I remember verbatim a conversation from 2002 with a well-respected skilled nursing facility pioneer. “Marc, I've been in this business for 32 years, and this is going to be the toughest year yet… of course I've said that every January for 32 years!” I hope with all my heart he'll have to retire that little joke after we mercifully close the books on 2020.
Thanksgiving is a time for gratitude and appreciation. But for those who endured the darkest days of this pandemic, it's an opportunity to process, in groups of no more than 10 people, hardships that were unthinkable way back in the age known as BCE (Before the COVID Era). Yet there is good reason for optimism with a vaccine just weeks away, so let's take a moment to reflect on what SNF operators can be thankful for this year.
Before I get into the amalgam of regression analyses and timestudies that shaped my worldview, I offer a heartfelt thank you to the army of tireless caregivers. “Frontline” is an apropos wartime reference, as so many are deserving of the Purple Heart. I am first and foremost thankful for your incredible courage and perseverance; as a distant second, I am thankful for perfect timing - for the October 1, 2019 arrival of PDPM.
Medicare's RUG-model was a statistical disgrace - I've never been shy about expressing that truth. It encouraged over-utilization, inflamed the sensibilities of the OIG and worse still, encouraged dozens of state Medicaid agencies to follow an out-of-tune Pied Piper into the river at “Ultra High” tide. Comparatively, PDPM, while far from perfect, is a statistical masterpiece. Yes, SNFs did well under the incredibly flawed RUG system, but think about how we would have fared had the Patient Driven Payment Model been delayed one year.
By the Numbers
While we can't recalculate 2020 PDPM scores to fit the RUG model, we can trend 2019 data for comparison. For purposes of this discussion, a typical SNF may have budgeted $590 per day through a RUG-priced pandemic. For national comparative integrity, $590 approximates the average 2019 RUG distribution using an inflated, urban-unadjusted rate set (i.e. 2020 MBI, wage index = 1.0), without netting co-payments.
We turn to CORE Analytics for PDPM data. Note that our figures are not based on Medicare's public Limited Data Set release. Many companies market their LDS compilations as “most up-to-date,” but that term is deceiving. As of October 31, the most recently available data covered claims processed through March 31, 2020. Not only does the LDS lack critical UB-04 detail to complete the patient profile, the COVID ICD-10 code did not enter use until April. CORE's data includes claims with dates of service through October 31, 2020; we don't call it “most up-to-date,” we call it “current & comprehensive,” big difference.
CORE released the industry's first look at PDPM claims last year, albeit from a relatively small sample size. Our client base at the time consisted of 587 SNFs. I stressed that our contributors were not randomly selected, and findings therefore should not be extrapolated to the national community. Most early adapters were well-prepared for the payment transition; as new providers were onboarded, their PDPM claims were retroactively uploaded. As anticipated, facilities that joined later reported lower rates during PDPM's onset relative to the launch group
There are now roughly 1,800 SNFs uploading monthly claims to CORE, with 2,231 individual facilities represented throughout our PDPM timeline. These are highly diverse providers in every relevant sense of the definition - more than appropriate to serve as a reasonable proxy for national benchmarking. Furthermore, this is an “apples-to-apples” comparison; every provider's performance is recalculated using the same, unadjusted urban rate set. If you're trying this at home - once you recalculate your unweighted average, the only qualifier to consider is average length of stay. After neutralizing the impact of the 1135 Waiver, CORE's ALOS settled in at around 27 days months ago - another indicator that our sample is sufficient for this discussion. Predictably, average per diem rates creeped higher as coding and capture accuracy improved through the year:
All data courtesy of CORE Analytics' proprietary database which does not reflect comprehensive industry claim submissions.
Performance differs from prior CORE releases as averages are recalculated when new facilities are added to the database
Clearly PDPM rates exceed RUG-IV's putative $590 PPD, but there's more to the story. For starters, therapy expenses were perverted by arbitrary RUG utilization thresholds (there, I said it). In order to achieve $590, a facility had to capture more than half its days at Ultra High, with (anecdotal) therapy costs per day significantly above 2020 spending. Resource-intensive nursing acuities were essentially irrelevant for rate setting - and therein lies the RUG, I mean RUB, I mean “there's the rub” (RUG-IV even confounded Hamlet).
The point is that our hypothetical 2020 RUG-IV “RUB” rate is $647 per day, but after netting direct rehab costs it's closer to $540. Now take the same resident in a COVID environment, non-medically complex, yet isolated per CDC (but not RAI) guidelines, and the RUG-IV rate drops to as low as $303 (PC1). Meanwhile, CORE's September 2020 average PDPM rate was $625, irrespective of direct therapy utilization or expenses.
PDPM did not result in the doomsday “rationing” scenario some feared . Pre-pandemic, therapy charges reported on CORE-analyzed UB-04s were down only slightly from 2019 levels; after a generation of therapy-centric payment, reimbursement was once again controlled by nursing. What happened next is the cause for thanks.
Therapy charges dropped precipitously as facilities were hobbled by the pandemic. Staffing shortages (sometimes exacerbated by therapists being reassigned to other departments) coupled with isolation protocols were the primary causes, but so was the 1135 Waiver. Much has been written about it from a compliance perspective, but nothing in context with RUG rates. As elective surgeries were suspended, long-term care residents replaced short-term subacute admissions as the driver of Medicare census.
While not a “ license to skill ,” Waiver patients often represented more than half of a facility's Medicare Part A billing - yet few of these residents would have been appropriate for two hours of daily therapy.
Many Waiver patients were skilled for “monitoring & observation,” often without intensive nursing needs. To make matters worse, protocols for capturing PDPM-darlings Respiratory Therapy and Depression were not broadly developed under RUGs, relegating many scores to the “lower 14” domain. Furthermore, nuevo-Swallowing Disorders and Mechanically Altered Diets had no “reimbursement-sensitivity” under RUGs without ST to drive the score. Those four items represent $180 of PDPM rate elasticity laid fallow under RUGs. Let's take a look:
Yes, isolation (ES1) paid a respectable RUG-IV rate without associated therapy expenses. However, our data suggests inconsistencies in MDS capture, as early confusion between CDC and RAI guidelines skews the analysis
To put things in perspective, we estimate that RUG-IV would have resulted in an average Medicare rate of about $420 per day throughout the pandemic, while PDPM generated approximately $620 for CORE's client base. This differential annualizes to roughly $90,000 in bottom line revenue per month for a facility with an average daily Medicare census of 15, operating under the conditions outlined within.
If that differential triggers your cognitive dissonance alarm on PDPM's budget neutral factor, hit the snooze button for now. There are so many mitigating factors to that equation. Comparing payment levels between 2019 (let alone 2017) and 2020 is like the impossible “GOAT” argument in sports - it's interesting to be sure, but eras are defined by distinct conditions making outcomes noncomparable; it's hard to believe our “BCE” ended only nine months ago.
Introducing MAPAX
Medicare Advantage, a growing threat as the “ FFS Attrition Rate ” rises, created a host of unique COVID revenue-cycle challenges that I'll politely avoid discussing today. That said, CORE Analytics' Medicare Advantage Post-Acute eXchange (MAPAX) is currently under development and scheduled for a March 2021 release. Aggregating and analyzing performance across payers, contract structures and platforms in a consistent, comparable way is far more challenging than doing so for Medicare claims. While CORE's FFS “Logic Tests” don't carry the same resonance, MA-SNF performance represents a “black hole” in the claimanalytics universe. This needs to change - and no time like the present.
Final Thoughts
As we approach America's favorite holiday, let's take some time to count our blessings. If you've forgotten what it means to be content, I offer this strategy: Stay safe, don't talk politics, say a prayer for the caregivers who sacrificed so dearly… and save a bit of gratitude for PDPM. It arrived just in time.
Happy ThankSNFing
Marc Zimmet
I remember verbatim a conversation from 2002 with a well-respected skilled nursing facility pioneer. “Marc, I've been in this business for 32 years, and this is going to be the toughest year yet… of course I've said that every January for 32 years!” I hope with all my heart he'll have to retire that little joke after we mercifully close the books on 2020.
Thanksgiving is a time for gratitude and appreciation. But for those who endured the darkest days of this pandemic, it's an opportunity to process, in groups of no more than 10 people, hardships that were unthinkable way back in the age known as BCE (Before the COVID Era). Yet there is good reason for optimism with a vaccine just weeks away, so let's take a moment to reflect on what SNF operators can be thankful for this year.
Medicare's RUG-model was a statistical disgrace - I've never been shy about expressing that truth. It encouraged over-utilization, inflamed the sensibilities of the OIG and worse still, encouraged dozens of state Medicaid agencies to follow an out-of-tune Pied Piper into the river at “Ultra High” tide. Comparatively, PDPM, while far from perfect, is a statistical masterpiece. Yes, SNFs did well under the incredibly flawed RUG system, but think about how we would have fared had the Patient Driven Payment Model been delayed one year.
By the Numbers
While we can't recalculate 2020 PDPM scores to fit the RUG model, we can trend 2019 data for comparison. For purposes of this discussion, a typical SNF may have budgeted $590 per day through a RUG-priced pandemic. For national comparative integrity, $590 approximates the average 2019 RUG distribution using an inflated, urban-unadjusted rate set (i.e. 2020 MBI, wage index = 1.0), without netting co-payments.
We turn to CORE Analytics for PDPM data. Note that our figures are not based on Medicare's public Limited Data Set release. Many companies market their LDS compilations as “most up-to-date,” but that term is deceiving. As of October 31, the most recently available data covered claims processed through March 31, 2020. Not only does the LDS lack critical UB-04 detail to complete the patient profile, the COVID ICD-10 code did not enter use until April. CORE's data includes claims with dates of service through October 31, 2020; we don't call it “most up-to-date,” we call it “current & comprehensive,” big difference.
CORE released the industry's first look at PDPM claims last year, albeit from a relatively small sample size. Our client base at the time consisted of 587 SNFs. I stressed that our contributors were not randomly selected, and findings therefore should not be extrapolated to the national community. Most early adapters were well-prepared for the payment transition; as new providers were onboarded, their PDPM claims were retroactively uploaded. As anticipated, facilities that joined later reported lower rates during PDPM's onset relative to the launch group
There are now roughly 1,800 SNFs uploading monthly claims to CORE, with 2,231 individual facilities represented throughout our PDPM timeline. These are highly diverse providers in every relevant sense of the definition - more than appropriate to serve as a reasonable proxy for national benchmarking. Furthermore, this is an “apples-to-apples” comparison; every provider's performance is recalculated using the same, unadjusted urban rate set. If you're trying this at home - once you recalculate your unweighted average, the only qualifier to consider is average length of stay. After neutralizing the impact of the 1135 Waiver, CORE's ALOS settled in at around 27 days months ago - another indicator that our sample is sufficient for this discussion. Predictably, average per diem rates creeped higher as coding and capture accuracy improved through the year:
Performance differs from prior CORE releases as averages are recalculated when new facilities are added to the database
Clearly PDPM rates exceed RUG-IV's putative $590 PPD, but there's more to the story. For starters, therapy expenses were perverted by arbitrary RUG utilization thresholds (there, I said it). In order to achieve $590, a facility had to capture more than half its days at Ultra High, with (anecdotal) therapy costs per day significantly above 2020 spending. Resource-intensive nursing acuities were essentially irrelevant for rate setting - and therein lies the RUG, I mean RUB, I mean “there's the rub” (RUG-IV even confounded Hamlet).
The point is that our hypothetical 2020 RUG-IV “RUB” rate is $647 per day, but after netting direct rehab costs it's closer to $540. Now take the same resident in a COVID environment, non-medically complex, yet isolated per CDC (but not RAI) guidelines, and the RUG-IV rate drops to as low as $303 (PC1). Meanwhile, CORE's September 2020 average PDPM rate was $625, irrespective of direct therapy utilization or expenses.
PDPM did not result in the doomsday “rationing” scenario some feared . Pre-pandemic, therapy charges reported on CORE-analyzed UB-04s were down only slightly from 2019 levels; after a generation of therapy-centric payment, reimbursement was once again controlled by nursing. What happened next is the cause for thanks.
Therapy charges dropped precipitously as facilities were hobbled by the pandemic. Staffing shortages (sometimes exacerbated by therapists being reassigned to other departments) coupled with isolation protocols were the primary causes, but so was the 1135 Waiver. Much has been written about it from a compliance perspective, but nothing in context with RUG rates. As elective surgeries were suspended, long-term care residents replaced short-term subacute admissions as the driver of Medicare census.
While not a “ license to skill ,” Waiver patients often represented more than half of a facility's Medicare Part A billing - yet few of these residents would have been appropriate for two hours of daily therapy.
Many Waiver patients were skilled for “monitoring & observation,” often without intensive nursing needs. To make matters worse, protocols for capturing PDPM-darlings Respiratory Therapy and Depression were not broadly developed under RUGs, relegating many scores to the “lower 14” domain. Furthermore, nuevo-Swallowing Disorders and Mechanically Altered Diets had no “reimbursement-sensitivity” under RUGs without ST to drive the score. Those four items represent $180 of PDPM rate elasticity laid fallow under RUGs. Let's take a look:
Yes, isolation (ES1) paid a respectable RUG-IV rate without associated therapy expenses. However, our data suggests inconsistencies in MDS capture, as early confusion between CDC and RAI guidelines skews the analysis
To put things in perspective, we estimate that RUG-IV would have resulted in an average Medicare rate of about $420 per day throughout the pandemic, while PDPM generated approximately $620 for CORE's client base. This differential annualizes to roughly $90,000 in bottom line revenue per month for a facility with an average daily Medicare census of 15, operating under the conditions outlined within.
If that differential triggers your cognitive dissonance alarm on PDPM's budget neutral factor, hit the snooze button for now. There are so many mitigating factors to that equation. Comparing payment levels between 2019 (let alone 2017) and 2020 is like the impossible “GOAT” argument in sports - it's interesting to be sure, but eras are defined by distinct conditions making outcomes noncomparable; it's hard to believe our “BCE” ended only nine months ago.
Introducing MAPAX
Medicare Advantage, a growing threat as the “ FFS Attrition Rate ” rises, created a host of unique COVID revenue-cycle challenges that I'll politely avoid discussing today. That said, CORE Analytics' Medicare Advantage Post-Acute eXchange (MAPAX) is currently under development and scheduled for a March 2021 release. Aggregating and analyzing performance across payers, contract structures and platforms in a consistent, comparable way is far more challenging than doing so for Medicare claims. While CORE's FFS “Logic Tests” don't carry the same resonance, MA-SNF performance represents a “black hole” in the claimanalytics universe. This needs to change - and no time like the present.
Final Thoughts
As we approach America's favorite holiday, let's take some time to count our blessings. If you've forgotten what it means to be content, I offer this strategy: Stay safe, don't talk politics, say a prayer for the caregivers who sacrificed so dearly… and save a bit of gratitude for PDPM. It arrived just in time.