To the Editor Following the discussion by Serghiou and Goodman,1 a closer examination of the application and reporting of the random-effects meta-analyses by Dr Busse and colleagues2 suggests additional caveats and questions.
Busse and colleagues2 did not explain their primary use of the DerSimonian-Laird method and their secondary use of the Hartung-Knapp-Sidik-Jonkman (HKSJ) method for the sensitivity analysis instead of using the superior HKSJ method throughout.3 Cornell et al4 summarized accumulated evidence that the DerSimonian-Laird method produces biased estimates. The biases would also affect the HKSJ method, which uses the same midpoint as the DerSimonian-Laird approach.
Hoaglin DC. Meta-analysis of Opioids for Chronic Pain. JAMA. 2019;321(19):1935–1936. doi:10.1001/jama.2019.2181
Customize your JAMA Network experience by selecting one or more topics from the list below.
Create a personal account or sign in to: