. 2020 Sep 25;22(9):e17831. doi: 10.2196/17831

Table 1.

Effect sizes regarding depressive symptoms in the meta-analysis of studies comparing computerized cognitive behavior therapy in adolescents and young adults with active treatment and passive controls at posttreatment.

Variable			N_com^a	N_par^b	Effect size		Heterogeneity		Grade^c	NNT^d
Variable			N_com^a	N_par^b	g	95% CI	I ²	95% CI	Grade^c	NNT^d
All studies
	Active treatment controls		5	403	−0.55	−1.18 to 0.08	87.52	73 to 94	++	3.31
	Passive controls		20	1604	0.52^e	0.33 to 0.71	68.69	50 to 80	++	3.5
		One outlier removed^f	19	1558	0.46^e	0.29 to 0.63	58.49	31 to 75	+++	3.91
Studies aimed at depression
	Active treatment controls		4	351	−0.70	−1.51 to 0.11	90.63	79 to 96	+	2.63
	Passive controls		13	1162	0.60^e	0.35 to 0.85	73.27	54 to 85	++	3.05
		One outlier removed^f	12	1116	0.51^e	0.30 to 0.72	61.76	28 to 80	++	3.55

^aN_com: number of comparisons.

^bN_par: number of participants.

^c+: very low quality; ++: low quality; +++: moderate quality.

^dNNT: number needed to treat.

^eP<.001.

^fOutlier Sethi (2013) excluded.