Total Per Capita Cost (TPCC) methodology proposed for the CY 2019 Physicians Fee Schedule


CMS continues punishing Geriatric, LTPAC, and Palliative Medicine for serving the frail elderly nearing the end of life.

This blog explores the systematic problems with existing methods for correctly predicting expenditures; as individuals approach the end of their lives. It’s based on comments we submitted to CMS on the draft 2019 Physician Fee Schedule.

America faces a unique problem in caring for its aging population; delivering that care is bankrupting us through the archaic design of traditional Medicare.  As a society, we are clinically effective at postponing death, often at great expense. The Medicare program covers most of that cost.  Congress, through the CMS, is attempting to delay the day of our fiscal insolvency by encouraging practitioners to join risk sharing payment models or requiring participation under MIPS.[i]

MIPS has many flaws, but the greatest is its total failure to focus on appropriate care for its most expensive beneficiaries – those nearing End of Life (EOL).  Instead, CMS continues to mandate the use of repurposed strategies that are useful in normally distributed (ambulatory) populations. MIPS and current A-APMs[ii] focus on preventing, postponing, or appropriately intervening in illnesses and/or injuries.

At the EOL, few, if any, of those options exist. Think about it, the Hospice Benefit was created because care at the EOL is different.  The facts are that many beneficiaries die outside of the Hospice Benefit, and Hospice itself is expensive.  We need a model of care that rewards Physicians and Advanced Practice Clinicians for cost effective, high quality, patient-centered care for beneficiaries nearing the EOL

My company is a technology vendor; We’re proud to serve medical practices that focus on LTPAC care settings.  Personally, I’ve read and channeled CMS’ regulations for more than40 years.  There is always some way to figure out how to satisfy mandatory reporting requirements – occasionally the solution isn’t cost justified, but there is a path.

Unfortunately, MIPS Eligible Clinicians who treat the frail and elderly are now guaranteed to flunk the CMS measures for Resource Use (formerly ‘Cost’ under VBP).  Beginning in 2018, Resource Use counts as 10% of a practitioner’s MIPS Score.  The draft Fee Schedule proposed it grow to 15% for 2019; Congress mandated that it count for 30% by 2022.  Over the same period, penalties grow from a current 5% rate to 9%.  Performance thresholds (minimum score to avoid a penalty) will also rise; the likely target is 50 out of the 100-point scale.

As a technology group, we know how to guide users to successfully managing their MIPS or ACO Quality Reporting.  We can also help avoid payment penalties secondary to poor CPT® coding.  What we cannot engineer out is the incredibly high cost of caring for individuals as they near the EOL. Why?  Because the CMS methodology for ‘risk-adjusting’ expenditures is a failure as an individual nears the end of life.  This is the population that practitioners working in Geriatrics, LTPAC, and Palliative Medicine specialize in serving.  It’s also the population that consumes ~25% of all CMS expenditures for Medicare A&B.

Today’s blog addresses the Total Per Capita Cost (TPCC) methodology proposed for the CY 2019 Physicians Fee Schedule.  It is the Resource Use measure that applies to Primary Care Practitioners working in Geriatrics, LTPAC, and/or Palliative Medicine.

The CMS states that the intention of MIPS scoring is to be both fair and accurate.  We believe that for MIPS Eligible Practitioners who care for the frailest beneficiaries, the TPCC methodology is neither fair, nor accurate.

According to the CMS document – Merit-Based Incentive Payment System (MIPS): Total Per Capita Costs for All Attributed Beneficiaries[iii],  the TPCC is driven by the CMS-HCC model.  That model uses diagnoses reported in the previous calendar year to risk adjust the current year’s expenditures for beneficiaries attributed to a Primary Care Physician.

“…rather than capturing the influence of treatment on a beneficiary’s health status, the risk adjustment model uses prior year (2017) risk factors to predict current year (2018) total per capita costs.” (pg. 7)

Beneficiaries who are approaching the end of life are increasingly served in special settings – Homecare, Assisted Living, or Nursing Facilities.  Medical practices are becoming more specialized – for example, based on a paper published in the JAMA[iv], titled Physicians and Advanced Practitioners Specializing in Nursing Home Care, 2012-2015 there was a 33% increase in practitioners specializing in Nursing Home Care; there are now more than 6,800 Practitioners in this special type of practice.  A private study based on the CMS Public Use Files indicates that in CY 2014, 80% of all nursing home care is delivered by less than 14,000 practitioners.

These practitioners, regardless of setting are specializing in treating the costliest Medicare Beneficiaries. According to a study published by the Henry Kaiser Foundation[v]

“Spending on Medicare beneficiaries in their last year of life accounts for about 25% of total Medicare spending on beneficiaries age 65 or older.”

Physicians and Advanced Practice Clinicians who focus on caring for beneficiaries as they near the end of life rarely have a long-standing relationship with the individual.  Community-based PCPs are not likely to follow their long-standing patients in institutional settings/programs.  Problems documented in the preceding year aren’t a reasonable predictor of current year’s costs at EOL.

The only ‘entry point’ for newly attributed beneficiaries for most these practitioners is at admission to an institutional LTC setting; always preceded by an Acute Care Episode in the case of a NF (POS 31) attribution, and frequently for a newly enrolled ALF resident.  The cost of this hospital episode is attributed to the Physician – but for which they had no participation.  However, under the Community CMS-HCC model, that practitioner is attributed 100% of that cost.

If 25% of all Medicare spending happens in the last year of life, and a practitioner specializes in treating those beneficiaries, how does the CMS-HCC model adequately risk-adjust?

The answer: The Model is a failure at correctly adjusting for the expenses in the last years of life.  In conjunction with AMDA – The Society for Post-Acute and Long-Term Care Medicine, the author has conducted studies of QRUR and now MIPS reports for multiple Long Term/Post-Acute care medical groups.  These studies cover the year’s 2012-2017.  The findings:

  • At a group level, 100% of the groups were ‘high cost’ under VBP, and now MIPS.
    • For MIPS, group scores were always the lowest @ 3 points.
    • When viewed at an individual NPI level a few individuals did achieve a maximum TPCC score of up to 4.5.

CMS policies for PACE[vi] programs address the inadequacy of the CMS-HCC model for the frailest adults. To quote from the on-line PACE manual[vii] (pg.5):

“A frailty factor is added to each individual’s risk score for PACE plan payment. Risk adjustment predicts (or explains) the future Medicare expenditures of individuals based on diagnoses and demographics. But risk adjustment may not explain all the variation in expenditures for frail community populations. The purpose of frailty adjustment is to predict the Medicare expenditures of community populations with functional impairments that are unexplained by risk adjustment. The frailty score added to the beneficiary’s risk score is calculated at the contract-level, using the aggregate counts of ADLs among HOS-M survey respondents enrolled in a specific organization. More information regarding the HOS-M can be found in Chapter 10, Section 30.7. Because the
CMS-HCC model has been designed to pay appropriately for the long-term institutionalized[viii] population, frailty adjustments are added to the risk scores only for community-based and short-term institutionalized enrollees (i.e., the frailty adjustment for long-term institutionalized enrollees is zero). Updated frailty factors are published in the Rate Announcement for the payment year in which they are first used.”

A 2004 RTI Study, commissioned by CMS, titled CMS Frailty Adjustment Model[ix] , clearly states that the CMS-HCC model is inadequate for predicting the expenditures of beneficiaries with deficits in 5 or more ADLs.

Similarly, CMS makes special adjustments for Medicare Advantage Special Needs Programs (SNPs).  There is a category for ‘newly enrolled’ beneficiaries which assigns additional risk adjustment points simply based on their SNP enrollment.  Premiums, and financial responsibility for payments, begin at enrollment date for MA-SNPs; the plan is not responsible for the cost of a hospital, and subsequent SNF, episode of care which occurred prior to enrollment. This contrasts with treatment of the MIPS Eligible Clinician serving an identical patient not enrolled in an MA plan – all the preceding episode(s) of care occurring in the same year are their responsibility.

On the day CMS closed the window for submitting comments on the draft 2019 Physicians Fee Schedule, the GAO published a report mandated by the IMPACT Act –

MEDICARE ADVANTAGE – Benefits and Challenges of Payment Adjustments Based on Beneficiaries’ Ability to Perform Daily Tasks[x]

The study showed individuals with ADL deficits were costlier to serve;

“Accounting for Functional Status Has Potential to Reduce Financial Disadvantages Associated with Enrolling and Caring for Beneficiaries with Functional Limitations.”

In addition, MA plans are paid on a month-to-month basis. If the patient dies, payments end. Under MIPS, if the patient expires, they are removed from the list of attributed patients.  Why is that significant?  For MA Plans, and traditional Medicare, the most significant cost containment practice is avoiding unnecessary hospitalizations. This is also usually in the patient’s best interest. Assume the MIPS EC is always successful in avoiding transfers to a hospital, and a subsequent death there – the result is no recognition under MIPS for all the extra effort required to achieve the desired outcome!

We agree that CMS has a vital interest in addressing the nearly uncontrolled expenditures that occur is the last year(s) of a beneficiary’s life. However, the current Resource Use model unfairly punishes the very practitioners CMS should be enlisting in that battle.

There is no simple solution. However, the role of practitioners who capture beneficiaries in LTC Settings is nearly identical to the enrollment criteria for SNPs, and I-SNPs in particular – 90+ days in a nursing facility, and eligibility for continuing Nursing Facility Care.

For this population, SNPs are only responsible for beneficiary expenses after the date of enrollment.  According to MEDPAC studies, SNP programs are popular, expanding, and profitable.  How can it be that practitioners treating the identical type of individual under traditional Medicare A+B are the worst performers under Resource Use?

Until the design of the MIPS risk adjustment is reworked to measure Resource Use appropriately, practitioners must focus on other opportunities to succeed when serving the EOL population.  Subsequent blogs in this series will discuss options for managing the diagnosis portion of Risk Adjustment (CMS-HCC), and efficient preparation/participation in MA and SNP programs.

[i] MIPS -Merit Based Incentive Payment System

[ii] A-APM: Advanced Alternate Payment Model


[iv] Ryskina KL, Polsky D, Werner RM. Physicians and Advanced Practitioners Specializing in Nursing Home Care, 2012-2015. JAMA. 2017;318(20):2040–2042. doi:10.1001/jama.2017.13378


[vi] PACE – Program for the All-inclusive Care of the Elderly


[viii] This refers to Medicare Advantage Institutional Special Needs Programs