Skip to main content
Fig. 1 | BMC Medical Research Methodology

Fig. 1

From: Predicting clinical events using Bayesian multivariate linear mixed models with application to scleroderma

Fig. 1

The latent disease state is ηis that causes observations comprising a KY -vector of biomarkers Yit and a binary KE vector of event indicators Eit at observation times \({t}_i=\left({t}_{i1},{t}_{i2},\dots, {t}_{i{n}_i}\right)\). ηis is affected by interventions and other covariates Xit through regression coefficients βη,X and by zero-mean random effects bi with covariance matrix D. The conditional distribution of the biomarkers given disease state is assumed to be Gaussian with mean Xitβη, X and covariance matrix Ση, Y

Back to article page