High-flow therapy: physiological effects and clinical applications

Abstract

Humidified high-flow therapy (HFT) is a noninvasive respiratory therapy, typically delivered through a nasal cannula interface, which delivers a stable fraction of inspired oxygen (F _IO2 ) at flow rates of up to 60 L·min^-1. It is well-tolerated, simple to set up and ideally applied at 37°C to permit optimal humidification of inspired gas. Flow rate and F _IO2 should be selected based on patients' inspiratory effort and severity of hypoxaemia. HFT yields beneficial physiological effects, including improved mucociliary clearance, enhanced dead space washout and optimisation of pulmonary mechanics. Robust evidence supports its application in the critical care setting (treatment of acute hypoxaemic respiratory failure and prevention of post-extubation respiratory failure) and emerging data supports HFT use during bronchoscopy, intubation and breaks from noninvasive ventilation or continuous positive airway pressure. There are limited data on HFT use in patients with hypercapnic respiratory failure, as an adjunct to pulmonary rehabilitation and in the palliative care setting, and further research is needed to validate the findings of small studies. The COVID-19 pandemic raises questions regarding HFT efficacy in COVID-19-related hypoxaemic respiratory failure and concerns regarding aerosolisation of respiratory droplets. Clinical trials are ongoing and healthcare professionals should implement strict precautions to mitigate the risk of nosocomial transmission.

Educational aims: Provide a practical guide to HFT setup and delivery.Outline the physiological effects of HFT on the respiratory system.Describe clinical applications of HFT in adult respiratory and critical care medicine and evaluate the supporting evidence.Discuss application of HFT in COVID-19 and aerosolisation of respiratory droplets.

Figure 1

Schematic of HFT equipment and settings display: temperature (in °C), flow rate 
(in L·min⁻¹) and F_IO2.

Figure 2

Changes in neural respiratory drive (quantified using EMG_para) in a patient with stable COPD. a) Neural respiratory drive quantified with EMG_para in a patient with very severe COPD (forced expiratory volume in 1 s <30% predicted). Neural respiratory drive index (the product of normalised EMG_para and respiratory rate, AU=arbitrary units) was measured at baseline and with HFT delivered at 10, 20, 30, 40 and 60 L·min⁻¹ at 37°C, F_IO2 0.21. Panels b) and c) illustrate raw EMG_para (mV, green), root mean square EMG_para (µV, orange) and respiratory rate (breaths·min⁻¹, blue). b) High neural respiratory drive observed at baseline. c) Optimal offloading of the respiratory muscle pump observed at 30 L·min⁻¹, with a significant reduction in neural respiratory drive.