Calibration of aerosol output from the Mefar dosimeter: Implications for epidemiological studies

Abstract

Standardized methods for the measurement of airway responsiveness may use the Mefar MB3 inhalation dosimeter to generate standard doses of methacholine aerosol. The manufacturer provides calibrated output data for every nebulizer, so that a standard output may be achieved by varying nebulization time. This output is, however, measured by weight loss (WL), which may over-estimate true aerosol output (AO) because of concomitant evaporation. We have used a chemical (fluoride) tracer method to measure AO directly from two batches of Mefar nebulizers (batch 1 n = 5, batch 2 n = 10) and compared results with manufacturer's quoted WL. Mean AO from batch 1 was 10.56 mg·s-1 (range 9.50-11.63, SD = 0.92 mg·s-1), and mean AO from batch 2 was 5.66 mg·s-1 (range 4.92-6.58, SD = 0.57 mg·s-1), implying that AO varied little within, but substantially between, the two nebulizer batches. Manufacturer's quoted WL does not reflect this near two fold difference: mean WL batch 1 = 14.0 mg·s-1 (range 13-15 mg·s-1); mean WL batch 2 = 11.1 mg·s-1 (range 11-12 mg·s-1). The median aerosol fractions (AO/WL) for batches 1 and 2 were 76% (range 65-83%) and 50% (range 43-60%), respectively. Similar results were obtained with our own measurement of weight loss. This implies that if the median nebulizers of batches 1 and 2 were calibrated (as recommended) by the manufacture's WL to deliver a presumed 100 μg methacholine dose, the actual doses delivered would be 76 μg and 50 μg, the range for all 15 nebulizers being 43-83 μg. We recommend that mean AO is measured for each nebulizer batch (or, where the batch source is not known, each dosimeter), and that measurements of airway responsiveness in multicentre studies are corrected accordingly.