[R-sig-ME] Meta-regression of left-censored data with the blme package
jpouzou at epixanalytics.com
Mon Aug 21 21:01:10 CEST 2017
Hello Mixed Modelers,
I'm using the blmer function within the blme package with a residual point prior of 1 to perform a random effects meta-analysis. The dependent variable is left-censored (due to a limit of detection), and at the moment I'm interpolating the log-transformed outcome using a linear regression before fitting the meta-regression separately. I'm looking for a way to fit a censored regression rather than separating the two steps, while still using the blmer function. Has anyone had any experience with blmer and censored data?
[[alternative HTML version deleted]]
More information about the R-sig-mixed-models