Weaning from mechanical ventilation in the operating room: a systematic review

Background Postoperative pulmonary complications (PPCs) are associated with postoperative mortality and prolonged hospital stay. Although intraoperative mechanical ventilation (MV) is a risk factor for PPCs, strategies addressing weaning from MV are understudied. In this systematic review, we evaluated weaning strategies and their effects on postoperative pulmonary outcomes. Methods Our protocol was registered on PROSPERO (CRD42022379145). Eligible studies included randomised controlled trials and observational studies of adults weaned from MV in the operating room. Primary outcomes included atelectasis and oxygenation; secondary outcomes included lung volume changes and PPCs. Risk of bias was assessed using the Cochrane Risk of Bias (RoB2) tool, and quality of evidence with the GRADE framework. Results Screening identified 14 randomised controlled trials including 1719 patients; seven studies were limited to the weaning phase and seven included interventions not restricted to the weaning phase. Strategies combining pressure support ventilation (PSV) with positive end-expiratory pressure (PEEP) and low fraction of inspired oxygen (FiO2) improved atelectasis, oxygenation, and lung volumes. Low FiO2 improved atelectasis and oxygenation but might not improve lung volumes. A fixed-PEEP strategy led to no improvement in oxygenation or atelectasis; however, individualised PEEP with low FiO2 improved oxygenation and might be associated with reduced PPCs. Half of included studies are of moderate or high risk of bias; the overall quality of evidence is low. Conclusions There is limited research evaluating weaning from intraoperative MV. Based on low-quality evidence, PSV, individualised PEEP, and low FiO2 may be associated with reduced postoperative pulmonary outcomes. Systematic Review Protocol PROSPERO (CRD42022379145).


Editor's key points
Intraoperative mechanical ventilation is a risk factor for postoperative pulmonary complications after general anaesthesia and surgery, but the impact of strategies for weaning from mechanical ventilation are understudied.In this systematic review, the authors evaluated weaning strategies and their effects on postoperative pulmonary outcomes.Based on 14 identified studies, strategies combining pressure support ventilation with positive endexpiratory pressure and low fraction of inspired oxygen (FiO 2 ) improved atelectasis, oxygenation, and lung volumes, but the quality of evidence was low.Additional studies focusing on weaning from mechanical ventilation are critical to support the potential benefits of such interventions in reducing postoperative pulmonary complications.
Postoperative pulmonary complications (PPCs) are a leading cause of increased mortality and morbidity in surgical patients, especially during the first postoperative week.1e3 Depending on severity, risk factors, and definitions, the incidence ranges between 6% and 40%. 3The development of even one mild PPC is significantly associated with increased ICU admission, early postoperative mortality, and prolonged length of stay in both hospital and ICU. 4 Atelectasis, the most frequent PPC, also leads to significant morbidity 4 and is associated with patient risk factors, general anaesthesia, mechanical ventilation, and type of surgery. 5Atelectasis worsens patient outcomes through a reduction in lung volumes and a decrease in partial pressure of oxygen in arterial blood (Pa O2 ) and in the ratio of partial pressure of oxygen in arterial blood to fraction of inspiratory oxygen concentration (Pa O2 /FiO 2 ). 6,7ltimately, atelectasis is associated with an increased risk of mortality, 4 and is the main cause of postoperative hypoxaemia leading to postoperative respiratory failure and the need for respiratory support ranging from noninvasive oxygen use to invasive mechanical ventilation. 8Finally, atelectasis is also considered a risk factor for development of other PPCs such as pneumonia and ventilator-induced lung injury. 5umerous strategies have been studied to reduce atelectasis, and in turn PPCs, during induction and maintenance of anaesthesia for major surgery.These studies have given rise to interventions including higher (fixed or individualised) positive end-expiratory pressure (PEEP), 9e11 low FiO 2 , 12,13 and recruitment manoeuvres. 14,15Most of these interventions decrease intraoperative atelectasis but fail to improve clinically relevant postoperative outcomes.16e18 Interestingly, imaging studies suggest similar postoperative atelectasis after extubation regardless of the intraoperative mechanical ventilation strategy used. 16There is paucity of research addressing the weaning and extubation phases.Conditions that contribute to atelectasis formation such as a high FiO 2 and low PEEP settings are often observed during emergence from anaesthesia.Thus, current approaches to weaning from mechanical ventilation could counteract intraoperative strategies, facilitating atelectasis and potentially increasing the incidence of PPCs.
Although most patients undergoing surgery under general anaesthesia are extubated in the operating room, there currently is no definitive standard for weaning of mechanical ventilation.Recommendations from an expert panel-based consensus published in 2019, addressing intraoperative lungprotective mechanical ventilation, generated only weak recommendations for the weaning and extubation phases.Suggested interventions, such as avoiding zero end-expiratory pressure and high inspiratory FiO 2 during the emergence phases, were based on limited evidence. 19In light of these limitations, we completed a systematic review to evaluate studies that proposed different weaning strategies from mechanical ventilation in the operating room and understand their effects on postoperative pulmonary outcomes.

Design
We conducted a systematic review with predetermined selection and outcome criteria.Our review protocol was registered on PROSPERO (CRD42022379145).We searched the following databases: Central, MEDLINE, PubMed, Cochrane Library, Scopus, and LILACS between January 1947 and March 2023.Eligible studies included RCTs and observational studies evaluating strategies addressing weaning from mechanical ventilation among adults undergoing surgery.Additionally, we searched the Clinical Trials Registry Database (https:// clinicaltrials.gov)for registered, unpublished, and ongoing studies evaluating weaning from mechanical ventilation in the operating room.We also searched bibliographies of included studies and review articles.Studies were restricted to English, French, Spanish, and Portuguese.For further information on the search strategy, please refer to Supplementary Appendix 1.
We included a study if it described at least one adult (>18 yr) patient being weaned from the ventilator in the operating room.Our primary outcomes were atelectasis measured by postoperative atelectasis and pulmonary aeration on computed tomography (CT) or lung ultrasound (LUS), and oxygenation through Pa O2 , Pa O2 /FiO 2 , estimated venous admixture, alveolarto-arterial oxygen gradient, and supplemental oxygen use.Secondary outcomes included lung volume changes measured by functional residual capacity through inert gas rebreathing, end-expiratory lung volume with opto-electronic plethysmography, end-expiratory and total lung volume with electrical impedance tomography (EIT), and PPCs measured by the incidence of PPCs.We report our review based on PRISMA guidelines (Supplementary Appendix 2). 20

Mechanical ventilation intervention and control groups
Intervention groups included strategies examining facets of mechanical ventilation such as FiO 2 , PEEP, mode of ventilation, and recruitment manoeuvres.Control groups were defined as receiving either common or non-personalised care.

Study selection
Two authors (MA and NS) independently reviewed retrieved abstracts and assessed eligibility using Covidence systematic review software (Veritas Health Innovation, Melbourne, VIC, Australia).A full-text review was conducted when either reviewer considered an abstract met inclusion criteria.Both reviewers agreed on full texts for inclusion, with an independent third reviewer (SMP) resolving disagreement.
Weaning from ventilation in operating room -425

Data extraction
Data from included studies were independently extracted by MA and SMP.The following data were extracted: study and patient characteristics, mechanical ventilation settings, study interventions, perioperative ventilatory management including induction, intraoperative, and weaning phases, and outcomes.Two independent authors (MA and SMP) independently assessed the risk of bias at the outcome level with regards to randomisation, deviations from intended interventions, missing outcome data, and selection of the reported result using the Cochrane Risk of Bias tool for randomised studies (RoB2). 21A third review author (MCS) resolved any discrepancy that arose in the assessment of the process.Quality of evidence for each study was assessed by MA and MCS using the GRADE framework and are reported by intervention. 22At the time of registration, we considered performing a meta-analysis; however, this was not feasible because of the low number of eligible studies and heterogeneity of interventions and outcomes studied.

Search results
Our search strategy identified 4082 citations that resulted in the inclusion of 14 studies based on screening eligibility (Fig. 1).

Characteristics of included studies
Overall, 14 studies were included; seven studies examined the weaning phase only (Table 1) and the other seven studies addressed induction and maintenance in addition to weaning (Table 2).

Studies that exclusively addressed the weaning phase
23e26 Control groups included varied strategies including zero end-expiratory pressure, intermittent manual assistance bag ventilation, and high FiO 2 .

Title of the study
Reference Population Weaning Postoperative outcome and results

Observation
Control Intervention 1 Intervention 2 The effect of increased FiO 2 before tracheal extubation on postoperative atelectasis PSV with PEEP and a low FiO 2 with a change in bed decubitus improved pulmonary aeration based on LUS scores and reduced the risk of oxygen supplementation (12% vs 58%). 26However, no difference was found in endexpiratory or total lung impedance among three different FiO 2 concentrations with a lung recruitment manoeuvre. 24A PEEP <5 cm H 2 O compared with a PEEP of 10 cm H 2 O did not lead to a change in alveolar-to-arterial pressure difference. 14wo studies examined a single intervention.In patients with chronic obstructive pulmonary disease (COPD), an FiO 2 of 0.3 compared with an FiO 2 of 1.0 improved Pa O2 (8.1 [0.93] vs 16.7 [0.93] vs kPa) in PACU immediately after extubation). 12In another study, zero end-expiratory pressure compared with a PEEP of 7 or 9 cm H 2 O did not reduce the area of postoperative atelectasis or improve the Pa O2 /FiO 2 ratio. 1027e31 Control groups included varied strategies: varied PEEP (ranging from 0 to 10 cm H 2 O); lung recruitment manoeuvres; high FiO 2 ; and both support and controlled ventilation modes.
Low FiO 2 was evaluated in four studies.Lower FiO 2 during induction, maintenance, and weaning led to improved LUS scores and less atelectasis (20% vs 39%). 31A lower FiO 2 during weaning also improved postoperative oxygenation (estimated venous admixture 8.1% vs 14.2%). 28However, the same group published another study where there was no change in atelectasis area when patients were weaned on 0.3 or 1.0 FiO 2 . 13High vs low FiO 2 during maintenance and weaning also did not result in differences in oxygenation or functional residual capacity. 27n improvement in end-expiratory lung volume was found when patients received a combination of lung recruitment manoeuvres and PEEP during maintenance and CPAP with an FiO 2 of 0.4 and with a pressure-limiting valve at 7 cm H 2 O throughout weaning compared with a group receiving FiO 2 1.0 and no CPAP for emergence (þ0.5 L lung volume vs e0.6 L). 30 When using PSV during the weaning phase, individualised PEEP compared with a PEEP of 4 cm H 2 O during maintenance and weaning phases showed a reduction in atelectasis on CT (6.2% nonaerated tissue vs 10.8%). 11In a large multicentre study, individualised PEEP intraoperative strategy with individualised postoperative CPAP suggested an association with reduced PPCs, an exploratory secondary outcome, compared with a standard low tidal volume intraoperative strategy with postoperative oxygen therapy (39% vs 48%). 29

Methodological quality of included studies
A summary of evidence by intervention for each outcome is outlined in Table 4.For the primary outcome, moderate-

Discussion
This systematic review on weaning from perioperative mechanical ventilation demonstrates that limited evidence guides this crucial component of clinical anaesthesia performed for nearly every patient undergoing surgery with general anaesthesia.Few studies have been conducted examining the weaning and emergence phases and their impact on postoperative pulmonary outcomes; only seven studies exclusively assessed the weaning phase, and seven other studies addressed the weaning phase in addition to the induction and maintenance phases.Of the strategies examined, PSV in combination with PEEP and low FiO 2 consistently improved atelectasis, oxygenation, and lung volumes based on    moderate to high quality of evidence and low risk of bias.Low FiO 2 alone improved atelectasis and oxygenation based on low to moderate quality of evidence with risk of bias, but did not demonstrate benefit on lung volumes.A fixed-PEEP strategy was not associated with improvements in atelectasis or oxygenation based on moderate-to high-quality evidence albeit with some concerns of bias; however, an individualised PEEP strategy suggested potential reduced incidence of PPCs but requires further study.Among the interventions studied, PSV in combination with PEEP and low FiO 2 consistently demonstrated improvement in atelectasis, oxygenation, and lung volumes.These findings were in line with its wide use during weaning from mechanical ventilation in ICU.32e34 Specifically, compared with unassisted ventilation, PSV increases driving pressure during inspiration, causing tidal recruitment and promoting lung expansion, while simultaneously decreasing work of breathing. 32,35Moreover, the level of support can be titrated to match ventilatory and oxygenation requirements. 35Among the studies included in this review, PSV was used to maintain PEEP but the level was applied variably: in one study, PSV was used with no clear indication on how the level was set. 11Another study set the level to generate a volume that was similar to the tidal volume used prior to emergence and was maintained until the trachea was extubated. 26Finally, one study applied a PSV intervention initially setting the level at 5 cm H 2 O and gradually adjusting this level according to the resulting respiratory rate and tidal volume. 25Although PSV was set differently and used in various combinations with other strategies, it improved atelectasis, 11,25,26 oxygenation, 25,26 and lung volumes. 30We propose that applying a PSV strategy to maintain tidal volume 25,26 could have a role during weaning from mechanical ventilation in the operating room.The combination of PEEP and low FiO 2 with PSV increases endexpiratory lung volume and counteracts airway closure, which might reverse or prevent atelectasis in patients undergoing surgery. 36vidence suggests that a higher FiO 2 contributes to absorption atelectasis 37 and surfactant impairment. 38Congruent with this, our review found that low FiO 2 alone reduced atelectasis and improved oxygenation but might not improve lung volumes.This is because lung volume, more specifically endexpiratory lung volume, is modified by pressure support and specifically PEEP levels. 30,39Two reasons could explain why low FiO 2 strategies improved atelectasis but not lung volumes.Firstly, we hypothesise that low FiO 2 is not sufficient alone and needs to be combined with another intervention to see positive results as shown with the application of low FiO 2 with PSV and PEEP (whereby PSV and PEEP would increase endexpiratory lung volumes).Secondly, EIT indirectly assesses changes in lung volumes by directly measuring changes in lung impedance, and therefore might not be as not sensitive as CT, which could explain why no significant differences were detected. 24The World Health Organization has previously issued a strong recommendation of an FiO 2 of 0.8 intraoperatively and for 6 h postoperatively for prevention of surgical site infection. 40However, these guidelines have drawn criticism as a result of their generation based on a subgroup analysis of patients under general anaesthesia with tracheal intubation, 40,41 evidence of increased mortality with liberal oxygen therapy in critically ill patients, 42 and their lack of description of harm when observational evidence has suggested increased risk of pulmonary complications with hyperoxia. 43Updated meta-analyses have suggested that evidence favouring high FiO 2 became weaker, 41 but also that there was no definitive evidence of harm from high FiO 2 . 44Our study provides support for lower FiO 2 in the weaning phases of mechanical ventilation in terms of reducing atelectasis and improving oxygenation.
Interventions that used a fixed or non-individualised PEEP strategy did not improve atelectasis or oxygenation.These PEEP strategies might not have adequately ameliorated lung collapse, especially as PEEP requirements vary vastly. 45Fixed PEEP strategies have been associated with inconclusive evidence of benefit. 17,18,39We found that strategies that incorporated individualised PEEP demonstrated improvements in atelectasis and oxygenation. 10,11The optimal PEEP strategy during weaning continues to be subject of debate.Inappropriately set PEEP levels have the potential to induce lung stress in non-dependent areas without effectively addressing atelectasis.Although one included trial examining an open-lung ventilation strategy incorporating individualised PEEP and postoperative CPAP among patients undergoing abdominal surgery suggested a reduction in PPCs, it was not powered for this secondary outcome and any effect should be considered hypothesis-generating.Further, differences in postoperative care could explain any possible effect as the treating team could not be blinded to the use of postoperative CPAP. 29ending support to the positive direct effect hypothesis of an individualised PEEP strategy on PPCs, a recent trial of individualised ventilation comprising a recruitment manoeuvre, individualised PEEP and postoperative respiratory support demonstrated lower risk of severe PPCs among patients undergoing lung resection. 46Further research and definitive trials are therefore required to address which patients benefit from individualised PEEP strategies, and how these strategies need to be implemented in anaesthetic practice both during the intraoperative and emergence phases.
Prior research has demonstrated that alveolar recruitment manoeuvres improved lung mechanics and collapsed lung. 19lthough no specific technique is currently recommended, bag-squeezing is discouraged and ventilator-driven alveolar recruitment manoeuvres should be performed instead. 19ecruitment manoeuvres should be applied using either the shortest effective time and lowest effective pressure or fewest number of breaths. 19In this review, recruitment manoeuvres combined with low FiO 2 improved atelectasis and oxygenation, 23 but not lung volumes, 24 and with PEEP they did not improve oxygenation. 14Our findings do not support their clinical utility; recruitment manoeuvres were not consistently associated with improved physiologic outcomes, and are not safe across all scenarios.A recent RCT in patients with acute respiratory failure demonstrated harm in patients who received a recruitment manoeuvre. 47he results of studies that combined interventions during different perioperative phases are consistent with studies limited to the weaning phase.Specifically, combinations including PSV and low FiO 2 were associated with improvements in oxygenation.Further, a mixed intervention that included PSV, low FiO 2 , and PEEP also improved lung volumes.These results imply that mechanical ventilation strategies during other phases of anaesthesia need to be combined with an intervention during the weaning phase to yield optimal outcomes.Resumption of active expiration without ventilatory support during emergence further decreased lung volumes 30 ; therefore, a combination of techniques to recruit the lungs, reduce atelectasis, and restore diaphragm tone, 48 with proper reversal of neuromuscular blocking agents, could Weaning from ventilation in operating room -433 minimise anaesthesia-induced physiologic changes to the respiratory system.
The strengths of this review include: an extensive literature search examining multiple databases including unpublished studies; inclusion of publications including intraoperative mechanical ventilation in addition to the weaning phase; a pre-registered protocol and analytic plan; and examination of the strength of evidence in addition to risk of bias.There are also several limitations.Firstly, this is an understudied subject and our review comprised only seven studies that addressed the weaning phase exclusively.As the seven other studies analysed included interventions during the induction and maintenance phases, it is possible that these led to significant heterogeneity in observed postoperative pulmonary outcomes.Secondly, our review identified lack of a standard weaning strategy.This lack of a consistent control group limited analyses, as potential facets of mechanical ventilation differed in both intervention and control groups.Thirdly, the limited research to date comprises heterogeneous patient and surgical populations.Patients studied included those with COPD who have compromised pulmonary function; those undergoing abdominal and chest surgery with greater postoperative respiratory derangements; and those undergoing laparoscopic surgery which is associated with marked cardiorespiratory impairment.Therefore, differences in outcomes could be influenced by confounding preoperative and intraoperative factors.Fourthly, outcome definitions were inconsistent, such as studies of oxygenation including Pa O2 , Pa O2 /FiO 2 , and estimated venous admixture.Similarly, studies evaluating atelectasis used both CT, the gold-standard, and LUS, which has limitations in inter-and intra-rater reliability.Fifthly, interventions studied varied from a single approach to multiple strategies in numerous combinations, and this limited the assessments of benefit of each individual intervention.Finally, contemporary definitions of PPC do not incorporate quantitative assessments of atelectasis. 49Consequently, a reduction in atelectasis as reported among included studies might not inherently correspond to reductions in PPC.Nonetheless, such quantitative assessments are useful as atelectasis is a risk factor for further lung injury and reduction in atelectasis has potential to improve PPC. 5,8iven the paucity of studies on this topic, further research is needed to elucidate physiological mechanisms of specific weaning strategies, and their clinical implications.The potential for improved patient outcomes demonstrated in our review emphasises the importance of developing interventions that target this phase of anaesthesia.RCTs focused on developing individualised mechanical ventilation strategies with focus on the weaning phase are urgently needed.In the interim, based on the limited evidence reviewed here, we suggest that clinicians consider weaning patients from mechanical ventilation after surgery with a combination of PSV, individualised PEEP, and low FiO 2 , pending more definitive studies.

Conclusions
This systematic review revealed limited data to guide weaning from intraoperative mechanical ventilation.The lack of a standard approach to weaning and limited evidence guiding this phase of anaesthesia result in considerable discrepancies in care.Based on low-quality evidence from randomised trials of modest sample sizes, the combination of PSV, individualised PEEP, and low FiO 2 during weaning from mechanical ventilation is potentially associated with diminished atelectasis and enhanced oxygenation.

Identification Screening Eligibility Included Fig
1. Identification of literature from search strategy based on inclusion and exclusion criteria.

Table 1
Interventions and physiological outcomes from studies that examined only the weaning period during mechanical ventilation, presented as Control vs Intervention 1 vs Intervention 2. *Values estimated from a figure 237no recruitment manoeuvre.23Patientsthatreceiveda combination of pressure support ventilation (PSV) and PEEP of 5 cm H 2 O demonstrated improved atelectasis (33% vs 57%) and oxygenation[12.3(3.5)vs11.1 (1.7) kPa].

Table 1 Continued
Title of the study

Table 2
Interventions and physiological outcomes from studies that examined the intraoperative and/or maintenance phases in addition to the weaning phase during mechanical ventilation, presented as Control vs. Intervention 1 vs. Intervention 2 vs. Intervention 3. *Values estimated from a figure.CPAP, continuous positive airway pressure; CT, computed tomography; EIT, electrical impedance tomography; FiO 2 , fraction of inspired oxygen; FRC, functional residual capacity; LUS, lung ultrasound; Pa O2 , partial pressure of oxygen in arterial blood; Pa O2 /FiO 2 , ratio of arterial oxygen partial pressure to fractional inspired oxygen; PEEP, positive end-expiratory pressure; PPCs, postoperative pulmonary complications; PSV, pressure support ventilation; VCV, volume-controlled ventilation; ZEEP, zero end-expiratory pressure.

Table 3
Characteristics of patients who underwent various surgical procedures in included studies.Data presented as median (range), n/total N (%), or mean (SD).ASA, American Society of Anesthesiologists; ENT, ear, nose, and throat.

Table 4
Summary of the evidence for each outcome studied in the intervention group compared with the control group with GRADE quality of evidence, risk of bias, and overall effect assessed.*Values estimated from a figure.A-a, alveolar-arterial oxygen pressure difference; ABG, arterial blood gas; CT, computed tomography; EIT, electrical impedance tomography; FiO 2 , fraction of inspired oxygen; FRC, functional residual capacity; LUS, lung ultrasound; PEEP, positive end-expiratory pressure; PPCs, postoperative pulmonary complications; PSV, pressure support ventilation; qLUSS, quantitative lung ultrasound score.