Replication and Extrapolation from RCTs to Observational Data to Predict Treatment Effects: Why, What, How
Author(s)
Schacht A1, Mulick A2
1Veramed, Frankfurt am Main, Germany, 2Veramed, Twickenham, UK
Presentation Documents
OBJECTIVES: RCT Duplicate has shown that we can replicate the results of an RCT using observational data. If such a replication is possible, it provides trust in the evidence coming from the used observational data, but it doesn’t leverage the strength of observational data. The replication also provides an estimation of bias introduced by moving from trial to observational data. This trust and the estimate of bias helps to extrapolate the findings beyond the RCT used, building on the richness of observational data.
METHODS: Under the assumption that the bias between RCT and observational study remains constant over the variables via which we extrapolate, we can predict treatment effects in 6 different ways.
- Estimations of treatment effect of additional comparators
- Analysis of further endpoints
- studying longer time periods
- investigating other patient populations
- researching further adaptations of the therapy including different doses and frequency of dosing
- understanding the impact of concomitant medication not allowed in the RCT.
RESULTS: We will strengthen the evidence of the extrapolation by understanding the robustness of the analyses. By assuming different models for predicting the treatment effects, using tipping point analyses to measure the impact of bias, or combining both approaches, we will measure the evidence provided. Graphical approaches play a vital role in understanding and communicating these sensitivity analyses.
CONCLUSIONS: Combining replication of existing clinical trials with the strengths of observational data through extrapolation leads to increased trustworthiness of the results from observational studies. Appropriate sensitivity analyses clarify the robustness of the conclusions and help decision makers.
Conference/Value in Health Info
Value in Health, Volume 25, Issue 12S (December 2022)
Code
MSR59
Topic
Clinical Outcomes, Methodological & Statistical Research, Real World Data & Information Systems
Topic Subcategory
Artificial Intelligence, Machine Learning, Predictive Analytics, Comparative Effectiveness or Efficacy, Confounding, Selection Bias Correction, Causal Inference, Reproducibility & Replicability
Disease
No Additional Disease & Conditions/Specialized Treatment Areas