CE Hours: 1.5

Seq# 381 - Moving Ahead with Causal Inference-Exploring Application in Dental Research

This session will be recorded and available through IADR CE On Demand after the meeting for Continuing Education Credit.  

Although the randomized controlled trial (RCT) is widely considered as the ‘gold standard’ experimental method in evaluating bio-medical interventions, alternative approaches are needed when RCTs are deemed unethical, unaffordable or inappropriate. The generalizability of RCT is also limited since participants’ characteristics are often different from the general population. A range of analytical methods for making causal inference from observational data have been used in the fields of epidemiology, public health, health policy, statistics and empirical economics for many years but these approaches have rarely been used in dental research.

This symposium aims to explore and highlight opportunities for using causal inference from observational data in dental research. State-of-the-art methods for causal inference include Propensity Score Matching (PSM), Instrumental Variable (IV), mediation analysis, and Difference-in-Differences (DiD). Microsimulation with Bayesian calibration can be used for better understanding the societal implications of (health) policy interventions. Our session will showcase the application of state-of-the-art methods for causal inference with particular attention on the exploitation of “natural experiments” such as natural disasters or health policy reforms. The session will also provide an example of simulation modelling, which evaluates the impact of policies/interventions yet to be implemented by incorporating future uncertainty via probabilistic approach. The symposium will show the advantages of making causal inference using observational data and their usefulness for evaluating the impacts of complex interventions to society.

Learning Objectives:

  • To provide an overview of state-of-the-art methods for causal inference
  • To highlight suitability of observational (cross-sectional) data for causal inference
  • To show usefulness of exploiting causal inferential methods and microsimulation for understanding the societal implications of (health) policy interventions