Transcript Cost-Effectiveness Analysis Overview Summary & COURAGE Case

University of Minnesota
Medical Technology Evaluation and
Market Research
Course: MILI/PUBH 6589
Spring Semester, 2013
Stephen T. Parente, Ph.D.
Carlson School of Management, Department of Healthcare Management
Have COURAGE
WSJ Article requires courage
The COURAGE Study as Case Exercise
•
Small Group Exercise
•
•
•
•
•
What are the competing technologies involved in the case?
What regulatory agencies are involved?
What reimbursement policies are involved?
Is comparative effectiveness as private initiative to bend the
cost curve down in jeopardy?
Going forward, if you were DHHS and you want to invest
research $$$, name 2 key objectives of the study?
Estimating the Costs of New
Medical Technologies
Why Measure Costs?
• CE Ratio = (Net change in Costs ) / Change
in QALYs
• Scenario:
– New drug that is prophylactic against heart
disease.
– You want health plans to cover this treatment?
– What are they going to want to know?
Opportunity Costs
• Key Cost Concept: Opportunity cost!
• What are opportunity costs?
• How do we identify these costs?
– Create a flow-chart outlining the events that
occur through the course of an illness.
• The role of ‘time costs’
Flow Chart of Course of Influenza
Influenza Seasons Starts
Person does not get flu shot
Stays well
Person gets flu Shot
Gets Sick
Mild Case
Moderate Case
Severe Case
Misses 2 days of work
Over the counter medicines
Kids get sick
Misses Week of work
Visits PhysicianOver the counter medicines
Kids get sick
Misses 2 weeks of work
Visits physician & ER visit
Over the counter medicines
Kids get sick
Three Steps to Measuring Costs
1. Identify resources used
•
After flowchart is constructed, for each cell
identify the costs associated with that state.
2. Measure the resources used.
•
•
Quantify the identified resources.
Example
3. Place a monetary value on resources used.
Whose Costs are we Measuring?
• Patient—out of pocket costs and inconvenience of
treatment.
• Hospitals providing care out of a fixed budget or
reimbursement system (e.g. Medicare). Increases in
demand from providing service.
• Insurers are concerned about the payments they
make to providers and the costs of processing
claims. Benefits are generally increases in demand
from providing service.
• Social Planner is concerned with all resources.
Direct and Indirect Costs
• Costs are generally divided into two categories: Direct and
Indirect
• Direct costs are those that are directly attributable to the
intervention.
• Drugs, tests, supplies, healthcare personnel and medical
facilities.
• Time costs can go in either the numerator or denominator—
probably best to put in the numerator.
• Productivity costs go into the denominator of the CE ratio
(e.g. QALY)
Costs to Consider
• Physician visits
• Hospital visits — includes
hospitalization and physician time
(billed separately)
• Medications
• Laboratory tests
•
•
•
•
Transportation
Long Term Care
Patient Time
Care Giver Time
Costs to Consider
Health Costs
Patient Costs
Hospitalization
Time in treatment/Care giver
time / Co-pays
Transportation to office / Copays
Ambulatory Care
Medications
OTC or Prescription? Co-pays
Long Term Care
Care giver time / insurance
Micro-Costing and Gross-Costing
• Aggregate costs obtained from the medical
literature are called gross costs.
• An Example
• The process of identifying each resource
used, measuring, valuing and adding up is
called micro-costing.
Using Micro and Gross Costs
• Gross costing is much easier and less time
consuming than micro-costing (e.g. hospital costs)
– Example: Obtaining the cost of a hospitalization for
appendicitis.
• 1st step is reduced to looking up the average costs for
appendicitis.
• Micro-costing should be used when the costeffectiveness analysis is centered on changes in the
way resources are delivered.
– Example from readings
Hospital and Ambulatory Costs
• Usually the biggest component of the cost of
treating a condition.
• Costs are easy to identify, difficult to measure.
• Value of claims data is very high here.
• Other sources of information:
– Health care Cost and Utilization Project (HCUP)—
AHRQ
– Medical Expenditure Panel Survey (MEPS)
– National Ambulatory Medical Care Survey (NAMCS)
Hospital Costs—Using Cost-toCharge ratios
• Charges do not reflect resources use. No one pays
them.
• Often, they are the only information available.
• Can deflate charges by cost-to-charge ratio to get an
approximation.
• Approximate cost = (charge for hospitalization) x
(total reimbursements/total charges)
• When is this ‘good enough’?
Transportation Costs
• Importance will vary with condition
• Transport costs = cost of a gallon x number
of gallons x hourly wage x number of hours
patients spends in transit.
• Example: Breast cancer treatments. Course of
treatment patients can travel 89 hours / 370
miles through course of treatment.
Time Costs
• Time costs are those costs associated with the time a patient
spends receiving a medical intervention— “Time in
treatment costs.”
• Duration of physician contract by disease and physician
specialty is available from the NAMCS.
• Average contact time 19.2 minutes.
• Valuing time costs —opportunity costs.
• Wage is a good proxy.
– Problems with wages.
– Earnings by profession: http://www.bls.gov/ncs/home.htm#data
– The mean physician earns $60 / hr.
Side Effects
• Side effects are valued via micro-costing.
• One medication / treatment may generate many
side effects—need to identify them all.
• Need to identify frequency at which they occur.
• Side effects are typically measured and valued based
on the ambulatory or hospital resources used in
treatment.
Medication Costs
• Medications consumed throughout the
course of an illness usually can be identified
through the medical literature or from clinical
practice guidelines.
• Prices of medicines can be obtained from
Drug Topics Red Book.
• Red Book lists the average wholesale price
for virtually every medication prescribed.
Fixed versus Variable Costs
• Fixed costs—Perhaps 20% of total provider
costs are fixed. Includes property, plant,
DME.
• These expenses are incurred not mater what.
• Variable costs—costs that vary with the
intervention.
• In general, do not include fixed costs in the
analysis—only use variable costs.
Future Costs
•
•
•
•
Identifying future costs can be important.
Perspective matters here (health plan, patient, society, etc.)
Examples.
Discounting—Need to discount future expenditures to make
comparable to current expenditures.
• Restricting cost analysis to a fixed time period may introduce
bias because most expenditures occur near death
T
c(t )
PDV  
t
(
1

r
)
t 1
Future Costs
(5% discount rate)
Future Years
Without
Intervention
With
Intervention
1
$1,000
$1,500
2
$2,000
0
3
$4,000
0
Total
$6,184.5
$1,425
Learning Curves
• Roughly, doubling cumulative output leads to
a 10-30 % reduction in the costs of
production.
• Saywell et al. (1989) reports a 50% reduction
in the costs of heart transplants at a single
hospital over 4 years.
• Adjusting costs measurements to account for
learning.
Learning Curves
• A new treatment has been performed 500 times and
the current unit cost is $10,000.
• There is ongoing learning.
• Doubling output is expected to lead to a 10%
reduction in costs ($1,000).
• One additional treatment (.2%) will lead to a $2
reduction in the cost of each future treatment.
• 200 treatments will be performed each year for the
next 10 years.
Learning Curves (cont)
• True marginal cost = Current production costs –
PDV of the reduction in future production costs.
• In example, 200 x $10,000=$2,000,000 total, naïve
production costs.
• The 501st unit reduces this production cost by .2%
= $4,000.
• So, true incremental cost is $10,000-$4,000 =
$6,000.
• Role of discounting.
Influenza Example
• Bridges, et al. (2000) JAMA
• Studying the cost-effectiveness of influenza
vaccines.
• Question: What are the costs and benefits of
administering flu shots?
Study Background
• N=1,200 for two years
• US manufacturing firm
• Direct Costs:
–
–
–
–
Flu shot ($24.70)
Drugs
$ spent on MD visits
$ spent on Hospitalizations
• Indirect Costs
– Hours lost for Sick days
– Hours lost for MD visits
Influenza Outcomes
Costs 1997-1998
Costs 1998-1999
Summary 1998-1999
Category
Unit Price
Cost/Person
Treatment
Cost/Person
Placebo
MD Visit
34.39
1.71
2.95
MD co-pay
10.00
.50
.86
Drugs
49.38
2.21
3.32
Drug co-pay
12.40
.55
.83
OTC
--
1.25
1.75
Hospitalization
s
7790
0.00
0.00
Lost workdays
235.1
19.40
28.43
Lost work
hours for MD
visits
29.39
1.11
2.12
Vaccination
24.70
24.70
0.00
Total
---
51.43
40.26