Table 4.

Results of Subgroup and Sensitivity Analyses for Sleep Onset Latency

Variable	Subgroup	Number of Studies	Point Estimate (min)	95% Confidence Interval	Heterogeneity (I2, %)	Deeks' χ2, P-Value
Age (y)	Children (0 to 18)	2	−16.7	−29.4, −4.0	Negligible (0)	.0008
	Adult (19 to 65)	7	−11.2	−27.7, 5.4	Substantial (84.0)
	Elderly (66 and older)	5	−10.3	−20.1, −0.6	Substantial (79.0)
Dosage (mg)	<1	2	−0.9	−5.4, 3.6	Negligible (0)	NA
	1 to 3	6	−9.6	−17.5, −1.7	Substantial (54.6)
	4 to 5	7	−13.8	−28.9, 1.3	Substantial (88.6)
Study duration (wk)	<1	1	−9.7	−20.5, 1.1	NA	.07
	1–2	5	−7.9	−17.5, 1.6	Negligible (0)
	3–4	8	−13.6	−22.0, −5.1	Substantial (88.9)
Primary diagnosis	Insomnia	12	−7.2	−12.0, −2.4	Substantial (60.5)	<.00001
	Delayed sleep-phase syndrome	2	−38.8	−50.3, −27.3	Negligible (0)
Study design	Parallel	3	−16.8	−29.0, −4.7	Negligible (0)	.008
	Crossover	11	−10.8	−17.9, −3.7	Substantial (85.9)
Jadad quality score	Moderate (Jadad score 2 to 3)	4	−5.4	−11.8, 0.9	Moderate (37.2)	.15
	High (Jadad score 4 to 5)	10	−14.2	−26.6, −1.7	Substantial (85.9)
Allocation concealment	Unclear	11	−10.1	−17.4, −2.8	Substantial (81.7)	.00006
	Adequate	3	−16.4	−24.3, −8.5	Negligible (0)

NA, not applicable