A MICROSIMULATION MODEL

1

• A MICROSIMULATION MODEL
• TO EVALUATE HYPOTHETICAL
• DRUG PLANS
• ___________________________________________
• Loan Nguyen and Anil Gupta
• Microsimulation Modelling and Data Analysis
  Division
• Applied Research and Analysis Directorate
• Health Canada

2
Background

• Canadas drug expenditure is the second largest
  component of its health care costs
• Average annual growth of 9.7 from 1985 to 2003
• Many Canadians with high drug cost and low income
  are reported at risk of stopping their drug
  treatment
• Pressure on the federal/provincial governments to
  provide relief to those individuals/families

3
Model Objectives

• Estimate drug benefits for families/ individuals
  and the government expenditure under the existing
  and new hypothetical drug and tax policies.
• Assess the distributional impacts on households
  and Federal/Provincial expenditure associated
  with a a suite of possible proposal changes
• Estimate the Federal/provincial income taxes and
  government transfers and credits
• Assess the interaction between the drug and tax
  policies due to any changes in these two policies
  

4
Data Requirements

• Model requires individual-level information on
• Socio-demographic characteristics
• Incomes
• Tax deductions/credits
• Drug coverage
• Prescription drugs dispensed
• Health related drug expenditures.

5
Data Sources and Synthesis
SLID - Individual File
Match Health Related Drug Expenditures to SLID
using Statistical Matching
SHS
Synthesized Database
SLID - Census Family File
Simulate information on Tax deductions /
credits using Monte Carlo technique
TAX Data
SLID Economic Family File
Growth Database
Impute Drug coverage Information For Families and
Individuals using Logistic Regression
Validation
NPHS
Simulation Model
Impute Drug Cost and Utilization For
Families /Individuals Using Monte Carlo
Technique
DRUG Data
6
Microsimulation Pharmasim Model
Family Unit
Micro Database
Current Drug Policies
Hypothetical Drug Policies
Option System
Base System
Hypothetical Tax Policies
Current Tax Policies

• Tax Module Calculation
• OAS/GIS, SA contribution
• Net Income, Taxable Income

Exit
Output Data

• Pharmacare Module Calculation
• Estimate Benefits for a family
• unit from individual plans

• Pharmacare Module Calculation
• Estimate Benefits for a family
• unit from family plan

Yes
Acceptable
Select the coverage that optimizes benefits for
the family
Option System Output
Benefits from the Drug Programs
Impact Analysis System
Base System Output

• Tax module Calculation
• Government Transfers
• Tax Credits

7
Application

• An example of income-based national drug coverage
  (a hypothetical plan) will be used to illustrate
  some of the functions in the model.
• Example
• Deductible
• Co-pay after reaching deductible
• Pay only to a maximum
• The above parameters will change with the ability
  to pay

8
Application A Hypothetical National Drug
Coverage Plan
9
Application - Questions

• 1) What are the total Provincial costs for
  this hypothetical drug coverage plan compared to
  the existing provincial drug coverage?
• 2) What are the impacts on the households
  as a consequence of implementing such a
  hypothetical plan, including winner/loser
  analysis ?

10
Application - Results

• Compared to the provincial drug plan, the
  hypothetical drug plan
• has 2.9 higher total Federal/Provincial drug
  expenditures
• covers 15.9 more individuals or 12.8 more
  families
• has lower average out-of pocket cost for families
  with income under 36,001 and higher for families
  with income over 36,000
• covers more families/individuals with low income
  (under 36,000) and fewer families/ individuals
  with higher income (over 36,000)
• has lower average OOP cost for families with
  total drug costs less than 100 and greater 750

11
Conclusions /Potential Developments

• Our current model
• Contains a reliable and representative data
  source
• A highly parameterized and robust policy tool
• Very flexible - can simulate a variety of health
  and tax policy options
• Useful - can asses the impacts on
  federal/provincial revenues and
  individual/household drug benefits, including
  winner/loser analysis
• Future work
• Develop the model for simulating drug benefits
  and costs for special diseases
• Incorporate health status in the comprehensive
  model data to estimate health and disease related
  future costs