I am doing a meta-analysis of the validity of immunotherapy as a prognostic factor. Different studies report different categories for drug timing, making them difficult to combine. For example, if a study reports hazard ratios for 3 categories e.g.
Group1: radiotherapy alone- HR= 1 (reference)
Group 2: radiotherapy after immunotherapy- HR= 1.89 (1.09, 3.33)
Group 3: radiotherapy before immunotherapy- HR=02.03 (1.67, 4.56),
how can I calculate a hazard ratio comparing those treated with radiotherapy alone and with radiotherapy and immunotherapy.
How to combine hazard ratio and 95% CI of group 2 and 3?
+1 for Reginald Quansah's answer.
I would suggest considering performing both random and fixed-effects inverse-variance meta-analyses to pool the data. Dependent on the software you are using, this is relatively straightforward. For example, in Stata you use and install the "admetan" command - I assume other packages have a similar command.
May I direct you to this post by David Fisher on the Statalist forums for further information:
"First of all, we need to enter the data into Stata. The code below is one way of doing it:
Code:
* Example generated by -dataex-. To install: ssc install dataex clear input str14 study float(hr lci uci) "Harris 2015a" .024 .003 .17 "Harris 2015b" .154 .027 .909 "Fristrup 2013a" .6 .39 .93 "Fristrup 2013b" .79 .51 1.25 end
Next, since a hazard ratio is measured on the ratio (exponential) scale, we need to take logarithms before combining in a meta-analysis:
Code:
gen double loghr = ln(hr) gen double loglci = ln(lci) gen double loguci = ln(uci)
Finally, we can run the meta-analysis and (re-) produce the forest plot:
Code:
admetan loghr loglci loguci, hr re study(study) forestplot(xlabel(.01 .1 1 10))"
https://www.statalist.org/forums/forum/general-stata-discussion/general/1453119-meta-analysis-and-forest-plot-for-hazard-ratio
I found to be quite useful, and you can change the xlabel parameters to give a scale which is suitable for your data.
Hope this helps.
Tommy
- Badges
- Science topic