Table 2.

Unadjusted and adjusted outcome parameters at the end of the intervention period

	Mean unadjusted, imputed values with SD		Mean adjusted, imputed values with 95%-CI
	BPT (n = 79)	CBT (n = 77)	BPT (n = 79)	CBT (n = 77)	AMD
Effect parameter
MADRS	15.4 (9.1)	18.1 (10.4)	15.8 [13.5; 18.1]	18.4 [16.1; 20.8]	-2.58 [−5.52; 0.25]
PHQ-9	9.0 (5.5)	10.4 (6.0)	9.4 [8.0; 10.7]	10.7 [9.3; 12.1]	-1.35 [−2.97; 0.40]
Cost parameter in €
Total costs	4624 (4793)	4199 (4395)	6019 [4255; 8629]	5266 [3684; 7972]	754 [− 1280; 2737]
Indirect costs	3726 (4705)	3418 (4396)	3410 [2493;4386]	4154 [3034; 5340]	− 744; [− 2247;660]
Direct costs	640 (563)	548 (591)	769 [581;1017]	510 [381;544]	259 [−3;544]
Intervention cost	257 (95)	233 (86)

Adjustment for baseline value, sex, depression severity, study site and wave

Effect estimates based on Generalized Linear Model with Gaussian-distribution,

Cost estimates for total and direct costs based on one-part gamma-distributed model, cost estimates for indirect costs based on two-part model with gamma distributed Generalized Linear Model as second part

Cost for inpatient treatment are included in the adjustment variables baseline direct costs and baseline total costs but did per definition not incur during the study period, hence adjusted costs deviate substantially from unadjusted costs

AMD Adjusted mean difference, CI Confidence interval, M Mean, MADRS Montgomery-Asberg Depression Rating Scale, PHQ-9 = Patient Health Questionnaire, SD Standard deviation