Bayesian joint modelling of longitudinal and time to event data: a methodological review

Table 2 Summary of longitudinal sub-models with multivariate longitudinal outcomes

	Number of articles (%)	Reference
Type of outcome
Continuous	8(36.4%)	[20, 21, 28, 32, 49, 63, 82, 83]
Rate, Ordinal, \ (or/and continuous), Continuous, Ordinal and Discrete^a	5(22.7%)	[14, 26, 34, 50, 84]
Continues and binary	2(9.1%)	[36, 57]
Continuous and ordinal	3(13.6%)	[29, 41, 45]
Continuous, ordinal and binary	4(18.2%)	[30, 85,86,87]
Model
GLM, Partially LME^a	2(9.1%)	[20, 32]
Multivariate GLM	4(18.2%)	[14, 34, 36, 57]
Multivariate mixed effect models	5(22.7%)	[21, 28, 63, 82, 83]
ZAB, Proportional-odds cumulative logit model^a	2(9.1%)	[26, 50]
GLM and CR mixed-effects model, Mixed-effect model and CR mixed-effects model, LME and continuous latent variable model, LME and a mixed-effects beta regression model, ZOIB^a	5(22.7%)	[29, 41, 45, 49, 84]
MLIRT	2(9.1%)	[30, 86]
MLLTM, MLTLM^a	2(9.1%)	[85, 87]
Random effect distribution
Normal	12 (54.5%)	[20, 26, 29, 30, 36, 45, 50, 82, 84,85,86,87]
Multivariate normal	7(31.8%)	[14, 21, 28, 41, 49, 63, 83]
Dirichlet process prior	3(13.7%)	[32, 34, 57]
Error distribution
Normal	12(63.2%)	[20, 29, 30, 36, 41, 45, 49, 57, 82, 83, 85, 87]
Multivariate normal SN	4(21.1%)	[14, 21, 34, 63]
Finite mixture of normal distributions, Multivariate SN, SN/I^a	3(15.7%)	[28, 32, 86]

^aThe order of the outcomes/models/distributions have the same order as in the reference