Variability of color doppler imaging parameters and consequences for sample size calculations

Purpose The aims of this study were: 1) to determine the intra‐observer variability of color Doppler imaging (CDI) measurements, 2) to compare the difference in variability of these measurements in healthy controls versus glaucoma patients, and 3) to provide calculations of minimum sample sizes for future CDI studies. Methods Patients with normal tension glaucoma (n=28), primary open angle glaucoma (n=19) and age‐matched healthy controls (n=22)underwent CDI on two occasions one month apart. Variability in CDI measurements was quantified using (within‐subjects) coefficients of variation. Based on this variability, minimum sample sizes were calculated to guide the design of future studies comparing CDI between and within groups. Results In general, within‐subject coefficients of variation for measurements one month apart were comparable to previously reported short‐term variations. Variability was higher in glaucoma patients than in healthy controls. The minimum sample size required for glaucomatous study populations is larger than for healthy controls. Smaller patient groups are required to detect change using the peak systolic velocities than using end diastolic velocities. Studies using a cross‐over design require smaller sample sizes than studies with a pre‐post or parallel design. Conclusion This study provides extensive information on long‐term intra‐observer variability of CDI measurements in healthy controls as well as glaucoma patients. Moreover, sample size calcualtions are provided for studies involving glaucoma patients as well as healthy individuals using three different study designs.