Online Program

Saturday, February 20
PS3 Poster Session 3 & Continental Breakfast sponsored by Capital One Sat, Feb 20, 8:00 AM - 9:15 AM
Ballroom Foyer

Calculating Power for an Ordinal Outcome in a Longitudinal Cluster-Randomized Clinical Trial Design with Differential Attrition (303248)

Michael L Berbaum, IHRP at UIC 
Molly A Martin, IHRP at UIC 
*Oksana Pugach, IHRP at UIC 
Zhi Yang, IHRP at UIC 

Keywords: power, sample size, longitudinal, cluster-randomized, clinical trial, dental health

The COordinated Oral health Promotion (CO-OP) Chicago Trial is a multi-level trial that aims to reduce early childhood caries in children through individual-, family-, and community-level interventions. The Co-OP Chicago Trial is a cluster-randomized study with several occasions of measurement of child oral health using the caries severity index (CSI). CSI is an ordinal response and requires a three-level ordinal generalized linear mixed model for estimation. Here, we discuss estimation of statistical power for two group comparison over time in longitudinal cluster-randomized clinical trials. The empirical power estimation via Monte Carlo simulation was developed into an R program. The program calculates the effect on power of changing various simulation conditions controlled by the researcher, such as sample size, within-cluster correlation, level of attrition, and smallest meaningful detectable difference. The tabulated results and graphs of the simulation provide a useful guide for a practitioner on a necessary sample size and power in this complex study design.