I am a pediatric cardiologist and have cared for children with heart disease for the past three decades. In addition, I have an educational background in business and finance as well as healthcare administration and global health – I gained a Masters Degree in Public Health from UCLA and taught Global Health there after I completed the program.
“You cannot expect to get anything useful by asking wizards to spangle machine learning magic on your business without some effort from you first.”
Cassie Kozyrkov, Head of Decision Intelligence
The following three aspects of executing an AI strategy in a healthcare organization need not be in sequence but can be pursued in parallel depending on level of commitment and resources that are available:
A small coalition (5-10 people) of the “AI” willing can be formed and maintained (“The Who”). Instead of a fragmented and decentralized overall approach to AI projects that may even be competing with one another, a cohesive and centralized structure of diverse stakeholders and expertise with oversight is preferred to encourage synergies amongst projects and esprit de corps between sectors in the healthcare organization. For instance, a coalition of AI champions can guide both efforts in the clinical realm (machine learning in risk profile for readmissions) as well as for the administrative suite (robotic process automation for revenue cycle management). An AI center or unit can also be the epicenter of an organizational focus in AI with a hub and spoke model for projects and accountability for progress.
An AI focus should be part of the overall long term vision for the organization (“The Why”). This AI movement within the organization should be nurtured from the top executive down to the associates as a meaningful transformation for the entire organization over a period of 10-20 years (think journey as opposed to short term focus). While John Kotter’s change management tactics are useful for this transformation, it is important to maintain a pace and direction that will be sustained in the midst of likely disappointments. Too often AI is used for tactical purposes (unsupervised learning in marketing or natural language processing in sentiment analysis for retailers) rather than a genuine long term strategy.
Think of AI as a new resource to solve existing old problems rather than a technology looking for a problem to solve (“The What”). One strategy for an organizational adoption of AI is to use a design thinking (”problem-first”) approach: seek out colleagues with ongoing issues or problems and offer the appropriate AI tool as a possible solution. The best solution is often, however, a combined AI and human cognition approach. This AI resource can also be in the form of a consult or “AI-as-a-service” that is coupled with AI education of various sizes to suit the individual needs of the group. While some or most prefer a relatively short introduction, a few may opt for a 2-5 day in-depth AI course and become an AI champion.
We at AIMed believe in changing healthcare one connection at a time. If you are interested in discussing the contents of this article or connecting, please drop me a line – [email protected]
