Meta-Analysis of Time-to-Event Outcomes
Author Information
Author(s): Siannis F, Barrett J K, Farewell V T, Tierney J F
Primary Institution: Department of Mathematics, University of Athens
Hypothesis
Can percentile ratios provide a better alternative to hazard ratios in meta-analysis of time-to-event outcomes?
Conclusion
The study suggests that percentile ratios can effectively summarize treatment effects in time-to-event outcomes.
Supporting Evidence
- The overall hazard ratio for time to death was found to be 1.18.
- The proposed methodology was illustrated using glioma data.
- Likelihood inference for percentile ratios was outlined.
Takeaway
This study looks at how we can compare survival times from different studies using a new method called percentile ratios instead of the usual hazard ratios.
Methodology
The study proposes a parametric approach for meta-analysis of individual patient data with survival endpoints, using percentile ratios as an alternative to hazard ratios.
Potential Biases
The study acknowledges potential biases in the selection of trials and data extraction.
Limitations
The study could not obtain permission to use data from 3 of the 12 trials.
Participant Demographics
Patients with high-grade glioma from 12 randomized controlled trials.
Statistical Information
P-Value
0.18
Confidence Interval
95% CI 1.09–1.28
Statistical Significance
p=0.18
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website