Table 4.

Results of meta-regression analysis

	Response (lnBR)				Depression symptoms (SMD)
	constant1	beta2	p	R²	constant1	beta2	p	R²
% early onset	0.34	−0.30	.564	.06	0.43	−0.63	.255	.19
baseline severity	−0.66	0.04	.160	.28	−0.89	0.05	.129	.32
diagnosis (reference dysthymia)	−0.05			.69	−0.18			.69
% chronic MD		1.44	.028			0.88	.252
% recurrent without remission		−1.16	.211			−0.88	.312
% double depression		0.40	.085			0.81	.111
mean age	1.46	−0.03	.463	.09	0.75	−0.02	.760	.01
% female	−0.55	1.13	.373	.13	−0.77	1.43	.372	.12
sessions	−0.13	0.02	.631	.03	−0.27	0.03	.560	.04
treatment duration (weeks)	0.31	−0.01	.521	.06	0.37	−0.02	.137	.19
setting outpatient (vs. inpatient)	0.39	−0.24	.348	.15	0.29	−0.20	.469	.07
type of PT (reference IPT)	0.07			.57	−0.01			.40
CBT		0.15	.555			0.06	.846
CBASP		0.33	.002			0.34	.028
BSP		−0.23	.365			−0.05	.807
type of medication SSRI (vs. others)	0.42	−0.40	≤.001	.72	0.53	−0.54	.003	.64
low risk of bias (vs. medium)	0.28	−0.12	.599	.05	0.21	−0.10	.754	.01
total N in study	0.23	−0.00	.756	.02	0.13	0.00	.980	.00
multiple regression analysis	0.33			0.79	0.38			.78
% chronic MD		0.51	.317			0.54	.237
CBASP		−0.66	.769			−0.01	.969
type of medication SSRI (vs. others)		−0.35	.021			−0.46	.035

¹ unstandardized regression constant (equals treatment effect reference category) ² unstandardized regression weight; positive values indicate superiority of the add-on psychotherapy to medication over medication alone; lnBR = natural logarithm of Benefit Risk; SMD = Standardized Mean Difference; p = level of significance; R² = explained variance; MD = Major Depression; PT = psychotherapy; IPT = Interpersonal Psychotherapy; CBT = Cognitive Behavioural Therapy; CBASP = Cognitive Behavioural Analysis System of Psychotherapy; BSP = Brief Supportive Psychotherapy.