Table 3.

Mixed effects model for the repeated measures analysis of change in total Positive and Negative Syndrome Scale scores^a.

Test vs reference		Value, N			Adjusted least square means (SE)			Comparison of adjusted least square means (test vs reference)
		Day	Test	Reference	Test	Reference	Difference, test–reference (SE; 90% CI)		P value
PEAR-004 (n=52) vs sham (n=54)
	Analysis 1	29	52	54	−1.6 (0.86)	−2.2 (0.84)	0.61 (1.200; −1.4 to 2.6)		N/Ab
	Analysis 2	57	49	47	−2.5 (0.9)	−3.3 (0.9)	0.80 (1.279; −1.3 to 2.9)		N/A
	Analysis 3	85	48	49	−2.6 (1.14)	−5.3 (1.13)	2.73 (1.611; 0.1 to 5.4)		.09

^aBaseline is the last measurement before treatment administration. Model: change from baseline of efficacy end point data was modeled using a mixed effects model with treatment, visit as fixed effects, baseline and disease duration at baseline as continuous covariates, treatment × visit, and baseline × visit interaction effects. The reported P value is 2-sided.

^bN/A: not applicable.