Describir: Generalized linear mixed model for binary outcomes when covariates are subject to measurement errors and detection limits.