Comparing pooled complication rates

Asked Apr 06 '17 at 14:23

Active Apr 06 '17 at 14:23

Viewed 90 times

I have started a meta analysis comparing the complication rates of different treatments of distal radius fractures. I have pooled the complication rates (n/N) for each treatment separately using the inverse variance method, the DerSimonian-Laird estimator for tau^2, logit transformation, Clopper-Pearson confidence interval for individual studies and used a continuity correction of 0.5 in studies with zero cell frequencies. Now I want to compare rates of different treatments with the golden standard of plaster cast fixation. How can I best compare the pooled rates, besides a descriptive comparison? I squared is high and will describe this in my paper as well as the biases in this meta-analysis.

asked Apr 06 '17 at 14:23

Rotterdam

1

Assuming you have another batch of studies on plaster casts then I think you are looking for meta-regression. – mdewey Apr 06 '17 at 17:13
Can I not compare the plaster cast with an other treatment separately with an other method? @mdewey – Rotterdam Apr 06 '17 at 19:51
What do you suggest? – mdewey Apr 06 '17 at 20:22
1

Why not using network meta-analytic methods (to do an indirect comparison given the star-shaped network)? – Giuseppe Biondi-Zoccai Apr 07 '17 at 14:27
A network meta-analysis will be very hard for me, as I have never done it and do not have the statistical computer program. Is meta-regression the best and only other way to compare 2 pooled complication rates with each other? @Joe_74 – Rotterdam Apr 07 '17 at 17:33
@user146922 Provide the dataset as R object and we will provide the suitable netmeta code – Giuseppe Biondi-Zoccai Apr 07 '17 at 21:00
@Joe_74 Where can I send the dataset to? – Rotterdam Apr 11 '17 at 14:44
1

@user146922 If it is not big, upload it on CV as R code/object here – Giuseppe Biondi-Zoccai Apr 11 '17 at 15:11
@Joe_74 Can I send the dataset to an email? – Rotterdam Apr 11 '17 at 19:39
@user146922 I am not sure this is appropriate for CV. You need to verify this. In any case, you can contact an EBM group such as http://www.cardiogroup.org for support – Giuseppe Biondi-Zoccai Apr 12 '17 at 08:14
@Joe_74 I had an approval by CV. Can you send me your email address? It would help me a lot if you could provide me the suitable code. – Rotterdam May 19 '17 at 13:58
I am voting to close this as off-topic since the OP is explicitly asking in the comments for code to run the analysis. – mdewey Aug 24 '18 at 14:04

Comparing pooled complication rates

0 Answers0