Implementing the DerSimonian and Laird inverse variance method and investigating heterogeneity in meta-analysis

Let’s look at why we would use the DerSimonian and Laird inverse variance method and how it relates to survival data meta-analysis. The DerSimonian and Laird method makes use of HRs and different variability metrics, such as variances and confidence intervals, to create a meta-analysis model based on inverse variances, that is, inverse variability metrics of the HRs. This works by adding weight to each study based on the inverse variances of the studies. In most meta-analysis approaches, if a study has less variability, it will be assigned more weight in the overall result. Finally, the overall HR and overall confidence interval will be calculated and presented using a forest plot that plots the individual studies.

We will be using the PythonMeta package. Load it as PMA:

Note

Make sure to run pip install PythonMeta before running the code.

import PythonMeta...