An application of a pattern-mixture model with multiple imputation for the analysis of longitudinal trials with protocol deviations

Table 5 Combined non-monotone and monotone data: Parameter estimates from de jure and de facto analyses

	Prednisolone			Time			Prednisolone x Time			ART			Prednisolone x ART			Age
Analysis	Est	s.e.	p-value	Est	s.e.	p-value	Est.	s.e.	p-value	Est.	s.e.	p-value	Est.	s.e.	p-value	Est.	s.e.	p-value
MI	-0.61	0.42	0.150	0.51	0.079	<0.0001	-0.004	0.115	0.974	1.50	0.394	<0.0001	0.21	0.581	0.718	-2.51	0.667	<0.0001
LMCF	-0.76	0.446	0.088	0.39	0.045	<0.0001	0.03	0.081	0.707	2.21	0.263	<0.0001	-0.04	0.406	0.924	-2.29	0.799	0.004
J2R (P-)	-0.36	0.486	0.455	0.47	0.047	<0.0001	0.04	0.076	0.561	2.03	0.258	<0.0001	-0.30	0.404	0.457	-2.25	0.779	0.004
J2R (P+)	-0.36	0.512	0.489	.45	0.052	<0.0001	0.03	0.080	0.737	2.06	0.321	<0.0001	-0.16	0.438	0.712	-2.25	0.795	0.001
CDR (P-)	-0.61	0.444	0.168	0.47	0.047	<0.0001	.05	0.083	0.577	2.04	0.259	<0.0001	-0.21	0.379	0.582	-2.27	0.783	0.0004
CDR (P+)	-0.59	0.471	0.216	0.46	0.051	<0.0001	0.02	0.083	0.849	1.93	0.288	<0.0001	-0.03	0.397	0.939	-2.24	0.790	0.005
CR (P-)	-0.64	0.438	0.143	0.47	0.047	<0.0001	0.04	0.075	0.644	2.04	0.258	<0.0001	-0.15	0.376	0.688	-2.29	0.782	0.003
CR (P+)	-0.66	0.451	0.142	0.46	0.050	<0.0001	0.02	0.079	0.792	1.91	0.288	<0.0001	-0.01	0.405	0.979	-2.22	0.792	0.001

ISSN: 1471-2288