You’ll never meet anyone that loves health data science more than Prashant Natarajan. He literally wrote the book on the subject (Check out Demystifying Big Data and Machine Learning for Healthcare to see why I mean literally). He recently gave a presentation on the 4 P’s of Innovation in Health Science which included this slide:
Sadly, I couldn’t find a recording of his presentation. However, this slide puts health data science in perspective. Prashant boiled it down to 4 simple points. The problem is that too many healthcare organizations are unable to really execute all 4 P’s in their health science innovation efforts.
No doubt each of these 4 P’s is challenging, but the most challenging one I see today is the first P: People.
I’m not sure all of the ways that Prashant addresses the people problem, but it’s somewhat ironic that people is the biggest problem with health science innovation. I see the challenge as two fold. First, finding people who have the health science mindset are hard to find. Competition for people with these skills is fierce and many of them don’t want to get into healthcare which is complex, regulated, and often behind.
The second major health science challenge revolves around the people who collect, aggregate, and enter the data. It’s easy for a front line person to not care about the downstream effects of them entering poor quality data. Not to mention being consistent in what you enter and how you enter it.
It’s somewhat apart of human nature for us to jimmy rig a solution to the problem we face. Those workaround solutions wreaked havoc downstream in your data science efforts. I recently heard the example of a hospital always choosing Mongolian for some setting because it was a setting that would never be used otherwise. The culture of the hospital just knew this is what to do. Once the data scientists started looking at the data they wondered why this Mongolian population kept coming up in their results. Every healthcare organization has their “mongolian” workaround that causes havoc on data science.
What do you think of these 4 Ps of Innovation in Health Science? Is there something missing? Do you see one of these as more important than another?