vlog

[Skip to Navigation]
Sign In
Comment & Response
September 26, 2024

Meta-Analysis Methods for Neoadjuvant Chemoimmunotherapy for Non–Small Cell Lung Cancer—Reply

Author Affiliations
  • 1Rosalind and Morris Goodman Cancer Institute, McGill University, Montréal, Quebec, Canada
  • 2Department of Human Genetics, McGill University, Montréal, Quebec, Canada
  • 3Faculty of Medicine and Health Sciences, McGill University, Montréal, Quebec, Canada
  • 4Department of Surgery, Faculty of Medicine and Health Sciences, McGill University, Montréal, Quebec, Canada
JAMA Oncol. 2024;10(11):1596-1597. doi:10.1001/jamaoncol.2024.4214

In Reply We thank Zhang and colleagues for their appraisal of our work.1 Regarding the first point, it appears that the authors have misinterpreted our statistical analysis. The authors state that our methods included the following sentence, “For each analysis, we included the 95% CIs, I2 statistic, τ2 statistic, and χ2; all tests were 2-sided, and a P value <.05.” In the methods section of our article, the following 2 separate sentences can be found, “For each analysis, we included the 95% CIs, I2 statistic, τ2 statistic, and χ2. All tests were 2-sided, and a P < .05 was considered significant unless otherwise indicated.”1 The first sentence regarding heterogeneity metrics is unrelated to the second sentence highlighting that P < .05 was used to assess significance among the efficacy outcomes. In our protocol (), we prespecified that heterogeneity was going to be assessed by I2; however, as Zhang et al notes, no threshold was set. Applying an a posteriori threshold of I2 greater than or equal to 50% as representing substantial heterogeneity, which aligns with the Cochrane handbook,2 none of the primary analyses conducted had substantial heterogeneity. Therefore, we do not believe that heterogeneity affects the interpretation of these results or requires further exploration.

Add or change institution
×