Hello, im relatively new to stata so please forgive me if there is something obvious im missing. Im conducting a meta-analysis of diagnostic test accuracy, I decided to use the more recent metadta, and there is handy help and FAQs, as well as the 2022 publication with the example code available online.
Despite all of these things, even if I use the code as advised, I get lots of errors.
The code is more complicated/nuanced than ive previously been used to.
My data is in 2x2 format.
Midas, meta and metandi commands seem much more simple. I have been using some midas commands as well as metadta. As far as I understand, I can't use the meta package for a dta review.
A few issues I have been encountering below:
1. Using metadta I can't seem to get my forest plots to look the way I want to and haven't found clear advice in the resources on how to either.
My plots display as sens/spec with the 95CI displayed on the right of the plot. Id like the study names to be listed on the left of the plot. Ive used a numeric id for the analysis as this is required, then converted to a nominal list of studies. In the graph editor I can't find a way to re-label after the graph displays.
2. The I2 statistic is not displayed on the plot as is seen when using other packages, and useful to have there, and I can't figure out how to display just sens or just spec on a plot, as spec is more relevant to my research question.
3. To aid sub-group analysis, and because I have the available IPD, I would like to display several forest plots together, by sub sets of participants. This might be where im getting confused/not using the package correctly as I want to compare the data across these different groups individually rather than using meta-regression, which I understand will need to be with study level covariates.
4. The last question is about whether it is appropriate to use the ipdmetan package or similar for AD, adjusting for patient level covariates, or to use these via the metadta command. I am not conducting an IPD review, but I have IPD openly provided by studies which will aid in the interpretation of the aggregate data, and (partly due to the characteristics of studies included), I dont think meta-regression using study level covariates provides enough information.
Thanks in advance!
Ffion
Related Posts with Trouble with metadta for diagnostic test accuracy systematic review
If condition with loopsHi, I would like to fill out the out of sample stock returns based on factor values (LogSize DY MOM…
Generating Mean Values for a variable based on 2 conditions in Panel DataHi, I have panel data of150+ companies for the years 2001-19. I have divided these years between pr…
MI for MAR Panel DataHello, I am hoping someone might be able to point me to code that would allow me to perform multipl…
Hosmer Lemeshow test for large dataHello, I am running logistic regression about 29k sample. How can I test the goodness of fit of the…
How to only keep the target icd code and its diagnose date for each obs?Hi, Below is part of my dataset. What I want to do is that for each id (row), I want to only keep …
Subscribe to:
Post Comments (Atom)
0 Response to Trouble with metadta for diagnostic test accuracy systematic review
Post a Comment