Extending access to a web-based mental health intervention: who wants more, what happens to use over time, and is it helpful? Results of a concealed, randomized controlled extension study

Table 4 Results of linear mixed effects models for primary and secondary outcomes after 3 months of extended Big White Wall access

Outcome	Extension Group, LS mean (SE) (n = 55)				Discontinuation Group, LS mean (SE) (n = 57)				Interaction term (time by group)			P-value
Outcome	n^a	Baseline^b	n^a	3 months	n^a	Baseline^b	n^a	3 months	v	d	F
Primary Outcome
RAS-r	55	82.0 (1.90)	39	82.0 (2.06)	57	83.3 (1.87)	32	81.0 (2.16)	1	77	1.02	.31
Secondary Outcomes
PHQ-9	55	13.5 (0.86)	36	12.0 (0.97)	57	11.2 (0.85)	32	11.6 (1.0)	1	77	2.76	.10
GAD-7	55	10.0 (0.73)	36	9.2 (0.84)	57	9.8 (0.72)	32	10.2 (0.88)	1	79	1.17	.28
EQ-VAS	55	53.7 (3.0)	36	57.6 (3.6)	57	59.2 (3.0)	30	59.2 (3.9)	1	76	0.52	.47
CIQ	55	13.7 (0.62)	36	14.1 (0.69)	57	13.7 (0.63)	30	14.0 (0.73)	1	69	0.00	.99

All models were adjusted for fixed effects of group, time, and recruitment site with random, subject-specific intercepts to account for repeated measures
LS mean Least squares mean estimated from linear mixed effects models, Numerator (v) and denominator (d) degrees of freedom corresponding with F test statistic, RAS-r Recovery Assessment Scale – revised, PHQ-9 Patient Health Questionnaire – 9 item, GAD-7 Generalized Anxiety Disorder Questionnaire – 7 item, EQ-VAS EuroQOL Visual Analog Scale, CIQ Community Integration Questionnaire
^aAfter missing data
^bRepresents baseline for the extension trial, which was at the time of re-randomization after the first 3 months of intervention access during the main trial

ISSN: 1471-244X