Effectiveness of a peer-led HIV prevention intervention in secondary schools in Rwanda: results from a non-randomized controlled trial

Table 2 Socio-demographic characteristics of respondents at baseline before and after adjusting for propensity score

Variable	Intervention allocation		P-Value	P-Value
	Control group	Intervention group
	Rwamagana	Bugesera	Before Adjustment*	After Adjustment**
Sex n (%)
Male	417 (47.55)	543 (50.70)	0.17	0.80
Age
Mean (sd)	17.60 (2.30)	18.41 (2.18)	<0.01	0.95
Ever had sex n (%)
Yes	196 (22.55)	302 (28.76)	<0.01	0.91
Condom at last intercourse n (%)
Yes	52 (6.39)	99 (10.25)	0.13	0.98
Never had sex	679 (83.42)	755 (78.16)	0.21	0.94
Intercourse in last 6 months n (%)
Yes	62 (7.09)	118 (11.24)	<0.01	0.98
Ever tested for HIV n (%)
Yes	369 (43.11)	564 (56.29)	<0.01	0.73
Don’t know	17 (1.99)	19 (1.90)	0.51	0.88
Socio-economic status n (%)
Middle	399 (45.65)	349 (32.80)	<0.01	0.63
High	47 (5.38)	38 (3.57)	<0.01	0.11
Live during the year n (%)
Boarding school, off campus	14 (1.64)	209 (20.47)	<0.01	<0.05
Parents/Family	283 (33.06)	430 (42.12)	<0.01	0.90
Baseline HIV knowledge
Mean (sd)	6.12 (1.80)	5.91 (1.90)	<0.05	0.94

*Calculated using simple regression models with the baseline characteristic as the outcome of interest, and the intervention group as the independent variable.
**Calculated using multiple regression models. Observations whose propensity scores fell outside of the minimum and maximum values for the opposite group were eliminated from the analysis. A model was fit with the baseline characteristic as the outcome of interest, and the intervention group as the independent variable. It was adjusted for the propensity score.

ISSN: 1471-2458