Recently, I found myself having a conversation with colleagues about AI in health care (AIH). We contemplated the inevitable involvement of regulatory bodies, like The Joint Commission (TJC) or Centers for Medicare & Medicaid Services (CMS), in guiding the utilization of AI in health care.
Low and behold, just a few days later, the TJC posted an article sharing insights from October’s RAISE conference, indicating a growing focus on how, when and why to implement AI, and the critical perspectives health systems should consider.
If you visit The Joint Commission’s website you’ll be greeted by a statement, “We Believe in Sustainable Healthcare.” This prompts a pertinent question: in a healthcare landscape challenged by staff and provider shortages and escalating costs, how do we define and achieve sustainable health care? From my perspective, sustainable health care, especially in the context of AI, revolves around longevity and scalability to enhance care delivery.
The publication,“To do no harm – and the most good – with AI in health care”, gave insight into a very promising avenue of sustainable health care1. The authors shared in depth commentary on who benefits and demonstrated a path forward by using AIH.
Some key points:
- Augment clinical practice with AI
- Establish guidelines for AI optimal use and education on how to use it
- Define clear outcome expectations that show AI is achieving the intended use
Let’s explore each of these in more detail.
Augment clinical practice with AI
While there are many ways you can augment clinical practice with AI, TJC emphasizes, “…prioritizing and analysis of test and imaging results…”
Earlier this month I discussed the national radiologists shortage that is only going to increase. Using AI to help radiologists, and care teams, by analyzing and prioritizing suspected abnormal findings contributes to faster access to care leading to better patient outcomes.
Using AI to augment clinical care team workflows creates efficiencies that improve productivity and have the downstream effect of lowering delivery costs. Clinical workflows augmented with AI are likely to contribute to sustainable health care initiatives as their functionality is embraced by providers and care teams alike.
Establish guidelines for AI optimal use and education on how to use it
Change management is an essential component of AI adoption and a high-quality AIH vendor should have programs, teams and tools to help healthcare organizations optimize their use of AI technology, such as Aidoc’s Clinical Workflow Optimization and Training program.
We know that “intuitive functionality does not equal intuitive awareness or adoption.” Therefore, education is key! Healthcare organizations should not be responsible for developing optimization and training programs for new AI technology – this is NOT sustainable. What healthcare organizations can, and should, do is establish their guidelines around optimal use and training requirements for their teams.
Define clear outcome expectations that show AI is achieving the intended use
Adopting AIH is a long-term commitment that requires a well-thought-out change management plan.
When integrating new technology into a clinical workflow, it’s imperative to identify what are the key metrics your team can measure to identify success with the AI. Metrics that show improved health outcomes, faster access to care, improved clinical workflows and increased productivity are among the ways to measure how AI is achieving sustainable healthcare goals.
Ultimately, the Goldberg publication stressed that adoption of AIH in avenues where it can address critical operational and care challenges is a matter of urgency. This point of view is interesting, especially when we continue to hear so many voices encouraging an AI slow down, not just in healthcare, but in many industries. Those at the RAISE conference understand there are immediate operational, financial and, above all else, patient care benefits with the right AI.
Lastly, a point that was not directly addressed within the overview from RAISE, but is becoming more obvious in a rapidly evolving AIH arena, is that sustainable AIH can not be achieved with point solutions or AIH marketplace vendors.
Why? The cost of acquiring, implementing and maintaining AI solutions that address limited use cases can become a resource drain on healthcare organizations and threaten sustainable healthcare initiatives.
AIH must be achieved through sustainable partnerships that address the holistic needs of the organization. With many different care teams and disease states treated, a platform is the only integration method that allows health systems to reliably deploy, measure and run AI at scale. The true promise of AIH.
- Goldberg, C. B., Adams, L., Blumenthal, D., Brennan, P. F., Brown, N. D., Butte, A. J., Cheatham, M., deBronkart, D., Dixon, J., Drazen, J. M., Evans, B. J., Hoffman, S. M., Holmes, C., Lee, P., Manrai, A. K., Omenn, G. S., Perlin, J. B., Ramoni, R., Sapiro, G., . . . Zhao, J. (2024). To do no harm — and the most good — with AI in health care. Nature Medicine. https://doi.org/10.1038/s41591-024-02853-7
