Hidden Cost of Healthcare AI: Why Advanced Prices Are Not Equal to Premium Results
I have had an undeniable impact throughout the healthcare industry: from nurse navigators using AI Triage Assistant to prioritize cases and respond faster, to tools that help providers record access to systems that automate duplication, resource-intensive management tasks. As use cases add up, so do investments in AI. But health care systems are starting to focus more carefully on whether they actually get the value of their own money.
Many organizations are paying high prices for AI tools that save minimal time without considering the true cost-effectiveness ratio. Many AI tools, which can only buy thousands of users every year, buy thousands of dollars a day, make the return on investment difficult to justify. Organizations are paying high prices for saving marginal time.
Organizations are fundamental disconnect between current AI tools, the limited value they offer and the more capable systems that emerge in the coming years. Organizations adopt AI because it sounds like something they should try, but the technology is still in its infancy. Today, many AI companies will not be able to survive because when renewal time comes, they will not be able to prove their value. As a result, organizations are hesitant and show signs of fatigue on some AI vendors.
According to Bain & Company and Klas Research’s 2024 Healthcare IT Spending Report, nearly half of healthcare providers see costs as the biggest pain point on the current technology stack. This means that AI tools with high-priced labels prove that limited ROI only add fuel to fires.
How did we get here?
Sometimes, AI tools work well in isolation during the pilot stage, but often have difficulties when squeezed out throughout the health system. This is especially true when these systems need to be integrated with complex workflows. In fact, some vendors claim that large customers are using their applications, when in reality, only one researcher can use their applications in one department. This means that these tools have not been proven and fine-tuned to work seamlessly on a large scale.
The best advice for today’s health systems is to ask the current supplier about their AI strategies before choosing a new company that offers only one solution. Mature companies that can integrate AI into existing workflows will outweigh the solutions that may not survive market corrections.
The same report found that regulatory and legal considerations were the biggest obstacles to implementing AI generation (38-43% of respondents). These complexities can create other barriers to slow adoption.
Meanwhile, many AI tools are built on public models like Chatgpt and then slightly customized and sold using healthcare brands. The apps may sound innovative, but the actual lift they provide is usually small.
Billions of dollars have been spent on healthcare AI so far, with billions of dollars coming. However, it is not clear that these investments actually translate into better care or meaningful time savings. “It's too early to generate AI in healthcare that can help, hurt or simply waste billions of dollars without improving people's lives,” the World Economic Forum said.
Four questions before next AI
While the promise of artificial intelligence is exciting, it's still early. Many of the tools that are available for the prime time are simply not ready. This uncertainty forces health system leaders to reevaluate their investments.
Then there is the pressure to manage multipoint solutions. Perhaps one tool handles documents, another tool can only handle billing, and patient follow-up requires another separate tool. The cost and complexity of these point solutions add up over time. Now, many CIOs are spending a lot of time evaluating new technologies.
This is one of the reasons why the health system is shifting its focus. Instead of buying from new vendors, they go back to their core platform and ask how AI can be integrated into the systems they already use. These solutions may not get the same spotlight as the latest startups, but they usually lead to a more reliable path with less disruption.
Dr. Daniel Yang of Kaiser Permanente is solving this problem in a thoughtful way. The organization applies system-wide governance to the entire AI efforts of research, clinical operations, education and management. He believes that AI should improve clinicians' judgment rather than changing clinicians. When Kaiser launches generated AI tools, it is supervised and designed on purpose.
For all these reasons, this is a good time for the health system to step back and ask some basic questions instead of rushing to the next AI purchase:
- What problems does this tool actually solve? Find tools that have specific operating bottlenecks with measurable results. Before any pilot, establish baseline metrics for the issues you want to solve. A good AI tool should improve efficiency and quality. If it only automates existing processes without improving patient outcomes or employee satisfaction, it may not be worth the investment.
- How much time and money does it actually save? Calculate the real minute cost savings, including implementation, training and ongoing support. If you spend more than $1,000 a year per year to save less than 15 minutes a day, the ROI may not justify the renewal fee. Tools that focus on eliminating the entire workflow steps, not just speeding up existing ones.
- Is this a pilot, or is it proved to be extended? Evidence for successful implementation is required in at least three different organizational sizes and settings. Find tools that can deliver consistent results across a wide patient population and always test in multiple clinical settings before deploying within the promised scope.
- Is this suitable for our existing system or do we add another layer to the already overloaded technology stack? Prioritize tools that directly integrate with EHR and reduce the number of systems employees need to navigate. Any AI solution that requires additional data entry, separate login or workflow interruption should be viewed suspiciously.
It's time for healthcare systems to take a more realistic approach to measuring the return on investment of AI investment. One that looks good in a demo that combines real value with limited use cases.
Photo: Phive2015, Getty Images
As CEO, Andy Flanagan is responsible for Iris Telehealth’s strategic direction, operational excellence and the company’s cultural success. With extensive experience in all aspects of the healthcare system in the U.S. and globally, Andy focuses on the success of patients and clinician Iris Telehealth, which can improve people's lives. Andy has worked at some of the largest global companies and has led several high-growth businesses, providing a unique perspective on behavioral health challenges in our world. Andy holds a Master of Health Informatics from Northwestern University’s Feinberg School of Medicine and a Bachelor of Science from the University of Nevada’s University of Reno. His previous experiences include three CEOs, including the establishment of SaaS and holding senior positions at Siemens Healthcare, SAP and Xerox.
This article passed Mixed Influencer program. Anyone can post a view on MedCity News' healthcare business and innovation through MedCity Remacence. Click here to learn how.
