Table 3.

Regression Models to Evaluate the Effects of RQP-MH on Patient Retention and Attendance

Coefficient	Retention Analyses		Attendance Analyses
	Four or More Visits During 6-Mo Period After Final Assessment (n = 188)*		At Least 1 Visit Scheduled During 6-Mo Period After Final Assessment (n = 188)*		Visits Attended Out of Total Scheduled During 6-Mo Period After Final Assessment (n = 171)†
	Odds Ratio	95% CI	Odds Ratio	95% CI	Rate Ratio	95% CI
Clinic (reference = comparison/clinic B)	1		1		1
Intervention (clinic A)	2.78	(1.33–5.79)	3.42	(1.02–11.41)	1.29	(1.16–1.43)
Age (yr)	1.01	(0.98–1.04)	1.03	(0.98–1.08)	1	(1.00–1.01)
Sex (reference = male)	1		1		1
Female	0.50	(0.19–1.36)	0.63	(0.13–3.10)	0.94	(0.84–1.05)
Race/ethnicity (reference = white)	1		Perfect Prediction‡		1
Latino	0.55	(0.15–1.96)			1.18	(1.01–1.38)
African American	0.44	(0.05–3.72)			1.04	(0.79–1.36)
Other	0.98	(0.13–7.53)			1.17	(0.92–1.50)
Education (reference = 11 yr or less)	1		1		1
12 yr or more	1.60	(0.76–3.35)	2.28	(0.76–6.85)	1.09	(0.99–1.20)
Veteran (reference)	1		1		1
Novice	0.41	(0.20–0.85)	0.19	(0.05–0.72)	1.03	(0.93–1.13)

^*Analyses included completers only (those with baseline and final assessment interviews); Data on attendance were not available for 6 cases (1 from Clinic A and 5 from Clinic B).

^†Patients without any scheduled appointments during follow-up were excluded from the analyses.

^‡Perfect Prediction as all non-Latino patients, except 1, had at least 1 visit scheduled during the 6-mo period after final assessment.