Brief psychological therapies for anxiety and depression in primary care: meta-analysis and meta-regression

Table 2 Results of univariate meta-regressions for the 13 studies of cognitive behaviour therapy

Variable	Category 1 (k; N)	Category 2 (k; N)	Regression coefficient	Standard error	95% CI	P*	I ²	Adj R ²
Dichotomous
Diagnosis (model a)	Anxiety = 0 (7; 497)	Depression = 1 (4; 275)	.72	.20	.28-1.16	.005	0%	91%
Diagnosis (model b)	Anxiety = 0 (7; 497)	Mixed anxiety & depression = 1 (2, 479)	.80	.18	.36-1.23	.003	2%	94%
Diagnosis (model c)	Depression = 0 (4; 275)	Mixed anxiety & depression = 1 (2; 479)	.07	.17	-.39-.53	.70	0%	0%
Country	UK = 0 (12; 1214)	Other = 1 (1; 37)	.36	.56	-.87-1.58	1.00	71%	0%
Type of data	Continuous = 0 (9; 931)	Dichotomous = 1 (4; 320)	-.11	.31	-.79-.58	1.00	70%	0%
Continuous	Range ( k ; N )
Number of sessions	3.7-9.8 (13; 1251)		-.03	.08	-.21-.15	1.00	65%	0%
Follow up (weeks)	4-31.6 (13; 1251)		.03	.02	-.003-.07	.46	67%	22%
Year of publication	1982-2004 (13; 1251)		-.02	.02	-.06-.02	.89	59%	16%
Number randomized	20-429 (13; 1251)		.001	.001	-.001-.004	.89	59%	7%
Attrition (intervention group)	0-37% (13; 1251)		-.01	.01	-.03-.02	1.00	67%	0%
Attrition (control group)	0-33% (13; 1251)		-.01	.01	-.04-.03	1.00	67%	0%
Study quality	1-4 (13, 1251)		.16	.22	-.31-.64	1.00	71%	0%

*P-values are adjusted for multiple testing, calculated using the Higgins and Thompson Monte Carlo permutation test (10,000 permutations), except for the diagnostic variables which are unadjusted.
Adj R ² = adjusted R ² (proportion of between-study variance explained by the covariate); k = number of studies; N = total sample size.

ISSN: 1741-7015