TABLE 2—

Results From Mixed-Effects Regression Models: White Coat, Warm Heart Intervention Trial; Yunnan and Fujian Provinces, China; October 2008–February 2010

	Baseline		6-Month Change From Baseline		12-Month Change From Baseline
Outcome Measures	Estimated Difference (SE)	P	Estimated Difference (SE)	P	Estimated Difference (SE)	P
Prejudicial attitude	0.493 (0.319)	.122	−2.400 (0.220)	< .001	−3.774 (0.267)	<.001
Avoidance intent	0.314 (0.236)	.184	−1.097 (0.174)	< .001	−1.856 (0.208)	<.001
Institutional support	−0.057 (0.279)	.838	0.390 (0.131)	.003	0.817 (0.160)	<.001

Note. Estimated difference = intervention − control. All models included the following covariates: age, gender, occupation, prior contacts with people living with AIDs, province, number of hospital beds (≤ 200, 201–500, > 500), and reported HIV cases in the hospital (none, 1–10, > 10). The group × visit interactions were significant for all 3 measures (P < .001). These models also included clinic-level random effects to account for dependence within clinics and a first-order autoregressive covariance structure to account for repeated observations for each provider.