Why traditional claims data is a bad yardstick.

Like it or not, your practice is already being measured.

It’s being measured by people who know nothing about physical therapy using limited, inaccurate data and unfair metrics. And, those flawed measurements are being used to select financial winners and losers.  
 
One of my favorite examples of the reporting of flawed measurements is the Propublica Treatment Tracker.

If you’re a Part B provider for Medicare, Propublica makes it easy to look up data that CMS has compiled on you and your practice. From the treatment tracker website, simply select your state, and then your specialty and your name.

At first glance, the data looks impressive. It includes services performed, number of patients, Medicare payments, average payment per patient, and average services performed per patient, and will rank you against your peers statewide.

When you dig into an individual provider’s data, Propublica ranks the relative complexity of each provider’s patient population along with the frequency of services provided by CPT, dollars billed to Medicare, and payments from Medicare.  
 
So what’s wrong with this picture? Plenty. 
 
Most importantly, it is all based on claims data, which is a bad yardstick.

Claims are produced after utilizing all kinds of arcane and complex rules for one overriding reason: to get paid.

Although other data is required on the claim, such as ICD-10 codes and CPT codes, those are of very limited value in discerning quality, and are often an after-thought for the provider. (Do you really list every comorbidity diagnosis code for your patients on your claims?)  
 
This problem with subpar, meaningless claims data especially rings true for rehab specialties.

As I’ve written in the past, ICD-10 codes describe diseases and injuries. By and large, we treat the sequelae of diseases and injuries, not the disease or injury itself.

CPT codes, published by the AMA, are primarily designed to account for technical procedures for our profession, not professional services, and don’t come close to doing an accurate job of describing the actual services provided.

Worse yet, the most valuable services we provide (thorough evaluations, patient management, counseling, education, etc.) are either not paid at all or are extremely under-valued.

As a result of this and as part of the game we all play in fee for service, we’re encouraged to spend more time providing timed technical services (ther ex, manual therapy, etc.) than untimed or non-reimbursed professional services (evaluations, counseling, education, management).   
 
So, we have codes that are placed on claim forms for the purposes of getting paid in a dysfunctional system that is being used today to make judgments about cost and quality. Worse yet, payers and employers are using that flawed data to make decisions about preferred provider networks. 

In addition to the flawed data problem, we have a well-known perverse incentive problem with fee for service.

The Incentive Alignment Problem
I recently had an opportunity to discuss this problem with one of the largest employers in the country wrestling with physical therapy utilization.

They’ve noticed that PTs tend to use every visit authorized, regardless of the number of visits authorized. In fact, they said “If our employees are authorized to get 20 visits of PT, they always seem to get 20 visits of PT.”

My response? “Of course!”

That’s the way incentives are aligned. PTs fight and scrape to get patients adequate treatment every day, and many of them don’t. When you have a rare instance of a generous pre-auth, it isn’t at all surprising that all of the visits tend to get used.

Evolution
The good news: CMS and commercial payers have gotten the message. Instead of relying on claims-based data from a fee for service system, there is a steady transition to measurements based on clinical quality.

And, more of that data will be available to the public than ever before.  

Among the mandates of the Affordable Care Act was helping consumers make informed decisions about the healthcare they receive. That mandate resulted in the development of the Physician Compare website.

From CMS: “Medicare patients and caregivers are able to use the Physician Compare website to search for and compare clinicians and groups who are enrolled in Medicare. The Physician Compare website also incentivizes clinicians to maximize their performance by making performance information publicly available.” 
 
Currently, the information available on the site isn’t very comprehensive, as it only includes demographic information, contact information, and educational background of providers, but very soon it will include quality data from the MIPS program. Exciting stuff. 
 
MIPS itself is changing.
There’s a definite shift away from “process based” measures that you may be familiar with from the old PQRS program (like medication list, falls risk, falls plan of care) to patient reported outcomes measures.

And, CMS is taking steps to make quality measures, activity improvement projects, cost, and interoperability more meaningful to clinicians through the MIPS Value Pathways.
 
The move from volume to value is real. The pace of change is accelerating. Measuring outcomes is changing from best practice to a requirement.  
 
Time to get a new yardstick. 
 
I’d love to hear your thoughts. Share them with me over on Twitter.

Jerry

P.S. I talked about the evolution of MIPS for 2020 with Holly, Keet’s fearless leader, in a recent webinar. Catch the recording here