Promoting STI testing among senior vocational students in Rotterdam, the Netherlands: effects of a cluster randomized study

Table 6 Results of multilevel analysis to assess intervention effects on STI testing: imputed data for missing cases at follow-up (n = 822)

	Model 1: random intercept for class, with individual variables and baseline		Model 2: Model 1 with intervention conditions
Measures of variation or clustering:
Class level variance	1.29			0.82
ICC Class level	0.28			0.20
Individual level variables:
	Beta (SE)	OR (95%CI)	p-value	Beta	OR (95%CI)	p-value
Sex (female vs male)	0.90 (0.33)	2.46 (1.29-4.70)	0.01	0.70 (0.32)	2.02 (1.08-3.80)	0.03
Ethnicity (Dutch vs non-Dutch)	0.26 (0.29)	1.29 (0.74-2.26)	0.37	0.33 (0.28)	1.39 (0.80-2.42)	0.24
Baseline test behaviour	0.90 (0.39)	2.46 (1.15-5.28)	0.02	-0.88 (0.39)	2.43 (1.13-5.22)	0.02
Intervention group 1^a				1.29 (0.54)	3.62 (1.26-10.40)	0.02
Intervention group 2^b				-0.26 (0.66)	0.77 (0.21-2.80)	0.69
Intervention group 3^c				0.71 (0.52)	2.04 (0.74-5.64)	0.17

^a = educational intervention and sexual health service, ^b = educational intervention, ^c = sexual health service

ISSN: 1471-2458