To the Editor:
Noninvasive ventilation (NIV) may control the incidence of postoperative complications in selected patients who develop hypoxemia after some elective thoracic and abdominal surgery. As a prophylactic intervention, NIV may be important in patients at risk of hypoxemia after cardiac surgery or lobectomy.1,2 After lung surgery, early NIV may reduce pulmonary dysfunction and improve respiratory function.3
Barbagallo et al4 used helmet for prophylactic CPAP following lung resection and found short-term improved PaO2/FIO2 and shorter hospital stay, but no difference in complications. These results may improve our understanding of postoperative ventilatory support for preventing postoperative complications, but we see 5 important issues with their methods and results.
First, the extent of lung resection (lung lobectomy) means that the postoperative lung function in their patient population (which had a mean FEV1 of > 85% of predicted and a mean FVC of > 90% of predicted) was not substantially affected, and this may explain the high success rate in the Barbagallo et al study.
Second, the preoperative arterial blood gas exchange is a key to understanding their results, but PaO2/FIO2 immediately before initiating helmet CPAP (hCPAP) was not reported, so we can't determine the benefit obtained after the first hCPAP cycle. The study could not be blinded, but were any similar postoperative supportive measures used in the control group?
Third, it is not known whether prophylactic hCPAP resulted in overtreatment of some subjects.
Fourth, the relationship between improved oxygenation and hospital stay is unknown, and hCPAP only transiently improved oxygenation and did not significantly influence complications, so it is doubtful that hCPAP influenced stay. We would like to know the relationship between a short-term improved oxygenation and the decision to discharge the patient earlier, which might have been influenced by physician bias, since the surgeon responsible for the discharge knew that the patient was given hCPAP.
Fifth, the relationship between the existence of minor complications in the hCPAP group versus the control group is interesting. Considering the absence of complications in the hCPAP group, the stay should have been compared to the stay of patients without major complications, and not to the stay of the control group, which had 4 patients with pneumonia and wound infection.
More studies are needed to determine hCPAP's effects on prognosis and postoperative complications. Also the presence of associated comorbidities at various postoperative periods needs to be studied.2,5–7 Despite these limitations and the necessity of randomized trials, we should consider helmet as a prophylactic and therapeutic tool to improve gas exchange in postoperative patients.2,8 We compliment them on their study.
Footnotes
The authors have disclosed no conflicts of interest.
- Copyright © 2013 by Daedalus Enterprises
References
The authors respond to: Noninvasive Mechanical Ventilation and Helmet After Lung Resection: Oxygenation Improvement: A Small Step or a Large Step?:
We thank Esquinas and Papadakos for their careful analysis of our paper.1 The authors of the letter are completely correct that our patient population had favorable features to undergo lung lobectomy. However, we investigated an unselected population that represents the mean standard population of lung cancer patients suitable for surgical resection. Nevertheless, the majority of our patients had mild to moderate COPD according to Global Initiative for Chronic Obstructive Lung Disease classification.2 Additionally, 86% of them (43/50) were active or former heavy smokers (median of 40 pack/years) and 62% (31/50) had cardio-circulatory diseases.
Regarding the blood gas values, they were collected at admission to ICU and immediately before and after the first helmet CPAP treatment, immediately before and after the second helmet CPAP treatment, and so on, according to time points scheduled. In the paper, Figure 2 nicely showed the evolution and trend of PaO2/FIO2 during the study period. After the first CPAP course a mild increase of PaO2/FIO2 was observed; it was also detected after the second course, but the difference was not statistically significant.
Regarding the hospital stay and the transient improvement of PaO2/FIO2, on one hand, our study showed that prophylactic use of helmet CPAP can progressively improve PaO2/FIO2, reaching a statistically significant higher value after the second course of CPAP, compared to the control group (P = .004). On the other hand, the hospital stay was statistically shorter in the helmet CPAP group than in the other group (P = .042). In our institution the overall median hospital stay after lobectomy is 7 days, which is in line with our results. The slight but significant difference between the 2 study groups, probably came from the 3 patients in the control group who developed pneumonia, even if that fact did not cause any significant difference in postoperative complications between the groups. So we can't be sure there is a correlation between the 2 variables. Nevertheless, hospital stay might have been influenced by various factors on which helmet CPAP had a positive impact. In any case, it would have been nonambiguous if the PaO2/FIO2 improvement had been long lasting; in that case, a convincing association could be hypothesized. Further study could focus on the continuation of postoperative CPAP in order to find a relationship between hospital stay and oxygenation improvement.
Thanks again to Esquinas and Papadakos for their important comments, which underline that our data give interesting insights into a prophylactic approach in the management of postoperative period after lung lobectomy.
Footnotes
The authors have disclosed no conflicts of interest.
