Invasive mechanical ventilation during general anaesthesia for surgery typically results in atelectasis as well as reduced lung volume due to a cephalad shift of the diaphragm and a decreased muscle tone after induction of anesthesia.1 In particular, in patients undergoing abdominal surgery, the risk of atelectasis increases the closer the incision is to the diaphragm.2,3 Although intraoperative atelectasis impairs intraoperative oxygenation,4 more importantly, atelectasis often continues into the postoperative period, changing the mechanics of regional lung aeration and impairing the postoperative recovery of pulmonary function.5 Accordingly, atelectasis could predispose to the development of postoperative pulmonary complications (PPCs), including hypoxemia and pneumonia,2 with an increased risk of postoperative morbidity and mortality.6
Postoperative regeneration of pulmonary function could depend, at least in part, on the intraoperative ventilation strategy. Indeed, a significantly greater reduction in perioperative and postoperative lung volumes is seen with general anaesthesia as compared with spinal anesthesia,7 and with controlled rather than with assisted modes of ventilation.8 Furthermore, so-called ‘protective intraoperative ventilation’ that uses a combination of low tidal volumes and positive end-expiratory pressure (PEEP) and recruitment manoeuvres (RMs) could prevent the development of PPCs.9–12 However, the protective role of PEEP in preventing PPCs was challenged recently.13,14
Despite protective intraoperative ventilation, PPCs occur in up to 39% of patients.9,10,13 Risk scores, using preoperative characteristics, for the development of PPCs and early recognition of patients who develop PPCs could contribute to an improved patient outcome.15,16 As with preoperative spirometry to predict PPCs,17,18 postoperative spirometry could be a useful tool to monitor postoperative recovery of lung function.7,9 Therefore, in this substudy of the international multicentre, randomised controlled ‘PROtective Ventilation using HIgh versus LOw PEEP’ (PROVHILO) trial,13,19 in which intraoperative ventilation with a high level of PEEP (12 cmH2O) and RMs was compared with a low level of PEEP (≤2 cmH2O) without RMs during general anaesthesia for planned open abdominal surgery in nonobese patients at risk of PPCs, we tested the hypothesis that postoperative spirometry results would be modified by the intraoperative level of PEEP. In addition, we compared postoperative spirometry test results in patients who did and who did not develop PPCs.
Materials and methods
Ethical approval and informed consent
This was a preplanned substudy of the recently published PROVHILO trial.13,19 This single-centre substudy was performed at the Düsseldorf University Hospital, Düsseldorf, Germany. Patients at our institution were included from November 2011 until January 2013. The original trial was approved by the Institutional Review Boards of the Academic Medical Center (AMC), Amsterdam, The Netherlands, and on 5 July 2011 by the Medizinischen Fakultät der Heinrich–Heine Universität Düsseldorf, Düsseldorf, Germany (Study number 3664, chairperson Prof. Kröncke), and registered at ClinicalTrials.gov NCT01441791. The latter additionally approved this substudy as an amendment. Participants had to give written informed consent prior to participation for any procedure related to the original trial and this substudy.
Design of the original trial
In the PROVHILO trial, nonobese patients with an intermediate or high risk of PPCs according to the Assess Respiratory Risk in Surgical Patients in Catalonia (ARISCAT) score15,16 and who were scheduled for open abdominal surgery under general anaesthesia were randomly assigned to intraoperative ventilation with high levels of PEEP and RMs (12 cmH2O; the ‘high PEEP group’) or ventilation with lower levels of PEEP without RMs (<2 cmH2O; the ‘low PEEP group’). In the high PEEP group, patients received RMs at the following times: after intubation at the start of ventilation; before tracheal extubation; after each accidental disconnection from the ventilator. RMs were performed as follows: peak inspiratory pressure limit is set at 45 cmH2O; tidal volume is set at 8 ml kg−1 predicted body weight (PBW), respiratory rate at 6 to 8 breaths min−1 (or lowest respiratory rate that the anaesthesia ventilator allows), and PEEP is set at 12 cmH2O; inspiratory to expiratory (I:E) ratio is set at 1 : 2; tidal volumes are increased in steps of 4 ml kg−1 PBW until a plateau pressure of 30 to 35 cmH2O is attained; three breaths are administered with a plateau pressure of 30 to 35 cmH2O; peak inspiratory pressure limit, respiratory rate, I:E ratio, and tidal volume are reset to the settings preceding each recruitment manoeuver. FiO2 remained unchanged during RMs.
Patients were excluded for the following reasons: a planned laparoscopic abdominal procedure, pregnancy, a BMI more than 40 kg m−2, severe cardiac or pulmonary or other comorbidities, and if they were participating in other interventional studies. This substudy had no additional exclusion criteria.
In both arms of the trial, patients were ventilated with a tidal volume of 8 ml kg−1 PBW, FiO2 was 0.4 or higher to maintain SpO2 at least 92%, the respiratory rate was adjusted to maintain end-tidal CO2 between 4.67 and 6.0 kPa; and the I:E ratio was 1 : 2. Patients, and postoperative investigators who assessed whether or not a patient developed one or more PPCs, were blind to the intraoperative ventilation strategy. As recommended by the protocol and according to institutional routine, in this substudy, extubation was undertaken without suctioning of the trachea. Patients received additional oxygen as deemed necessary by the attending anaesthesiologist and were typically positioned supine with heads elevated to a maximum of 30° when extubated.
The primary endpoint of the PROVHILO trial was a composite of PPCs within the first five postoperative days. These PPCs consisted of an unexpected need for supplementary oxygen, severe hypoxemia, bronchospasm, suspected pulmonary infection, a pulmonary infiltrate, aspiration pneumonitis, development of ARDS, atelectasis, pleural effusion, pulmonary oedema caused by cardiac failure or pneumothorax. The definition of every complication is presented in the online Supplement Table 1, http://links.lww.com/EJA/A115.
Design of the present substudy
Preoperative and postoperative spirometry on postoperative days 1, 3 and 5 was performed after detailed instructions to participating patients. Patients were requested to rate their pain, while at rest in the supine position with 30° upper body elevation, on a numeric rating scale of 0 to 10 (from 0, no pain, to 10, maximum pain). Spirometric testing was only performed if pain scores at rest were 3 or less; otherwise, analgesia was optimised before spirometric measurements. According to institutional protocol, a continuous infusion of ropivacaine 0.2% through a thoracic epidural catheter was used for analgesia. Additional bolus doses and rate adjustments were made by the pain service according to the patients’ needs. For patients without epidural catheters, piritramid was administered intravenously as bolus doses or by patient-controlled analgesia pumps. Spirometry was performed in accordance with the American Thoracic Society's standards20 using a single pneumotachograph (SpiroPro, Jaeger, Würzburg, Germany) with the patient in the supine position with 30° upper body elevation. Each measurement was performed three times at each timepoint and the best value was selected for inclusion in the analysis. The postoperative investigators who performed the spirometry were blind as to the intraoperative ventilation strategy.
The primary endpoints of this substudy were the postoperative time-weighted averages (TWAs) of both the forced expiratory volume in 1 s (FEV1) and the forced vital capacity (FVC), up to postoperative day 5. TWAs were calculated for each patient as the area under the curve for FVC and FEV1 measurements divided by the follow-up duration in hours.
We intended to include all the participants within our centre from the original trial into this substudy. Consequently, the number of patients who could be included was restricted to the recruitment period of the original trial. An a priori sample-size estimate indicated that a minimum of 57 patients per group would provide an 80% chance of detecting a 20% difference in the TWAs of FVC and FEV1 from a presumed postoperative TWA of FVC of 1.6 ± 0.5 l with a corresponding TWA of FEV1 of 1.2 ± 0.4 l, with an alpha error level of 2.5% for combined outcomes. On the same basis, a minimum of 25 patients per group would provide an 80% chance of detecting a 30% difference.
We first compared results of postoperative spirometric measurements between patients ventilated with high PEEP with RM to those ventilated with low PEEP without RM. Then, postoperative spirometry results were compared between patients who developed PPCs and those who did not.
In two posthoc analyses, we evaluated whether intraoperative pulmonary compliance or the site of the surgical incision in combination with high PEEP and RM or low PEEP without RM influenced outcomes in our substudy. Therefore, we first compared spirometric results in patients with an intraoperative pulmonary compliance more than 50 ml cm−1 H2O (high compliance) to those with a compliance 50 ml cm−1 H2O or less (low compliance). The cut-off was based on the median intraoperative pulmonary compliance in all substudy patients. We then further subdivided the spirometric results of patients with high or low compliance by ventilation strategy, that is low PEEP without RM versus high PEEP with RM. Secondly, we compared postoperative spirometry results of patients who had the incision closer to the diaphragm (i.e. upper abdominal surgery) with those with the incision at a distance from the diaphragm (i.e. lower abdominal surgery) and also further subdivided these spirometric results by ventilation strategy, that is low PEEP without RM or high PEEP with RM.
Data are presented as absolute values, means with standard deviation or medians with interquartile range, as appropriate. Analyses were performed on an intention-to-treat basis. We used the Kolmogorow–Smirnow test to test the distribution of data and the two-tailed Fishers’ exact test, Student's t-test or Mann–Whitney U tests as appropriate for comparison between groups. For analysis within groups, the Wilcoxon rank-sum test was performed.
The IBM SPSS Statistics Versions 21 and 22 (IBM Deutschland GmbH, Ehningen, Germany) were used. To take account of two primary endpoints, FEV1 and FVC, a Bonferroni-corrected P value less than 0.025 was considered to be statistically significant for the two primary outcomes TWA of FEV1 and FVC. For secondary outcomes, which were exploratory, P value less than 0.05 was considered to be statistically significant.
Substudy patients and occurrence of postoperative pulmonary complication
All 63 patients enrolled in the PROVHILO trial in Düsseldorf participated in this substudy; 31 and 32 patients were randomised to the high and the low PEEP group, respectively (Fig. 1). One patient in the high PEEP group received ventilation with a PEEP of 5 cmH2O for 3 out of 4 hours by mistake; PEEP of 12 cmH2O was thus only applied for the last hour. According to the intention-to-treat analysis, this patient remained in the high PEEP group. Six patients were excluded from the analysis because postoperative spirometry results could not be obtained, leaving 27 in the high PEEP group and 30 in the low PEEP group for the final analysis (Fig. 1).
The occurrence of PPC in the substudy was high (24/57 = 42%) but comparable to that found in the original trial (346/880 = 39%, Chi-squared P = 0.677). A comparison between characteristics of patients enrolled in the original trial and patients in the substudy is provided in the online supplement (Supplement Table S2, http://links.lww.com/EJA/A115).
Among patients participating in the substudy, baseline characteristics, including preoperative spirometry results, did not differ between the two randomisation groups ventilated with high or low PEEP (Table 1). The level of PEEP and peak inspiratory pressure levels were different between the randomisation groups, as was the pulmonary compliance during intraoperative ventilation (Supplement Table S3, http://links.lww.com/EJA/A115). PPC did not differ between the randomisation groups (Supplement Table S4, http://links.lww.com/EJA/A115). Postoperatively, the ratio between FEV1/FVC remained within the normal range in the majority of patients [TWA FEV1/FVC = 76 (68 to 80)%].
Association between intraoperative ventilation strategy and spirometry results
Spirometry results were unaffected by the intraoperative ventilation strategy: TWA of FVC and FEV1 were not different between the high and low PEEP group [TWA FVC = 1.8 (1.6 to 2.4) versus 1.7 (1.2 – 2.4) l (P = 0.792) and TWA FEV1 = 1.2 (1.1 to 2.5) versus 1.2 (0.9 to 1.9)] l (P = 0.497). There were also no differences in FVC or FEV1 between randomisation groups on individual postoperative days (Fig. 2).
Association between the occurrence of postoperative pulmonary complication and spirometry results
In patients who developed PPCs, FEV1 and FVC values on postoperative day 5 were about 30% lower than patients who did not develop PPCs (Fig. 3). Compared with patients who did not develop PPCs, patients who developed PPCs had longer surgery (supplement Table 5, http://links.lww.com/EJA/A115), received higher tidal volumes, higher minute ventilation volumes and more intravenous fluids during surgery (supplement Table 6, http://links.lww.com/EJA/A115).
Association between intraoperative pulmonary compliance and spirometry results
On the first postoperative day, spirometry results were about 40% higher in patients with high pulmonary compliance, but unaffected by PEEP and RM (Fig. 4a, b). Patients with high intraoperative pulmonary compliance were not different from those with a low compliance (supplement Table 7, http://links.lww.com/EJA/A115), but more frequently received intraoperative ventilation with high PEEP (supplement Table 8, http://links.lww.com/EJA/A115), but incidence of PPCs was not different (supplement Table 9, http://links.lww.com/EJA/A115).
Association between location of incision and spirometry results
Patients who had upper abdominal surgery were current smokers more frequently and had a longer duration of surgery than patients who had lower abdominal surgery (supplement Table 10, http://links.lww.com/EJA/A115) and received more fluids and transfusions (supplement Table 11, http://links.lww.com/EJA/A115). The incidence of PPCs, however, was not different (supplement Table 12, http://links.lww.com/EJA/A115). Postoperative spirometry showed no differences between the high and low PEEP groups, neither in patients who had upper abdominal surgery nor in patients who had lower abdominal surgery (Fig. 4c, d).
The results of this substudy of a larger randomised controlled trial comparing high with low PEEP during intraoperative ventilation in nonobese patients at risk of PPCs and scheduled for open abdominal surgery can be summarised as follows: In patients ventilated with a tidal volume of 8 ml kg−1 PBW, postoperative spirometry results are no different between patients receiving ventilation with high PEEP and RM and patients receiving ventilation with low PEEP without RM: postoperative spirometry results are abnormal up to postoperative day 5: occurrence of PPCs seems to be associated with a change in postoperative spirometry results on postoperative day 5.
To our knowledge, this is one of the largest prospective randomised controlled studies investigating the association between postoperative spirometry changes in patients undergoing major abdominal surgery and at risk of developing PPCs.
This substudy stopped when the PROVHILO trial completed recruitment, so we did not recruit the total number of patients required according to the sample size calculation, and we had less than 80% power to show a 20% statistically significant difference between the two groups. A comparison of the median TWAs of the two treatment groups, high PEEP with RM versus low PEEP without RM, suggests no difference between FEV1 and a difference of only 6% in FVC. The latter would not be considered to be of clinical relevance. We calculated a potential effect size based on the means and standard deviation of each treatment group. FVC TWA in the high PEEP group was 1.89 ± 0.99 versus 1.95 ± 0.68 in the low PEEP group. FEV1 TWA in the high PEEP groups was 1.46 ± 0.8 versus 1.29 ± 0.5 I in the low PEEP group. Thus, dCohen effect size for FVC TWA would be 0.07 [95% confidence interval, 95% CI -0.4 to 0.59] and 0.24 [95% CI -0.7 to 0.27]. Our initial hypothesis and sample size calculation was built on a much stronger effect size of 0.5, which we consider to be clinically relevant.
However, we detected significant differences in spirometric test results between patients who developed PPCs and those who did not. Although this might not seem to be a surprising result, to our knowledge, postoperative spirometry is not used commonly as a tool to detect or predict PPC. It is important to note that our study was not designed to show a direct or timely correlation between spirometric results and the development of PPCs. Further studies are needed to determine, whether spirometric results could predict or indicate the development of PPCs at an early stage such that this would allow the initiation of preventive or early therapeutic measures. Postoperative spirometry per se might be a useful as a tool to detect PPCs. However, technical and practical reasons limit its utility as a postoperative monitor. For instance, pain needs to be adequately controlled and patients need to be fully awake and compliant.
In this substudy and preplanned analysis, we had a unique opportunity to determine the effect of two different levels of PEEP and RM during intraoperative ventilation on postoperative lung function test results. Its prospective design, the completeness of follow-up and the fact that occurrence of PPCs was scored by assessors who were blind to the intraoperative ventilation strategy helped reduce bias. In addition, the definition of PPCs was defined a priori and the patients were similar with regard to their clinical characteristics and type of surgery. Lastly, all patients were ventilated with tidal volumes of 8 ml kg−1 PBW; thus, we were able to assess the effect of PEEP and RMs on postoperative lung function.
FVC and FEV1 decreased by more than 50% compared with preoperative values in both randomisation groups. This restrictive ventilatory pattern has long been recognised after upper abdominal surgery and results from reduced ventilatory muscle activity, diaphragmatic dysfunction and decreased lung compliance and is also influenced by pain levels.21 Although we found intraoperative dynamic lung compliance to be significantly higher in the high PEEP group, this did not protect against a decline in postoperative lung function. These findings are consistent with the overall results of the PROVHILO trial, in which the occurrence of PPCs was high, but not different between patients who received high PEEP or low PEEP during intraoperative ventilation.13
The results of the present study support the information that came from two preceding trials of intraoperative ventilation.9,22 In an Italian single-centre, randomised controlled trial of patients scheduled for open abdominal surgery lasting more than 2 h, the FVC and FEV1 on postoperative day 1 were also approximately 50% lower than preoperative values.9 However, in that trial, recovery of lung function was better in patients ventilated with a lung-protective ventilation strategy (a PEEP of 10 cmH2O, a low tidal volume of 7 ml kg−1 PBW and RM) compared with patients ventilated with a conventional ventilation strategy (no PEEP, a tidal volume of 9 ml kg−1 PBW, without RM).9 In a German single-centre, randomised controlled trial of patients undergoing upper abdominal surgery, in which all patients received a similar level of PEEP and the same RM, postoperative changes in spirometry results were not different in patients ventilated with 6 ml kg−1 PBW versus 12 ml kg−1 PBW.22 On the basis of the results of these two preceding trials and the results from the present study, we suggest that postoperative spirometry changes, specifically in the time course of lung function recovery, might be affected by a combination of the two parameters ‘size of intraoperative tidal volume’ along with ‘PEEP’ and ‘RM’, but not solely by changes either in ‘tidal volume’ or ‘PEEP’ alone.
Since publication of the Italian trial mentioned above, two other randomised trials of intraoperative ventilation have been published.10,11 In both trials, patients were randomly assigned to lung-protective ventilation with low tidal volumes and high PEEP or conventional ventilation with high tidal volume and no PEEP. Both trials found fewer pulmonary complications with lung-protective ventilation. The results from the original trial of this substudy, the PROVHILO trial,13 suggest that high levels of PEEP with RM do not protect against development of PPCs in patients ventilated with low tidal volumes.
Accordingly, a differentiated algorithm for protective intraoperative mechanical ventilation has recently been proposed.23 In nonobese patients without acute respiratory distress syndrome undergoing open abdominal surgery, mechanical ventilation should be performed with tidal volumes of 6 to 8 ml kg−1 PBW combined with a low PEEP of 2 cmH2O or less. If hypoxemia develops and hypotension, hypoventilation or other causes have been excluded, inspiratory oxygen fraction should be increased first, followed by increase of PEEP and recruitment manoeuvers.23
Of note, high PEEP with RM failed to affect postoperative spirometry results in two subgroups of patients in which more benefit of high PEEP could be expected, that is in patients with a lower pulmonary compliance during intraoperative ventilation and patients who underwent upper abdominal surgery.
Our study was restricted to patients at risk of PPCs who were scheduled to undergo open abdominal surgery. The majority of patients in our substudy received thoracic epidural anaesthesia both intraoperatively and postoperatively. Therefore, the results could be different in other patient groups. We detected differences in spirometric test results between patients who developed PPCs and those who did not. However, our study was not designed to show a direct or timely correlation between spirometric results and the development of PPCs. Further studies are needed to determine whether spirometric results could predict or indicate the development of PPCs at an early stage, and whether this would allow preventive or early therapeutic measures to be initiated. However, even though postoperative spirometry per se might prove to be useful as a tool to detect PPCs, technical and practical reasons could limit its utility: postoperative pain needs to be adequately controlled, and patients need to be fully awake and compliant. Another limitation of our study relates to intra-abdominal pressure. Intra-abdominal pressure in the postoperative period could interfere with lung function and hence spirometry results. We did not measure intra-abdominal pressure, and thus cannot evaluate, whether this influenced our results.
With the knowledge of our results, the question may arise as to whether patient management during the emergence phase of anaesthesia could influence lung function to such an extent that the consequences of several hours of intraoperative ventilation become negligible. We do not know whether extending the application of PEEP into the postoperative period, or prohibiting use of 100% oxygen during extubation would have changed our results. Other trials suggest that if there is an effect due to how the emergence phase of anaesthesia is managed, this would only have minor consequences.24,25 Interestingly, as part of a protective ventilation strategy, the beneficial effect of RMs might also be questioned.26 The focus of our study was to compare the effects of several hours of ventilation using high levels of PEEP along with RMs with similar periods of ventilation without PEEP and RMs.
In conclusion, postoperative spirometry test results are not affected by the PEEP level during intraoperative ventilation during anaesthesia for open abdominal surgery in patients at high risk of PPC. Spirometry test results on postoperative day five are associated with the development of PPCs during this time period.
Acknowledgements relating to this article
Assistance with the study: we thank Renate Babian for her assistance with the study.
Financial support and sponsorship: this work was supported by the Department of Anesthesiology, Düsseldorf University Hospital, Düsseldorf, Germany.
Conflicts of interest: none.
1. Tiefenthaler W, Pehboeck D, Hammerle E, et al. Lung function after total intravenous anaesthesia or balanced anaesthesia with sevoflurane. Br J Anaesth
2. Qaseem A, Snow V, Fitterman N, et al. Risk assessment for and strategies to reduce perioperative pulmonary complications for patients undergoing noncardiothoracic surgery: a guideline from the American College of Physicians. Ann Intern Med
3. Joris J, Kaba A, Lamy M. Postoperative spirometry after laparoscopy for lower abdominal or upper abdominal surgical procedures. Br J Anaesth
4. Hedenstierna G, Tokics L, Strandberg A, et al. Correlation of gas exchange impairment to development of atelectasis during anaesthesia and muscle paralysis. Acta Anaesthesiol Scand
5. Westerdahl E, Lindmark B, Eriksson T, et al. Deep-breathing exercises reduce atelectasis and improve pulmonary function after coronary artery bypass surgery. Chest
6. Sabaté S, Mazo V, Canet J. Predicting postoperative pulmonary complications: implications for outcomes and costs. Curr Opin Anaesthesiol
7. von Ungern-Sternberg BS, Regli A, Reber A, Schneider MC. Comparison of perioperative spirometric data following spinal or general anaesthesia in normal-weight and overweight gynaecological patients. Acta Anaesthesiol Scand
8. Zoremba M, Kalmus G, Dette F, et al. Effect of intra-operative pressure support vs pressure controlled ventilation on oxygenation and lung function in moderately obese adults. Anaesthesia
9. Severgnini P, Semo g, Lanza C, et al. Protective mechanical ventilation during general anesthesia for open abdominal surgery improves postoperative pulmonary function. Anesthesiology
10. Futier E, Constantin JM, Paugam-Burtz C, et al. A trial of intraoperative low-tidal-volume ventilation in abdominal surgery. N Engl J Med
11. Ge Y, Yuan L, Jiang X, et al. [Effect of lung protection mechanical ventilation on respiratory function in the elderly undergoing spinal fusion]. Zhong Nan Da Xue Xue Bao Yi Xue Ban
12. Futier E, Jaber S. Lung-protective ventilation in abdominal surgery. Curr Opin Crit Care
13. Hemmes SNT, Gama de Abreu M, Pelosi P, Schultz MJ. PROVE Network Investigators for the Clinical Trial Network of the European Society of Anaesthesiology. High versus low positive end-expiratory pressure during general anaesthesia for open abdominal surgery (PROVHILO trial): a multicentre randomised controlled trial. Lancet
14. Serpa Neto A, Hemmes S, Barbas S, et al. Protective versus conventional ventilation for surgery: a systematic review and individual patient data meta-analysis. Anesthesiology
15. Canet J, Gallart L, Gomar C, et al. Prediction of postoperative pulmonary complications in a population-based surgical cohort. Anesthesiology
16. Canet J, Sabate S, Gallart L, et al. Development and validation of a score to predict postoperative respiratory failure in a multicentre European cohort: a prospective, observational study. Eur J Anaesthesiol
17. van Huisstede A, Biter L, Luitwieler R, et al. Pulmonary function testing and complications of laparoscopic bariatric surgery. Obes Surg
18. Jeong O, Ryu SY, Park YK. The value of preoperative lung spirometry test for predicting the operative risk in patients undergoing gastric cancer surgery. J Korean Surg Soc
19. Hemmes SNT, Severgnini P, Jaber S, et al. Rationale and study design of PROVHILO - a worldwide multicenter randomized controlled trial on protective ventilation during general anesthesia for open abdominal surgery. Trials
20. American Thoracic Society. Standardization of Spirometry, 1994 update. Am J Respir Cri Care Med
21. Dureuil B, Cantineau JP, Desmonts JM. Effects of upper or lower abdominal surgery on diaphragmatic function. Br J Anaesth
22. Treschan TA, Kaisers W, Schaefer M, et al. Ventilation with low tidal volumes during upper abdominal surgery does not improve postoperative lung function. Br J Anaesth
23. Güldner A, Kisss T, Neto A, et al. Intraoperative protective mechanical ventilation for prevention of postoperative pulmonary complications: a comprehensive review of the role of tidal volume, positive end-expiratory pressure, and lung recruitment maneuvers. Anesthesiology
24. Edmark L, Auner U, Lindbäck J, et al. Post-operative atelectasis - a randomised trial investigating a ventilatory strategy and low oxygen fraction during recovery. Acta Anaesthesiol Scand
25. Lumb AB, Greenhill SJ, Simpson MP, Stewart J. Lung recruitment and positive airway pressure before extubation does not improve oxygenation in the post-anaesthesia care unit: a randomized clinical trial. Br J Anaesth
26. Defresne AA, Hans GA, Goffin PJ, et al. Recruitment of lung volume during surgery neither affects the postoperative spirometry nor the risk of hypoxaemia after laparoscopic gastric bypass in morbidly obese patients: a randomized controlled study. Br J Anaesth
* PROVE network = the PROtective VEntilation Network (www.provenet.eu).