Skip to main content

Table 1 Percentage point difference detectable assuming control group base rate and intra-cluster correlations

From: Continuous quality improvement (CQI) in addiction treatment settings: design and intervention protocol of a group randomized pilot study

Outcomes (1) Intra-cluster correlation (2) Base rate (for control arm) (3) Percentage point difference detectable
Length of stay > 90 days 0.008 50.0% 10.7
Satisfaction:% rating treatment experience as ‘Very Good’ 0.110 52.0% 24.0
% giving counselor a 'Very good’ rating 0.038 78.0% 11.4
Positive compliance 0.002 60.0% 6.6
% weekly or more frequent substance use* 0.048 10.9% 8.3
Employed or in school* 0.150 64.4% 26.0
Positive employment activities* 0.010 16.0% 7.2
Emergency medical care* 0.009 8.2% 4.2
Attended self-help groups* 0.008 75.4% 6.9
  1. * Measured at baseline and follow-up. The power calculations reflect the assumption that the correlation between baseline and follow-up observations of the same measure is 0.7 and that the baseline measure is a covariate in the analytic model.