GenAI is already transforming the healthcare industry

Generative AI (genAI) is quickly transforming healthcare, according to a survey by consultancy McKinsey & Co. released today. The survey of 150 healthcare industry organizations found 85% are exploring or had already adopted genAI.

The organizations, which included insurers, health systems, and health services and technology (HST) groups, found that 40% or more from each group had already implemented genAI. Not surprisingly, HST organizations have the highest rate of genAI implementations at 57%. The least? Healthcare providers with a 40% implementation rate.

GenAI has already been proven in studies using historical patient data to be more accurate at diagnosing illnesses and other conditions. For example, OpenAI’s GPT-4 Turbo model not only outperformed physicians, but also outdid every single AI system developed for healthcare over the last 50 years, according to the lead doctor in one study at Beth Israel Medical Center in Boston. And it did so without any medical training.

Healthcare stakeholders are also using genAI to explore ways to create value and reduce costs, despite ongoing challenges such as evolving regulations and capability gaps.

Of the respondents that have already implemented genAI tools, 64% reported that they anticipated or had already quantified positive ROI, “suggesting high expectations for genAI technology,” McKinsey said.

Many organizations are also forming partnerships to access external talent while customizing AI solutions, and partnerships with hyperscalers could ensure successful implementations, McKinsey said.

“According to our survey, risk concerns was the top cited barrier to implementing Gen AI among respondents. Within risk concerns, survey respondents ranked security, inaccuracies and biases, and regulatory compliance as their top concerns,” said Drew Ungerman, a senior partner with McKinsey.

Healthcare is still slightly behind most other industries when it comes to Gen AI adoption, though the gap could be closing, Ungerman continued.

In another recent McKinsey survey, 63% of healthcare respondents stated their organizations were using AI in at least one function, compared to 68% of respondents in consumer goods and retail. Compared to financial services organizations, however, the difference was far less — only 2% higher use in financial services, he said.

Early genAI use cases have focused on improving administrative efficiency, dealing with IT gaps, and boosting clinical productivity. As capabilities grow, other uses could expand to patient engagement and quality-of-care improvements. Leaders recognize the importance of AI risk management and governance for safe implementation.

Organizations that have developed their genAI capabilities are seeing success with large-scale implementations, b ut future success will depend on a value-driven strategy, strong execution, and effective management, McKinsey said.

Most respondents reported their organizations had implemented or were developing genAI use cases, with more in the implementation phase than the proof-of-concept stage. However, 15% had not started proof-of-concept work.

Other prominent findings include:

Nearly half (47%) of payers, health systems and HST groups have already rolled out genAI tools. 

About three-quarters of healthcare leaders believe the technology could help most with administrative efficiencies and clinical productivity — and more than half (55%) believe it has value for patient/member engagement and IT/infrastructure. 

Sixty-one percent of payers and providers expect to collaborate with partners on building out their genAI capabilities; 20% said they’d build their own; 19% said they’d buy a program. 

GenAI may create “tremendous value” in areas that could fundamentally improve patient experience and streamline operations to generate cost savings, the report said.