Low reproducibility of maximum urinary flow rate determined by portable flowmetry.

To evaluate the reproducibility in maximum urinary flow rate (Qmax) in men with lower urinary tract symptoms (LUTSs) and to determine the number of flows needed to obtain a specified reliability in mean Qmax, 212 patients with LUTSs (mean age, 62 years) referred to the University Hospital Nijmegen, with various degrees of obstruction on pressure-flow studies, used a portable home-based uroflowmeter with 12 disposable beakers. Voided volume and maximum flow rate were recorded continuously during micturition. Flows with voided volumes of at least 100 ml and without possible artifacts were included. All analyses were repeated while excluding flows with voided volumes <150 ml. A coefficient of variation (CV) was calculated for each patient. The CV represents the standard deviation relative to the mean. All individual CVs were subsequently pooled into a population mean CV. This parameter was used to estimate the number of flows required to obtain a mean Qmax with specified reliability for an individual patient. All analyses were repeated, while successively excluding the first, the first two, and the first three flows, to assess a possible learning curve. A total of 1,854 flows was available for analyses, yielding an average of nine flows per patient. Mean Qmax was 13.2 ml/sec; the mean CV was 24%. To allow, for instance, a 10% deviation from the true mean Qmax (e.g., 15 ml/s +/- 1.5 ml/s), approximately 25 flows are necessary. The actual number of flows needed is in fact even higher due to the presence of small and artifactual flows. Using a 150 ml volume cutoff point, somewhat fewer flows are required, but the total number of flows needed (that is, valid, small, and artifactual flows) increases. There was no evidence of a learning curve. The boundaries of a confidence interval around a single Qmax measurement that is likely to contain the true mean Qmax, lie approximately 50% below or above that single Qmax measurement. To reduce this proportion down to 10%, approximately 25 flows are needed. Thus, to obtain reliable mean Qmax values, considerably more flows are required than are normally performed in urologic practice.