Use of AI in Healthcare Picking Up Momentum, Report Shows

Ken Terry

October 10, 2019

Interest in artificial intelligence (AI) is soaring in the healthcare industry, according to a recent Optum survey of 500 US healthcare executives from hospitals, health plans, life sciences organizations, and pharmaceutical and device companies.

The survey found an 88% increase this year in the number of healthcare leaders who said their organizations are implementing an AI strategy compared to 2018. Sixty-two percent of the respondents said they had implemented such a strategy, up from 33% last year.

The average organization expected to invest $39.7 million in AI over the next 5 years ― $7.3 million more than was estimated last year, Optum found.

Many respondents also anticipated a return on investment (ROI) in the near term. Fifty percent of the healthcare executives said they expect to see ROI on their AI applications within 3 years, compared to 31% in 2018. That includes 55% of hospitals and 52% of health insurers.

The healthcare leaders showed a clear preference for administrative over clinical AI applications. Of the top-ranked applications in the survey, 4 of 5 were administrative, and 50% of the respondents said their organizations will invest first in automating business processes, such as administrative tasks or customer service.

More than a third of the organizations (36%) will invest in personalizing clinical care recommendations, such as drug therapies, and the same percentage also will invest in accelerating research for new therapeutic or clinical discoveries, according to the report.

Many of the healthcare executives said they hoped to use AI to solve operational challenges, such as automating prior authorizations (51%), managing electronic health records (45%), and selecting appropriate care settings (38%).

About half of the respondents expected AI to create more work opportunities. Most of the executives agreed that hiring people with experience in AI technology is a priority for their organization.

There was also broad agreement that up to 50% of new roles for employees will require experience working with AI and that their employees are not being trained fast enough in this area.

Two Different Views of AI

Most news reports about AI and machine learning focus on how the new technology is starting to improve clinical care. For example, the Ochsner Health System recently announced it is using AI to predict patient deterioration outside the intensive care unit with 90% accuracy, according to the journal Health Data Management.

The same journal reported that Duke University Hospital has introduced a machine learning program that can identify patients in the early stages of sepsis so that rapid response teams can intervene. And a Frost & Sullivan report predicted that 35% to 40% of hospital operating rooms in the United States will be integrated with AI and virtual reality applications by 2022.

The Optum survey, in contrast, indicates that most of the AI action is on the administrative side, where healthcare organizations can quickly achieve ROI.

Laura Marston, a principal with ECG Management Consultants, told Medscape Medical News that AI investments can be recouped much more rapidly than the 3-year time horizon in the Optum survey.

"You may get ROI from using AI in claims follow-up, eligibility verification, or preauthorizations," she said. "Those are very repetitive processes, and they're very tied to cash in the door and your bottom line. If you can eliminate errors in those three processes, I've seen organizations get ROI much sooner."

Will this mean layoffs? Marston doesn't think so. Healthcare organizations will need fewer people to do repetitive work in revenue cycle management, she said, and in some cases they may not replace employees when they leave.

"But I find that more often than not, you can refocus your workforce on tasks that require higher-level thinking," Marston explained.

She also believes that organizations will increasingly adopt AI applications that can improve patient care, reduce length of stay, and eliminate the need for patients to receive more costly interventions. "Everybody cares about quality outcomes and patient experience," she noted.

Physicians will also benefit from the use of AI applications, Marston said. "If I can predict that a patient will have a certain outcome or will need a certain intervention, if I can use historical clinical data to help address something sooner, I think that's where it will have the impact."

As AI spreads from big healthcare systems to smaller organizations, it will also be used to automate revenue cycle management in physician groups, she added. "AI has not just a clinical but a business impact as well."

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: