In a recent interview with Andy Bartley, senior solutions architect in the Health and Life Sciences Group at Intel, we dove into the topic of predictive analytics and the role it plays in healthcare today and into the future.
During the interview, Andy framed the discussion well when he said that hospitals have made investments in EMR and other digital clinical applications and these data stores represent a rich opportunity for healthcare. Data is the precursor to predictive analytics and healthcare is primed to take advantage of this opportunity.
There are a wide variety of predictive analytics already seeing good results in healthcare. Some of those analytics include predicting hospital acquired conditions like sepsis and activating a rapid response team to be able to handle the clinical situation. Another operational example is around hospital staffing ratios and predicting staffing needs so your organization is not overstaffed or understaffed. Andy also suggested that revenue cycle management was seeing success with predictive analytics that recommends the best course of action when collecting payments.
Given these use cases, the benefits are obvious: improved revenue, improved outcomes from hospital acquired conditions, decreased staffing costs, avoiding lost revenue as patients are routed to other hospitals, and increased patient throughput to name a few.
I also asked Andy if these predictive analytics were only going to be available for the largest organizations or whether they would scale down to smaller hospitals as well. He gave a mixed response suggesting that some AI and precision health efforts are going to require large data sets that will require a smaller hospital to partner with the larger hospitals. However, Andy did say that many use cases will apply to hospitals of all sizes since many of them just rely on EHR data which the majority of hospitals have in place.
The reality is that predictive analytics are never “done” since they can always be improved by adding more data over time. Therefore, instead of focusing on perfection, a hospital should focus on if the predictive model’s result was good enough to show improvement over the status quo. Essentially, hospitals shouldn’t be afraid of the data they don’t have, but instead, they should focus on the data sets and models they do have which will show improvement over the status quo.
That just gives you a small flavor of the insights Andy shared in the full video interview you can watch below. Additionally, Andy dives into the barriers affecting predictive analytics including the need to get the technology, people, and process right. He also dives into how to push these analytics down to the point of care and ensure they are actionable analytics that benefit the provider who receives them. This is essential to building providers’ trust in analytics. Finally, Andy provides a look at what organizations can do to leverage predictive analytics over the next five to ten years.
Watch the full video interview with Andy Bartley to learn all the details about what’s happening with predictive analytics today and things you can do to prepare for it in the future:
Where are you at in your predictive analytics journey? In what ways has it generated value for your organization? What are you doing to prepare for it in the future?