Accuracy and reproducibility of the Assess peak flow meter.

The variability of peak flow measurements was studied in 24 patients with airway obstruction, using 12 Assess peak flow meters and a Fleisch pneumotachograph type 4 as a standard. The Assess peak flow meter gave systematic under-readings of 19-34% (low accuracy). The value scale of the Assess peak flow meter should be adjusted. As systematic under-readings do not influence the variability in peak flow measurements, variability was only caused by the random error of the instrument. The random error of the Assess peak flow meter was relatively low: about 4% of the measured peak flow value (high precision). The contribution of the Assess peak flow meter to the total variability of peak flow measurements showed a linear relationship (r = 0.53) with the value of the peak flow rate itself and varied from 9% (at 3 l.s-1) to 86% (at 10.8 l.s-1). Reading errors of the peak flow meter were responsible for about 1% of the total variability. The remaining variability is probably caused by the correctness of the peak flow performance and the motivation of the patient.