Introduction to the Theory and Practice of Economic Evaluation in Health Care: empirical methods, September 15-16, 2022


Introduction to the Theory and Practice of Economic Evaluation in Health Care: empirical methods

September 15-16, 2022

Finnish Society for Health Economics will organize the course “Introduction to the theory and practice of economic evaluation in health care”, concentrating on empirical methods applied in health economic evaluation research. The course covers experimental and quasi-experimental methods for policy evaluation, such as difference-in-differences, regression discontinuity, propensity score matching and instrumental variable methods. It is offered to PhD/Master students in health economics but practitioners from health care (e.g. doctors, decision-makers) can also participate. The course is a continuation to the first part of the course organized in spring 2020 covering the theory and principles of economic evaluation. The course is provided free of charge. Lectures will be given in English.

Professor Heather Brown from Lancaster University UK will teach the course. The lectures will be given at

Finnish Institute for Health and Welfare (THL)

Mannerheimintie 166

00271 Helsinki


The time schedule is as follows:

DAY 1: Thursday, September 15, 2022, 9:00-16:00 (incl. 1 hour lunch break; lunch is not paid by the Finnish Society for Health Economics), J-Auditorium, Rooms 1-3, THL

Session 1: Introduction to policy evaluation (60 minutes)

  • Brief overview for market failures in health (and how this relates to policy evaluation)
  • Why do we need an economic perspective in health
  • What is policy evaluation?
  • RCTs and why they may not always be feasible
  • How does policy evaluation relate to economic evaluation?
  • What is public health economics and how can it help you in your work.

Group Activity 1:  Give an example of when you would use economic evaluation and when you would use another type of policy evaluation method? (20 minutes)

This session introduces students to policy evaluation methods and economic evaluation to show how they are different but complementary and some of the challenges associated with evaluating policy. 

Session 2: An introduction to econometrics for public health (60 minutes)

  • Why is it important to be able to correctly interpret statistical findings from public health studies in the media and academic journals?
  • 5 Steps for correctly interpreting statistical studies of public health

            1) Understand the claim

            2) Correlation vs causation

            3) Averages and Medians

            4) Putting the results in perspective

            5) Beware statistical significance

This session maps how evidence is used to assess the effectiveness of population health and well-being.  Errors in interpretation can impact on patient safety.  The importance of effective communication of statistical results is also emphasised in this session.  Honest and integrity of presenting statistical results is emphasised. 

Group Activity 2: How would you evaluate the impact of the sugar tax on childhood obesity (20 minutes)

Session 3: Propensity Score Matching (60 minutes) *Note we will start this and take a break for lunch and come back

  • What is it and how does it work?
  • Identifying an appropriate comparison group
  • 4 different matching methods
    • Nearest Neighbour
    • Calipher Matching
    • Kernel Based Matching
    • Mahalanobis Matching
  • Interpreting the Coefficients
  • Example: Impact of training programmes on wages

Group Activity 3: Trying an example in STATA-different ways of matching and how you can interpret the results across different matching methods (50 minutes)

Session 4: Regression Discontinuity (RD) Approach  (30 minutes)

  • What is it?
  • Types of RD
    • Fuzzy
    • Sharp
  • A simplified example: Relationship between mental health and a change in the retirement age
  • Defining an outcome measure
  • Graphically representing the data
  • Defining the bandwidths
  • Steps for estimation
  • Generalisability of the results

Group Activity 4: Can you think of any examples of when you would use  a Sharp RD design (15 minutes)

Activity:  Trying an example with STATA  (60 minutes)

DAY 2: Friday, September 16, 2022, 9:00-12:00 (3 hours), J-Auditorium, Rooms 1-3, THL

Session 5: Interrupted Time Series (45 minutes)

  • What is it?
  • How does it differ to RD
  • When to use?
  • Simple example of evaluating the sugar tax on number of dental caries in 7-10 year olds.
  • Some examples using administrative data

Group Activity 5: What is the difference between RD and ITS? (15 minutes)

  • Give an example of when you would use RD and when you would ITS?
  • How would you convince your team one method is preferred over the other?

Activity: Trying an example in R  (30 minutes)

Session 6: Difference-in-Difference Approach (45 minutes)

  • Why use this approach
  • How to find an appropriate control group
  • What types of data can be used with this estimation approach
  • Two examples using DiD
  • Robustness Checks
  • Some extensions of the basic DiD framework

Activity: Trying an example with STATA (40 minutes)

Sessions 3, 4, 5, and 6 outline different quasi-experimental methods for evaluating policy and public health interventions.  Practical examples with code to be used in STATA and R are provided to give students hands on experience of using each method and provide a starting point for using the methods in their own research. 

Session 6: Conclusion: Bringing it all together (20 minutes)

You can register to the course from here: