Appendix C.

Staff participation outcomes in training activities among the subset of 70 programs that had any participation by intervention status^*

	RS		CEP		Group effect
Variable	Estimate	95% CI	Estimate	95% CI	t	df	p
Primary outcome
Total hours in training, mean	.46	0.17-0.75	2.47	1.1-3.85	4.1	69	<.001
Within type of programs
Healthcare programs	.58	.20-.97	2.01	.25-3.78	2.2	69	.031
Primary care	.31	−.01 to .62	.71	−.05-1.47	1.4	69	.152
Mental health services	.53	−.13 to 1.19	4.08	−1.37-9.52	2.9	69	.005
Substance abuse	.43	−.17-1.03	4.03	1.52-6.53	2.6	69	.011
Social-community programs	.27	.07-.48	3.24	1.93-4.55	5.9	69	<.001
Homeless services	.17	.00-.35	5.06	−1.12-11.23	7.3	69	<.001
Community-based programsa	.26	.08-.43	3.49	1.84-5.15	6.1	69	<.001
Secondary outcomes-collaborative care component
Any CBT Training,b %	7.6	2.69-12.51	12.34	4.96-19.71	1.1	69	.268
Total hours in CBT training, mean	.19	.04-.34	1.08	.28-1.89	3.1	69	.003
Any CM training,c %	4.57	1.41-7.74	19.36	9.66-29.07	3.5	69	<.001
Total hours in CM training, mean	.11	.03-.2	.8	.41-1.18	4.7	69	<.001
Any med training,d %	2.91	−.23 to 6.04	8.45	3.86-13.04	2	69	.053
Total hours in MED training, mean	.02	−.01 to .05	.08	.03-.13	2	69	.053
Any other types of training,e %	6.85	3.62-10.08	11.15	4.79-17.51	1.4	69	.177
Total hours in other types of training, mean	.11	.03-.18	.6	.23-.96	3.6	69	<.001

^*Logistic regression models for binary variables or log-linear regression models for total hours adjusted for community and accounted for the design effect of the cluster randomization; full sample analyses were also controlled for sector.

^asocial services agencies, faith-based, senior centers, parks and recreation, barber shops.

^bcognitive behavioral therapy.

^ccare management.

^dantidepressant medication management and education.

^eadministrator training in team building, grant writing for programs, active listening, “resiliency-class”.