Skip to content

corneliushennch/validate

 
 

Repository files navigation

Comparing effect estimates in randomized trials and observational studies in the same population

ABSTRACT

Introduction

To understand where results from observational studies and randomized trials are comparable, we carried out an observational emulation of a target trial designed to ask similar questions as the VALIDATE randomized trial within the same population as the trial. VALIDATE compared the effect of bivalirudin and heparin on the risk of death, myocardial infraction, and bleeding in individuals undergoing percutaneous coronary intervention across Sweden.

Methods

We specified the protocol of a target trial that was similar to the VALIDATE protocol, then emulated the target trial in the period before the trial took place using data from the SWEDEHEART register; the same register in which the trial was undertaken.

Results

The emulation of the target trial and the VALIDATE trial both estimated no difference in the effect of bivalirudin or heparin on the risk of death or myocardial infarction by 180 days (emulation of target trial risk ratio for death 1.21, 95% confidence interval (CI) 0.88, 1.54; VALIDATE hazard ratio for death 1.05, 95% CI 0.78, 1.41). However, the observational data were not adequate to capture less severe cases of bleeding, which resulted an inability to define a bleeding outcome similar to the trial, or account for intractable confounding early in follow up.

Conclusion

Our findings show that using real world data to emulate a target trial can deliver accurate long term effect estimates in certain situations. However, even with high-dimensional data, it is not always possible to estimate the short term effect of interventions, or the effect on outcomes that cannot be well defined. If registers included detailed information on reasons for making treatment decisions, researchers may be better positioned to identify important potential confounders.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • R 100.0%