Skip to main content
. 2011 Jul 17;343:d4065. doi: 10.1136/bmj.d4065

Table 3.

 Comparison of Cohen-Mansfield agitation inventory (CMAI) total score between control and intervention (stepwise protocol for treatment of pain) groups using repeated measures analysis of covariance (ANCOVA)*

Week Mean (SD) CMAI total score Effect of intervention on CMAI total† Intracluster correlation coefficient‡
Control group Intervention group Estimate (95% CI) P value
0 56.2 (16.1), n=177 56.5 (15.2), n=175 0.162
2 53.9 (17.0), n=161 52.0 (19.5), n=158 −3.6 (−0.5 to −6.7) 0.022 0.261
4 52.5 (16.3), n=160 49.4 (19.0), n=148 −4.1 (−0.9 to −7.4) 0.012 0.231
8 52.8 (16.8), n=157 46.9 (18.7), n=147 −7.0 (−3.7 to −10.3) <0.001 0.226
12 52.5 (16.0), n=152 50.3 (20.3), n=142 −3.2 (0.1 to −6.4) 0.058 0.253

*Baseline score as covariate and least squares weighted by number of patients within cluster; P value from multivariate test of intervention was 0.002, and cross effect between week and intervention was <0.001.

†Variable estimate by week of effect of intervention on CMAI score from estimated model.

‡Proportion of total variance between clusters, and measured within framework of ANCOVA.