For patients who met exclusion criteria, 2 were due to switching the patient during surgery to the lateral decubitus position and 3 were due to the patient having a body mass index lower than 35 on the day of surgery. Fio2 indicates fraction of inspired oxygen; PEEP, positive end-expiratory pressure.
aThe number of patients assessed for eligibility is not reported because it was not collected at all sites.
bRecruitment maneuvers are an increase in airway pressure through a stepwise increase of tidal volume and eventually PEEP.
The data marker sizes are proportional to the numbers of patients entering the analysis. PEEP indicates positive end-expiratory pressure.
aCalculated as weight in kilograms divided by height in meters squared.
Amendments and changes to the trial protocol
Statistical analysis plan
Lists of committees, investigators, and study sites
eTable 1. The Assess Respiratory Risk in Surgical Patients in Catalonia (ARISCAT) score
eTable 2. Rescue strategies for intraoperative hypoxemia
eTable 3. Types of intraoperative IV fluids
eTable 4. Types of intraoperative vasoactive drugs
eTable 5. Types of anesthetic, muscle paralyzing and reversal agents
eTable 6. Further characteristics of surgery
eTable 7. Assessment of postoperative pain and dyspnea
eTable 8. Per-protocol analysis
eTable 9. Sensitivity analysis for the primary outcome
eFigure 1. Effect size (Z) according to the enrollment of patients in the trial
eFigure 2. Tidal volume during anesthesia
eFigure 3. Positive end-expiratory pressure during anesthesia
eFigure 4. Peak pressure during anesthesia
eFigure 5. Peripheral oxyhemoglobin saturation during anesthesia
eFigure 6. Driving pressure during anesthesia
eFigure 7. Fraction of inspiratory oxygen during anesthesia
eFigure 8. Probability of postoperative pulmonary complications in the first 5 postoperative days
eFigure 9. Probability of severe postoperative pulmonary complications in the first 5 postoperative days
eFigure 10. Probability of postoperative extra-pulmonary complications in the first 5 postoperative days
eFigure 11. Probability of death in the first 5 postoperative days
eFigure 12. Results of the sensitivity analysis for the primary endpoint
Data sharing statement
Customize your JAMA Network experience by selecting one or more topics from the list below.
Writing Committee for the PROBESE Collaborative Group of the PROtective VEntilation Network (PROVEnet) for the Clinical Trial Network of the European Society of Anaesthesiology. Effect of Intraoperative High Positive End-Expiratory Pressure (PEEP) With Recruitment Maneuvers vs Low PEEP on Postoperative Pulmonary Complications in Obese Patients: A Randomized Clinical Trial. JAMA. 2019;321(23):2292–2305. doi:10.1001/jama.2019.7505
Does a high level of positive end-expiratory pressure (PEEP) with alveolar recruitment maneuvers decrease postoperative pulmonary complications in obese patients undergoing surgery compared with a low level of PEEP?
In this randomized trial of 1976 obese adults undergoing noncardiac, nonneurological surgery under general anesthesia, the rate of pulmonary complications was 21.3% among those randomized to a strategy of mechanical ventilation combining alveolar recruitment maneuvers and a higher level of PEEP compared with 23.6% among those randomized to a strategy with a lower level of PEEP without alveolar recruitment maneuvers; however, the difference was not statistically significant.
An intraoperative mechanical ventilation strategy with a higher level of PEEP and alveolar recruitment maneuvers did not reduce postoperative pulmonary complications in obese patients.
An intraoperative higher level of positive end-expiratory positive pressure (PEEP) with alveolar recruitment maneuvers improves respiratory function in obese patients undergoing surgery, but the effect on clinical outcomes is uncertain.
To determine whether a higher level of PEEP with alveolar recruitment maneuvers decreases postoperative pulmonary complications in obese patients undergoing surgery compared with a lower level of PEEP.
Design, Setting, and Participants
Randomized clinical trial of 2013 adults with body mass indices of 35 or greater and substantial risk for postoperative pulmonary complications who were undergoing noncardiac, nonneurological surgery under general anesthesia. The trial was conducted at 77 sites in 23 countries from July 2014-February 2018; final follow-up: May 2018.
Patients were randomized to the high level of PEEP group (n = 989), consisting of a PEEP level of 12 cm H2O with alveolar recruitment maneuvers (a stepwise increase of tidal volume and eventually PEEP) or to the low level of PEEP group (n = 987), consisting of a PEEP level of 4 cm H2O. All patients received volume-controlled ventilation with a tidal volume of 7 mL/kg of predicted body weight.
Main Outcomes and Measures
The primary outcome was a composite of pulmonary complications within the first 5 postoperative days, including respiratory failure, acute respiratory distress syndrome, bronchospasm, new pulmonary infiltrates, pulmonary infection, aspiration pneumonitis, pleural effusion, atelectasis, cardiopulmonary edema, and pneumothorax. Among the 9 prespecified secondary outcomes, 3 were intraoperative complications, including hypoxemia (oxygen desaturation with Spo2 ≤92% for >1 minute).
Among 2013 adults who were randomized, 1976 (98.2%) completed the trial (mean age, 48.8 years; 1381 [69.9%] women; 1778 [90.1%] underwent abdominal operations). In the intention-to-treat analysis, the primary outcome occurred in 211 of 989 patients (21.3%) in the high level of PEEP group compared with 233 of 987 patients (23.6%) in the low level of PEEP group (difference, −2.3% [95% CI, −5.9% to 1.4%]; risk ratio, 0.93 [95% CI, 0.83 to 1.04]; P = .23). Among the 9 prespecified secondary outcomes, 6 were not significantly different between the high and low level of PEEP groups, and 3 were significantly different, including fewer patients with hypoxemia (5.0% in the high level of PEEP group vs 13.6% in the low level of PEEP group; difference, −8.6% [95% CI, −11.1% to 6.1%]; P < .001).
Conclusions and Relevance
Among obese patients undergoing surgery under general anesthesia, an intraoperative mechanical ventilation strategy with a higher level of PEEP and alveolar recruitment maneuvers, compared with a strategy with a lower level of PEEP, did not reduce postoperative pulmonary complications.
ClinicalTrials.gov Identifier: NCT02148692
Up to 18% of obese patients undergoing surgery have postoperative pulmonary complications,1 which is almost twice the risk among normal weight or overweight patients.2,3 Postoperative pulmonary complications prolong hospitalization and increase mortality.3,4 In 2012, it was estimated that more than 310 million surgical procedures were conducted worldwide.5 Given that the global prevalence of obesity is increasing,6,7 the burden of postoperative pulmonary complications will increase if the number of surgical procedures remains unchanged over the coming decades.
Protective intraoperative mechanical ventilation has been associated with reduced incidence of postoperative pulmonary complications.8,9 Among normal weight and overweight patients, low tidal volumes and low levels of positive end-expiratory pressure (PEEP) with alveolar recruitment maneuvers reduced the risk of major pulmonary and extrapulmonary complications compared with high tidal volumes and no PEEP.10 However, when low tidal volumes were used in different ventilator strategies, a higher level of PEEP with alveolar recruitment maneuvers did not reduce the incidence of postoperative pulmonary complications compared with a lower level of PEEP.11,12
Obesity is associated with increased risk of atelectasis and impaired respiratory function during general anesthesia.13 An approach using an intraoperative high level of PEEP and alveolar recruitment maneuvers prevented those alterations,14,15 and has been proposed for routine intraoperative mechanical ventilation in obese patients.13,16 Whether this approach improves postoperative outcomes remains uncertain.
The Protective Intraoperative Ventilation With Higher Versus Lower Levels of Positive End-Expiratory Pressure in Obese Patients (PROBESE) trial was conducted to test whether an intraoperative mechanical ventilation strategy with a higher level of PEEP and alveolar recruitment maneuvers reduces the incidence of postoperative pulmonary complications during the initial 5 postoperative days compared with a lower level of PEEP without alveolar recruitment maneuvers in obese patients undergoing surgery who are at increased risk for these complications.
This was an international, investigator-initiated, assessor-blinded randomized clinical trial. The protocol was published17 and appears in Supplement 1. Amendments and changes to the trial protocol appear in Supplement 2. The final statistical analysis plan that was written prior to locking the database appears in Supplement 3. The institutional review board at each site approved the protocol. Written informed consent was obtained from all participating patients. A data and safety monitoring committee oversaw the conduct of the study and reviewed blinded safety data. Onsite monitoring for adherence to the trial protocol and completeness of data was conducted at the sites that included more than 60 patients.
Patients were included if they had a body mass index (calculated as weight in kilograms divided by height in meters squared) of 35 or greater, were scheduled for a laparoscopic or nonlaparoscopic surgery that was expected to exceed 2 hours under general anesthesia, and had an intermediate to high risk of developing postoperative pulmonary complications as indicated by an Assess Respiratory Risk in Surgical Patients in Catalonia score18 of 26 or greater (eTable 1 in Supplement 4).
Patients were excluded if they were younger than 18 years, previously had lung surgery, had received invasive mechanical ventilation for longer than 30 minutes within the last 30 days prior to surgery, or had received chemotherapy or radiotherapy within 2 months prior to surgery. Additional exclusion criteria included cardiac and neurological surgery, intraoperative one-lung ventilation, planned reintubation after surgery, need for intraoperative prone or lateral decubitus positioning during surgery, or current participation in another interventional study.
In addition, patients were excluded if pregnant or had persistent hemodynamic instability or intractable shock, severe chronic obstructive pulmonary disease, severe cardiac disease, concurrent acute respiratory distress syndrome expected to require prolonged postoperative mechanical ventilation, severe pulmonary hypertension, intracranial injury or tumor, or neuromuscular disease (eMethods in Supplement 4).
Randomization was based on a computer-generated allocation sequence and was performed using a password-protected, encrypted web interface. The 1:1 allocation sequence used permuted, random block sizes of 4, 6, and 8 and was stratified by study site.
Patients received volume-controlled mechanical ventilation with a tidal volume of 7 mL/kg of predicted body weight and were assigned to either (1) a PEEP level of 12 cm H2O with alveolar recruitment maneuvers performed after endotracheal intubation, which were repeated every hour after any disconnection from the mechanical ventilator and before the end of surgery (high level of PEEP group) or (2) a PEEP level of 4 cm H2O (low level of PEEP group) (Figure 1).
Alveolar recruitment maneuvers were standardized17 and consisted of a stepwise increase of tidal volume and, if necessary, PEEP level was increased until an airway plateau pressure between 40 and 50 cm H2O was achieved (eMethods in Supplement 4). All patients received the lowest fraction of inspired oxygen (Fio2), but not less than 0.4, that maintained greater than 92% peripheral oxygen saturation as measured by pulse oximetry (Spo2).
Tidal volume was set based on predicted body weight, which was calculated using standard formulas.19 When Spo2 decreased to 92% or lower, the general strategy was to increase Fio2 in the low level of PEEP group and to increase PEEP in the high level of PEEP group (eTable 2 in Supplement 4).
Other aspects of perioperative care were managed according to each study site’s routine practice; however, optional recommendations also were provided (eMethods in Supplement 4).
The investigators who were responsible for assessing the primary outcomes were blinded to study group assignment. However, the attending anesthesiologists, intraoperative nursing staff, and intraoperative assessors were not blinded to study group assignment.
The primary outcome was a composite of postoperative pulmonary complications and was defined as having occurred if any preselected complication developed within the first 5 postoperative days. The preselected complications included mild, moderate, and severe respiratory failure; acute respiratory distress syndrome; bronchospasm; new pulmonary infiltrates; pulmonary infection; aspiration pneumonitis; pleural effusion; atelectasis; cardiopulmonary edema; and pneumothorax.
The 9 secondary outcomes included (1) the composite of severe postoperative pulmonary complications, (2) postoperative extrapulmonary complications (systemic inflammatory response, sepsis, severe sepsis, septic shock, extrapulmonary infection, coma, acute myocardial infarction, acute kidney failure, disseminated intravascular coagulation, gastrointestinal failure, and hepatic failure), (3) impaired postoperative wound healing, (4) the unexpected need for intensive care unit admission or readmission, (5) the number of hospital-free days at postoperative day 90, the intraoperative adverse events of (6) hypoxemia (defined as oxygen desaturation with Spo2 ≤92% for >1 minute), (7) hypotension (defined as systolic arterial pressure <90 mm Hg for >2 minutes), and (8) bradycardia (defined as heart rate <50 beats/min or a decrease >20% if the heart rate was <50 beats/min before a recruitment maneuver), and (9) in-hospital mortality.
Post hoc outcomes included 5-day mortality, the need for rescue due to desaturation, and the need for vasoactive drugs.
Anticipating a rate of postoperative pulmonary complications of 40% in the low level of PEEP group18,20 and assuming a dropout rate of 5%, it was originally determined that an enrollment of 748 patients would provide 80% power to detect a relative risk of 0.75 for the incidence of postoperative pulmonary complications at a 2-sided α level of .05. Among normal weight and overweight patients undergoing abdominal surgery, an intraoperative protective mechanical ventilation strategy consisting of low tidal volume and low level of PEEP with alveolar recruitment maneuvers was associated with a relative risk of postoperative pulmonary complications between 0.19 and 0.69 compared with a nonprotective strategy.14 Because the present study focused on the effects of PEEP with alveolar recruitment maneuvers and because tidal volume was protective in both groups, a more conservative relative risk of 0.75 was considered to be appropriate by the steering committee while still being clinically relevant.
After blinded data review of 618 patients by the data and safety monitoring committee, the pooled incidence of postoperative pulmonary complications was approximately 20%. Sample size could be recalculated without affecting the type I error rate.21 Therefore, it was conservatively assumed that the rate of postoperative pulmonary complications would be 20% in the low level of PEEP group. Accordingly, 2013 patients would provide 80% power to detect a relative risk of 0.75 for the primary end point at a 2-sided α level of .05, assuming a dropout rate of 5%, and interim analyses for efficacy, harm, and futility at 50%, 75%, and 100% of the total number of patients for which a nonbinding sequential design with stopping rules was used (eFigure 1 in Supplement 4). The data and safety monitoring committee recommended continuation of the trial on the basis of these analyses.
Patients were analyzed on an intention-to-treat basis according to their randomization group. The analysis data set included all patients who were randomized and had general anesthesia for eligible surgery. Because there were no missing data for the primary outcome, only complete case analysis was performed. All patients were followed up for the duration of the trial unless they withdrew consent. In such cases, data were censored at the time that consent was withdrawn. Additional analyses were performed in the per-protocol population that excluded patients with missing mechanical ventilation data and either receiving (1) a PEEP level greater than 4 cm H2O and who had an Fio2 of less than 1.0 in the low level of PEEP group or (2) a PEEP level of less than 12 cm H2O in the high level of PEEP group.
The effect of the intervention on the primary outcome is reported as number and percentage and estimated with risk ratios (RRs) and 95% CIs that were calculated using the Wald likelihood ratio approximation test and the χ2 test for hypothesis testing. The 2-sided α level for the primary outcome was .044 to account for the interim analyses. Time until postoperative pulmonary complications was assessed using Kaplan-Meier survival curves, and reported as hazard ratios and 95% CIs estimated from a Cox proportional hazards model. The Schoenfeld residuals against the transformed time was used to test the proportional hazards assumptions. As a sensitivity analysis, the effect of the intervention on the primary outcome was reestimated using a generalized linear mixed-effects model with a stratification variable (study site) as the random effect. Because the primary outcome was a composite outcome, sensitivity analyses also were performed.
For other binary outcomes, the effect of the intervention was assessed with RRs and 95% CIs that were calculated using the Wald likelihood ratio approximation test and the χ2 test for hypothesis testing. The effect of the intervention on hospital-free days at day 90 was estimated using the t test and reported as the mean difference between groups. The effect of the intervention on 5-day mortality was estimated using Kaplan-Meier curves, and the hazard ratios and 95% CIs were calculated using Cox proportional hazards models without adjustment for covariates. The Schoenfeld residuals against the transformed time were used to test the proportional hazards assumptions. For the secondary outcomes, a significance level of .05 was used without adjusting for multiple comparisons. Because of the potential for type I error due to multiple comparisons, the findings from the analyses of the secondary end points should be interpreted as exploratory.
The treatment effects were analyzed according to the following prespecified subgroups: (1) nonlaparoscopic vs laparoscopic surgery; (2) body mass index less than 40 vs 40 or greater; (3) baseline Spo2 of less than 96% vs 96% or greater; (4) peripheral vs upper abdominal procedures; and (5) waist-to-hip ratio less than 1.0 vs 1.0 or greater. The analyses for the heterogeneity of effects across subgroups used treatment × subgroup interaction terms added to a generalized linear model considering a binomial distribution.
Complete case analysis was performed for all outcomes. Post hoc analyses comparing the number of procedures for rescue due to hypoxemia, and the need for vasoactive drugs in both groups were performed.
Baseline characteristics were reported as counts and percentages, means and standard deviations, or medians and interquartile ranges whenever appropriate. Hypothesis tests were 2-sided at an α level of .05. All analyses were performed using R version 3.4.1 (R Foundation for Statistical Computing).
From July 2014 through February 2018, a total of 2013 adults were randomized (mean age, 48.8 years; 1381 [69.9%] women; 1778 [90.1%] underwent abdominal operations) at 77 sites in 23 countries (a list of the sites appears in Supplement 4). Twenty-nine patients were excluded after randomization, resulting in 1984 patients who met the criteria for the intention-to-treat analysis. Another 8 patients were lost to follow-up after surgery (4 patients in each treatment group). Final follow-up occurred during May 2018.
Therefore, data from 1976 patients were included in the intention-to-treat analysis. Data from 1829 patients were included in the per-protocol analysis. Baseline characteristics of the 2 groups appear in Table 1.
Intraoperative variables appear in Table 2 and in eTables 3-6 in Supplement 4. Tidal volumes were comparable between groups (eFigure 2 in Supplement 4). The mean level of PEEP was 12.0 cm H2O (SD, 1.1 cm H2O) in the high level of PEEP group compared with the mean level of 4.0 cm H2O (SD, 0.5 cm H2O) in the low level of PEEP group (P < .001; eFigure 3 in Supplement 4). In the high level of PEEP group, alveolar recruitment maneuvers were performed in 968 patients (97.9%) after intubation, in 951 patients (96.2%) during the first hour of surgery, and in 968 patients (97.9%) during the last hour of surgery. In the low level of PEEP group, alveolar recruitment maneuvers were performed for rescue purposes in 11 patients (1.1%).
Compared with the low level of PEEP group, peak pressure and Spo2 increased and the driving pressure (ie, plateau pressure minus level of PEEP) and Fio2 decreased in the high level of PEEP group (eFigures 4-7 in Supplement 4). The following did not significantly differ between groups: need for fluids, transfusion of blood products, characteristics of anesthesia, use of epidural analgesia, management of neuromuscular blockade, duration of surgery, and duration of anesthesia. Postoperative pain and dyspnea were comparable between groups (eTable 7 in Supplement 4).
Postoperative pulmonary complications within the first 5 days following surgery occurred in 211 patients (21.3%) in the high level of PEEP group compared with 233 patients (23.6%) in the low level of PEEP group (difference, −2.3% [95% CI, −5.9% to 1.4%]; RR, 0.93 [95% CI, 0.83 to 1.04]; P = .23) (Table 3; eFigure 8 in Supplement 4). The most common postoperative pulmonary complication was mild respiratory failure, which was reported in 135 patients (13.7%) in the high level of PEEP group compared with 154 patients (15.6%) in the low level of PEEP group (difference, −1.9% [95% CI, −5.1% to 1.2%]; RR, 0.92 [95% CI, 0.80 to 1.05]; P = .22).
Pleural effusion occurred in 43 patients (4.3%) in the high level of PEEP group compared with 21 patients (2.1%) in the low level of PEEP group (difference, 2.2% [95% CI, 0.7%-3.8%]; RR, 1.35 [95% CI, 1.14-1.62]; P = .005). The rates of the other components of the primary end point did not significantly differ between the groups. The effect of PEEP level on the occurrence of postoperative pulmonary complications was consistent across subgroups (Figure 2), including nonlaparoscopic vs laparoscopic surgery, body mass index less than 40 vs 40 or greater, baseline Spo2 less than 96% vs 96% or greater, peripheral vs upper abdominal incision procedures, and waist-to-hip ratio less than 1.0 vs 1.0 or greater.
Postoperative secondary outcomes appear in Table 3 and in eFigures 9-11 in Supplement 4. Severe postoperative pulmonary complications, postoperative extrapulmonary complications, unexpected admission to the intensive care unit, the number of hospital-free days at day 90, and mortality during hospital stay did not significantly differ between groups. During the intraoperative period, hypoxemia was less common in the high level of PEEP group, and hypotension and bradycardia were less frequent in the low level of PEEP group.
The results of the per-protocol and the intention-to-treat analysis did not significantly differ (eTable 8 in Supplement 4). The results were unaffected by adjustment for randomization factors (eTable 9 in Supplement 4). Additional sensitivity analyses using different statistical assumptions yielded similar results (eTable 9 and eFigure 12 in Supplement 4).
During the intraoperative period, post hoc analyses showed that the need for a rescue strategy for desaturation was less common in the high level of PEEP group, whereas in the low level of PEEP group, the need for vasoactive drugs was lower (Table 3). In addition, 5-day mortality did not significantly differ between groups.
Among obese patients undergoing surgery, intraoperative mechanical ventilation with a high level of PEEP and alveolar recruitment maneuvers did not reduce postoperative pulmonary complications compared with a low level of PEEP. During the intraoperative period, hypotension was more frequent in patients randomized to the high level of PEEP group, whereas hypoxemia was more common in patients randomized to the low level of PEEP group.
Use of an intraoperative high level of PEEP and alveolar recruitment maneuvers may prevent the development of lung atelectasis,22 decrease the driving pressure,23 homogenize ventilation,15 and minimize the repetitive opening and closing of lung units, which could mitigate the development of pulmonary complications.24 However, use of a high level of PEEP and alveolar recruitment maneuvers can also have adverse effects, including increased static stress and strain,25 inflammation,26 impaired hemodynamics,11 and decreased lung lymphatic drainage.27
The choice of a PEEP level of 4 cm H2O in the low level of PEEP group was based on data that was already available when the study was designed.3 Because a PEEP level of 4 cm H2O was unlikely to provide enough stability for a substantial proportion of lung units being kept open after alveolar recruitment maneuvers, such maneuvers were not routinely performed in that group. The design of the intervention in the high level of PEEP group was consistent with the concept of increasing the availability of lung units for gas exchange and stabilizing them at expiration, while limiting the effect on hemodynamics.22 For this reason, but also because chest wall elastance may increase during surgery, alveolar recruitment maneuvers were repeated at intervals of 1 hour in the high level of PEEP group.
Previous studies reported that PEEP and alveolar recruitment maneuvers improved intraoperative pulmonary function.15,22 However, those studies were inadequately powered to address postoperative pulmonary complications. The finding of the present study that ventilation with a higher level of PEEP and alveolar recruitment maneuvers did not reduce postoperative pulmonary complications is consistent with results among normal weight patients who underwent abdominal surgery, in whom high levels of PEEP with alveolar recruitment maneuvers also did not prevent postoperative pulmonary complications.11,12 Taken together, evidence from major trials indicates that intraoperative mechanical ventilation strategies aiming to reduce atelectasis do not prevent postoperative pulmonary complications compared with a strategy allowing higher degrees of atelectasis (also known as permissive atelectasis).
The finding that intraoperative hypotension and bradycardia were more frequent in patients randomized to higher levels of PEEP is consistent with previous reports among obese patients undergoing bariatric surgery15 and among normal weight patients undergoing abdominal surgery.11 Theoretically, preoperative optimization of intravascular volume might have decreased hemodynamic adverse events. However, this practice is not universally accepted in surgical patients, and given the pragmatic nature of this trial, it was not recommended to be incorporated at the sites. In addition, the observation that hypoxemia occurred more frequently and the rescue strategy for desaturation was needed more often in the lower level of PEEP group is consistent with recent findings from patients who underwent bariatric surgery, had lung reexpansion, and individual titration of PEEP15 and among normal weight and overweight patients.11,12
Therefore, the data from the present study confirm that intraoperative PEEP exerts concurring effects on lung function and circulation. A decrease in the driving pressure among patients in the higher level of PEEP group compared with the lower level of PEEP group did not result in improved clinical outcome measures, which was expected based on previous studies involving patients with acute respiratory distress syndrome28 and conducted in the operating room.8,29 However, such studies were not interventional, suggesting that the driving pressure is a reliable marker for clinical outcome; however, its usefulness as a therapeutic target is still unclear.
The observed incidence of postoperative pulmonary complications was within the range predicted by the Assess Respiratory Risk in Surgical Patients in Catalonia score and comparable with that reported in a previous study focusing on intraoperative mechanical ventilation in obese patients.1 Mild respiratory failure was the most frequent pulmonary complication and also was reported in several studies addressing the incidence of postoperative pulmonary complications.1-4,11,18 Mild respiratory failure was associated with prolonged hospitalization and mortality in the general population4 and in obese patients undergoing surgery1; therefore, it is clinically important.
The overall incidence of pleural effusion was lower than in a previous trial in normal weight and overweight patients,11 but was higher in the higher level of PEEP group compared with the lower level of PEEP group. In a ventilatory strategy with high levels of PEEP, increased hydrostatic forces across the lung capillaries (associated with raised venous pressures and impaired lymphatic drainage) may result in interstitial fluid sequestration.30 Given comparable rates of postoperative pulmonary complications, clinicians can titrate intraoperative PEEP level to optimize oxygenation or to maintain blood pressure as indicated in particular patients.
This study has several strengths. A composite outcome of postoperative pulmonary complications was selected because the complications have been associated with prolonged hospitalization and increased mortality. The interventions were based on the current practice in the lower level of PEEP group,1 and on recommendations for clinical practice in the higher level of PEEP group.13,16 Bias was minimized by using concealed allocation, blinding of outcome assessors, an intention-to-treat analysis, and avoiding loss to follow-up. The sample size was readjusted after a recommendation from the data and safety monitoring committee, maintaining the power to detect clinically significant differences between the groups.
Additional strengths of this trial were that patients were enrolled during a relatively short period, minimizing the influence of changes in clinical practice. Furthermore, patients were enrolled at 77 sites in 23 countries and several types of surgery were included. The present results are thus generalizable.
This study has several limitations. First, intraoperative anesthesiologists could not be blinded to the interventions. However, patients and postoperative assessors were fully blinded to the intraoperative period.
Second, the alveolar recruitment maneuver was based on stepwise increases in tidal volume, which is accompanied by transient increases in the driving pressure. How best to perform alveolar recruitment maneuvers during anesthesia remains unclear, but it is unlikely that results would differ with an alternative approach.
Third, because the trial was pragmatic, individual titration of PEEP level was not attempted. In obese patients undergoing surgery, a PEEP level of 10 cm H2O for open surgery and of 14 cm H2O for laparoscopic surgery represent a reasonable compromise between lung overdistension and collapse.31 A PEEP level of 12 cm H2O in the higher level of PEEP group was chosen as a compromise.
Fourth, respiratory management during emergence and during the immediate postoperative period was not harmonized among sites, but recommendations were given following international standards. Nevertheless, the results of the present study still apply only to intraoperative ventilation management.
Fifth, the composite outcome of postoperative pulmonary complications included events with different degrees of severity. However, even so-called minor pulmonary complications are associated with clinically relevant outcomes.
Sixth, it was not possible to differentiate upper from lower abdominal incisions, which may be associated with different degrees of pulmonary complications. However, in an analysis of patients who underwent abdominal visceral surgery, the interaction between the surgical approach (ie, open vs laparoscopic) was not significant.
Among obese patients undergoing surgery under general anesthesia, an intraoperative mechanical ventilation strategy with a higher level of PEEP and alveolar recruitment maneuvers, compared with a strategy with a lower level of PEEP, did not reduce postoperative pulmonary complications.
Published Online: June 3, 2019. doi:10.1001/jama.2019.7505
Correction: This article was corrected on November 12, 2019, to fix the names for collaborators that were incorrect or missing at the end of the article along with Supplement 4 to fix the names and affiliations that were added or corrected.
Accepted for Publication: May 17, 2019.
Corresponding Author: Marcelo Gama de Abreu, MD, MSc, PhD, DESA, Department of Anesthesiology and Critical Care Medicine, Pulmonary Engineering Group, University Hospital Carl Gustav Carus, Technische Universität Dresden, Fetscherstrasse 74, 01307 Dresden, Germany (firstname.lastname@example.org).
Writing Committee for the PROBESE Collaborative Group: The following investigators take authorship responsibility for the study results: Thomas Bluth, MD; Ary Serpa Neto, MD, MSc, PhD; Marcus J. Schultz, MD, PhD; Paolo Pelosi, MD, FERS; Marcelo Gama de Abreu, MD, MSc, PhD, DESA.
Affiliations of Writing Committee for the PROBESE Collaborative Group: Department of Anesthesiology and Critical Care Medicine, Pulmonary Engineering Group, University Hospital Carl Gustav Carus, Technische Universität Dresden, Germany (Bluth, Gama de Abreu); Department of Critical Care Medicine, Hospital Israelita Albert Einstein, Sao Paulo, Brazil (Serpa Neto); Department of Intensive Care, Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands (Schultz); Department of Surgical Sciences and Integrated Diagnostics, University of Genoa, Policlinico San Martino, Genoa, Italy (Pelosi).
Author Contributions: Drs Bluth and Gama de Abreu had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs Bluth, Serpa Neto, Schultz, Pelosi, and Gama de Abreu contributed equally to this article.
Concept and design: All authors.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: All authors.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: All authors.
Obtained funding: Schultz, Gama de Abreu.
Administrative, technical, or material support: Bluth, Gama de Abreu.
Supervision: All authors.
Conflict of Interest Disclosures: Dr Bluth reported receiving personal fees from Comen Eletronics Technology Co Ltd. Dr Serpa Neto reported receiving personal fees from Drager. Dr Gama de Abreu reported receiving grants and personal fees from Drägerwerk AG and GlaxoSmithKline and receiving personal fees from GE Healthcare. No other disclosures were reported.
Funding/Support: The Clinical Trials Network of the European Society of Anaesthesiology provided financial support for the steering committee meetings, onsite visits to participating sites, for the building of the electronic data capture system, and for the advertising of the study. The Technische Universität Dresden provided logistical support for the coordinating site. The Conselho Nacional de Desenvolvimento Científico e Tecnológico provided financial support for insurance in Brazil. The Association of Anaesthetists of Great Britain and Ireland and the Northern Ireland Society of Anaesthetists provided financial support for the participating sites in the United Kingdom.
Role of the Funder/Sponsor: The funders/sponsors had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
PROBESE Collaborative Group:Steering Committee: Thomas Bluth, Ilona Bobek, Jaume C. Canet, Gilda Cinella, Luc de Baerdemaeker, Cesare Gregoretti, Göran Hedenstierna, Sabrine N. T. Hemmes, Michael Hiesmayr, Markus W. Hollmann, Samir Jaber, John Laffey, Marc-Joseph Licker, Klaus Markstaller, Idit Matot, Gary H. Mills, Jan Paul Mulier, Christian Putensen, Rolf Rossaint, Jochen Schmitt, Mert Senturk, Paolo Severgnini, Juraj Sprung, Marcos Francisco Vidal Melo, Hermann Wrigge, Ary Serpa Neto, Marcus J. Schultz, Paolo Pelosi, and Marcelo Gama de Abreu (chair). Study Coordinating Center: Thomas Bluth and Marcelo Gama de Abreu. PROBESE Investigators: The names are listed in alphabetical order and the affiliations appear in Supplement 4. Fernando Abelha, Sühayla Abitağaoğlu, Marc Achilles, Afeez Adebesin, Ine Adriaensens, Charles Ahene, Fatima Akbar, Mohammed Al Harbi, Rita Al Khoury al Kallab, Xavier Albanel, Florence Aldenkortt, Rawan Abdullah Saleh Alfouzan, Reef Alruqaie, Fernando Altermatt, Bruno Luís de Castro Araujo, Genaro Arbesú, Hanna Artsi, Caterina Aurilio, Omer Hilmi Ayanoglu, Alessandro Bacuzzi, Harris Baig, Yolanda Baird, Konstantin Balonov, Jaume Balust, Samantha Banks, Xiaodong Bao, Mélanie Baumgartner, Isabel Belda Tortosa, Alice Bergamaschi, Lars Bergmann, Luca Bigatello, Elena Biosca Pérez, Katja Birr, Thomas Bluth, Ilona Bobek, Elird Bojaxhi, Chiara Bonenti, Iwona Bonney, Elke M. E. Bos, Sara Bowman, Leandro Gobbo Braz, Elisa Brugnoni, Sorin J. Brull, Iole Brunetti, Andrea Bruni, Shonie L. Buenvenida, Cornelius Johannes Busch, Giovanni Camerini, Jaume C. Canet, Beatrice Capatti, Javiera Carmona, Jaime Carungcong, Marta Carvalho, Anat Cattan, Carla Cavaleiro, Davide Chiumello, Stefano Ciardo, Mark Coburn, Umberto Colella, Victor Contreras, Pelin Corman Dincer, Elizabeth Cotter, Marcia Crovetto, William Darrah, Simon Davies, Luc de Baerdemaeker, Stefan De Hert, Enrique Del Cojo Peces, Ellise Delphin, John Diaper, Paulo do Nascimento Junior, Valerio Donatiello, Jing Dong, Maria do Socorro Dourado, Alexander Dullenkopf, Felix Ebner, Hamed Elgendy, Christoph Ellenberger, Dilek Erdoğan Arı, Thomas Ermert, Fadi Farah, Ana Fernandez-Bustamante, Cristina Ferreira, Marco Fiore, Ana Fonte, Christina Fortià Palahí, Andrea Galimberti, Marcelo Gama de Abreu, Najia Garofano, Luca Gregorio Giaccari, Fernando Gilsanz, Felix Girrbach, Luca Gobbi, Marc Bernard Godfried, Nicolai Goettel, Peter A. Goldstein, Or Goren, Andrew Gorlin, Manuel Granell Gil, Angelo Gratarola, Juan Graterol, Cesare Gregoretti, Pierre Guyon, Kevin Haire, Philippe Harou, Göran Hedenstierna, Antonia Helf, Sabrine N. T. Hemmes, Gunther Hempel, María José Hernández Cádiz, Björn Heyse, Michael Hiesmayr, Markus W. Hollmann, Ivan Huercio, Jasmina Ilievska, Samir Jaber, Lien Jakus, Vijay Jeganath, Yvonne Jelting, Minoa Jung, Barbara Kabon, Aalok Kacha, Maja Karaman Ilić, Arunthevaraja Karuppiah, Ayse Duygu Kavas, Gleicy Keli Barcelos, Todd A. Kellogg, Johann Kemper, Romain Kerbrat, Suraya Khodr, Peter Kienbaum, Bunyamin Kir, Thomas Kiss, Selin Kivrak, Vlasta Klarić, Thea Koch, Ceren Köksal, Ana Kowark, Peter Kranke, Bahar Kuvaki, Biljana Kuzmanovska, John Laffey, Mirko Lange, Marília Freitas de Lemos, Marc-Joseph Licker, Manuel López-Baamonde, Antonio López-Hernández, Mercedes Lopez-Martinez, Stéphane Luise, Mark MacGregor, Danielle Magalhães, Julien Maillard, Patrizia Malerbi, Natesan Manimekalai, Michael Margarson, Klaus Markstaller, Archer K. Martin, David P. Martin, Yvette N. Martin, Julia Martínez-Ocon, Ignacio Martin-Loeches, Emilio Maseda, Idit Matot, Niamh McAuliffe, Travis J. McKenzie, Paulina Medina, Melanie Meersch, Angelika Menzen, Els Mertens, Bernd Meurer, Tanja Meyer-Treschan, Changhong Miao, Camilla Micalizzi, Morena Milić, Gary H. Mills, Norma Sueli Pinheiro Módolo, Pierre Moine, Patrick Mölders, Ana Montero-Feijoo, Enrique Moret, Jan Paul Mulier, Markus K. Muller, Zoe Murphy, Pramod Nalwaya, Filip Naumovski, Paolo Navalesi, Lais Helena Navarro e Lima, Višnja Nesek Adam, Claudia Neumann, Christopher Newell, Zoulfira Nisnevitch, Junaid Nizamuddin, Cecilia Novazzi, Michael O’Connor, Günther Oprea, Mukadder Orhan Sungur, Şule Özbilgin, Maria Caterina Pace, Marcos Pacheco, Balaji Packianathaswamy, Estefania Palma Gonzalez, Fotios Papaspyros, Sebastián Paredes, Maria Beatrice Passavanti, Juan Cristobal Pedemonte, Paolo Pelosi, Sanja Peremin, Christoph Philipsenburg, Daniela Pinho, Silvia Pinho, Linda M. Posthuma, Vincenzo Pota, Benedikt Preckel, Paolo Priani, Christian Putensen, Mohamed Aymen Rached, Aleksandar Radoeshki, Riccardo Ragazzi, Tamilselvan Rajamanickam, Arthi Rajamohan, Harish Ramakrishna, Desikan Rangarajan, Christian Reiterer, J. Ross Renew, Thomas Reynaud, Rhidian Rhys, Eva Rivas, Luisa Robitzky, Rolf Rossaint, Francesca Rubulotta, Humberto S. Machado, Catarina S. Nunes, Giovanni Sabbatini, Jon D. Samuels, Josep Martí Sanahuja, Pasquale Sansone, Alice Santos, Mohamed Sayedalahl, Maximilian S. Schaefer, Martin Scharffenberg, Eduardo Schiffer, Nadja Schliewe, Jochen Schmitt, Raoul Schorer, Marcus J. Schultz, Roman Schumann, Gabriele Selmo, Mar Sendra, Mert Senturk, Ary Serpa Neto, Paolo Severgnini, Kate Shaw, Mirjana Shosholcheva, Abdulrazak Sibai, Philipp Simon, Francesca Simonassi, Claudia Sinno, Nukhet Sivrikoz, Vasiliki Skandalou, Neil Smith, Maria Soares, Tania Socorro Artiles, Diogo Sousa Castro, Miguel Sousa, Savino Spadaro, Juraj Sprung, Emmanouil Stamatakis, Luzius A. Steiner, Andrea Stevenazzi, Alejandro Suarez-de-la-Rica, Mélanie Suppan, Robert Teichmann, José Maria Tena Guerrero, Bram Thiel, Raquel Tolós, Gulbin Tore Altun, Michelle Tucci, Zachary A. Turnbull, Žana Turudić, Matthias Unterberg, Jurgen Van Limmen, Yves Van Nieuwenhove, Julia Van Waesberghe, Marcos Francisco Vidal Melo, Bibiana Vitković, Luigi Vivona, Marcela Vizcaychipi, Carlo Alberto Volta, Anne Weber, Toby N. Weingarten, Jakob Wittenstein, Hermann Wrigge, Piet Wyffels, Julio Yagüe, David Yates, Ayşen Yavru, Lilach Zac, and Jing Zhong.
Data Sharing Statement: See Supplement 5.
Additional Contributions: We are indebted to all multidisciplinary team members at the study sites for their enthusiastic support and for diligently following the study interventions. We also thank the members of the data and safety monitoring committee: Daniel Sessler, MD (chair), Jennifer Hunter, MBE, MB ChB, PhD, Jeanine Wiener-Kronish, MD, Jean-Louis Vincent, MD, PhD, and Andreas Hoeft, MD, PhD. We also thank Edward Mascha, PhD (Cleveland Clinic Lerner College of Medicine at Case Western Reserve University, Cleveland, Ohio), for support in the study design and in the sample size estimations. None of the persons listed in this section received compensation.