Effect of exhalation flow rates and level of nitric oxide output on accuracy of linear approximation of pulmonary nitric oxide dynamics

The method of Tsoukias and George (T and G) is a commonly used linear approximation of pulmonary nitric oxide (NO) dynamics that can be used to calculate bronchial NO output (JawNO) and alveolar NO concentration (CANO). We aimed to investigate how flow rate range in exhaled NO measurements and levels of pulmonary NO parameters affect the accuracy of the T and G method. This study has three parts. (a) A theoretical part demonstrating how different exhalation flow rates and NO parameter levels affect the accuracy of the T and G method, (b) testing how exhalation flow rate range affects the method in a sample of asthmatic and healthy subjects, and (c) a meta-analysis of published literature to test whether minimum flow rate has an association with the NO parameter values. We found that both the chosen exhalation flow rates and magnitude of the pulmonary NO parameters affect the accuracy of the T and G method. Underestimation of JawNO increased with lower flow rates and higher bronchial diffusion factor of NO (DawNO), while overestimation of CANO increased with higher DawNO and bronchial wall NO concentration (CawNO) and lower CANO. Of the NO parameters, CANO was the most prone to bias and high DawNO was the most significant factor causing the bias. Furthermore, we found that using 40 ml s-1 as the lowest flow rate in our sample and 50 ml s-1 in the meta-analysis compared to 100 ml s-1 resulted in higher CANO, but JawNO was not statistically significantly affected. We have provided objective evidence that not only the flow rates used but also the magnitude of NO output in the test subjects affect the accuracy of the T and G method. We suggest that flow rates below 100 ml s-1 should not be used with the T and G method to maintain accuracy.Peer reviewe