Diapositiva 1

1
Development and Validation of a Microsimulation
Economic Model to Evaluate the Disease Burden
Associated with Smoking in Argentina Phase II,
Extravascular Pathology and Extrapulmonary
Tumors Andres Pichon-Riviere MD MSc Institute
for Clinical Effectiveness and Health Policy
(IECS Argentina)
PRELIMINARY RESULTS Calibration of the model was
done by comparing model results with key
parameters extracted from Argentinas vital
statistics. Validation was performed with 25
cohorts simulations of 50,000 subjects in each
age category (40, 50, 60 and 70 years). In figure
1 and 2 we schematically show the model
validation process. Validation showed an
excellent external validity (/-5 of predicted
values). Table 1 and Graphic 1 show the main
results.
METHODS (cont.) The main advantages were the
possibility of having memory of each subjects
history (as opposed to the memorylessness of
Markov models) incorporating the possibility of
simultaneously experiencing different health
states during follow-up (which would render a
Markov model nearly unmanageable) and being
intrinsically probabilistic. A spreadsheet-based
model using Microsoft Excel was programmed using
Visual Basic based Macro functions in order to
simulate individual subjects experiences. The
original model calculates the rate of occurrence
of five medical conditions coronary heart
disease (myocardial infarction and unstable
angina), non-ischemic cardiovascular disease,
cerebrovascular disease, chronic obstructive
pulmonary disease and lung cancer. As part of
this project we are working to extend this
original model in order to include extravascular
pathology and extrapulmonary tumors. To populate
the model it was necessary to obtain the absolute
risk of occurrence of events for the non-smoking
subpopulation. Once the baseline risk is derived,
the absolute risk for smokers and former smokers
could be determined, using the known relative
risks. Absolute risks in non-smokers were
estimated using local vital statistical data and
epidemiological data, and using international
figures when data were not available (Formula
1). For the estimation of the probability of a
diagnosis of cancer conditions, once the absolute
risk of death in the non smokers' population was
obtained from the previous formula, this
information was utilized to estimate the
probability of a diagnosis of lung cancer
(Formula 2)

• BACKGROUND
• Tobacco control policies have been poor in our
  region. A contributing factor is the lack of
  local research on the health and economic
  consequences of tobacco, and an inappropriate
  belief of policy makers that tobacco control
  interventions are less urgent than action on
  other diseases.
• Although international evidence about disease
  burden of tobacco is extensive, it is well known
  that health economic studies results are not
  directly transferable from one country to
  another. There is still a need to more precisely
  understand the disease burden associated with
  tobacco measured in terms of clinical
  consequences (illnesses related to smoking), in
  years and quality of life as well as economic
  aspects to inform more efficiently tobacco
  control policies.
• The origin of this project is a multi-country
  collaboration among seven countries in Latin
  America with the aim of explore each countrys
  context (decision makers attitudes, availability
  and quality of epidemiological data) in order to
  select the most suitable methodological framework
  and elaborate a common health economic evaluation
  model to be applied in the participant countries
  as well as in other settings where information is
  scarce (Research Grants from INCLEN Trust and
  IC-Health).

Graphic 1 . Model Validation Mean differences
between Predicted and Expected death rates. Mean
differences are expressed as percentage over
expected rates.
Table 1. Model Validation Predicted vs Expected
(Local vital statistics) death rates 
OBJECTIVEThe aim of this project is to provide
relevant data on the disease burden associated
with smoking in Argentina, including effects on
health and economic impact.
EXPECTED OUTCOMES The study will provide
information on the impact of tobacco use on
health from epidemiological, social and economic
perspectives in relation to the costs incurred in
all health sub-sectors, for the diseases and
conditions related to the smoking habit. The
results will be presented as consequences of
smoking on Mortality, Disability-adjusted life
years (DALYs), Quality-adjusted life years
(QALYs), Number of events and Total costs. The
data produced will serve to hasten the
implementation of programs and effective policies
for monitoring and reducing tobacco use. They
will also serve as input for other researchers in
Argentina to objectively assess the impact and
the potential health and economic benefits of the
measures that promote a reduction of smoking.
These results can be essential strategic inputs
in convincing the politicians and decision-makers
in Argentina to finally ratify the FCTC.
METHODS After performing a systematic review of
existing health economic models of tobacco we
decided to build a microsimulation based model.
This model incorporates the natural history,
costs and quality of life of the most important
tobacco-related diseases.
Contact Andres Pichon-Riviere MD MSc,
apichon_at_iecs.org.ar - Institute for Clinical
Efectiveness and Health Policy (IECS
Argentina), www.iecs.org.ar