How are payers, providers, and pharma extracting value from GenAI?

Generative AI (GenAI) has become an overnight sensation in the healthcare industry, capturing the fascination (and investment dollars) of leaders across the payer, provider, and pharmaceutical sectors. 

Don’t let healthcare’s harrowing history with the health IT hype cycle fool you, however: this appears to be the real thing. Or at least it could be, if organizational leaders can learn from past mistakes and pay close attention to where, when, why, how, and how fast they adopt this new generation of AI-powered tools. 

There does seem to be more attention focused on these factors than in previous cycles of digital innovation. Every day, there’s a new survey or industry report designed to increase transparency and provide some level-setting to help leaders make more informed decisions about bringing GenAI into the fold – something that seemed, anecdotally, to be much slower, less comprehensive, and less frequent in the regulatory-driven phases of EHR adoption during the meaningful use era.  

The latest data paints a generally positive picture of the burgeoning GenAI landscape.  McKinsey & Company recently stated that 85% of organizations were exploring or have already adopted GenAI tools in some capacity. Google Cloud revealed that three-quarters of organizations using GenAI are seeing quick return on investment. And a Panda Health report found that organizations are making an effort to develop quantifiable metrics to gauge the success of AI implementations earlier in the process, and are simultaneously increasing their scrutiny of vendor claims to make sure they’re investing wisely.   

So how exactly are organizations helping to make GenAI a success without falling into the pitfalls of the past? Yet another industry report, this time from Bessemer Venture Partners, provides some intriguing clues. 

It’s not a fight to secure executive buy-in

The C-suite is all-in on GenAI. The survey of payer, provider, and pharma stakeholders shows that C-suite members are controlling about 70% of the AI decisions, particularly around financial use cases, and these executives are largely willing to fund their favored initiatives with money from the organization’s centralized budget instead of leaving individual departments to pay for experimentation. 

With the green light to spend big bucks on GenAI tools, and the freedom to choose projects without being subject to regulatory requirements, organizations can actively work to get strategic about leveraging GenAI to secure a competitive edge in the market without first spending their precious resources convincing their leaders that it’s worthwhile. 

A consensus is forming around the highest-value use cases

GenAI seems like it’s in everything already, but organizations are actually trying to take a beat and think about what jobs can really benefit from an AI lift.   

For payers, these potential use cases include: 

• Claims-related tasks: Benefits verification; approvals and denials; coordination of benefits, and appeals management 
• Network management: Provider contracting, credentialling, and enrollment; provider quality/outcomes management; prior authorizations and concurrent utilization monitoring 
• Member services: Member enrollment and engagement; care gap analysis and care navigation; call center operations; social determinants of health (SDOH) tracking 

 In the pharma sector, organizations are looking closely at: 

• Preclinical tasks: Identifying promising molecules; indication selection; biomedical literature review and mining 
• Clinical tasks: Optimizing clinical trial recruitment; protocol design; regulatory intelligence; pharmacovigilance monitoring 
• Marketing and sales: Patient services; provider engagement; competitive intelligence; pricing, budgeting, and forecasting; medical affairs team enablement 

And for providers, GenAI has strong applications on both sides of the clinical/financial divide:  

• Patient care optimization: Patient and staff scheduling; care gap identification; documentation support; calculating and reporting quality metrics; risk adjustment; referral management 
• Revenue cycle management: Payer enrollment, contracting, and denials management; prior authorizations; medical coding and clinical documentation improvement 

There’s still more work to be done, however, as only half of surveyed organizations said they have a formal GenAI roadmap in place and only 57% have an AI governance committee to ensure safe, ethical, and uniform deployment of AI-powered tools. 

Plentiful pilots might raise the odds of hitting the jackpot  

With so many options for use cases, there’s a fine balance between adopting a full-on “spray-and-pray” attitude toward launching pilot programs across these areas and intentionally spreading one’s bets to cover the most ground in the hope of parallel processing one’s successes. Right now, organizations are definitely walking that line with dozens of pilots running simultaneously across multiple areas of the enterprise.  

Nearly half (45%) of projects remain in the ideation or testing stages, the report says, with payers and pharma surprisingly lagging a little bit behind the provider community in their progress. This could be because a large number of providers have access to “easy win” technologies, like ambient scribes and GenAI documentation assistance rolled out through Epic, which they haven’t had to fully develop themselves.  

For example, the survey shows that 30% of providers already have system-wide deployments of ambient scribes, while another 22% are in implementation and a further 40% are actively piloting these solutions. 

With this high-priority and relatively well-tested use case in hand, providers can spend more of their funding and attention on other projects, and progressively narrow their efforts on those that show the most promise until they fully zero in on the ROI they’re looking for.  

Commitment to powering through challenges and roadblocks

There are, of course, some familiar barriers to contend with. More than half of respondents said security concerns are giving them pause, with payers among the most likely to be worried about this issue. Payers were, however, the least likely to say they don’t have enough in-house AI expertise to make things happen. Just 41% of payers said this might be a challenge, compared to 48% of providers and 52% of pharma companies. 

Other big issues that could slow down the GenAI adoption cycle are the costs of integration with other members of the care continuum, especially with payers, and broader challenges with preparing AI-ready data, which seemed to be the most problematic for pharma (47%) and less of a focus for providers (41%) and payers (39%). 

But with the majority of respondents saying money was no object when it comes to pursuing GenAI goals, it’s likely that organizations will find a way to work through these speedbumps and continue to pursue AI enablement across their enterprises. 

Combined with strong executive interest in getting things done with GenAI, as well as a relatively large number of pilots and proof of concept projects in progress, it’s likely that GenAI will be increasingly deeply integrated into the payer, provider, and pharma environments at a healthy clip.  

Success will depend on how quickly and thoroughly these organizations can extract meaningful insights from the experimentation phase and iterate on those learnings to prioritize the use cases that make most sense for staying competitive in their markets and achieving ROI. 

Jennifer Bresnick is a journalist and freelance content creator with a decade of experience in the health IT industry.  Her work has focused on leveraging innovative technology tools to create value, improve health equity, and achieve the promises of the learning health system.  She can be reached at [email protected].