Table 1. GLMMs summary for PTSD and depression symptom severity (clinical outcomes).

GLMMs were used to test the predictors of NETfacts (no direct vs. direct involvement), time (baseline vs. post follow up), trauma (number of traumatic events), and their interaction on PTSD and depression symptom severity. Marginal and conditional R² fit, and dispersion test estimates are shown for each outcome. Standardized model estimates, standard errors, standardized confidence intervals, z and p values are shown for the predictor terms. Random terms are represented as variance, standard deviation and sample size. Note that only the count model part of the zero-inflated GLMMs is shown here. Full GLMMs summaries including post hoc Tukey test for PSS-I are shown in S3 and S4 Tables.

PTSD symptom severity (PSS-I)
[Zero-inflated truncated Poisson GLMM; R2 = .20/.99; dispersion = .77, p = 0.192]
Predictor terms	ß	SE	CI	z	p
Intercept	2.19	.21	[1.78: 2.59]	10.60	< 0.001
NETfacts	-.08	.17	[-.41: .25]	-.45	0.65
Time	.00	.12	[-.23: .24]	.02	0.98
Trauma	.06	.18	[-.29: .40]	.33	0.75
Interaction term
NETfacts: time: trauma	-.43	.13	[-.67: -.18]	-3.43	0.001
Random terms	variance	SD	n
Participant	.39	.63	200
Interviewer	.14	.37	17
Depression symptom severity (PHQ-9)
[Zero-inflated truncated Poisson GLMM; R2 = .30/.96; dispersion = .97, p = 0.880]
Predictor terms	ß	SE	CI	z	p
Intercept	1.71	.14	[1.44: 1.98]	12.39	< 0.001
NETfacts	.19	.11	[-.02: .40]	1.74	0.08
Time	.03	.04	[-.05: .11]	.73	0.47
Trauma	.16	.04	[.07: .24]	3.61	< 0.001
Interaction terms
NETfacts: time	-	-	-	-	ns
NETfacts: time: trauma	-	-	-	-	ns
Random terms	variance	SD	n
Participant	.16	.40	199
Interviewer	.07	.26	17

ns: non-significant interaction effect.