On accuracy and precision of flowmeters used for oxygen therapy in a veterinary teaching hospital

To determine the accuracy and precision of flowmeters used for oxygen therapy in a veterinary teaching hospital.

An observational study.

A total of 50 flowmeters used for oxygen therapy were evaluated using Defender 530 gas flow analyzers to measure flow. For each flowmeter, a minimum of seven flow settings were tested in random order and in triplicate. Flow measured at ambient conditions was converted to standardized flow specifications (21.1 °C and 760 mmHg) and analyzed using general linear mixed models. Flowmeters were considered accurate at a given flow setting when the targeted mean flow was within the corresponding 95% confidence interval. Precision of flow was characterized based on the magnitude of variance component estimates.

Flowmeters of 1.0, 3.5 and 8.0 L minute^-1 were considered accurate across flow settings corresponding to their capacity range. Flowmeters of 7.0 and 15.0 L minute^-1 were accurate at flow settings ≤2.0 L minute^-1. For flow settings ≥3.0 L minute^-1, average oxygen flow was consistently below reference values. Precision varied with the capacity of the flowmeter, ranked by decreasing precision as 1.0 > 3.5 > 8.0 > 7.0 > 15.0 L minute^-1.

A flowmeter of the smallest maximum capacity within the desired flow range is more appropriate for smaller patients where accurate, precise flow delivery is needed. Although 15.0 L minute^-1 flowmeters were accurate at flow settings ≤2.0 L minute^-1, the graduated increments do not allow exact flow settings <0.5 L minute^-1. Flowmeters of 15 L minute^-1 capacity should be useful for high-flow oxygen delivery for which accuracy and precision are not critical.