[R-meta] Fwd: Question on Umbrella review and other types
Hi Imran, I think the publication of new studies justifies an updated review, particularly if the existing reviews have inconsistent or inconclusive results. You should, however, avoid to be selective with the studies, let alone selective based on the results! This would be publication bias. What you can do if you suppose there is a time effect is a metaregression with "year published" as a covariate. Best, Gerta UNIVERSIT?TSKLINIKUM FREIBURG Institute for Medical Biometry and Statistics Dr. Gerta R?cker Guest Scientist Stefan-Meier-Stra?e 26 ? 79104 Freiburg gerta.ruecker at uniklinik-freiburg.de https://www.uniklinik-freiburg.de/imbi-en/employees.html?imbiuser=ruecker -----Urspr?ngliche Nachricht----- Von: Im via R-sig-meta-analysis <r-sig-meta-analysis at r-project.org> Gesendet: Montag, 22. April 2024 20:46 An: R Special Interest Group for Meta-Analysis <r-sig-meta-analysis at r-project.org> Cc: Im <aligners79 at gmail.com> Betreff: Re: [R-meta] Fwd: Question on Umbrella review and other types Dear All! Many thanks indeed for your much valuable replies. My research is based on "Prognosis in relation to age in a carcinoma". The study results are "young age patients have better prognosis", "old age have worst prognosis", "age has not affect on prognostication" etc.... Is it possible to be selective in the studies in terms of time of studies published etc., so that there can be some change in the analysis than previous ones ? With such a trivial number (only two) of previous SRMA, it seems better to conduct a new SRMA instead. But I am bit concerned whether it is Ok to conduct a new SRMA when the previously SRMA have been published with a gap of just couple of years (2020 and 2021)? So, if there is no problem in conducting a new SRMA, then I would be add the studies and proceed with the routine "meta" package in R studios for forest plots and funnel plots ? best regards, Imran On Mon, Apr 22, 2024 at 6:35?PM Dr. Gerta R?cker via R-sig-meta-analysis <
r-sig-meta-analysis at r-project.org> wrote:
Yes,I have noted that the word "umbrella" has experienced a change of meaning over the last years. In former times (that is, before network meta-analysis became popular) it meant to conduct a number of pairwise meta-analyses for different treatments for the same condition in the same review. Since this is now better done using network meta-analysis, they now use the term for something different, namely overview of overviews, and this is what the original poster meant. As I understand, in the described situation, a re-analysis is planned, thus I am not sure whether a special R package is needed. Best, Gerta UNIVERSIT?TSKLINIKUM FREIBURG Institute for Medical Biometry and Statistics Dr. Gerta R?cker Guest Scientist Stefan-Meier-Stra?e 26 ? 79104 Freiburg gerta.ruecker at uniklinik-freiburg.de https://www.uniklinik-freiburg.de/imbi-en/employees.html?imbiuser=ruecker -----Urspr?ngliche Nachricht----- Von: Michael Dewey <lists at dewey.myzen.co.uk> Gesendet: Montag, 22. April 2024 17:23 An: R Special Interest Group for Meta-Analysis < r-sig-meta-analysis at r-project.org> Cc: Dr. Gerta R?cker <gerta.ruecker at uniklinik-freiburg.de> Betreff: Re: [R-meta] Fwd: Question on Umbrella review and other types There is a package https://cran.r-project.org/package=metaumbrella on CRAN which performs umbrella reviews. Whether what they mean by umbrella is what you mean is open to question. Michael On 22/04/2024 15:44, Dr. Gerta R?cker via R-sig-meta-analysis wrote:
Hi (whoever you are), If you want to repeat the old meta-analysis anyway, assuming that all
studies adhere to some common well-defined inclusion criteria, then why not conducting a meta-analysis including the union of all studies included in the two older reviews and the new studies? This would simply be an updated review, not necessarily an umbrella review.
Best, Gerta UNIVERSIT?TSKLINIKUM FREIBURG Institute for Medical Biometry and Statistics Dr. Gerta R?cker Guest Scientist Stefan-Meier-Stra?e 26 ? 79104 Freiburg gerta.ruecker at uniklinik-freiburg.de
-----Urspr?ngliche Nachricht----- Von: Im via R-sig-meta-analysis <r-sig-meta-analysis at r-project.org> Gesendet: Montag, 22. April 2024 15:47 An: r-sig-meta-analysis at r-project.org Cc: Im <aligners79 at gmail.com> Betreff: [R-meta] Fwd: Question on Umbrella review and other types ---------- Forwarded message --------- From: Im <aligners79 at gmail.com> Date: Mon, Apr 22, 2024 at 12:56?PM Subject: Question on Umbrella review and other types To: <r-sig-meta-analysis at r-project.org> Hello! I am planning to do an Umbrella review. The number of systematic/meta analysis are two. Rest of the studies are
3-4
new studies published after or almost at same time of these two reviews. My question is can I conduct the Umbrella review ? If not then is there
any
other type of review that could be performed ? I want to include the
previous systematic/meta reviews. I also plan to repeat already performed
meta-analysis. Can I add the meta analysis studies HRs and new studies to
make another meta-analysis ?
Please guide me which review could be the best in this situation.
best regards
[[alternative HTML version deleted]]
_______________________________________________ R-sig-meta-analysis mailing list @ R-sig-meta-analysis at r-project.org To manage your subscription to this mailing list, go to: https://stat.ethz.ch/mailman/listinfo/r-sig-meta-analysis _______________________________________________ R-sig-meta-analysis mailing list @ R-sig-meta-analysis at r-project.org To manage your subscription to this mailing list, go to: https://stat.ethz.ch/mailman/listinfo/r-sig-meta-analysis
-- Michael
_______________________________________________ R-sig-meta-analysis mailing list @ R-sig-meta-analysis at r-project.org To manage your subscription to this mailing list, go to: https://stat.ethz.ch/mailman/listinfo/r-sig-meta-analysis
_______________________________________________ R-sig-meta-analysis mailing list @ R-sig-meta-analysis at r-project.org To manage your subscription to this mailing list, go to: https://stat.ethz.ch/mailman/listinfo/r-sig-meta-analysis