HEA PTP: M212 Economic Evaluation 1

1
Session 7 Defining Assessing Benefits for
Economic Evaluation

• 1. Why, what and how of benefits.
• 2. Benefit assessment for CEA.
• 3. Benefit assessment for CUA.
• 4. Practical exercise in estimating benefits for
  CUA.

2
Why Measure Benefits?
Efficiency Maximise benefits for given
resources
3
Key Features of Economic Evaluation

• Economic evaluation is
• The comparative analysis of alternative
• courses of action in terms of both their
• costs and consequences in order
• to assist policy decisions.
• 1. Costs and consequences - efficiency!
• 2. Comparative - relative efficiency

4
Benefit Categories
5
Should Changes in Productivity be Included?

• May depend upon viewpoint (govt., societal, NHS)
• Main issues are level of true loss and
  comparability
• Measurement of value of loss (gross wage,
  friction cost)
• Double-counting, especially with CUA/CBA
• Comparability with health focus (viewpoint
  again)
• Comparability with other studies (applies to
  other variables also)
• Solution?
• Provide a good reason why they should be
  measured/included
• Report separately from other results
• Differentiate measurement and valuation

6
Should Benefits be Discounted?

• Why not discount?
• Health, unlike resources, cannot be traded over
  time
• Inter-generational equity (cf environmental
  economics)
• If are discounted, may be different rate to cost
• Why discount?
• Inconsistent treatment costs and benefits
• Inconsistent policy, especially in comparison
  with other sectors
• Counter-intuitive conclusions for investment. eg
  always postpone!
• Individuals do trade health over time
  ((dis)invest in health)

7
Negative And Positive Benefits (and Costs!)
C/E ratio net cost/net benefits Net cost
positive cost negative cost Net benefit
positive benefit negative
benefit Negative cost cost saving, eg
reduced LoS Negative benefit reduced health,
eg adverse event
8
Types of Economic Evaluation
Type of Analysis
Result
Consequences
Costs
Dollars
Cost Minimisation
Least cost alternative.
Identical in all respects.
Cost Effectiveness
Different magnitude of a common measure eg., LYs
gained, blood pressure reduction.
Dollars
Cost per unit of consequence eg. cost per LY
gained.
Cost Utility
Single or multiple effects not necessarily
common. Valued as utility eg. QALY
Dollars
Cost per unit of consequence eg. cost per QALY.
Cost Benefit
Dollars
As for CUA but valued in money. eg
willingness-to-pay
Net cost benefit ratio.
9
How Can Health Be Measured?

• Length of life
• Mortality (numbers, rates, SMRs)
• Life expectancy
• Life years lost
• Quality of life
• Numerous QoL measures (generic and specific)
• SF-36, Nottingham Health Profile, Guttman Scale,
  Rotterdam Symptom Checklist, Hospital Anxiety and
  Depression scale etc.

10
Process of Benefit Assessment

• 1. Identification
• 2. Measurement
• 3. Valuation

• Mortality.
• Quality of life.
• Measure in natural physical units (eg. number of
  deaths averted).
• Value benefits if appropriate ie. if performing
  CUA or CBA.

11
Issues in Assessing Benefits for CEA

• 1. Efficacy vs effectiveness vs efficiency.
• 2. Intermediate versus final outcome.
• 3. Sources of data for CEA.

12
Efficacy Vs Effectiveness Vs Efficiency

• Efficacy measure of effect under ideal
  conditions.
• Effectiveness effect under real life
  conditions.
• Efficacy does not imply effectiveness
• Efficiency relationship between costs
  benefits.
• Effectiveness does not imply efficiency

13
Intermediate Vs Final Outcome Measures

• Final change in health (status)
  resulting from the
• programme.
• Intermediate change in clinical indicator
  resulting from the
  
  
  
  programme.
• Need to establish causal link between
• intermediate and final outcome measure.

14
Examples of Intermediate Vs Final Outcomes
Indicators (PBAC (PBS) Oz)
Condition being Final outcome Surrogate Outcome I
ndicators treated indicator Coronary
thrombosis Quality-adjusted Number
surviving Number with specified Number achieving
coronary (thrombolysis survival level of left
ventricular re-perfusion function Stable
angina Quality-adjusted Number with Number who
can walk Number with adequate (various
interventions) survival acceptable a specified
distance relief of pain quality of
life Asthma Quality-adjusted Number
surviving Number with adequate Number achieving a
target (various drugs) survival control of
bronchial level of airways functions hyperreact
ivity Depression Quality-adjusted Number
avoiding Quality of life (may be Number achieving
a target (various drugs) survival suicide improved
by drugs) Hamilton or Montgomery- Asberg
Depression Rating Scale Hypertension Quality-a
djusted Number avoiding Quality of life (may
be Number achieving a target (various
drugs) survival a stroke worsened by
drugs) blood pressure
15
Sources of Effectiveness Data

• 1. Clinical trials, eg RCTs.
• 2. Epidemiological studies, eg cohort studies.
• 3. Synthesis methods, eg meta-analyses.
• 4. Use of modelling.

16
Randomised Controlled Trials

• Gold standard - minimal bias and confounding.
• Disadvantages
• 1. Often establishes efficacy, not effectiveness.
• 2. Selective subjects used.
• 3. Limited opportunity to conduct.
• 4. Limited time horizon.
• 5. Costly to conduct.
• 6. Often unethical and/or unfeasible.

17
Epidemiological Studies

• Real life setting - establish effectiveness
• Disadvantages
• 1. Potential for significant bias and
  confounding.
• 2. Causal link can be weak.

18
Decision Rules CEA

• CEA result CEI (c/e). eg cost per LY gained
• Decision rule adopt lowest CEI
• Application technical efficiency
• Qst addressed Should we undertake program X
  or
• program Y to treat condition A?

19
Limitations of Measurements/Need for Valuation

• Ambiguity in assessing overall improvement or
  detriment in health
• Allocative efficiency - value of benefits gt
  (opportunity) cost

20
Valuation Versus Measurement

• Value is determined by benefits sacrificed
  elsewhere (weighted preference)
• Valuation requires a trade-off between benefits
  - measurement does not

21
Methods of Valuing Health

• Utility or preference assessment
• Quality-Adjusted Life Years (QALYs)
• Variants on QALY - Years of Health Life (YHL),
  Health-Adjusted Person Years (HAPY),
  Health-Adjusted Life expectancy (HALE)
• Healthy-Year Equivalents (HYEs) (based on
  sequence of SG)
• Saved-young-life equivalent (SAVE) (based on PTO)
• Monetary terms eg WTP
• Willingness-to-pay (WTP)
• Human Capital

22
Quality Adjusted Life Years(QALYs)

• Adjusts data on quantity of life years saved to
  reflect a valuation of the quality of those years
• If healthy QALY 1
• If unhealthy QALY lt 1

23
Qol Profile
24
QALY Procedure

• Identify possible health states - cover all
  important and relevant dimensions of QoL
• Derive weights for each state
• Multiply life years (spent in each state) by
  weight for that state

25
Utility Weight

• Utility satisfaction/well-being - reflects a
  consumers (weighted) preferences
• Utility weights are necessarily subjective -
  they elicit an individuals preferences for, or
  value of, one or more health states.
• Must 1. Have interval properties
• 2. Be anchored at death and
• good health

26
Techniques For Measuring Utility

• Variety of techniques available, including
• Time Trade off
• Person Trade Off
• Standard Gamble
• Rating Scale

27
Obtaining Utility Weights

• Two means of obtaining utility weights
• 1. Evaluation specific/holistic measures -
  develop evaluation specific (holistic)
  description of health state and then derive
  weight for that specific state directly by
  population survey
• 2. Use generic or multi-attribute
  instruments - use predetermined weights, based
  on combination of dimensions of health
  yielding a finite number of health states/values

28
Evaluation Specific/holistic Measure

• Advantages 1. Sensitive
• 2. Account for wider QoL (eg
  process, duration, prognosis)
• Disadvantages 1. Cost and time intensive
• 2. Lack of comparability

29
Generic (MAU) Instruments

• Advantages 1. Supply weights off the shelf
• 2. Comparability
• Disadvantages 1. Insensitive to small changes in
  health
• 2. Dimensions may not be
• sufficiently comprehensive
• 3. Weights may not be
• transferable across groups

30
Some Other Issues

• Choosing respondents for utility estimation -
  whose values count?
• What constitutes a correct health state
  description?
• What is the appropriate measurement technique?
• Aggregation of values?
• Biases - ageist, life enhancing versus
  life-saving etc.

31
Decision Rules CUA

• CUA result CEI (c/e). eg cost per QALY gained
• Decision rule adopt lowest CEI
• Application 1. technical efficiency
• 2. possibly allocative efficiency within
• health care sector
• Qst addressed 1. Should we undertake program
  X
• or Y to treat condition Z?
• 2. Should we treat condition A or B?

32
Decision Rules Issues

• 1. Perspective - Health Care Sector
• - Purchaser/Provider
• - Societal
• 2. Comparator
• 3. Budget constraint/indivisibility
• 4. NPV vs BCI
• 5. Limited nature of economic evaluation

33
CUA and Rationing

• Market system - price mechanism establishes
  equilibrium (efficient allocation)
• Non-market system - absence of price as
  allocative tool leads to other, non-price,
  techniques
• Issue is one of (i) philosophical basis for
  rationing and (ii) applied technique for
  rationing

34
Methods of Explicit Rationing
(Coast et al, Priority setting the health care
debate, John Wiley, 1996)
35
Explicit Rationing Technical Methods

• Single principle
• Little distinction between setting priorities at
  different levels
• Examples
• maximising health gain
• need-based rationing
• lotteries
• age-based rationing

36
Technical Method QALY League Tables

• Economic evaluation produces information on
  cost-effectiveness
• If using comparable outcomes (eg QALY) can rank
  according to c/e
• Can use resultant league table to allocate
  resource to most c/e first

37
League Tables Handle With Care!

• Studies show differences in methodology
• choice of discount rate
• method of estimating utility values
• range of costs included
• choice of comparator
• Requires consistent methodology, admission
  criteria for inclusion, applicability in local
  decision context

38
The Oregon Plan

• 1987 - decision to stop funding for organ
  transplantation
• 1989 - Oregon Health Services Commission begins
  work
• 1990 - List 1
• 1991 - List 2
• 1994 - plan begins

39
Oregon List Version 1

• Efficiency principle
• 1600 condition/treatment pairs
• Cost/QALY gained
• social values
• outcome
• cost

40
Oregon List Version 1

• ... looked at the first two pages of that list
  and threw it in the trash can
• ... the presence of numerous flaws, aberrations
  and errors

(Harvey Klevit, member, Oregon Health Services
Commission)
41
Oregon List Version 2

• Equal treatment for equal need
• 709 condition/treatment pairs
• Method
• Development ranking of categories
• Ranking C/T pairs within categories
• Public preferences
• Outcome
• Professional judgement

42
Oregon List Version 2

• Top Five C/T pairs
• 1 Pneumonia - medical
• 2 Tuberculosis - medical
• 3 Peritonitis - medical/surgical
• 4 Foreign body - removal
• 5 Appendicitis - surgical

• Bottom Five C/T pairs
• 705 Aplastic anaemia - medical
• 706 Prolapsed urethral mucosa - surgical
• 707 Central retinal artery occlusion -
  
  
  
  
  paracentesis of aqueous
• 708 Extremely low birth weight, lt 23
  
  
  
  
  weeks - life support
• 709 Anencephaly - life support

43
Summary

• 1. Benefits must be assessed to establish
  efficiency.
• 2. Breadth and depth of benefits measured (
  valued) varies across type of economic
  evaluation.
• 3. Difference between valuation and measurement.
• 4. Debate on role of CUA ( CEA) in allocative
  efficiency
• 5. Beware league tables!