Aspiration pneumonia combined with respiratory failure is a common cause of death in patients that have experienced a stroke. The primary treatment of this condition is to rapidly correct the hypoxia state, improve the oxygen supply to the tissue, and use antibiotics to prevent MODS from hypoxia . HFNC is a new type of oxygen therapy that not only can provide relatively stable oxygen concentrations and highly efficient airway humidification but also can regulate the flow velocity to provide positive end-expiratory pressure in the respiratory tract, similar to that which would be provided by a ventilator. HFNC is also better tolerated and has a better treatment effect than traditional invasive modes of oxygen therapy [16,17]. Studies from many countries have shown that, in addition to significantly improving oxygenation status and tolerance, HFNC can reduce the incidence of re-intubation in patients with chronic obstructive pulmonary disease [18,19]. HFNC has also played an important role in the treatment of viral pneumonia, including MERS-CoV pneumonia, H1N1 pneumonia, and novel coronavirus pneumonia. It has become part of a relatively standardized treatment protocol and has achieved a good therapeutic effect . The traditional mode of oxygen therapy often has a poor effect for patients that experience complications from aspiration pneumonia after a stroke: Injury to their central nervous system leads to limited respiratory function and a reduced ability to discharge phlegm autonomously, and these patients will inevitably require invasive auxiliary ventilation, increasing their medical burden. In addition, the prognosis of patients that also experience ventilator-associated pneumonia is very poor.
To improve the oxygenation status of patients and avoid tracheal intubation, it is necessary to identify a non-invasive mode of oxygen therapy that is more effective than traditional modes of oxygen therapy. Through the retrospective analysis of 87 patients with post-stroke aspiration pneumonia and respiratory failure, it was found that HFNC significantly improved the patients’ oxygenation status and was especially effective for the early and rapid correction of hypoxia. As shown by the contour map (Figure 1), the oxygenation indices of the HFNC group began to improve approximately 4 h after therapy, and the improvement in oxygenation indices after 8–24 h was significantly greater than in the traditional Venturi mask treatment group. Repeated measurement statistics demonstrated that HFNC therapy contributed to the improvement of oxygenation indices at a rate of 75.1%.
Meanwhile, we found that there were different degrees of CO2 retention in the two groups of patients, in addition to the pulmonary infection during admission, some patients with COPD also aggravated the probability of CO2 retention. Interestingly, although the proportion of patients in the HFNC group combined with COPD was significantly higher than that in the Venturi group, with the implementation of the two oxygen therapies, the CO2 retention rate of the HFNC group compared with Venturi group was significantly improved at about 4h - 12h. We mainly considered two factors: firstly, HFNC was a new type of nasal humidification oxygen supply method, which could output constant oxygen concentration in the range of 21% - 100% under the condition of constant temperature of 37 ℃ and relative humidity of 100%, so as to ensure the normal function of airway mucosa cilia, promote sputum dilution, and effectively ensure the discharge of airway secretions. The maximum oxygen flow output of HFNC could reach 60 L / min, which could effectively increase the alveolar ventilation volume of patients, and reduce the ventilation volume of the physiological invalid alveolar cavity, at the same time, it could reduce the respiratory power consumption in the inspiratory phase, quickly and effectively alleviate the symptoms of dyspnea. The oxygen inhalation device could also generate a certain positive airway pressure in the expiratory phase, similar to the positive end expiratory pressure（PEEP : 2 – 5 cm H2O）, which could promote the expansion of atelectasis, increase the residual capacity of the lung and counteract endogenous PEEP, while correcting hypoxemia, CO2 retention was also reduced. However, the Venturi mask was made according to the Venturi principle. When oxygen entered the mask through a narrow hole, a negative pressure was generated around the jet airflow, and a certain amount of air was carried into the mask from the open edge. The size of the mask edge seam changed the ratio of air to oxygen. Its main advantage was that it could humidify oxygen and provide stable oxygen concentration (up to 50%), and high flow rate gas can effectively avoid repeated inhalation of CO2. Therefore, compared with the two oxygen therapy methods, HFNC may be more effective in improving lung function, promoting oxygenation and reducing CO2 retention in the same oxygen concentration [16,17]. Secondly, the traditional view indicated that the improvement of neurological function mainly occurs in the 12-24 weeks after stroke and tends to be stable. However, recent studies have shown that if timely and effective rehabilitation treatment was given after stroke, about a quarter of ischemic stroke patients could continuously improve their neurological function within 12 to 48 weeks after stroke, thus improving the 5-year clinical outcomes of stroke patients . On the contrary, if the patients could not get effective rehabilitation treatment, or the recovery of neurological function was not ideal after treatment, as the disease progresses, the incidence of aspiration pneumonia caused by stroke induced systemic immune suppression and improper feeding was significantly increased. And with the extension of stroke duration, the patient's condition would develop into severe oxygenation disorder due to pulmonary infection, then develop into multiple organ failure, and the mortality rate of 1-5 years was significantly increased. [22-23]. In this study, there was a significant difference in duration of stroke (≥48 weeks) between the two groups, which might have an impact on the oxygenation state and proportion of patients with multiple organ failure. In addition, we compared the incidence of invasive ventilation after receiving endotracheal intubation at different time cutoff points (12h, 24h, 48h, 72h) and the total incidence of invasive ventilation with endotracheal intubation in 72 hours between the two groups. The results showed that before 12h, there was no significant difference in the incidence rate of invasive ventilation between the two groups. In 24 - 48h, the invasive ventilation rate in the Venturi group was higher than that in the HFNC group; in 48h-72h, the incidence of invasive ventilation was not statistical different. This result is consistent with the improvement trend of oxygenation in patients (Fig. 1).
In addition, it was also found that, within 72 hours after treatment, a significantly lower proportion of patients from the HFNC group required invasive ventilation than in the Venturi group, where the incidence of invasive ventilation in the HFNC group was 0.406 times than of the Venturi group. This suggests that HFNC could significantly reduce the incidence of transition to invasive auxiliary ventilation. The possible reason was that with HFNC treatment, the atelectasis of lung tissues were effectively reversed, and the oxygenation status was improved, however, some severe non-functional lung tissues did not have the basis for functional improvement under the existing time and treatment conditions. However, as far as the overall observation window (72h) was concerned, HFNC could significantly reduce the incidence of invasive ventilation compared with Venturi treatment in such patients.
Although the incidence of multiple organ failure was significantly different between the two groups（ Venturi group：29 (64.4%) vs HFNC : 15 (35.7%)）, the 28-day mortality rate of the two groups showed no obvious statistical difference. The reasons may be as follows: for the patients with multiple organ failure, we not only give them positive respiratory function support, but also give multiple organ comprehensive treatment measures including continuous blood purification and nutritional support, which may play a positive role in reducing the mortality of patients.
The analysis of the general data for the patients in the two groups also revealed differences in terms of organ failure. A higher proportion of patients experienced MODS in the Venturi group than in the HFNC group. A study by He Ping et al.  found that the improvement of 90-day neurological function in patients that had experienced an acute stroke could be attributed to the improvement of oxygenation of organs, including brain tissue, during the early post-stage. However, the study did not specifically analyze the time at which the patients’ hypoxic state was significantly corrected, and all of the enrolled patients were in the acute stage of stroke (within 24 hours of onset) and had not experienced organ failure. By contrast, the patients enrolled in the present study were all in the post-stroke sequelae stage, and some were experiencing MODS. As a result, this study found significantly higher rates of mortality and invasive auxiliary ventilation than were reported in the study mentioned above.
Since the 90’s, noninvasive ventilation (NIV) has been widely used in COPD exacerbation but with controversial results aspiration pneumonia combined with respiratory failure . NIV can create a much higher gas flow rate and reduce inspiratory effort through positive pressure. Compared to NIV, HFNC is simpler to use and apply, at the same time, HFNC is more comfortable and delivers high fraction of inspired oxygen, and HFNC generates a low level of positive pressure . HFNC can also provide irrigation of dead space in the upper respiratory. Therefore, HFNC is more preferred than NIV in the treatment of patients with aspiration pneumonia combined with respiratory failure [27-28].
The present study had some limitations. First, the results may have been impacted by the sample size. Second, the results may have been biased because the study was not stratified according to the specific organs that failed, the number of organs that failed, or the duration of stroke sequelae. The results must be verified by a stratified randomized controlled study.