This may sound crazy, but it seems that science is now catching up and statistically indicating that there really may be differences between clinicians and clinics.
Let me take a step back real quick. Whenever the outcome of your service is predicted, you want your data risk adjusted. Whenever your services are compared, you want risk adjusted data used for comparison purposes. I mean, our patients aren't one big family of twins are they?
Last week I had a little lesson in statistics. Both Mark Werneke and Linda Resnik were my "tutors" so to speak. Although the tweet that snagged my interest seems from long ago, I couldn't understand one small twist in this study well enough to convey the impact it could have on the rehabilitation industry.
If you aren't familiar with tweets, here's the recent publication in Health Sciences Research that perked my interest. What grabbed my attention? "Clinic and therapist effects explained 11.6 percent of variation in FS. Clinics ranked in the lowest quartile had significantly different outcomes than those in the highest quartile (p < .01)."
I was lucky enough to get a crash course on the statistics used in this study. You know what, I still don't perfectly understand what Mark and Linda were exactly conveying. For some reason, Pedro, Linda and Benjamin decided to statistically account for patients who did not have complete data sets. At the same time, the researchers used hierarchical 3-level linear regression modeling which mathematically indicated "clustering" happened with their sample. Clustering meant that patients may not have randomly chosen a clinic... patients may not have been randomly assigned a clinician.
Currently the FOTO risk adjustment model accounts for 35% variance. (Although 35% doesn't seem like a very large percent, it's pretty high and appears to have room for improvement based on this current study.) The current risk adjustment model is totally focused on the patient and patient factors. What excites me is this new study brings the impact of our interactions into the picture and alludes to the value our interactions can bring to the results of our care. In other words, we matter too!
I am jumping the gun here because I am making a huge assumption based on quantitative data. The current study does not convey why clinic and therapist effects explained 11.6% of the variation. The exciting aspect, as we learn more and learn in what ways our interactions affect our patients, we can definitely alter our interactions to have the most positive effect possible.
I believe this is the first study in the rehabilitation world that statistically captured the effect we may have on the outcomes of our patients. Definitely correct me if I've missed a previously published study!
Until next time,