CostEffectiveness of health programs Nov. 16, 2000

1
Cost-Effectiveness of health programsNov. 16,
2000
Presented by   Donald S. Shepard, Ph.D. Schneider
Institute for Health Policy Heller School, Room
122A, Mail Stop 035 Brandeis University Waltham,
MA 02454-9110 Tel [phone] Fax
781-736-3965
Web http//ihp.brandeis.edu/shepard E-mail
Shepard_at_Brandeis.edu   Brandeis University Nov.
16, 2000
2
Practical information

• Teaching assistant
• Jsuaya_at_Brandeis.edu
• Administrative assistant
• Linda Purrini, Next to library in Heller
• 781-736-3930
• Purrini_at_Brandeis.edu
• Cost of packet 4.50

3
Cost minimization

• Ancestor of cost-effectiveness
• Goal find the least expensive approach to
  generating the same output.
• Illustrated by the refrigerator costing exercise,
  in which you attempt to minimize the cost of
  keeping vaccines refrigerated.

4
Measuring effectiveness

• Two fundamental policy issues in debating policy
• Prediction what will happen
• Valuation an appropriate measure of output
• Zeckhauser and Shepard (1976) article discusses
  these issues.

5
Von Neumann Morgenstern Utility 1

• Suppose we have a good outcome which has a
  utility of 1 (e.g. living year in perfect health)
• Bad outcome has a utility of 0 (e.g. dying at the
  start of the year)
• Intermediate outcome (I) has utility of x (e.g.
  living with a health limitation), x u(I)

6
Von Neumann Morgenstern Utility 2

• Suppose I is indifferent between probability p of
  good outcome, and 1-p of bad outcome
• The utility is von Neumann-Morgernstern if x
  equals p.

7
Ways of assessing utilities

• Probability approach breakeven probability
• Time tradeoff - breakeven time

8
Time tradeoff

• Time tradeoff - find the breakeven time
• Suppose we were to live 10 years with an impaired
  state, I
• Instead, we could live Y years with perfect
  health.
• Suppose the the value y would make us just
  indifferent.
• Then the utility of I is y / 10.
• E.g., if y is 7, then the utility is 7/10 0.70.

9
Burden of disease
10
Assessing utilities in the conservation sector

• Identify most important attributes, i.e.,
• Assessing awareness about conservation issues
• Assessing behavior of villagers around
  conservation
• Develop a way of measuring each
• Examine the tradeoffs between them

11
Assessing awareness aboutconservation issues

• Interview a sample of villagers, with a
  questionnaire, e.g. Can you identify a
  particular endangered species
• Is it good or bad to cut large trees for
  firewood?
• How can we prevent the cutting of trees?
• Scale identify 5 domains about conservation
  practices, and topics within those domains.
  Perfect score is a correct answer for all items
• Worst wrong score on all items. 100 questions
  (score 0 to 100).

12
Assessing behavior of villagers around
conservation

• Surveying villagers about their self reported
  behavior
• How many trees were planted during the past year
• Observe the number of seedlings
• Forest protecting behaviors
• Scale worst clear all trees in the area
  density is 0
• No harm cut no trees.
• Best Raise the density of trees up to 2.5 meter
  spacing in square pattern
• Best brings density up to 1600 trees in 100 m x
  100 m.
• Tree scale proportion of the maximum possible of
  1600

13
Discounting of health impacts

• Impacts in future years should be discounted,
  just as money is discounted
• Discounting arises due to time preference
• We are anxious to get good things soon.
• If we have to wait, them good is less highly
  valued now.

14
Discount factors
15
Graph of discount factors
16
Table of DALYs
17
DALYs