Hospitals share lessons learned from AI ‘workers’
Hospitals have increasingly adopted artificial intelligence tools to improve the productivity of their back-office staff. That has brought benefits and some challenges, say hospital finance leaders.
Based off their own experience, hospital executives from three organizations identified ways that new adopters of AI help can get the most benefit and minimize problems. Their insights came during an Aug. 14 session of the HFMA Digital Annual Conference titled “Adapting to the New Normal: Accelerating Productivity and Revenue With AI.”
Key lessons to smooth the addition of AI tools included:
- Addressing employee fears of job losses
- Implementing AI tools broadly from the beginning
- Finding the funding first to speed implementation
- Avoiding overanalysis of its efficacy before spreading to other departments
- Obtaining wide organizational participation to eliminate implementation barriers
Susan Nelson, MedStar Health
Karen Testman, MemorialCare Health System
Employee fears addressed
Melanie Wilson, vice president of Revenue Cycle for Essentia Health, an integrated healthcare system with facilities in Minnesota, Wisconsin and North Dakota, said she made sure to address employees’ concerns that their jobs were threatened during her organization’s nine AI “implementations” within revenue cycle operations.
“You just cannot overcommunicate something like this,” Wilson said. “Our intent was to redeploy resources. We wanted people to be operating as humans, having the human element, using their critical thinking skills and a lot of the things were looking to automate were just repetitious work.”
Susan Nelson, executive vice president & CFO at MedStar Health, a health system in the Baltimore–Washington metropolitan area, said employees there also were concerned about job losses during the roll out of its 18 AI implementations.
The health system assured employees “that this was an opportunity for them to move into a different area, to understand a new aspect of the process, to expand their skill sets,” Nelson said. “They managed very carefully during this implementation time to create an environment where we had other positions and vacancies that they could explore; almost a job-fair-type approach for people that were going to be looking at a different part of the work. That really opened up people’s mindsets to ‘Hey, maybe this is a good thing for me, too.’”
Karen Testman, CFO at MemorialCare Health System, in Los Angeles, said leadership’s role was critical to address employee concerns and encouraging employee support for the adoption of AI.
“The fear is still there; it hasn’t gone away entirely,” Testman said.
To address that, human resources has committed to finding employees whose roles were affected by the introduction of AI with finding other positions within the organization.
“It’s one of those issues of recognizing it, acknowledging that it is a sensitivity that is there that has to be addressed and working then through that,” Testman said.
Sean Lane, CEO of Olive, which provides AI systems to healthcare organizations, urged hospitals to use it move employees from menial, repetitive tasks to those that take advantage of their reason and personality.
“The great irony of creating an AI workforce is that it is really an investment in your humans and it’s really believing in their potential,” Lane said.
An improvement in employee productivity may allow organizations ultimately to increase their personnel as that improved productivity allows the organization to take on new healthcare missions, he said.
Melanie Wilson, Essentia Health
Sean Lane, Olive
Other lessons learned
Other lessons Nelson learned from her organization’s adoption of AI systems included the need to consider, from the beginning, broad adoption of AI within a department. Limited implementations can curtail the financial and efficiency benefits from such tools.
“If we had looked at it more broadly, we probably could have automated more of it the first time around,” Nelson said. “If I had known that at the beginning, I could have factored that in.”
Finance leaders who are considering implementation of AI assistance need to prioritize early in the planning process how they will fund the launch of such systems, Testman said about her experience.
“We’re very cost-conscious as an organization; our resources tend to be really stretched,” Testman said. “And adding something like this to their plate tends to slow some things down.”
Wilson emphasized avoiding overanalysis of the results of small-scale implementations before wider expansions. because that would delay needed savings for organization.
“Don’t take too long to overanalyze, really get going and widely utilize it in multiple areas,” Wilson said.
To ensure AI is not limited to the revenue cycle, Wilson underscored the importance of broad participation in AI implementation from across the organization. Their implementation team included leaders from finance, personnel and IT.
“That initial engagement really helps prevent barriers and roadblocks downstream,” Wilson said.