The authors reply
First, our systematic review reported that noninvasive ventilation (NIV) could reduce intubation rate in acute hypoxemia nonhypercapnic respiratory failure excluding chronic obstructive pulmonary disease exacerbation and cardiogenic pulmonary edema (CPE) patients compared with standard oxygen therapy (2). However, the question raised by Dr. Cheng (1) is concerned that who will benefit from NIV in acute respiratory failure. It is clear that our meta-analysis was not designed to explain this overly complex question.
Second, it is arbitrary and not rigorous to compare the risk ratio of each study included in the meta-analysis as Dr. Cheng (1) did (Fig. 1). Actually, we did subgroup analysis according to PaO2/FIO2 (P/F) ratio in our meta-analysis (Supplemental Fig. S4) (2). The results showed that NIV could decrease the intubation rate not only in 200 mm Hg < P/F ≤ 300 mm Hg (p = 0.0001) group but also 100 mm Hg ≤ P/F ≤ 200 mm Hg (p = 0.003) group, which is not associated with decreasing the ICU mortality. However, as clearly stated in our article, we emphasized that our subgroup analysis about severity of acute hypoxemic respiratory failure (AHRF) was according to the average P/F ratio from patients’ characteristics of all RCTs. Therefore, we cannot emphasize these results or give any firm recommendation to clinicians. We have discussed this in our discussion part to alert readers about this issue (2). In order to avoid further misleading, we did not try to do another subgroup analysis according to 150 mm Hg as cutoff value.
Third, as Dr. Cheng (1) noticed, it is true that researches by Ferrer et al (3) (30/105, 28.5%) and Antonelli et al (4) (9/40, 22.5%) enrolled a proportion of patients with CPE. However, we had excluded this part of patients with CPE (Supplemental Table 1) in our meta-analysis carefully from the beginning (2), which would not cause any bias conclusion.
In addition, Dr. Cheng (1) cited a recent study by Bellani et al (5), who reported that compared with invasive mechanical ventilation, NIV was associated with higher ICU mortality in P/F ration less than 150 mm Hg in acute respiratory distress syndrome patients from the Large Observational Study to UNderstand the Global Impact of Severe Acute Respiratory FailurE (LUNG-SAFE) study. Thille et al (6) also suggested that NIV may be used in AHRF patients with P/F ratio greater than 150 mm Hg. However, it still needs more studies to assess if 150 mm Hg (P/F ratio) could be the perfect cutoff point for AHRF patients switching from NIV to intubation and invasive mechanical ventilation.
Finally, we wish to thank Dr. Cheng (1) for his emphasis on careful patient selection when initiating NIV. We would like to reiterate that we did not recommend NIV application in AHRF patients with 100 mm Hg ≤ P/F ≤ 200 mm Hg according to our meta-analysis.