Cost Drivers in Oncology - Community Oncology Alliance

Download Report

Transcript Cost Drivers in Oncology - Community Oncology Alliance

Cost Drivers in
Oncology
Bundled Payments
Fee for Service
Episodes of Care
Oncology Medical Home
Pathways
ACO’s
The Problem

What we hear and read:
 Oncology
is expensive
 Cost increases are unsustainable
 We

have to bend the cost curve
All true, but:
 What
is the cost curve?
 What parts can be bent?
 What are the premises behind any specific payment
reform model? Does it solve the right problem?
The Cost Curve

Typically (almost always) oncology costs are discussed one to
two components at a time

Drug pricing – 25% of oncology costs

True costs can only be understood when all are considered
together

Main issue for this session

Biases

Payer experts in the audience

Much is unknown

But that itself is informative

We have to understand the limits of our knowledge to be truly
informed
What are we spending?

Oncology Care
 $89
 5%
billion in 2007
of total healthcare spending
 0.8%
of GDP
Causes of the Rising Cost Curve

Demographics


Likely most powerful factor
More intense treatment

More lines of appropriate therapy

Increased amount of time on active therapy

More labs and imaging to monitor therapy

Rising cost of therapeutics

Change in site of care
Top Ten Causes of Death
Jones et. al. N Engl J Med 2012 366;25
Top Ten Causes of Death
Jones et. al. N Engl J Med 2012 366;25
Demographics

Tangka 2010: Cancer Treatment Costs in the United States

Compared data from the Medical Expenditure Panel Survey
(2001-2005) with it predecessor, the National Medical
Expenditure Survey (1987)

Increase in cost driven mainly by increase in number of cases
rather than the cost per treated case

Cancer costs doubled over the study period but so did other
health costs

Cancer costs 4.8% of overall medical expenditures in 1987; 4.90%
between 2001-2005
Tangka et al. Cancer Treatment Costs in US. Cancer 2010.
Demographics

Mariotto 2010: Projections of the Cost of Cancer 20102020

Dynamic projection of cost of cancer care

27% increase in costs expected due to US population
changes only - independent of incidence, survival, and
cost of care per case

Model:
 2%
annual cost increases per case – 39% increase in costs
 5%
annual cost increases per case – 66% increase in costs
Mariotto et al. Projections of the cost of cancer care in the United States: 2010-2020. J Natl Cancer Inst 2011; 103:117-128
Demographics

Key cost driver in oncology is non-modifiable
 Cannot

bend this part of the curve
Good data is hard to find
 Either
old or based on modeling and
projections
 Payers
have the newest data
More Intense Treatment

Metastatic Colon Cancer

Survival
 Increased
from 8 months to 2 years over past two decades

Drug cost of Mayo regimen of 5-FU/leucovorin for eight
weeks- $63

Newer drugs
 Irinotecan,
oxaliplatin, capecitabine, bevacizimab,
cetuximab, panitumimab, regorafenib, aflibercept
Lines of Therapy
What is Treatment Intensity?








Not just drug
Multiple lines of therapy: what else do we do along the way?
Labs, imaging, clinic visits, supportive care, possibility of
hospitalization for each line of therapy
 Maintain quality of life
 Is it working?
 Complications of therapy
Surgical resection of primary site/liver disease
Stereotactic body radiation
Chemoembolization
Radioembolization
Clinical Trials
Payer Spending
What About Drugs?
Source: The US Oncology Network
2012 FDA Approvals

11 of 34 drugs approved by the FDA were oncology drugs

Axitinib

Vismodegib

Pertuzumab

Carfilzomib

Aflibercept

Enzalutamide

Regorafenib

Bositinib

Omacetaxine

Cabozantinib
Drug Development Cost

Costs have to be modelled (with all the inherent
shortcomings of this approach)

“Expected value terms”
 Have
to include cost of failing drugs with drugs
ultimately approved
 Basic
research and three phases of clinical
development
DiMasi et al. The Price of Innovation: New Estimates of Drug Development Costs. Journal of Health Economics. 2003; 22:151-185
Drug Development Cost

Out-of-pocket cost allocated over a timeline

Expenditures capitalized at a discount rate to account
for capital invested over time

$403 million out-of –pocket cost per approved drug in
the US in 2003

$802 million fully capitalized cost (discount rate of 11%)

Cancer drugs likely more expensive than other drugs
DiMasi et al. The Price of Innovation: New Estimates of Drug Development Costs. Journal of Health Economics. 2003; 22:151-185
Drug Development Cost

No restriction on the initial price of a drug

“Whatever the market will bear”

Afterwards, price subject to mandated 340b and
Medicaid discounts

Price movement influenced by ASP

“just price” for drugs

OK, but what about everything else?

Your high prices are a problem, mine are OK
Site of Care Shifts
Site of Care Shifts
Cost of Consolidation:
Milliman 2011 & Avalere 2012 Studies


Milliman 2011 study on Medicare costs by site-of-service

$6,500 annualized higher chemo treatment costs in outpatient
hospitals versus MD community cancer clinics

$650 annualized higher out-of-pocket costs for Medicare beneficiaries
Avalere 2012 on private payer costs by site-of-service

Up to 76% higher chemo treatment costs in outpatient hospitals versus
clinics

24% higher on average in outpatient hospitals
Sources:
Site of Service Cost Differences for Medicare Patients Receiving Chemotherapy, Milliman, October 2011
Total Cost of Cancer Care by Site of Service: Physician Office vs Outpatient Hospital, Avalere, March 2012
Site of Care

UNH costs for commercially insured patients
 Chemotherapy
 Inpatient and outpatient facility services
 Physician
 Payments
 Private
services
24%
54%
22%
for amounts for chemotherapy
practice oncologists
Medicare + 22%
 Hospital employed oncologists Medicare + 146%
Site of Care
Payment Amounts for Single Dose of Aloxi
$1,200
$1,000
$800
$600
$400
$200
$0
Lake Norman Oncology
Other Independent
Oncologists
Series 1
Series 2
Hospital and Hospital
Owned Facilities
Series 3
Paying a Premium for Cancer Drugs. Charlotte Observer 2012.
Series 4
Site of Care
Bevacizumab

Examined treatment patterns and cost differences for patients
receiving bevacizumab by site of administration for metastatic
colorectal and lung cancer

Claims database from 2005-2012 for commercial and Medicare
Advantage health plan members

Episode of care longer in physician office setting (OFF) vs hospital
outpatient (HOP) for colorectal cancer patients but not lung
patients

Number of bevacizumab infusions per EOC were greater in
physician office setting for both lung and colorectal patients
Engel-Litz et al. The American Journal of Managed Care. 2014;20(11):e515-522.
Site of Care
Bevacizumab


Cost per month (all cause, patient adjusted) higher in HOP vs OFF

mCRC
38% higher

LC
31% higher
Cost for entire episode of care (all cause, patient adjusted) higher in
HOP vs OFF


mCRC:

HOP $161k

OFF
$134k
LC

HOP $176k

OFF
$118K
Site of Care
Bevacizumab
Actual infusion day cost
$20,000
$18,000
$16,000
$14,000
$12,000
$10,000
$8,000
$6,000
$4,000
$2,000
$0
Colorectal Cancer
Lung Cancer
OFF
HOP
Weekly dose of bevacizumab lower in the HOP setting
Is Fee-for-Service a Cost
Driver?
United Health Care
Episode of Care Pilot

Study predicated on theory that fee-for-service provides
theoretical incentives for overuse and the selection o f
expensive branded drugs

Episode of care payment removed any reimbursement
related to drug selection or treatment vs non-treatment
decision

Primary metric: total medical cost per episode of care
United Health Care
Episode of Care Pilot

Total cost per episode
 Predicted
fee-for-service
 Actual


$98 million
$64 million
Chemotherapy drug cost
 Predicted
$7.5 million
 Actual
$21 million
Paradoxical outcome based on rationale for study
Fee-For- Service

Examined effect of the Medicare Modernization Act on
chemotherapy usage from 2003-2006 in the FFS setting vs integrated
health networks (IHNs)

Lung and colon cancer

Hypothesis:

Decline in use of drugs with lowered reimbursement rates after MMA
were greater in FFS setting than in IHNs

Change in reimbursement did not have a clear impact on
prescribing patterns in FFS

Introduction of new drugs and clinical evidence appeared to play
a role
Hornbrook M et al. Did Changes in Drug Reimbursement After the Medicare Modernization Act Affect Chemotherapy Prescribing? J Clin Oncol 2014;32:1-13.
ASP + 0?

OK, but why would I take the risk and manage the capital
outlay to purchase drug inventory?

Remembered CAP, the Competitive Acquisition Program

Lone vendor at ASP + 4%

Private offices make it work for ASP + 4.3% under sequester

Where are the savings?

IV chemotherapy in the office is more reliable than specialty
pharmacy drug dispensing

Physician offices can compound and infuse complex drugs more
reliably than specialty pharmacy can but pills in a bottle and
print a Fed Ex label
Conclusions

Oncology costs are high and we do need to find
solutions

We need to distinguish between what is not modifiable
(demographics, population aging) and what is

If we are trying to find savings, then paying more for the
same thing (site of care shift) is nonsensical (and
unethical if we also limit patient choices to save money)

If you think drug prices are too high, then don’t double
them
Conclusions



Questionable if we have accomplished any net cost
savings since MMA

Patients have been shifted into structurally higher cost
settings

Drug pricing, which it was hoped MMA would indirectly
address, has not been impacted
To approve a new drug, rigorous scientific proof must be
rendered
Payment system can be changed and new policies
implemented based on weak evidence, bias, and
conjecture
OK
Fire Away…