Estimating the Power of Indirect Comparisons: A Simulation Study
Author Information
Author(s): Edward J. Mills, Isabella Ghement, Christopher O'Regan, Kristian Thorlund
Primary Institution: University of Ottawa
Hypothesis
How fragile are confidence interval estimations and hypothesis testing relying on indirect comparisons?
Conclusion
Indirect comparisons typically suffer from low power, especially when the comparisons are unbalanced.
Supporting Evidence
- Indirect comparisons are increasingly used in medical treatment evaluations.
- Confidence interval estimation often suffers from under-coverage.
- Low power in hypothesis testing is common with indirect comparisons.
Takeaway
This study shows that comparing treatments indirectly can be tricky and often doesn't give strong results, especially if there aren't enough trials.
Methodology
A simulation study was conducted to examine the performance of indirect comparisons in estimating treatment effects.
Potential Biases
Potential biases may arise from individual trial differences and the conduct of the indirect model.
Limitations
The study used simulations instead of real data, which may not fully capture the complexities of actual trials.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website