Read on Twitter
Tweetorial on our @JAMASurgery paper looking at the contribution of RVUs to differences in physician compensation across specialties.
Many have suggested RVUs are the PRIMARY reason... but it really hasn't been studied: https://jamanetwork.com/journals/jamasurgery/fullarticle/2763980
Many have suggested RVUs are the PRIMARY reason... but it really hasn't been studied: https://jamanetwork.com/journals/jamasurgery/fullarticle/2763980
Premise...
There are large differences in compensation across specialties.
Median salary...
Neurosurg: $730K
Family practice: $240K
There are large differences in compensation across specialties.
Median salary...
Neurosurg: $730K
Family practice: $240K
The RVU scale is often pointed to as a primary driver...
Eg: One study looked at cataract surgery, colonoscopy, and office visits and found procedures were compensated at ~ 4x higher rates than office visits https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/1754364
Eg: One study looked at cataract surgery, colonoscopy, and office visits and found procedures were compensated at ~ 4x higher rates than office visits https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/1754364
Cherry picking CPT codes distorts the story.
GI doesn't spend all day performing colonoscopies. Nor do ophthalmologists spend all day performing cataract surgery.
GI doesn't spend all day performing colonoscopies. Nor do ophthalmologists spend all day performing cataract surgery.
We sought to calculate average compensation rates (wRVUs/min) for an entire specialty over the course of an entire year at the national level...
e.g. when you look at all the care provided by all general surgeons in the course of a year, what is the average payment rate?
e.g. when you look at all the care provided by all general surgeons in the course of a year, what is the average payment rate?
We used publicly available CMS files and a 100% Medicare FFS file
What did we find?
Any given specialty performs a WIDE variety of procedures...
General surgeons, for example, billed over 4000 different CPT codes over the course of 2017
Any given specialty performs a WIDE variety of procedures...
General surgeons, for example, billed over 4000 different CPT codes over the course of 2017
When you average it all out here is the data....
There are certainly differences across specialties, but these differences are quite small...
Highest... EM (0.057 wRVUs/min)
Lowest... Path (0.029 wRVUs/min)
Highest... EM (0.057 wRVUs/min)
Lowest... Path (0.029 wRVUs/min)
In fact, we found no statistical difference between the average wRVUs for medical specialties (0.040 wRVUs/minute) and surgical specialties (0.043 wRVUs/minute)
Why the relatively narrow range?
It comes down to where each specialty spends the majority of their time and how each of those tasks is compensated...
It comes down to where each specialty spends the majority of their time and how each of those tasks is compensated...
Here is the distribution of each specialty's time...
Note that surgical specialists spend relatively little time on "intra-service" (i.e. actually operating)...
Eg. Ophthalmologists, as a specialty, only spend 2.4% of their time on cataract surgery
Eg. Ophthalmologists, as a specialty, only spend 2.4% of their time on cataract surgery
Far more time is spent on other activities, such as office visits, or postoperative rounding... time that is compensated at relatively fixed rates across specialties:
What about the concerns that RVU valuations are inaccurate?
A very valid concern...
We have published about this in the past
Most of the literature has focused on surgical CPT codes b/c there are large registries that have objective measures on LOS and operative time
A very valid concern...
We have published about this in the past
Most of the literature has focused on surgical CPT codes b/c there are large registries that have objective measures on LOS and operative time
It is likely that operative times and postoperative LOS are overestimated in RVU valuations... and therefore our estimates of compensation rates were too low...
We corrected for this...
And there were still quite small differences between medical and surgical provider compensation rates(~ 25%). And, importantly, this ASSUMES medical CPT code valuations are accurate (which is likely not true either...)
And there were still quite small differences between medical and surgical provider compensation rates(~ 25%). And, importantly, this ASSUMES medical CPT code valuations are accurate (which is likely not true either...)
But what about Dr. X who makes $2.5M doing skin biopsies all day?
There is likely MASSIVE variation within a specialty - likely more so than there is across specialties (sounds like a future study to me...). We can't cherry pick outliers to assume they are the norm.
There is likely MASSIVE variation within a specialty - likely more so than there is across specialties (sounds like a future study to me...). We can't cherry pick outliers to assume they are the norm.
We need to think about the myriad of other things that contribute to compensation differences...
- Work hours
- Payor mix
- % of time spent on non-compensated activity
- Industry payments
- Research
- Work hours
- Payor mix
- % of time spent on non-compensated activity
- Industry payments
- Research
Big picture?
We need to stop approaching the RVU system with anecdote... We are certainly NOT defending the current RVU system... we are the first ones to point out its many flaws. But we are pro evidence...
We need to stop approaching the RVU system with anecdote... We are certainly NOT defending the current RVU system... we are the first ones to point out its many flaws. But we are pro evidence...
RVU valuations need to be driven by objective data
Assessment of equity of the RVU system should be based on empiric evaluations
Assessment of equity of the RVU system should be based on empiric evaluations
