Medicine Sellers for Prevention and Control of Sexually Transmitted Infections: Effect of a Quasi-Experimental Training Intervention in Bangladesh

Table 3

Effect of training on three major outcome variables based on difference-in-differences (DID) estimation.

Outcome variable	Without covariate			With covariates
	DID estimator (SE)		value	DID estimator (SE)		value
Referred patients with STI symptoms in the last one month	0.114 (0.082)	0.165	0.116	0.052 (0.061)	0.132	0.369
Advised patients with STI symptoms to use condoms	0.121 (0.041)	0.131	0.003	0.093 (0.027)	0.110	0.04
Advised patients with STI symptoms for partner treatment	0.259 (0.035)	0.313	0.001	0.191 (0.041)	0.259	0.009

The difference-in-differences estimator was calculated using linear regression model, adjusted for education, religion and sites (brothel or truck stand), age, and training of the MSs.