Synthetic control methodology as a tool for evaluating population-level health interventions

Area of Science:

• Epidemiology
• Health Services Research
• Biostatistics

Background:

• Evaluating public health interventions often relies on observational data when randomized controlled trials are impractical.
• Existing methods like interrupted time series analysis and regression discontinuity have limitations, particularly in selecting appropriate counterfactuals.
• The synthetic control method (SCM) offers a novel approach by constructing a counterfactual from a weighted combination of control units.

Purpose of the Study:

• To explain the synthetic control method (SCM) and its application in health research.
• To outline the advantages, assumptions, and limitations of SCM.
• To demonstrate SCM implementation through a case study on life expectancy post-German reunification.

Main Methods:

• The synthetic control method involves creating a "synthetic" control unit by taking a weighted average of potential control units.
• This synthetic control aims to mimic the trends of the treated unit prior to the intervention.
• The method was applied to a case study examining life expectancy changes in East Germany after reunification.

Main Results:

• SCM is advantageous when dealing with a small number of treated and control units and does not require parallel pre-intervention trends.
• The credibility of SCM hinges on achieving a strong pre-intervention fit between the treated unit and the synthetic control.
• A significant post-intervention outcome difference, given a good pre-intervention fit, can be attributed to the intervention effect.

Conclusions:

• Synthetic control methods represent a valuable tool for evaluating public health interventions where randomization is not feasible.
• SCM is currently underutilized in public health research and warrants broader application.
• Applying SCM in conjunction with other methods can help assess the robustness of findings to underlying assumptions.