In our last blog we covered the concept of the Minimum Clinically Important Difference (MCID) as a way to use PROs in the clinical setting. The idea of using MCID as a clear metric to see if someone is improving is fantastic and will really be helpful to integrate PROs into patient care. I have high hopes for using MCID in this capacity, but unfortunately it may not be ready just yet. There are challenges with MCID and here are the top three (3):
1. Time Point & Patient Population
I’ve heard a lot of people say things like ‘the MCID for the KOOS survey is 20′. This is a mistake – a PRO tool does not have a specific MCID. The MCID varies at different time points. Similarly, the MCID varies depending on the patient population. For example, a KOOS MCID for a patient’s 3-month post-operative Total Knee Replacement surgery could be 20 and KOOS MCID for an ACL patient’s 1-year post-operative could be 10. Put simply, patients’ expectations change over the course of their post-operative recovery and vary depending on the procedure. So BEWARE when you hear the term MCID used to blanket cover a PRO tool – that is not how it works folks. If you try to incorporate that into your practice, it will fail you.
2. Math Fail
Calculation using Distribution Method: Using the distribution method to calculate the MCID doesn’t always work well. With this method, the number of patients in your sample size changes what the MCID is and the larger the sample, the less meaningful it becomes. I know, it’s super weird – I’m going to save you the math-nerd speak, but basically it uses the variation in PROs to get an MCID, so as the number of patients gets bigger the variation (standard variation if you want to be hardcore) gets smaller and actually approaches zero. That is just a mathematical property and has no clinical relevance. And really, nothing can be done to fix it – it’s math.
3. Ceiling Effect
Calculation Using Anchor Method: The anchor method utilizes Quality of Life (QoL) questions as a surrogate or anchor to evaluate overall patient improvement. I actually like this method a lot, but there is one big issue with it: ceiling effect is a thing. Let’s say the PRO tool is a 100-point scale and the MCID is 25 points. Now imagine you have a patient, sweet Bob, who has a pre-operative score of 77. Bob is a tough guy with a high pain tolerance at baseline and he doesn’t let his aching hip get in the way of his day-to-day life. He powers through things, hence the pre-operative score of 77. That being said, sweet Bob can’t do the one thing that brings him so much joy because of pain – golf. So Bob has his hip replaced. A few months post-operative Bob is thrilled to be back playing golf and is very happy with the results. But according the MCID, Bob’s doing poorly. Even though Bob answered “yes my quality of life has improved” on the anchor question, he did not hit the MCID because his starting point was 77. Math sabotage again!
The Take Home
In my humble opinion, this is a scary thing because insurance companies are using MCID to say that the procedure was a success or was not a success. Think back to sweet Bob, who is reading this blog while waiting to tee off on the 18th hole…imagine if patients like Bob can get denied surgery in the future because predictive analytics say based on his pre-operative score of 77 he will not hit the MCID. So sad, right? The take home is that the higher the MCID, does not always mean the better outcome. In fact there is a lot of data, specifically Total Joint Arthroplasty, that supports the opposite being true – you can let your arthritis get so bad that there is a “point of no return” and therefore you should have the intervention earlier on in the disease process.
Where to go from here? MCID is a helpful starting point to integrate PROs into clinical practice. It it perfect? No – at this point in time, the concept needs to be fine-tuned. However, understanding the MCID for patient population and interval in time, then integrating that +/- number into clinical practice, is proving to be a valuable tool for providers. There is a lot of amazing and promising work being done to refine this potential gem into something real shiny – stay tuned for our next blog on “The future of MCID” – coming soon!