POWER DETERMINATION FOR GEOGRAPHICALLY CLUSTERED DATA USING GENERALIZED ESTIMATING EQUATIONS

Study designs in public health research often require the estimation o f intervention effects that have been applied to a cluster of subjects in a common geographic area, rather than randomly assigned to individ ual subjects, and where the outcome is dichotomous. Statistical method s that account for the intracluster correlation of measurements must b e used or the standard errors of regression coefficients will be under estimated. Generalized estimating equations (GEE) can be used to accou nt for this correlation, although there are no straightforward methods to determine sample-size requirements for adequate power. A simulatio n study was performed to calculate power in a GEE model for a proposed study of the effect of an intervention, designed to reduce lower-back injuries among nursing personnel employed in nursing homes. Nursing h omes will be randomly assigned to either an intervention or control gr oup and all employees within a nursing home will be treated alike. His torical injury data indicates that the baseline-injury risk for each h ome can be reasonably modelled using a beta distribution. It is assume d that the risk for any individual nurse within a nursing home follows a Bernoulli probability distribution expressed as a legit function of fixed covariates, which have values of odds ratios determined from pr evious studies which represent characteristics of the study population , and a random-intercept term which is specific for each home. Results indicate that failure to account for intracluster correlation can lea d to overestimates of power as well as inflation of type I error by as much as 20 per cent. Although the GEE method accounted for the intrac luster correlation when present, estimates of the intracluster correla tion were negatively biased when no intracluster correlation was prese nt. In addition, and possibly related to the negatively biased estimat es of intracluster correlation, we also found inflated type I error es timates from the GEE method.