Monday, Oct' 8: Modeling in CE analysis

1
Monday, Oct. 8Modeling in CE analysis
by  Donald S. Shepard, Ph.D. Schneider Institute
for Health Policy Heller School, MS 035 Brandeis
University Waltham, MA 02454-9110 USA   Tel
781-736-3975 Fax [phone] Web
http//www.sihp.brandeis.edu/shepard E-mail
Shepard_at_Brandeis.edu
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 6.00

3
Cost-effectiveness of alternative peri-natal AZT
regimens
4
Cost-Effectiveness of Sildenafil, 1
Abstract Background Coverage of sildenafil by
health insurance plans is a contentious issue.
Objective To evaluate the cost-effectiveness
of sildenafil treatment for erectile dysfunction.
Design A Markov decision model to estimate the
incremental cost-effectiveness of sildenafil
compared with no drug therapy.
5
Cost-Effectiveness of Sildenafil, 2
Data Sources Values for the efficacy and safety
of sildenafil and quality-of-life utilities were
obtained from the published medical literature.
Base-case values were chosen to bias against
sildenafil use. Target Population Men 60 years
of age with erectile dysfunction.
6
Cost-Effectiveness of Sildenafil, 3
Time Horizon Lifetime. Perspective Societal
and third-party payer. Intervention Sildenafil
or no treatment in identical hypothetical
cohorts. Outcome Measures Cost per
quality-adjusted life-year (QALY) gained.
7
Markov model for erectile dysfunction
 
8
(No Transcript)
9
(No Transcript)
10
Cost-Effectiveness of Sildenafil, 4
Results of Base-Case Analysis The cost per
QALY gained for sildenafil treatment compared
with no therapy was 11 290 from the societal
perspective and 11 230 from the third-party
payer perspective.
11
Sildenafil, sensitivity analysis
12
Cost-Effectiveness of Sildenafil, 5
Results of Sensitivity Analysis From the
societal perspective, the cost per QALY gained
associated with sildenafil was less than 50,000
if treatment-related morbidity was less than 0.8
per year, mortality was less than 0.55 per year,
treatment was successful in more than 40.2 of
patients, or sildenafil cost less than 244 per
month. The results were sensitive to variation of
erectile dysfunction utilities, but cost per QALY
gained was less than 50,000 if successful
treatment increased utility values by 0.05 or
more on a scale of 0 (death) to 1 (perfect
health).
13
Cost-Effectiveness of Sildenafil, 6
Conclusions In an analysis biased against use
of sildenafil, the cost-effectiveness of
sildenafil treatment compared favorably with that
of accepted therapies for other medical
conditions.
14
Cost-Effectiveness of Dengue Control
Source Shepard, D.S. and Halstead, S.B. Dengue
(with notes on yellow fever and Japanese
encephalitis). In Disease Control Priorities
for Developing Countries, Jamison, D.T., Mosley,
W.H., and Measham, A.R., eds. Bobadilla J.L.,
New York Oxford University Press for the World
Bank, pp. 303-320, 1993.
15
Model for dengue case management
Dengue infections with no interventions (CASES)
 
SHOCK.R
Dengue Hemorrhagic Fever Dengue Shock
Syndrome DHF/DSS
No serious illness
FATAL
Would die without improved case management
SALVAGE
Die despite improved case management
Recover
16
Model for dengue vaccine
Healthy, no infection (STAND.POP)
 
COVERAGE
Receive vaccination
No vaccination
VACC.EF
Dengue infection
No dengue infection
No dengue infection
Dengue infection
CASES
Dengue infection (CASES)
17
Dengue model variables Morbidity and mortality
STAND.POP The number of persons in the standard
population (an arbitrary size) to which the model
is applied. Here STAND.POP is one million persons
of all ages. CASES Number of dengue infections
without vaccination or vector control in the
hypothetical birth cohort (all births within the
standard population in one year). SHOCK.R (Shock
rate) Proportion of dengue infections that
progress to dengue shock syndrome. FATAL
Case-fatality rate of DDS 1960-65. CLINICAL
Proportion of dengue infections which are
clinically apparent. DUR Average duration of
clinical illness, expressed in disability-adjusted
life years. COHORT Number of persons in one
year's birth cohort in the standard population.
18
Dengue model variables Effectiveness
YEAR.D Discounted remaining life expectancy of a
person at the average age of death of a fatal
dengue case. SALVAGE Proportional reduction in
case fatality rate of DHF/DSS after improved case
management. SHORTEN Proportional reduction in
duration of illness among hospitalized cases
after improved case management. VACC.EF (vaccine
efficacy) Proportional reduction in number of
cases. COVERAGE Proportion of birth cohort
vaccinated. VCTRC.EF (vector chemical efficacy)
Proportional reduction in number of cases from
chemical vector control. VCTRE.EF (vector
environmental efficacy) Proportional reduction
in number of cases from environmental vector
control.
19
Dengue modelModel for single interventions
At the baseline (policy BASE) D.BASELINE CASES
? SHOCK.R ? FATAL. With case management improved
(policy C) DEATHS CASES ? SHOCK.R ? FATAL (1
-SALVAGE). With immunization or vaccination
(policy I) DEATHS CASES (1 - VACC.EF ?
COVERAGE) ? SHOCK.R ? FATAL. For each strategy,
the number of deaths averted is D.AVERTED
D.BASELINE - DEATHS.
20
Dengue modelTwo-way combinations of
interventions
With vaccination and case management (policy
IC) DEATHS CASES ( 1 - VACC.EF COVERAGE)
SHOCK.R (1 - SALVAGE) FATAL.
21
Table 2. Efficacy and Costs of Inter-ventions for
Dengue (per 1 million pop.)
22
Table 3. Incremental Cost-Effectiveness of
Interventions for Dengue Control
23
Legend to Dengue CE graphs

• Efficient policies Inefficient policies
•  Note BASE Baseline Efficient policies C
  Improved case management, VC Vector chemically
  controlled and case management, IC Vaccination
  and case management IVC Vaccination, case
  management, and vector chemically controlled, EC
  Environmental vector control and case
  management, ICE Vaccination, case management,
  and environmental vector control Inefficient
  policies V Vector chemically controlled, I
  Immunization, IV Immunization and vector
  chemically controlled, E Environmental vector
  control, IE Immunization and environmental
  control.

24
Figure 4. DALYs saved with vaccine in developed
health system
25
Fig 5. Deaths averted with vaccine in developed
health system
26
Final Assignment of CE module, 1
Purpose See how cost-effectiveness analysis can
be applied to a subject of interest to
you.   Background Select and cite an article or
report of interest to you that describes the
effectiveness or cost-effectiveness of a clinical
intervention or program in a human services
sector. For suggestions or help in identifying
articles, see the instructor or teaching
assistant. For example, the instructors report,
Economic analysis of anti-viral therapy in
Botswana provides data on the cost-effectiveness
of AZT for pregnant women. A copy is available
on the web site under downloads. Two types of
papers are possible.   Alternative A Perform a
preliminary cost-effectiveness analysis. If the
study is an effectiveness study only, perform a
preliminary cost-effectiveness analysis. Clearly
indicate the alternatives you wish to compare and
the effectiveness measure, and provide
illustrative results. Use existing data where
readily available, and explicitly assume values
for other the unknown data. Interpret the
findings. Discuss what kinds of data would be
needed to convert the preliminary study into a
more valid cost-effectiveness study, and where
would they be obtained? Which items do you think
are most critical?  
27
Final Assignment of CE module, 2
Alternative B Discuss a cost-effectiveness
analysis. If the study is a cost-effectivenessstu
dy, discuss the appropriateness of the data,
analysis, and interpretation, and put the results
in a larger context. For example, compare the
intervention under study with other interventions
for the same disease or the same
population.   Submission The length of the paper
should be about 6 double spaced pages, including
any tables or figures per student. Two or more
students may work together and submit a joint
paper that is proportionally more extensive, and
longer if desired (6 pages per student). Please
submit to Linda Purrini, administrative
assistant, Heller G6, tel [phone],
Purrini_at_Brandeis.edu. Also, please include a
copy a summary, abstract, or full copy of the
article with your submission. The paper may also
be submitted electronically to Ms.
Purrini.   Schedule Due Thursday, Oct. 25, 2001
(10 days after the last class).   Evaluation
The module will be graded according to each
students registration (audit, pass-fail, or
letter grade) based on this final exercise, other
exercises, and class participation. In addition,
the instructors plan to give written comments on
this exercise. The instructors look forward to
your papers.
28
Final class session

• Monday, Oct. 15, 6 pm 9 pm

29
Final class, Oct. 15

• Help us identify illustrative, empirical,
  preferably intervention studies outside of the
  health sector
• Collect and read the selected studies
  (distributed later this week)
• Be prepared to discuss the cost-effectiveness
  framework
• What are the alternative(s)?
• What are the costs?
• What are the outcomes (effectiveness)?
• How do we interpret cost-effectiveness results?