health providers are not making the rules issued by insurers. That's a silly conjecture.
AI has allowed maximizing charges under insurer rules. In lay terms, AI has streamlined padding the bill.
The healthcare costs being cited are not about using assimilated knowledge to deliver medical care. AI assuming administrative tasks, such as billing, is not about delivery of medical care. The seeded article focuses on the aspect of increased costs experienced by insurers. But AI is not limited to billing insurance companies. AI performing the administrative task of billing will increase costs for medical care not covered by insurance, too.
AI deployed for the purpose maximizing billable charges is not a 'common good'.
From this: "Health providers, apparently, are using AI to maximize billable coding for the medical care that has been delivered. AI is allowing billing for any request for information or complaint mentioned by a patient."
The AI does not make the rules, but this is worded to make it look as though it does.
AI did not make the decisions.
Again, AI is a mechanism. It does not make the rules. Human beings still call the shots. AI performing the mechanical operations should be a cost-saving due to increased efficiency.
Okay.
Are you sure? The article suggests AI is interpreting a provider's session records and is making decisions about billable charges.
AI would be an autonomous mechanism. AI is capable of making rules if it is tasked to make rules. AI would be capable of creating workplace rules for medical providers that would maximize charges for the least expenditure of resources. What the article describes extends beyond automated efficiency. The AI deployed by medical providers has been incorporated into the decision making process for billing.
Yes. AI is a tool that is created by human beings, purchased by human beings, and deployed by human beings. It does not call the shots. Any decision-making authority it exercises has been delegated by human beings and organizations and they remain accountable. If human beings deploying a software product (such as AI) accept the rules that are embodied by said product, that is their choice.
AI is still a (very impressive, awesome even) machine. It does not reason (yet) like a human being; it can follow patterns for solving problems and drawing conclusions but does not have the experience, judgment, moral agency, etc. of a human being. It follows examples of human reason and human-derived knowledge and uses it to produce impressive automated responses and (with agents) actions.
AI is a powerful tool, not an independent authority.
So AI is going to take everyone’s job because people are more expensive than computers - unless it’s about healthcare, where every single aspect of AI related tangents that you can dream up will be pretzled into a reason why insurance will need to get more expensive. We are heading to a future where extremely wealthy capitalists will only be able to cannibalize other extremely wealthy capitalists, because everyone else will be reduced to such poverty that they have nothing left to steal from. This (and the birth of the country’s first trillionaire) is the expected end result of abject capitalism.