Independent lines of evidence re: rad onc workload circa 2020 vs long ago

[scarbrtj]

A human by the name of @wandering star came up with a neat way of looking at things yesterday estimating about 175 new starts per year per rad onc.

A while back I estimated 100 de novo patient new starts per rad onc each year. (It's very easy to prove this number.) Got criticism re: this number (too low) but some didn't understand these were the recently diagnosed "first line treats." (Even if palliative, and de novo diagnosed, would fit in this.) If the proportion of a service that's treating patients diagnosed >1y ago (palliatives, recurred, retreat etc) is 50%, this would mean ~200 new starts a year. If it's 40%, 167 new starts a year; if 33%, 150 new starts a year, etc.

A fellow named Graypeace did an even easier to cite/understand analysis that the number is 200 new starts per year.

Thus if we take the numbers (175, 200, 167, 150, 200) the mean is about 178 with a 99% CI of the mean of about 145-205.

Graypeace also showed that in 2003 rad oncs averaged about 350 new starts a year. If we do something (statistically questionable) and scale (175, 200, 167, 150, 200) about 350 we'd get 350, 400, 334, 300, 400.

[175, 200, 167, 150, 200] vs [350, 400, 334, 300, 400];

These two samples are different at p=0.01*

*Mann-Whitney

---

Members don't see this ad.

 

[Mandelin Rain]

 

[Lamount]

350 new starts a year seems like a lot...

In 2003, I presume many of those cases involved drawing a polygon on a plain film and calling it a day.

7 new (mostly) IMRT plans a week is alright for a few weeks, but I wouldn't want to make a career out of it. I am sure many people here routinely do this but, given the amount of time that is required for modern treatment planning, I wouldn't expect this to be the norm.

 

[medgator]

350 new starts a year seems like a lot...

Really?

 

[RickyScott]

350 new starts a year seems like a lot...

In 2003, I presume many of those cases involved drawing a polygon on a plain film and calling it a day.

7 new (mostly) IMRT plans a week is alright for a few weeks, but I wouldn't want to make a career out of it. I am sure many people here routinely do this but, given the amount of time that is required for modern treatment planning, I wouldn't expect this to be the norm.

It’s very doable with mixed pt load and 40hour week with plenty of internet surfing. In past it would amount to upper 20s on treatment.

 

[Lamount]

Really?

It’s very doable with mixed pt load and 40hour week with plenty of internet surfing. In past it would amount to upper 20s on treatment.

Fair enough.

In retrospect, I am 60% FTE... and in the COVID era, I am home on my non-clinic days doing mostly research, so it probably is not a fair comparison. Most of my department is the same, so I forget sometimes that a lot of people spend more time in clinic.

...I have also made a practice out of unnecessarily complex patients and likely overthink my treatment planning (which is another conversation entirely )

 

[wandering star]

If we want to compare patient volumes across time we can do so. I based my analysis on the CMS 2018 data, but the data goes back 6 years to 2012.

In 2012 there were 3849 rad onc MDs seeking CMS reimbursement for a total of 296,688 unique beneficiaries under the "management of 5 treatments" code. In 2018 there were 4263 rad onc MDs seeking reimbursement for 305,667 unique beneficiaries under the same code.

81% of the cohort in 2012 was again present in 2018. About 800 left (presumably retired) and about 1200 members entered. Of the MDs present in both cohorts, we can correlate volume and produce a scatter plot, with volume in 2012 on x-axis and volume in 2018 on y-axis:

There is no data available if there were 10 or fewer submissions. Maybe this is a privacy or PHI thing. The R^2 of a linear fit (with intercept at 0) is 79%. On average there is 16% less volume per matched MD as measured by this metric. Obviously broader trends at play here (differing insurance or new treatment paradigms) make the comparison more complex.

 

[radiation]

I think one thing we will likely have to let go of is reminiscing what our practices used to look like and trying to get back to the way things used to be. I can imagine a huge part, perhaps more than >50% of many rad onc practices having their bread and butter be metastatic pts and retreats.

When you think of med oncs giving 2nd, 3rd, 4th lines of (mostly) ineffective chemo, faced with decreasing total unique pts, a natural reflex is going to be lowering the threshold for palliative radiation for an established pool of pts. There will be more incentive to follow definitive pts who end up developing mets and treating as they come up vs. relying med oncs to send them to you. What percent of your stage III lung cancer pts will receive at least two courses of radiation? Gotta be at least half of them between brain mets and bone mets amirite?

 

[scarbrtj]

If we want to compare patient volumes across time we can do so. I based my analysis on the CMS 2018 data, but the data goes back 6 years to 2012.

In 2012 there were 3849 rad onc MDs seeking CMS reimbursement for a total of 296,688 unique beneficiaries under the "management of 5 treatments" code. In 2018 there were 4263 rad onc MDs seeking reimbursement for 305,667 unique beneficiaries under the same code.

81% of the cohort in 2012 was again present in 2018. About 800 left (presumably retired) and about 1200 members entered. Of the MDs present in both cohorts, we can correlate volume and produce a scatter plot, with volume in 2012 on x-axis and volume in 2018 on y-axis:

View attachment 331177

There is no data available if there were 10 or fewer submissions. Maybe this is a privacy or PHI thing. The R^2 of a linear fit (with intercept at 0) is 79%. On average there is 16% less volume per matched MD as measured by this metric. Obviously broader trends at play here (differing insurance or new treatment paradigms) make the comparison more complex.

If I'm viewing it right, from 2012 to 2018 radiation "demand" was remarkably stable but radiation (MD) supply went up. This is becoming a meme.

I think one thing we will likely have to let go of is reminiscing what our practices used to look like and trying to get back to the way things used to be. I can imagine a huge part, perhaps more than >50% of many rad onc practices having their bread and butter be metastatic pts and retreats.

Reading "2021" advances today. There's a radiotherapeutic tittynope in there about a 63 patient study in metastatic nasopharyngeal where the primary and nodes were treated w/ XRT. Yay. That was it for XRT in 2020. More inconvenient truths: the increase in rad onc MDs (and MD/PhDs) and residents the last 15 years has been associated with a decrease in XRT advances... at least as viewed through the eyes of the JCO. (Lori Pierce's name is front & center in the 2021 report though.) Rad onc got referred to in the 2006 text ("shout outs" = RT mentions) more than 4 times as much as the 2021 text. "The way things used to be" indeed.

I encourage you to read through the 2021 report. The improvements in OS and DFS from "chemotherapy" and exponential complexification of the (non-radiotherapeutic) oncological landscape give the rad onc some pause. We've literally decreased treatments (OK, fractions... but we've lost some indications too) the last 15 years while the profound opposite has happened in med onc.

 

[radiation]

If I'm viewing it right, from 2012 to 2018 radiation "demand" was remarkably stable but radiation (MD) supply went up. This is becoming a meme.



Reading "2021" advances today. There's a radiotherapeutic tittynope in there about a 63 patient study in metastatic nasopharyngeal where the primary and nodes were treated w/ XRT. Yay. That was it for XRT in 2020. More inconvenient truths: the increase in rad onc MDs (and MD/PhDs) and residents the last 15 years has been associated with a decrease in XRT advances... at least as viewed through the eyes of the JCO. (Lori Pierce's name is front & center in the 2021 report though.) Rad onc got referred to in the 2006 text ("shout outs" = RT mentions) more than 4 times as much as the 2021 text. "The way things used to be" indeed.

I encourage you to read through the 2021 report. The improvements in OS and DFS from "chemotherapy" and exponential complexification of the (non-radiotherapeutic) oncological landscape give the rad onc some pause. We've literally decreased treatments (OK, fractions... but we've lost some indications too) the last 15 years while the profound opposite has happened in med onc.

Shouldn't the takeaway from this be that whatever we are doing for the last 15 years has not advanced our field, and new approaches need to be taken if we want to change things? Lamenting the past is not likely to fix the future. Residency expansion alone isn't the culprit for this lack of advances, but certainly has exacerbated the economics and job prospects problem.

 

[RickyScott]

Shouldn't the takeaway from this be that whatever we are doing for the last 15 years has not advanced our field, and new approaches need to be taken if we want to change things? Lamenting the past is not likely to fix the future. Residency expansion alone isn't the culprit for this lack of advances, but certainly has exacerbated the economics and job prospects problem.

What if it is not fixable? That is my belief.

 

[TheWallnerus]

What if it is not fixable? That is my belief.

I heard kind of a "wow" piece of info person-to-person from the vice chair of a leading academic place in NYC today. He said that 30% of their on-treatment volume are SBRTs for mets now. He seemed proud. Inside I died a little. Not that this is bad per se. I just don't know if it's good. Hope the data holds up for us. This is such a huge change from "what is a radiation oncologist?" versus when I trained (now it's very short treatment schedules... with minimal side effects). Of course I finished training when Jesus was in 3rd grade. Oh another fun fact. This same leading academic place in NYC? When they transitioned to tele-health at the beginning of COVID they were doing the video chat thing for initial consults. One patient complained that he got a bill for $5000 for what was a 15 minute chat. Wild stuff going on out there!

 

[OTN]

I heard kind of a "wow" piece of info person-to-person from the vice chair of a leading academic place in NYC today. He said that 30% of their on-treatment volume are SBRTs for mets now. He seemed proud. Inside I died a little. Not that this is bad per se. I just don't know if it's good. Hope the data holds up for us. This is such a huge change from "what is a radiation oncologist?" versus when I trained (now it's very short treatment schedules... with minimal side effects). Of course I finished training when Jesus was in 3rd grade. Oh another fun fact. This same leading academic place in NYC? When they transitioned to tele-health at the beginning of COVID they were doing the video chat thing for initial consults. One patient complained that he got a bill for $5000 for what was a 15 minute chat. Wild stuff going on out there!

If the 30% on-treatment mets volume is on top of what they were already treating, then that's overall good for the field. Probably isn't, though.

Price transparency (mandated, or at least backed up by a real fine) would fix the last part of that.

 

[wandering star]

NEJM published a remarkable "sounding board" perspective article on CMS innovation a few days ago. The "free preview" version of their central figure is below (not sure if we are allowed to share the version behind the paywall here). This is a list of all the innovations they have tried to reduce cost and/or improve quality.

The red bars are the innovations that have generated losses. The blue ones are the ones that have generated savings. OK, so the real article shows a few more bars that are cropped off that are blue, but the overwhelming trend is that they got a bunch of smart people together, tried our their best ideas, and nearly all of their ideas fell flat.

As an outsider it is easy to imagine that healthcare value would be so much better if [X] were done (e.g. price transparency) but on the whole our system seems remarkably resilient to easy fixes.

 

[drewdog1973]

Who says people working for the federal government are “smart”? Amirite @Chartreuse Wombat ??

 

[Chartreuse Wombat]

Who says people working for the federal government are “smart”? Amirite @Chartreuse Wombat ??

The federal government has no monopoly on talent that's for sure. With the government huge failures (like those described in the article) never have real consequences like dissolution of a program; more money is thrown at the problem because there "has to be a way to succeed" especially when we use "the best and brightest".

The most chilling statement for me is "Although voluntary models allow providers to choose whether to participate, the Center has found that it is very difficult for voluntary models to lower costs."

Which brings us to RO-APM which is the model for the "Innovation Center" going forward. Mandatory participation and capitation. The way for the federal government to lower its (our) costs is to require participation and negotiate payments. This is the future dreamed of by single-payor enthusiasts. Perhaps the majority of US citizens desire this but IMHO they really don't know what they are getting.

 

[TheWallnerus]

"Although voluntary models allow providers to choose whether to participate, the Center has found that it is very difficult for voluntary models to lower costs."

And EVERY rad onc in the USA should rage, rage against the distorting of the truth here.

RAD ONC HAS LOWERED ITS OWN COSTS, VOLUNTARILY.

CMS Rad Onc Part B, 2011: $1.8 billion

CMS Rad Onc Part B, 2018: $1.5 billion

I don't think a single specialty in the history of CMS has shown a percent decrease in cost like this over such a short period of time.

 

[medgator]

And EVERY rad onc in the USA should rage, rage against the distorting of the truth here.

RAD ONC HAS LOWERED ITS OWN COSTS, VOLUNTARILY.

CMS Rad Onc Part B, 2011: $1.8 billion

CMS Rad Onc Part B, 2018: $1.5 billion

I don't think a single specialty in the history of CMS has shown a percent decrease in cost like this over such a short period of time.

Yet we still got APM'd. That's an indictment of ASTRO's leadership

 

[GI_RadOnc]

If we want to compare patient volumes across time we can do so. I based my analysis on the CMS 2018 data, but the data goes back 6 years to 2012.

In 2012 there were 3849 rad onc MDs seeking CMS reimbursement for a total of 296,688 unique beneficiaries under the "management of 5 treatments" code. In 2018 there were 4263 rad onc MDs seeking reimbursement for 305,667 unique beneficiaries under the same code.

81% of the cohort in 2012 was again present in 2018. About 800 left (presumably retired) and about 1200 members entered. Of the MDs present in both cohorts, we can correlate volume and produce a scatter plot, with volume in 2012 on x-axis and volume in 2018 on y-axis:

View attachment 331177

There is no data available if there were 10 or fewer submissions. Maybe this is a privacy or PHI thing. The R^2 of a linear fit (with intercept at 0) is 79%. On average there is 16% less volume per matched MD as measured by this metric. Obviously broader trends at play here (differing insurance or new treatment paradigms) make the comparison more complex.

Question: Should one include the other treatment management charges for single fraction, as well as up to 5 fraction SBRT/SRS? I would think that there would be more patients in 2018 under those codes versus 2012. Mind you; I'm not doubting the less patient per rad onc; but just being a nerd on methods