British Journal of Anaesthesia, 133 (5): 1073e1084 (2024)

doi: 10.1016/j.bja.2024.08.005
Advance Access Publication Date: 11 September 2024
Clinical Investigation

RESPIRATION AND THE AIRWAY

Association between inspired oxygen fraction and development of

postoperative pulmonary complications in thoracic surgery: a
multicentre retrospective cohort study
Nicholas J. Douville1,2,3 , Mark E. Smolkin4 , Bhiken I. Naik5 , Michael R. Mathis1,2 ,
1,2 1,2 5 6
Douglas A. Colquhoun , Sachin Kheterpal , Stephen R. Collins , Linda W. Martin ,
7 8 5,
Wanda M. Popescu , Nathan L. Pace , Randal S. Blank * on behalf of the Multicentre
Perioperative Clinical Research Committeey
1
Department of Anesthesiology, Michigan Medicine, Ann Arbor, MI, USA, 2Institute of Healthcare Policy & Innovation,
University of Michigan, Ann Arbor, MI, USA, 3Computational Medicine and Bioinformatics, University of Michigan, Ann
Arbor, MI, USA, 4Department of Public Health Sciences, Division of Biostatistics, University of Virginia, Charlottesville,
VA, USA, 5Department of Anesthesiology, University of Virginia School of Medicine, Charlottesville, VA, USA, 6Division of
Thoracic and Cardiovascular Surgery, Department of Surgery, University of Virginia School of Medicine, Charlottesville,
VA, USA, 7Department of Anesthesiology, Yale School of Medicine, New Haven, CT, USA and 8Department of
Anesthesiology, The University of Utah, Salt Lake City, UT, USA

*Corresponding author. E-mail: rsb8p@uvahealth.org

y
The authors gratefully acknowledge the valuable contributions to protocol and final manuscript review by the Multicenter Perioperative Outcomes
Group (MPOG) collaborators, including: Michael Aziz (Department of Anesthesiology and Perioperative Medicine and Oregon Health & Science
University), Justin D. Blasberg (Department of Surgery, Yale New Haven Hospital), Andrew C. Chang (Department of Surgery, University of Michigan),
Robert E. Freundlich (Department of Anesthesiology, Vanderbilt University Medical Center), Vikas O’Reilly-Shah (Department of Anesthesiology &
Pain Medicine, University of Washington) and Robert B. Schonberger (Department of Anesthesiology, Yale School of Medicine). Full list of
collaborators affiliations and ICMJE-defined activities can be found in Supplementary Appendix S1.

Abstract
Background: Limited data exist to guide oxygen administration during one-lung ventilation for thoracic surgery. We
hypothesised that high intraoperative inspired oxygen fraction during lung resection surgery requiring one-lung venti-
lation is independently associated with postoperative pulmonary complications (PPCs).
Methods: We performed this retrospective multicentre study using two integrated perioperative databases (Multicenter
Perioperative Outcomes Group and Society of Thoracic Surgeons General Thoracic Surgery Database) to study adult
thoracic surgical procedures using one-lung ventilation. The primary outcome was a composite of PPCs (atelectasis,
acute respiratory distress syndrome, pneumonia, respiratory failure, reintubation, and prolonged ventilation >48 h).
The exposure of interest was high inspired oxygen fraction (FiO2), defined by area under the curve of a FiO2 threshold >
80%. Univariate analysis and logistic regression modelling assessed the association between intraoperative FiO2 and
PPCs.
Results: Across four US medical centres, 141/2733 (5.2%) procedures conducted in 2716 patients (55% female; mean age 66
yr) resulted in PPCs. FiO2 was univariately associated with PPCs (adjusted OR [aOR]: 1.17, 95% confidence interval [CI]:
1.04e1.33, P¼0.012). Logistic regression modelling showed that duration of one-lung ventilation (aOR: 1.20, 95% CI:
1.03e1.41, P¼0.022), but not the time-weighted average FiO2 (aOR: 1.01, 95% CI: 1.00e1.02, P¼0.165), was associated with
PPCs.
Conclusions: Our results do not support limiting the inspired oxygen fraction for the purpose of reducing postoperative
pulmonary complications in thoracic surgery involving one-lung ventilation.

Received: 3 March 2024; Accepted: 7 August 2024

© 2024 British Journal of Anaesthesia. Published by Elsevier Ltd. All rights are reserved, including those for text and data mining, AI training, and similar
technologies.
For Permissions, please email: permissions@elsevier.com

1073
 1074 - Douville et al.

Keywords: hyperoxia; inspired oxygen fraction; lung resection; one-lung ventilation; postoperative pulmonary compli-
cations; protective ventilation; single-lung ventilation; thoracic surgery

associated with postoperative pulmonary complications after

lung resection surgery.
Editor’s key points
 The dose and duration of higher inspired oxygen
fraction required during one-lung ventilation for Methods
thoracic surgery is poorly evidence-based. Study design
 The authors examined whether high intraoperative
inspired oxygen fraction during lung resection sur- This retrospective observational cohort study was approved
gery requiring one-lung ventilation may account for by the University of Virginia Institutional Review Board
postoperative pulmonary complications, using a (21039) and the REporting of studies Conducted using
large mutlicentre database. Observational Routinely-collected health Data (RECORD)
 Postoperative pulmonary complications occurred in guidelines, an extension of the existing STrengthening the
141/2733 (5.2%) patients. Reporting of OBservational studies in Epidemiology (STROBE),
 Logistic regression modelling failed to find any link were followed.10 Informed patient consent was waived
with time-weighted average FiO2 and postoperative because no patient care interventions were involved in the
pulmonary complications. conduct of the study. Additionally, the Institutional Review
 By considering the individual effects of duration of Board of each contributing organisation approved aggrega-
exposure and FiO2, these results do not support tion of this limited dataset. Study methods including data
limiting the inspired oxygen fraction for the purpose collection, outcomes, and statistical analyses were estab-
of reducing postoperative pulmonary complications. lished a priori, reviewed and approved by the Multicenter
Perioperative Outcomes Group (MPOG) peer-review commit-
tee (July 13, 2020).11
Oxygen is one of the most widely used therapies during
anaesthesia and mechanical ventilation but might also
contribute to lung injury and postoperative pulmonary com- Inclusion criteria
plications via a number of potential pathways.1 In general, Adult patients (18 yr) who underwent a lung resection of at
clinical evidence directly linking alveolar hyperoxia to post- least 60 min duration and utilising one-lung ventilation be-
operative pulmonary complications is mixed and there is a tween 2012 and 2020 were included in this observational
paucity of evidence in the thoracic surgical population.2,3 cohort study. Additionally, included cases had to meet a
Large observational studies of surgical patients have identi- previously validated Perioperative Research Standard for data
fied inspired oxygen fraction (FiO2) as a risk factor for respi- completeness and quality (Supplementary Appendix S2)12
ratory complications4 and mortality5 as well as acute and have an accompanying STS General Thoracic Surgical
respiratory distress syndrome (ARDS)6 in some but not all in- Database (STS-GTSD) record. In addition to patient charac-
vestigations.7 Recent meta-analyses of randomised trials8,9 teristic elements (age, sex, ASA physical status) and validated
failed to demonstrate a link between higher FiO2 and pulmo- case elements (anaesthesia start and end times), the Periop-
nary complications; however, this previous work did not focus erative Research Standard requires submission of both labo-
on the thoracic surgery population. Patients presenting for ratory results and discharge diagnoses, which not every
pulmonary resection surgery may serve as a sensitive clinical MPOG institution submitted at all points in our study window
model for the genesis of hyperoxia-related complications for a (Figure 1).
number of reasons. These include pre-existing pulmonary
disease, resection of functional lung parenchyma, the detri-
Exclusion criteria
mental effects of one-lung ventilation including oxidative
stress and the potential for ventilation-induced lung injury. Exclusion criteria were (i) prior one-lung ventilation within 90
Best-practice strategies for administration of intraoperative days, (ii) procedures requiring cardiopulmonary bypass or
FiO2 during thoracic surgery remain unknown. extracorporeal circulation, and (iii) lung transplantation.
To assess the association between high FiO2 during Pneumonectomies were also excluded from the analysis, as
thoracic surgery and postoperative pulmonary complications these were a major cause of the numerical imbalance and
we performed a retrospective multicentre cohort study. By involved a drastically different set of procedural consider-
utilising both the Multicenter Perioperative Outcomes Group ations (duration of surgery, severity of disease, and physio-
(MPOG) and Society of Thoracic Surgeons General Thoracic logic stressors) than the other classes. The inclusion/
Surgery Databases (STS-GTSD), we were able to integrate exclusion criteria for the study were designed to capture a
intraoperative parameters with adjudicated registry out- wide sample of intraoperative oxygenation practices within
comes. We hypothesised that intraoperative inspired oxygen the thoracic surgical population, and therefore included a
fraction (assessed as area under the curve of FiO2 above the diverse set of lung resection procedures to allow meaningful
80% threshold) during one-lung ventilation is independently evaluation of clinical practice.
 Alveolar hyperoxia and postoperative pulmonary complications - 1075

Exposure of interest
Inspired oxygen concentration
16 028 Adult patients undergoing surgical procedure
(duration ≥ 60 min) utilizing one-lung The primary metric for inspired oxygen concentration was
ventilation at participating MPOG institution
defined a priori as area under the curve (AUC) of theFiO2 above
and clinical data available in the STS-GTSD
the 80% threshold (percentage*minutes).5 AUC of FiO2 above
registry
80% was selected a priori, in an effort to most accurately cap-
ture the cumulative exposure experienced by a patient (dose
1086 Multiple qualifying cases times duration). This definition is analogous to that used in a
recent large multicentre observational study of hyperoxia in
Does not meet MPOG research surgeries excluding those utilising one-lung ventilation.5 The
2936 standard; CPB/ECMO; spine; base unit for the AUC of FiO2 above 80% logistic regression
pneumonectomy; transplant models is 1000%*minutes, meaning FiO2¼90% for a 60-min
case (minus the FiO2 80% threshold) would translate to a
3360 One-lung ventilation start or stop base unit of 10% times 60 min, or 600%*minutes, or 0.6 in terms
time not available of 1000%*minutes.
Because the primary exposure variable intrinsically con-
PPC outcome missing; limited tains both FiO2 and duration of one-lung ventilation, we
5913
contribution health system cannot directly resolve the independent effect of each; there-
fore, alternative definitions for high FiO2, specifically, time in
2733 Cases included in study minutes with FiO2 >80% and peripheral capillary oxygen
saturation (SpO2) >98% (time-discretionary high FiO2), and
time-weighted average FiO2 were also tested in sensitivity
Fig 1. Derivation of the study cohort. MPOG, Multicenter Peri- analyses. The three definitions for the exposure variable: (i)
operative Outcomes Group; STS-GTSD, Society of Thoracic AUC for FiO2 above 80% threshold (Model 1), (ii) time-
Surgeons General Thoracic Surgical Database. discretionary high FiO2 (Model 2), and (iii) time-weighted
average FiO2 (Model 3) are illustrated in Figure 2.
Additionally, sensitivity analyses were performed using the
primary high FiO2 variable (AUC of FiO2 above 80%) (i) during
Data sources the entire period of mechanical ventilation (as opposed to just
the period of one-lung ventilation) and (ii) excluding patients/
We collected study data from two sources: the MPOG registry13
cases who experienced intraoperative desaturation (SpO2
and the STS-GTSD14 that were integrated as previously
<88% for more than three consecutive minutes). Within the
described15 (Supplementary Tables S1 and S2). The STS data
high FiO2 metric, discretionary high FiO2 was calculated as
were used to obtain information for candidate covariates
time above the concurrent FiO2 >80% and SpO2 >98%
based on previously published thoracic surgery models16e19
threshold. Although PaO2 data are generally only available for
and for postoperative outcome data. GTSD records were
patients with indwelling arterial cannulae, we characterised
linked to MPOG records using patient-level identifiers at each
the association between both (i) FiO2 vs PaO2 and (ii) SpO2 vs
participating site as previously described.15
PaO2 in the subset of patients for whom PaO2 data were avail-
able. To assist in visualisation of these relationships PaO2
Primary outcome measurements were classified based on FiO2 (80%) and SpO2
(98%) threshold pairings.
The primary outcome was a composite of postoperative pul-
monary complications from the STS Thoracic Database,
including pneumonia, atelectasis, ARDS, respiratory failure, Covariates
reintubation, and ventilator support for >48 h. Diagnostic Patient characteristics, preoperative, anaesthetic, surgical,
criteria from the STS-GTSD Data Dictionary for each of the comorbidity, and exposure data were derived from the MPOG
composite postoperative pulmonary complications can be Database (Supplementary Table S2a) and STS Thoracic Data-
found in Supplementary Appendix S3. These outcome vari- base (Supplementary Table S2b). Additional comorbidities
ables were chosen based on known or suspected relationships were derived using Elixhauser classifications20 applied to In-
to the exposure (high FiO2) or to any of the other endpoints ternational Classification of Diseases billing information ob-
related to the exposure. tained from the MPOG Database21 (Supplementary Table S2c).
Because atelectasis differs markedly with regard to severity Additional details on how each covariate was defined
and clinical significance, when compared with the other out- (including specifications from the STS-GTSD Data Dictionary
comes (pneumonia, ARDS, respiratory failure, and prolonged and standardised MPOG phenotypes) can be found in
ventilator support), post hoc sensitivity analysis was also per- Supplementary Appendix S4, including how missing forced
formed using criteria for composite postoperative pulmonary expiratory volume in 1 s (FEV1) data were handled.22e26
complications that did not include the diagnosis of atelectasis.

Statistical analyses
Secondary outcomes
It should be noted that 806 (29.5%) of the cases in our cohort
We also recorded non-pulmonary postoperative complica- have been previously reported in a 2021 study on low tidal
tions including death (Supplementary Table S1). volume during one-lung ventilation.15 There is no patient
 1076 - Douville et al.

a
100 100

80% FiO2 threshold

Patient FiO2 administered
90 90
Patient SpO2 measured
Model 1
AUC of FiO2 above 80%
FiO2 administered

80 80 (652% * min)
Model 2

SpO2
Time with FiO2 > 80% &
SpO2 > 98% (7 min)
70 70 Model 3
Model 3
Time–weighted average
FiO2 (67%)

60 60

50 Model 2 50

0 50 100 150
Duration of one-lung ventilation (min)

b
100 100

80% FiO2 threshold

90 90
Patient FiO2 administered
Model 3
Patient SpO2 measured
Model 1
FiO2 administered

80 80 AUC of FiO2 above 80%

(1475% * min)
SpO2
Model 2
70 70 Time with FiO2 > 80% &
SpO2 > 98% (12 min)
Model 3
Time–weighted average
60 60 FiO2 (86%)

Model 2
50 50

0 50 100 150
Duration of one-lung ventilation (min)

Fig 2. Exposure variables for hypothetical cases of desaturation during one-lung ventilation. (a) Illustration of exposure variables for a
hypothetical case depicting marginal oxygen saturation at the initiation of one-lung ventilation followed by improvement and subsequent
weaning of FiO2. (b) Illustration of exposure variables for a hypothetical case depicting severe oxygen desaturation (such as might be due to
lung isolation device malposition), followed by correction and subsequent less aggressive weaning of FiO2. AUC, area under the curve; FiO2,
inspired oxygen fraction; SpO2, peripheral capillary oxygen saturation.

overlap with an earlier 2016 study.19 Perioperative and intra- MPOG (Supplementary Fig. S1). All analyses were conducted
operative characteristics were summarised using medians using SAS 9.4 (SAS Institute, Cary, NC, USA) or R version 3.6.0
and interquartile ranges for continuous variables and counts (R Core Team, Vienna, Austria).
and percentages for categorical covariates. Comparisons of
continuous data were made using ManneWhitney U tests and
Multivariable logistic regression: inspired oxygen
categorical data were compared using Pearson chi-squared
fraction (primary model)
tests. P-value <0.05 was selected a priori to denote statistical
significance. Logistic regression modelling was performed for We assessed the association between intraoperative FiO2 and
multivariable assessment of the primary outcomes, with postoperative pulmonary complications using logistic regres-
additional covariates selected a priori using all data available in sion modelling (Model 1: AUC of FiO2 above 80%). Association
Table 1 Comparison of preoperative covariates. AUC, area under the curve; FEV1, forced expiratory volume in 1 s; FiO2, inspired oxygen fraction; MPOG, Multicenter Perioperative
Outcomes Group.

Variable Level n Overall 0e25% of AUC 25e50% of AUC 50e75% of AUC 75e100% of P-value
n¼2716 n¼679 n¼678 n¼680 AUC n¼679

Total area (%*min) above 2716 814.0 (287.5, 1862.2) 65.5 (3.6, 168.3) 537.8 (416.5, 680.5) 1227.4 (993.5, 1549.2) 2722.5 (2248.0, 3310.3) -
FiO2¼80%: primary
exposure variable
Age(yr) 2716 66.0 (57.0, 73.0) 65.0 (55.0, 73.0) 65.0 (55.0, 72.0) 66.0 (58.0, 73.0) 68.0 (60.0, 74.0) <0.001
Sex Female 2716 1504 (55.4%) 397 (58.5%) 384 (56.6%) 363 (53.4%) 360 (53.0%) 0.128
Male 1212 (44.6%) 282 (41.5%) 294 (43.4%) 317 (46.6%) 319 (47.0%)
Race American Indian 2716 5 (0.2%) 2 (0.3%) 1 (0.1%) 0 (0.0%) 2 (0.3%) <0.001
or Alaska Native
Asian or Pacific Islander 147 (5.4%) 48 (7.1%) 44 (6.5%) 35 (5.1%) 20 (2.9%)
Black, not of 113 (4.2%) 38 (5.6%) 23 (3.4%) 25 (3.7%) 27 (4.0%)
Hispanic origin
Hispanic, Black 1 (0.0%) 0 (0.0%) 0 (0.0%) 1 (0.1%) 0 (0.0%)
Hispanic, White 55 (2.0%) 17 (2.5%) 16 (2.4%) 15 (2.2%) 7 (1.0%)
Middle Eastern 1 (0.0%) 0 (0.0%) 0 (0.0%) 1 (0.1%) 0 (0.0%)

Alveolar hyperoxia and postoperative pulmonary complications

Unknown race 139 (5.1%) 42 (6.2%) 33 (4.9%) 36 (5.3%) 28 (4.1%)
White, not of 2255 (83.0%) 532 (78.4%) 561 (82.7%) 567 (83.4%) 595 (87.6%)
Hispanic origin
MPOG Institution MPOG Institution 1 2716 708 (26.1%) 110 (16.2%) 160 (23.6%) 177 (26.0%) 261 (38.4%) <0.001
MPOG Institution 2 39 (1.4%) 8 (1.2%) 9 (1.3%) 10 (1.5%) 12 (1.8%)
MPOG Institution 3 315 (11.6%) 138 (20.3%) 65 (9.6%) 47 (6.9%) 65 (9.6%)
MPOG Institution 4 1654 (60.9%) 423 (62.3%) 444 (65.5%) 446 (65.6%) 341 (50.2%)
Year of surgery 2012 2716 121 (4.5%) 15 (2.2%) 33 (4.9%) 29 (4.3%) 44 (6.5%) <0.001
2013 182 (6.7%) 18 (2.7%) 52 (7.7%) 52 (7.6%) 60 (8.8%)
2014 150 (5.5%) 30 (4.4%) 22 (3.2%) 41 (6.0%) 57 (8.4%)
2015 498 (18.3%) 113 (16.6%) 118 (17.4%) 120 (17.6%) 147 (21.6%)
2016 638 (23.5%) 163 (24.0%) 157 (23.2%) 159 (23.4%) 159 (23.4%)
2017 400 (14.7%) 136 (20.0%) 93 (13.7%) 89 (13.1%) 82 (12.1%)
2018 316 (11.6%) 93 (13.7%) 90 (13.3%) 72 (10.6%) 61 (9.0%)
2019 390 (14.4%) 107 (15.8%) 110 (16.2%) 110 (16.2%) 63 (9.3%)
2020 21 (0.8%) 4 (0.6%) 3 (0.4%) 8 (1.2%) 6 (0.9%)
ASA physical status 1 2716 4 (0.1%) 2 (0.3%) 2 (0.3%) 0 (0.0%) 0 (0.0%) 0.007
2 401 (14.8%) 92 (13.5%) 110 (16.2%) 123 (18.1%) 76 (11.2%)
3 2181 (80.3%) 551 (81.1%) 532 (78.5%) 528 (77.6%) 570 (83.9%)
4 130 (4.8%) 34 (5.0%) 34 (5.0%) 29 (4.3%) 33 (4.9%)
Height (cm) 2657 167.0 (160.0, 175.0) 166.5 (160.0, 174.0) 167.0 (160.0, 175.1) 167.0 (160.0, 175.3) 165.5 (159.0, 175.3) 0.653
Weight (kg) 2672 77.0 (65.0, 90.0) 70.8 (62.0, 84.5) 76.9 (64.7, 90.0) 79.2 (66.2, 91.4) 80.3 (68.1, 94.6) <0.001
BMI (kg m2) 2656 27.3 (24.0, 31.1) 25.8 (22.6, 29.1) 27.4 (24.1, 30.8) 27.7 (24.4, 31.3) 28.7 (25.3, 33.1) <0.001
Predicted body weight (kg) 2656 59.6 (52.4, 69.7) 59.3 (52.4, 68.7) 60.1 (52.4, 69.6) 60.5 (52.4, 70.7) 59.3 (52.4, 70.7) 0.703
Tobacco use 1 Non-smoker 2715 845 (31.1%) 244 (36.0%) 244 (36.0%) 202 (29.7%) 155 (22.8%) <0.001
2 Prior smoker 1653 (60.9%) 368 (54.3%) 389 (57.4%) 424 (62.4%) 472 (69.5%)
3 Current smoker 217 (8.0%) 66 (9.7%) 45 (6.6%) 54 (7.9%) 52 (7.7%)
Preoperative comorbidities N 2714 2128 (78.4%) 552 (81.4%) 539 (79.6%) 535 (78.7%) 502 (73.9%) 0.007
Y 586 (21.6%) 126 (18.6%) 138 (20.4%) 145 (21.3%) 177 (26.1%)

-
FEV1 missing N 2716 2466 (90.8%) 599 (88.2%) 595 (87.8%) 619 (91.0%) 653 (96.2%) <0.001

1077
Y 250 (9.2%) 80 (11.8%) 83 (12.2%) 61 (9.0%) 26 (3.8%)
FEV1 predicted 2466 92.0 (78.0, 105.0) 93.0 (80.0, 105.0) 93.0 (79.0, 106.0) 93.0 (81.0, 106.0) 90.0 (73.0, 103.0) <0.001

Continued
 1078 - Douville et al.

of the outcome with this exposure was evaluated for the

period of one-lung ventilation, and for the entire period of
P-value

<0.001

<0.001

<0.001
mechanical ventilation (AUC of FiO2 > 80%). Model discrimi-
0.017
nation is reported using the c-statistic. Measures of effect for
model covariates are reported as adjusted odds ratios with
95% confidence intervals. With regard to treatment of missing
values at the modelling stage, no imputation was performed
and only patients with complete data for all variables used in a
75e100% of
AUC n¼679

1.3 (0.9, 1.8)

particular model were included in that model. Details on the
585 (86.2%)

496 (73.0%)

107 (15.8%)
482 (71.0%)
197 (29.0%)
94 (13.8%)
31 (4.6%)

45 (6.6%)

missing number (and percentage) of cases for each of the

variables included in our final regressions is provided in
Supplementary Table S3. FEV1 was included as a risk predictor
in our models; however, as absence of these data is likely to be
non-random, missing FEV1 data was included as a covariate in
regression models when FEV1 data were unavailable. We
50e75% of AUC

handled missing FEV1 as previously described.19 This

0.9 (0.7, 1.2)
611 (89.9%)

265 (39.0%)

349 (51.3%)
556 (81.8%)
124 (18.2%)

approach leaves the FEV1 coefficient identical to the condition

69 (10.1%)
11 (1.6%)

55 (8.1%)

in which data were restricted to only patients with non-

n¼680

missing FEV1 and separately models risk for those patients

with missing FEV1 data. In each model, medical centre was
treated as a fixed effect with multiple categories rather than a
25e50% of AUC

random effect due to the limited number of centres (n¼4)

available in our study.
0.8 (0.6, 1.2)
604 (89.1%)

252 (37.2%)

359 (52.9%)
582 (85.8%)
74 (10.9%)

96 (14.2%)
11 (1.6%)

56 (8.3%)
n¼678

Sensitivity analyses
Preplanned sensitivity analyses included evaluation of the
impact of time-discretionary high FiO2 (Model 2) and evalua-
tion of a time-weighted average (Model 3) plus four further
0e25% of AUC

models (Models 4e7), including a post hoc sensitivity analysis

1.0 (0.7, 1.4)
576 (85.0%)
102 (15.0%)

289 (42.6%)

307 (45.2%)
551 (81.1%)
128 (18.9%)

during the review process, are provided in the Supplementary

16 (2.4%)

67 (9.9%)
n¼679

material.

Results
Study population characteristics
A total of 2733 procedures were included in the final analysis
2715 2376 (87.5%)

1302 (47.9%)

1122 (41.3%)
2716 2171 (79.9%)

2716 1.0 (0.7, 1.4)

339 (12.5%)

545 (20.1%)
223 (8.2%)

(Fig. 1), undertaken in 2716 patients (Table 1). Sixty cases were
2716 69 (2.5%)
n¼2716
Overall

included within the categorical grouping of ‘missing’ as a

result of the absence of recorded/available BMI data.27 Patients
were ventilated with a median dynamic driving pressure12 of
16.7 cm H2O (interquartile range [IQR] 13.9e19.6) and median
n

PEEP 5.0 cm H2O (4.5e5.2). Further ventilatory details are pro-

vided in Supplementary Table S4.

Distribution of exposure of interest: FiO2 >80%

Wedge resection
Segmentectomy

The median of AUC of FiO2 > 80% (primary exposure variable,

Bilobectomy
Lobectomy

Model 1) across all cases was 814.0%*minutes (IQR:

287.5e1862.2%*minutes). Illustrative perioperative traces
Level

(Fig. 2) and the distribution of the exposure variables used in

N

N
Y

each model are shown in Supplementary Figure S2.

Very high FiO2 levels were used early in one-lung ventila-
Prior chemotherapy or radiation

tion and FiO2 declined over the case duration (Fig. 3a). Mean
FiO2 at reinflation following completion of one-lung ventila-
tion was 82.7%. Discretionary high FiO2 was common and
frequent during lung resection, with 60% of cases using FiO2
Total fluid input (L)

>0.80 (despite SpO2 >98%) at the start of one-lung ventilation

Table 1 Continued

(Fig. 3b). Intraoperative PaO2 values were only available in 198

Thoracotomy

cases.
Procedure
Variable

Primary outcome
A postoperative pulmonary complication was experienced
by141/2733 (5.2%) (Table 2).
 Alveolar hyperoxia and postoperative pulmonary complications - 1079

Univariate analyses
a The incidence of postoperative pulmonary complications
100 (P¼0.005) and major morbidity (P<0.001) increased with each
quartile of higher FiO2 exposure (Supplementary Table S4).
The highest FiO2 exposure quartile featured higher driving
90 pressure and tidal volumes, compared with the lowest quar-
tile. There were no differences in PEEP utilisation between FiO2
exposure quartiles (P¼0.198). The time above 80% FiO2 and
FiO2

80 SpO2 of 98% during one-lung ventilation (P<0.001) and time-

weighted average FiO2 also increased with increasing AUC
above the FiO2 80% threshold (Supplementary Table S4). A
detailed comparison between the cohort developing pulmo-
70
nary complications and those not developing pulmonary
complications can be found in Supplementary Table S5a
(preoperative covariates) and Supplementary Table S5b
60 (intraoperative events and outcomes).
0 30 60 90 120 150 180
One-lung ventilation time (minutes) Multivariable logistic regression models
25th percentile 75th percentile
Primary definition for inspired oxygen concentration (area
50th percentile
under the curve above FiO2¼80% threshold during one-
b lung ventilation) (Model 1)
100 Exposure to supplemental oxygen (AUC above FiO2 80%
% of cases meeting criteria

90 threshold) was associated with postoperative pulmonary

80 complications (adjusted OR [aOR]: 1.17, 95% confidence inter-
70 val [CI]: 1.04e1.33, P¼0.012; Table 3). ASA Physical Status
Classification System (base unit: 1-point increase in ASA sta-
60
tus) (aOR: 2.27, 95% CI: 1.40e3.68, P<0.001), past smoking his-
50
tory (aOR: 1.97, 95% CI: 1.16e3.35, P¼0.012), thoracotomy (aOR:
40 2.47, 95% CI: 1.63e3.74, P<0.001), FEV1 in litres) (aOR: 0.98, 95%
30 CI: 0.97e0.99, P<0.001), and patient’s comorbidities (aOR: 1.55,
20 95% CI: 1.04e2.32, P¼0.032) were also associated with post-
10 operative pulmonary complications. The AUC of FiO2 above
80% model had good discrimination (c-statistic ¼ 0.769).28
0
0 30 60 90 120 150 180
One-lung ventilation time (minutes) Sensitivity analyses
Discretionary oxygen administration There was no association between time above the FiO2¼80%
(SpO2 > 98%) and SpO2¼98% thresholds during one-lung ventilation (Model
2; Supplementary Table S6) or time-averaged FiO2 (Model 3;
FiO2 > 0.4 FiO2 > 0.6 FiO2 > 0.8
Supplementary Table S7). When the primary definition for
high FiO2 was extended to the entire period of mechanical
ventilation as opposed to just the period of single-lung venti-
Fig 3. Trends in FiO2 administration as a function of time after
lation (Model 4), AUC above FiO2 80% threshold) remained
the start of one-lung ventilation. (a) Discretionary FiO2 adminis-
significant (Supplementary Table S8). Excluding any cases
tration (25%, median, and 75%) as a function of time after the
with intraoperative desaturation (defined as SpO2 <88% for
start of one-lung ventilation. (b) Percentage of cases using
discretionary high FiO2 (SpO2 >98% at different FiO2 thresholds)
more than three consecutive minutes) demonstrated results
as a function of time after the start of one-lung ventilation. For similar to that of the primary analysis (Model 5;
any given time point on the x-axis, the height of the lines (25%, Supplementary Table S9). AUC above FiO2 80% threshold
median, and 75%) utilises the measurements of only those pa- remained associated with postoperative pulmonary compli-
tients whose one-lung ventilation period lasted to that point or cations (Model 6), using a revised definition for postoperative
beyond. Therefore, any downward trends may be due to FiO2 pulmonary complications that did not include atelectasis
levels decreasing over time for individual patients as one-lung (Supplementary Table S10) with similar discrimination char-
ventilation continues, or the one-lung ventilation period ending acteristics to Model 1 (c-statistics ¼ 0.769). Post hoc analyses
for those patients receiving higher FiO2, or both. FiO2, inspired requested during the review process to further clarify the
oxygen fraction; SpO2, peripheral capillary oxygen saturation. relationship between FiO2 and duration of exposure are
detailed in Model 7 (Supplementary material).

Secondary outcomes Discussion

Major morbidity was experienced by 854//2733 (31.3%) and 17/ We found that AUC above FiO2 80% threshold is associated
2733 (0.6%) died within 30 days of surgery (Table 2; with postoperative pulmonary complications. Because this
Supplementary Fig. S3a). exposure variable intrinsically contains both FiO2 and
 1080 - Douville et al.

Table 2 Outcomes table. Tracheal reintubation was originally proposed as a separate pulmonary metric in the composite outcome;
however, reintubation was included within the respiratory failure outcome on STS versions 2.3 and 2.41. For v2.2, respiratory failure
and reintubation were defined as separate outcomes, and both were included within the primary outcome composite. The outcomes
included in each composite are not mutually exclusive. For example: the total number of postoperative pulmonary complications is
141 (however, the number of atelectasis (43) þ pneumonia (72) þ acute respiratory distress syndrome (ARDS) (10) þ respiratory failure
(42) þ prolonged ventilator support (7) ¼ 174). This is because a single patient can only qualify for the composite outcome once, despite
having multiple qualifying complications.

Complication Numerator Population (n¼2733) Percentage

Denominator

Postoperative pulmonary complications

Atelectasis requiring bronchoscopy 43 2731 1.6
Pneumonia 72 2733 2.6
ARDS 10 2731 0.4
Respiratory failure 42 2731 1.5
Initial ventilator support >48 h 7 2731 0.3
Composite postoperative pulmonary complication 141 2733 5.2
Major morbidity
Postoperative events 820 2733 30.0
Atrial arrhythmia (requiring treatment) 186 2731 6.8
Ventricular arrhythmia (requiring treatment) 5 2731 0.2
Myocardial infarction 5 2731 0.2
Deep venous thrombosis (requiring treatment) 9 2731 0.3
Ileus 14 2731 0.5
Empyema (requiring treatment) 11 2731 0.4
Surgical site infection 15 2601 0.6
Sepsis 8 2603 0.3
Other infection (requiring i.v. antibiotics) 13 2731 0.5
New central neurological event 9 2731 0.3
Delirium 22 2731 0.8
Renal failure (RIFLE criteria) 26 2731 1.0
Readmission (within 30 days of discharge) 159 2723 5.8
Mortality
Death at discharge 10 2733 0.4
30-day mortality 17 2733 0.6
Composite major morbidity 854 2731 31.3

duration of one-lung ventilation, we could not directly resolve other disease states (and not inspired oxygen concentration)
the independent effects of each based upon the positive as- may play a larger role in the pathogenesis of postoperative
sociation shown in the primary model. However, by consid- pulmonary complications. Because atelectasis and the other
ering the individual effects of duration of exposure and FiO2, outcomes (pneumonia, ARDS, respiratory failure, and pro-
our results do not support limiting the inspired oxygen frac- longed ventilator support) differ markedly with regard to
tion for the purpose of reducing postoperative pulmonary severity and clinical significance, inclusion of atelectasis
complications. within the composite definition remains controversial. Our
Sensitivity models were designed to better understand the findings were confirmed in a post hoc sensitivity analysis using
impact of time exposure (Model 2) and FiO2 (Model 3). Because a definition that excluded cases classified as having a post-
average FiO2 is independent of time (unit: percentage); dura- operative pulmonary complication solely upon diagnosis of
tion of ventilation could be controlled for in Model 3. This atelectasis.
sensitivity analysis revealed duration of ventilation, but not Very high FiO2 levels are used early in one-lung ventilation
average FiO2 was associated with postoperative pulmonary and, to a moderate extent, decline gradually during the one-
complications. Additionally, examination of discretionary lung ventilation period in lung resection surgery. This likely
high FiO2dthat is, the combination of FiO2 levels >80% reflects the need to compensate for the obligatory intra-
concomitant with SpO2 values >98%drevealed no significant pulmonary shunt resulting from one-lung ventilation and is
association with the primary outcome. Overall, these results perhaps an intentional effort by practitioners to facilitate ab-
are consistent with those of prior studies in thoracic surgery sorption atelectasis in the nonventilated lung. Subsequent
which demonstrated that duration of exposure to mechanical decreases in FiO2 likely reflect the effects of hypoxic pulmo-
ventilation and duration of one-lung ventilation3 were inde- nary vasoconstriction and perhaps surgical factors leading to
pendent risk factors for postoperative pulmonary complica- decreased shunt. However, high FiO2 levels are used, even
tions,3,29 but do not unambiguously support a relationship when not strictly necessary based on the degree of systemic
between inspired oxygen concentration and postoperative oxygenation (Fig. 3b); 60% of cases use FiO2 >0.80 (despite SpO2
pulmonary complications in this clinical context. >98%) at the start of one-lung ventilation and high SpO2 is
Discretionary high FiO2 was also not associated with post- typically maintained during the period of one-lung ventilation
operative pulmonary complications, further suggesting that (Fig. 3b). The detailed characterisation of practice patterns for
prolonged ventilation, baseline pulmonary dysfunction or oxygen administration during lung resection surgery are of
 Alveolar hyperoxia and postoperative pulmonary complications - 1081

Table 3 Multivariable logistic regression for primary definition for alveolar hyperoxia (Model 1). C-statistic: 0.769. Note: the adjusted
odds ratio for the variable: ‘Presence of missing FEV1 data’ is relative to those with ‘FEV1 predicted’ both non-missing and equal to 0.
FEV1, forced expiratory volume in 1 s; FiO2, fraction inspired oxygen concentration; MPOG, Multicenter Perioperative Outcomes Group;
SpO2, peripheral capillary oxygen saturation.

Odds ratio 95% confidence P-value

interval

Area under the curve above FiO2¼80% (1000%*min) 1.17 1.04 1.33 0.012
MPOG Institution 1 (reference: 4) 0.59 0.369 0.95 0.029
MPOG Institution 2 (reference: 4) 0.47 0.096 2.26 0.344
MPOG Institution 3 (reference: 4) 0.50 0.256 0.97 0.040
Age (yr) 1.01 0.99 1.03 0.198
Female sex 0.98 0.67 1.41 0.894
BMI (kg m2) 0.99 0.96 1.02 0.379
ASA physical status 2.27 1.40 3.68 <0.001
Past smoking history (reference: non-smoker) 1.97 1.16 3.35 0.012
Current smoker (reference: non-smoker) 1.58 0.71 3.50 0.258
Prior history of chemotherapy, radiation, or both 1.00 0.60 1.70 0.986
Bilobectomy/lobectomy (reference: wedge resection) 1.58 1.01 2.48 0.045
Segmentectomy (reference: wedge resection) 1.16 0.54 2.46 0.708
Thoracotomy (reference: video-assisted thoracoscopic surgery) 2.47 1.63 3.74 <0.001
Presence of preoperative comorbidities 1.55 1.04 2.32 0.032
FEV1 predicted 0.98 0.97 0.99 <0.001
Presence of missing FEV1 data 0.211 0.070 0.64 0.006
Total fluid input (L) 1.16 0.93 1.46 0.186

possible utility in alerting providers to the potential for FiO2 complications.2 This contrasts with our primary finding in the
reduction. thoracic population that, although duration of ventilation was
Use of high FiO2 during one-lung ventilation has been neither time-weighted average FiO2 nor time above FiO2,
previously reported in a very small sub-cohort of thoracic thresholds were independently associated with postoperative
surgical cases,30 although practice in this regard has not been pulmonary complications.
previously well characterised. Practice patterns related to The study of higher vs lower FiO2 in the thoracic surgical
inspired oxygen are of interest in thoracic surgery because of population is, of course, limited by need for generally higher
the tension between an increased oxygen requirement FiO2 during one-lung ventilation because of the obligatory
because of intrapulmonary shunting and the known or sus- right to left intrapulmonary shunt which in turn limits prac-
pected relationships between high inspired oxygen concen- tice variation. Consequently, present study findings should be
tration and oxidative stress, ischaemia reperfusion injury, and viewed in light of the somewhat restricted range of clinical
alveolar damage. High oxygen concentrations are known to be practice related to oxygen administration. Although substan-
toxic to lung tissue,31 increasing pulmonary capillary perme- tial experimental and circumstantial evidence implicate
ability32 and contributing to oxidative protein damage in pa- inspired oxygen as a risk factor and potential aetiologic factor
tients undergoing lung resection.33 Oxidative stress is related in the genesis of lung injury after major thoracic surgery, the
both to the duration of one-lung ventilation and to the risk of clinical significance remains incompletely elucidated. None-
postoperative complications.34 Further, experimental evi- theless, substantial expert opinion supports efforts to mini-
dence supports the contention that oxidative stress is exac- mise this exposure, particularly for lung cancer surgery.37
erbated by hyperoxia1 suggesting that minimising excessive Expert consensus recommendations for protective ventila-
oxygen exposure might attenuate this stress response and tion of the surgical patient include the use of the lowest
associated injury. inspired oxygenation concentration compatible with adequate
In general, clinical evidence directly linking alveolar oxygenation,38 consistent with guidelines of the British
hyperoxia to postoperative pulmonary complications is Thoracic Society.39 Results of the present study indicate that
mixed. A paucity of evidence in the thoracic surgical popula- further reduction of FiO2, consistent with optimal systemic
tion (which involves unique stressors including one-lung oxygenation during one-lung ventilation for lung resection, is
ventilation, direct surgical trauma, and loss of functional achievable. We have demonstrated that the incidence of
lung parenchyma) limits direct comparison with other studies. discretionary high FiO2 during one-lung ventilation remains
Randomised trials comparing FiO2 30% vs 80% in abdominal35 high, particularly during the later stages of lung resection,
and colorectal surgery36 did not demonstrate a difference in including during lung reinflation (mean FiO2 at time of re-
the incidence of postoperative pulmonary complications. In a inflation¼82.7%). However, evidence from this study also
study in the nonthoracic surgery population, Staehr-Rye and suggests that practitioners are making an effort to reduce FiO2
colleagues4 utilised a definition of hyperoxia similar to that during one-lung ventilation. Overall, during lung resection
used in our secondary analysis and found that median FiO2 surgery in this study, the mean FiO2 decreased from 89.3% at
was associated with respiratory complications in a dose- the start of one-lung ventilation to 79.4% at the 3-h time point.
dependent manner, even when adjusted for duration The reported incidence of discretionary high FiO2, that is high
of anaesthesia. Another prospective study demonstrated FiO2 relative to the level of systemic oxygenation, may have
an association between time-weighted average FiO2 during utility in alerting anaesthesiologists to the potential for FiO2
one-lung ventilation and postoperative pulmonary reduction.
 1082 - Douville et al.

Our study has notable strengths, as well as several limita- Michigan Medicine for creative and technical support on study
tions. This large multicentre study was well positioned to figures.
assess the impact of FiO2 on postoperative pulmonary com-
plications across a range of practice patterns and geographi-
cally diverse set of institutions. Furthermore, outcome data Declaration of interest
were obtained from well-validated, abstractor-adjudicated The authors declare that they have no conflicts of interest
national databases utilising standardised metrics and defini- beyond those described in the funding statement.
tions.40 The retrospective nature of the study has inherent
limitations, including the possibility of yet unidentified con-
founding variables. As the study approach does not enable us Funding
to account for how different clinical situations may influence Foundation for Anesthesia Education and Research (FAER) e
FiO2 selection, we attempted to address this concern via two Mentored Research Training Grant (MRTG-02-15-2020-Dou-
approaches: (i) modelling discretionary high FiO2 and (ii) per- ville) (to NJD). National Institute of Diabetes and Digestive and
forming a sensitivity analysis where cases with desaturation Kidney Diseases (K08 DK131346 to NJD). National Heart,
were excluded (defined as SpO2 <88% for more than three Lung, and Blood Institute Grants (K01HL141701 to MRM,
consecutive minutes). To minimise bias, we have excluded K08HL159327 to DAC). This project was supported by depart-
cases from institutions not meeting a minimum research mental funding from University of Virginia and University of
standard. Further, we have controlled for a number of vari- Michigan Anesthesiology departments. The content is solely
ables which might otherwise introduce bias, including patient the responsibility of the authors and does not necessarily
characteristic variables, comorbidities, physiologic trespass of represent the official views of the National Institutes of Health.
surgery (resection type), institution, and additional risk pre-
dictors. Additionally, the decision to use a missing indicator
for FEV1 pulmonary function data, as opposed to a technique MPOG funding statement
such as multiple imputation, can introduce bias into the
Funding was also provided by departmental and institutional
analysis.41 Finally, SpO2 measurements may introduce racial
resources at each contributing MPOG site. In addition, partial
biases based on skin pigmentation, a variable not controlled in
funding to support underlying electronic health record data
the present study. Black patients exhibited much higher rates
collection into the Multicenter Perioperative Outcomes Group
of occult hypoxemia not detected by pulse oximetry compared
registry was provided by Blue Cross Blue Shield of Michigan/
with White patients.42
Blue Care Network as part of the Blue Cross Blue Shield of
In summary, in this multicentre observational cohort
Michigan/Blue Care Network Value Partnerships program.
study, alveolar hyperoxia was not unambiguously associated
Although Blue Cross Blue Shield of Michigan/Blue Care
with postoperative pulmonary complications after lung
Network and Multicenter Perioperative Outcomes Group work
resection surgery. Our results do not support limiting the
collaboratively, the opinions, beliefs and viewpoints expressed
inspired oxygen fraction for the purpose of reducing post-
by the authors do not necessarily reflect the opinions, beliefs,
operative pulmonary complications. Further study is neces-
and viewpoints of Blue Cross Blue Shield of Michigan/Blue
sary to better elucidate the relationship between oxygen
Care Network or any of its employees.
exposure and postoperative pulmonary complications.

Data availability statement

Authors’ contributions
The dataset is governed by the MPOG Data Use Agreement,
Study conception: NJD, MES, RSB which allows it only to be shared with other MPOG DUA
Study design: NJD, MES, BIN, MRM, DAC, SK, SRC, LWM, WMP, holders. A limited dataset would be available to other parties
NLP, RSB after publication upon execution of a Data Use Agreement and
Data analysis: MES fulfilment of other regulatory requirements (including IRB
Data interpretation: NJD, MES, BIN, MRM, DAC, SK, SRC, LWM, approval).
WMP, NLP, RSB
Writing the manuscript: NJD, RSB
Critical revision of the manuscript: NJD, MES, BIN, MRM, DAC, Appendix A. Supplementary data
SK, SRC, LWM, WMP, NLP, RSB Supplementary data to this article can be found online at
https://doi.org/10.1016/j.bja.2024.08.005.

Acknowledgements
References
The authors acknowledge Shelley Vaughn and Tomas Medina
Inchauste (Department of Anesthesiology, University of 1. Olivant Fisher A, Husain K, Wolfson MR, et al. Hyperoxia
Michigan Medical School, Ann Arbor, MI, USA) for their con- during one lung ventilation: inflammatory and oxidative
tributions in data acquisition and electronic search query responses. Pediatr Pulmonol 2012; 47: 979e86
programming for this project. Additionally, we thank MPOG 2. Okahara S, Shimizu K, Suzuki S, Ishii K, Morimatsu H.
contributors Marcel E. Durieux (Department of Anesthesi- Associations between intraoperative ventilator settings
ology, University of Virginia School of Medicine, Charlottes- during one-lung ventilation and postoperative pulmonary
ville, VA, USA) and Stephen Patrick Bender (Department of complications: a prospective observational study. BMC
Anesthesiology and Critical Care Medicine, The University of Anesthesiol 2018; 18: 13
Vermont Medical Center, Burlington, VT, USA). We thank 3. Lai G, Guo N, Jiang Y, Lai J, Li Y, Lai R. Duration of one-lung
Elizabeth Jewell from the Department of Anesthesiology at ventilation as a risk factor for postoperative pulmonary
 Alveolar hyperoxia and postoperative pulmonary complications - 1083

complications after McKeown esophagectomy. Tumori 19. Blank RS, Colquhoun DA, Durieux ME, et al. Management of
2020; 106: 47e54 one-lung ventilation: impact of tidal volume on complica-
4. Staehr-Rye AK, Meyhoff CS, Scheffenbichler FT, et al. High tions after thoracic surgery. Anesthesiology 2016; 124: 1286e95
intraoperative inspiratory oxygen fraction and risk of ma- 20. Quan H, Sundararajan V, Halfon P, et al. Coding algo-
jor respiratory complications. Br J Anaesth 2017; 119: 140e9 rithms for defining comorbidities in ICD-9-CM and ICD-10
5. McIlroy DR, Shotwell MS, Lopez MG, et al. Oxygen administrative data. Med Care 2005; 43: 1130e9
administration during surgery and postoperative organ 21. Douville NJ, Jewell ES, Duggal N, et al. Association of
injury: observational cohort study. BMJ 2022; 379, e070941 intraoperative ventilator management with postoperative
6. Blum JM, Stentz MJ, Dechert R, et al. Preoperative and oxygenation, pulmonary complications, and mortality.
intraoperative predictors of postoperative acute respira- Anesth Analg 2020; 130: 165e75
tory distress syndrome in a general surgical population. 22. Liu M, Wei L, Zhang J. Review of guidelines and literature
Anesthesiology 2013; 118: 19e29 for handling missing data in longitudinal clinical trials
7. Kivik P, Oganjan J, Aun A, Sormus A. Epidemiology, with a case study. Pharm Stat 2006; 5: 7e18
practice of ventilation and outcome for patients at 23. Lang KM, Little TD. Principled missing data treatments.
increased risk of postoperative pulmonary complications: Prev Sci 2018; 19: 284e94
LAS VEGAS e an observational study in 29 countries. Eur J 24. Poullis M, McShane J, Shaw M, et al. Prediction of in-
Anaesthesiol 2017; 34: 492e507 hospital mortality following pulmonary resections:
8. Wetterslev J, Meyhoff CS, Jørgensen LN, Gluud C, improving on current risk models. Eur J Cardiothorac Surg
Lindschou J, Rasmussen LS. The effects of high perioper- 2013; 44: 238e42. discussion 242e3
ative inspiratory oxygen fraction for adult surgical pa- 25. Chiou SH, Betensky RA, Balasubramanian R. The missing
tients. Cochrane Database Syst Rev 2015, CD008884 indicator approach for censored covariates subject to limit
9. Mattishent K, Thavarajah M, Sinha A, et al. Safety of 80% of detection in logistic regression models. Ann Epidemiol
vs 30e35% fraction of inspired oxygen in patients under- 2019; 38: 57e64
going surgery: a systematic review and meta-analysis. Br J 26. Kayembe MT, Jolani S, Tan FES, van Breukelen GJP. Impu-
Anaesth 2019; 122: 311e24 tation of missing covariates in randomized controlled tri-
10. Benchimol EI, Smeeth L, Guttmann A, et al. The REporting als with continuous outcomes: simple, unbiased and
of studies Conducted using Observational Routinely- efficient methods. J Biopharm Stat 2022; 32: 717e39
collected health Data (RECORD) statement. PLoS Med 27. Purnell J.Q. Definitions, Classification, and Epidemiology of
2015; 12, e1001885 Obesity. 2023 May 4. In: Feingold K.R., Anawalt B., Blackman
11. Perioperative Clinical Research Committee (PCRC). Avail- M.R., Boyce A., Chrousos G., Corpas E., de Herder W.W.,
able from https://mpog.org/events/perioperative-clinical- Dhatariya K., Dungan K., Hofland J., Kalra S., Kaltsas G.,
research-committee-pcrc/(accessed 16 June 2021). Kapoor N., Koch C., Kopp P., Korbonits M., Kovacs C.S.,
12. Douville NJ, McMurry TL, Ma JZ, et al. Airway driving Kuohung W., Laferre re B., Levy M., McGee E.A., McLachlan
pressure is associated with postoperative pulmonary R., New M., Purnell J., Sahay R., Shah A.S., Singer F., Sperling
complications after major abdominal surgery: a multi- M.A., Stratakis C.A., Trence D.L., Wilson D.P., Endotext
centre retrospective observational cohort study. BJA Open [Internet]. South Dartmouth (MA): MDText.com, Inc.; 2000
2022; 4, 100099 e. PMID: 25905390.
13. Colquhoun DA, Shanks AM, Kapeles SR, et al. Considerations 28. Hosmer DW, Lemeshow S. Applied logistic regression. New
for integration of perioperative electronic health York, USA: John Wiley & Sons, Inc; 2000. https://doi.org/
records across institutions for research and quality 10.1002/0471722146
improvement: the approach taken by the Multicenter 29. Licker M, de Perrot M, Spiliopoulos A, et al. Risk factors for
Perioperative Outcomes Group. Anesth Analg 2020; 130: acute lung injury after thoracic surgery for lung cancer.
1133e46 Anesth Analg 2003; 97: 1558e65
14. Shahian DM, Jacobs JP, Edwards FH, et al. The society of 30. Suzuki S, Mihara Y, Hikasa Y, et al. Current ventilator and
thoracic Surgeons national database. Heart 2013; 99: oxygen management during general anesthesia: a multi-
1494e501 center, cross-sectional observational study. Anesthesiology
15. Colquhoun DA, Leis AM, Shanks AM, et al. A lower tidal 2018; 129: 67e76
volume regimen during one-lung ventilation for lung 31. Heerdt PM, Stowe DF. Single-lung ventilation and oxida-
resection surgery is not associated with reduced post- tive stress: a different perspective on a common practice.
operative pulmonary complications. Anesthesiology 2021; Curr Opin Anaesthesiol 2017; 30: 42e9
134: 562e76 32. Royer F, Martin DJ, Benchetrit G, Grimbert FA. Increase in
16. Kozower BD, Sheng S, O’Brien SM, et al. STS database risk pulmonary capillary permeability in dogs exposed to 100%
models: predictors of mortality and major morbidity for O2. J Appl Physiol 1988; 65: 1140e6
lung cancer resection. Ann Thorac Surg 2010; 90: 875e81. 33. Williams EA, Quinlan GJ, Goldstraw P, Gothard JW,
discussion 881e3 Evans TW. Postoperative lung injury and oxidative dam-
17. Shapiro M, Swanson SJ, Wright CD, et al. Predictors of major age in patients undergoing pulmonary resection. Eur
morbidity and mortality after pneumonectomy utilizing the Respir J 1998; 11: 1028e34
society for thoracic Surgeons general thoracic surgery 34. Misthos P, Katsaragakis S, Theodorou D, Milingos N,
database. Ann Thorac Surg 2010; 90: 927e34. discussion 934e5 Skottis I. The degree of oxidative stress is associated with
18. Wright CD, Gaissert HA, Grab JD, O’Brien SM, Peterson ED, major adverse effects after lung resection: a prospective
Allen MS. Predictors of prolonged length of stay after lo- study. Eur J Cardiothorac Surg 2006; 29: 591e5
bectomy for lung cancer: a Society of Thoracic Surgeons 35. Li X-F, Jiang D, Jiang Y-L, et al. Comparison of low and
General Thoracic Surgery Database risk-adjustment high inspiratory oxygen fraction added to lung-protective
model. Ann Thorac Surg 2008; 85: 1857e65 ventilation on postoperative pulmonary complications
 1084 - Douville et al.

after abdominal surgery: a randomized controlled trial. 39. Shovlin CL, Condliffe R, Donaldson JW, Kiely DG, Wort SJ,
J Clin Anesth 2020; 67, 110009 British Thoracic Society. British Thoracic Society clinical
36. Cohen B, Ruetzler K, Kurz A, et al. Intra-operative high statement on pulmonary arteriovenous malformations.
inspired oxygen fraction does not increase the risk of Thorax 2017; 72: 1154e63
postoperative respiratory complications: alternating 40. Stewart JM. Abstraction techniques for the STS national
intervention clinical trial. Eur J Anaesthesiol 2019; 36: 320e6 database. J Extra Corpor Technol 2016; 48: 201e3
37. Grocott HP. Oxygen toxicity during one-lung ventilation: is 41. Knol MJ, Janssen KJM, Donders ART, et al. Unpredictable
it time to re-evaluate our practice? Anesthesiol Clin 2008; bias when using the missing indicator method or com-
26: 273e80 plete case analysis for missing confounder values: an
38. Young CC, Harris EM, Vacchiano C, et al. Lung-protective empirical example. J Clin Epidemiol 2010; 63: 728e36
ventilation for the surgical patient: international expert 42. Sjoding MW, Dickson RP, Iwashyna TJ, Gay SE, Valley TS.
panel-based consensus recommendations. Br J Anaesth Racial bias in pulse oximetry measurement. N Engl J Med
2019; 123: 898e913 2020; 383: 2477e8

Handling Editor: Gareth Ackland