AI Powered Behavior Change in Healthcare is Not Just for Patients

Most of the stories we read about changing behavior in healthcare are about patients. Not this one. In this article, we are focusing on how Artificial Intelligence (AI) coupled with a better approach to training can lead to measurable improvement for clinicians.

Why is it so hard to change behavior?

Megan Call, Associate Director of U of U Health’s Resiliency Center, answered this question succinctly in her 2020 article: Behavior change is complicated and complex because it requires a person to disrupt a current habit while simultaneously fostering a new, possibly unfamiliar, set of actions. This process takes time—usually longer than we prefer.

In a study published in 2017 by the University College London, researchers found that people are biased towards perceiving anything challenging to be less appealing. In other words, people are likely to stick with the status quo rather than change because change requires more effort which we have a natural bias against.

To explore the topic of behavior change in healthcare and the role technology could play in accelerating it, Healthcare IT Today sat down with Dr. Andres Jimenez, Chief Medical Information Officer at CorroHealth. For over 10 years, Dr. Jimenez has studied and researched behavior change.

Traditional approach to behavior change

“We’ve had one model for behavior change in healthcare that has worked for a long time,” explained Dr. Jimenez. “It’s the apprentice model – see one, do one, teach one. We have used it for decades. The knee jerk reaction to behavior change in healthcare is to just provide one-size-fits-all training.”

If you think about it, the advances we have made with technology have only improved the delivery of training. We have moved from classrooms to CD-ROMs to on-demand videos with self-tests, to live-streams, but we are still teaching new behaviors using a paradigm that is centuries old.

“We have to step away from the old paradigm,” said Dr. Jimenez. “We need to move past digital lectures. With the advances we have made in technology, we can literally wipe the slate clean and build something that will actually change behavior.”

Improving starts with measuring

Dr. Jimenez believes that the key to improving training and by extension, behavior change, is better measurement. Traditionally, training has been measured by things like: number of butts-in-seats, hours delivered, test scores and CE credits earned. Unfortunately, these only measure activity and not outcomes.

Only by tracking outcomes can we truly gauge the effectiveness of training and the depth of behavior change achieved. Lucky for us in healthcare, there are now standardized metrics that we can use for this purpose: CMS quality and patient experience metrics.

“These measures can be used to gauge the impact of training and behavior change initiatives,” stated Dr. Jimenez. “After we train, do we see a corresponding improvement in the quality outcome metrics? If we do then we know it has been successful. If we don’t, then we know we have to go back to the drawing board.”

Measuring the change in outcomes at an organizational level can help improve behavior change initiatives at the macro level. Measuring the change at an individual level can help improve at the micro level.

How can this be done? The answer according to Dr. Jimenez is technology: “Technology can help you clearly see if the desired outcome has been achieved. Whether that’s the improvement in quality patient outcomes or improvement in patient experience. It can be quantitatively measured in near real-time now. This information can help you pinpoint where improvements need to be made and who needs help to improve.”

The surprising role technology can play

One of the most exciting ways that technology can help change behaviors is by improving the relevance and personalization of training for individuals. Instead of a one-size fits all approach, imagine if training could be customized to the unique needs of each clinician or administrator in a healthcare organization.

After all, everyone learns differently. Some are visual learners. Some learn by doing. Some need constant positive encouragement. Some need almost no instruction at all.

Dr. Jimenez envisioned a future where a physician’s practice patterns and learning preferences could be factored into the creation of a hyper-personalized training program. In addition, data from the physician’s EHR and other systems could be used to measure the effectiveness of the training program – thus creating a closed loop.

For example, say you wanted to change the way a physician documented encounters in order to reduce the number of rejected claims or the amount of back-and-forth with the coding team. A “smart-system” would know how this physician likes to learn (visually through repetition), it would also recognize that the physician has a full patient schedule each day. Based on this, the system would come up with a unique training program – maybe one that involved 10min videos that showed specific examples of better documentation that the physician could watch during lunch, followed by guided tutorials that would guide them through practice with different fictitious patient encounters.

To top it off, the system could even trigger an alert to a system expert who could spend an hour with the physician each day for a week to answer any questions they have.

If this sounds like Artificial Intelligence (AI), you would be right. Dr. Jiminez’s vision is all about applying AI to understand the patterns of individual physicians and using what it learns to build those personalized training programs. AI even could be used to predict where errors may occur in the future, based on observed behavior and metrics drawn from existing electronic systems.

A personalized approach

CorroHealth’s Clarity product is built on this model of behavior change – data-based, AI-powered behavior change using personalized training. They have applied this specifically in the area of coding and RCM improvement.

Their system analyzes a physician’s coding pattern and if it detects outlier behavior, it will send that physician an email that let’s them know that they are experiencing more errors than their peers and providing examples.

“This is really powerful,” explained Dr. Jimenez. “The first question a physician will ask is ‘what patients are you talking about?’, followed by ‘what did I do wrong?’ Having specific examples really helps.”

Based on what is detected, the system would recommend a personalized training program to address the issue. That program would be considerate of the physician’s workload, learning preferences and past behavior. The system then monitors the EHR and HIM systems to determine if the program is successful and adjust as needed.

I have to admit that this all sounds fantastical, but the technology exists. More importantly, this hyper-personalized approach to training isn’t new. It has been discussed and documented for many years. Dr. Jimenez and the CorroHealth team have just adapted it successfully in this area of healthcare.

What strikes me most is how much sense this approach makes. If we want physicians to change behaviors that have been ingrained for years, we can’t expect that a few hours of lectures is going to be effective. Just think of the training physicians go through. They don’t become fully licensed immediately following medical school – they have to go through several years of residency on top of their years of lectures.

It’s time we retire the lecture-only approach to behavior change in healthcare. It’s time for something new.

Watch the full interview with Dr. Jimenez to learn more about:

  • How early automobile designs mirror the current way we teach new behaviors (or try to)
  • How AI coupled with CDI and Rev Cycle data could predict areas of improvement
  • Practical steps that a healthcare organization can take to change behavior

For more information about CorroHealth, visit https://corrohealth.com/

CorroHealth is a supporter of Healthcare IT Today

Listen and subscribe to the Healthcare IT Today Interviews Podcast to hear all the latest insights from experts in healthcare IT.

And for an exclusive look at our top storiessubscribe to our newsletter.

Tell us what you think. Contact us here or on Twitter at @hcitoday.

About the author

Colin Hung

Colin Hung is the co-founder of the #hcldr (healthcare leadership) tweetchat one of the most popular and active healthcare social media communities on Twitter. Colin speaks, tweets and blogs regularly about healthcare, technology, marketing and leadership. He is currently an independent marketing consultant working with leading healthIT companies. Colin is a member of #TheWalkingGallery. His Twitter handle is: @Colin_Hung.

   

Categories