0

As part of a meta-analysis of a disease in R, I received some feedback on the paper of which it is a part by external consultants. I cannot contact them again, and am confused about the advice they have given. This is part of an undergraduate project, and I am relatively new to meta analysis in general. I have been given 4 datasets, namely age, sex(male), traction(Bronchiactis) and Honeycombing. Each dataset contains the following parameters:
Hazard ratio(HR),Lower control limit(LCL), Upper control limit(UCL) and the Standard Errors(SE)[of Ln(HR)] values, and I used the SE values to calculate Variances(var in the datasets). One of the meta analysis on one of the above datasets can be found here, (the dataset along with the code I used and the results of the model.)

The parts of the feedback which confused me were,

1.Conduct a Sensitivity analysis and a Subgroup analysis.

What exactly are sensitivity analysis and Subgroup analysis? Are there any packages in R which will help get these analyses, and are the 4 datasets, containing the parameters I stated above enough for the analyses, or is more data/more parameters needed?

2.In the multivariate analysis, add information about the adjust factor.

I looked this up and could not find any information about this. What exactly is an adjust factor and how can I calculate it?

P.S: Sorry, cannot include the paper here, for copyright reasons.

iFoundNemo
  • 21
  • 2
  • I think this is impossible to answer without knowing what they are really asking. There are many, many ways of doing sensitivity analyses. At a complete guess the subgroup analysis involves using all the datasets together with a four-level factor for dataset as a moderator. I suspect when they say multivariate they mean multivariable. – mdewey Mar 05 '22 at 13:36
  • @mdewey got some clarification, I need to do the sensitivity analysis of publication bias. I ran a standard metafor funnel, and got some results, [here][1]. But how do I proceed from here, to do the sensitivity analysis, for the publication bias? [1]: https://stats.stackexchange.com/questions/566619/doubt-regarding-funnel-plot-in-meta-analysis-using-metafor-in-r – iFoundNemo Mar 05 '22 at 20:24

0 Answers0