Content area
Background
The potential clinical value of driving pressure (DP) and mechanical power (MP) as digital biomarkers for predicting postoperative pulmonary complications (PPC) has been emphasized. Although several studies have explored this association, evidence from clinical cohorts involving large patient populations remains limited.
Methods
A retrospective cohort study was conducted between October 2004 and May 2023 on patients who underwent OLV surgery. The association between time-weighted median dynamic DP or MP and PPC was analyzed using multivariate logistic regression models adjusted for confounders. Additionally, risk threshold analysis was conducted to propose thresholds for an increased risk of PPC.
Results
Among the 3386 (using plateau pressure; Pplat) or 4951 (using peak inspiratory pressure; PIP) patients included, PPC occurrence was 19.31 % and 17.43 %, respectively. The mean of the time-weighted median DP derived from Pplat (DP-Pplat) and MP derived from Pplat (MP-Pplat) were 14.6 cmH
Conclusions
In this OLV surgery population, a DP-Pplat-limited mechanical ventilation strategy of 15 cmH
1 Introduction
Postoperative pulmonary complications (PPC) are associated with poor patient outcomes, such as prolonged hospital stay, increased morbidity, and higher mortality rates [ 1, 2]. A significantly increased incidence of PPC (20–70 %) has been reported in thoracic surgeries compared to a 4 % incidence of PPC in orthopedic or urologic surgeries [ 3], highlighting the need for effective management and prevention strategies [ 4]. Intraoperative lung-protective mechanical ventilation has been recognized as an important measure to decrease the incidence of PPC [ 5]. However, one-lung ventilation (OLV) presents unique challenges in implementing these strategies. OLV often requires high tidal volumes and pressures to maintain adequate oxygenation and ventilation, making it difficult to achieve lung-protective ventilation.
Driving pressure (DP), defined as the difference between plateau pressure (Pplat) and positive end-expiratory pressure (PEEP) in mechanically ventilated patients, has emerged as a key parameter for assessing lung mechanics and guiding ventilation strategies [ 6]. In addition to DP, mechanical power (MP), a comprehensive measure of the energy delivered to the lungs per breath, has gained much attention in clinical settings [ 7]. MP combines variables, such as tidal volume, respiratory rate, and DP, to quantify the total energy applied to the lungs during mechanical ventilation [ 8]. Clinically, DP and MP indicate lung compliance and the stress exerted on the lungs during mechanical ventilation [ 9, 10]. Adequate monitoring and adjustment of these parameters may help clinicians make informed decisions regarding ventilatory settings, potentially reducing the risk of PPCs, especially in sensitive scenarios, such as OLV, where there are limited ventilatory options, and patients are more prone to lung injury. However, the specific relationships among DP, MP, and PPC development in the context of OLV surgery remain insufficiently explored. Existing research has highlighted the general association between elevated DP levels and adverse pulmonary outcomes [ 11, 12], but these findings have not been extensively studied in OLV scenarios.
This study aimed to explore the associations among DP, MP, and PPC in patients undergoing OLV surgery. We hypothesized that higher DP and MP would be associated with an increased incidence of PPC in these patients. The association was analyzed using two calculation methods, Pplat and peak inspiratory pressure (PIP), and risk thresholds for the associated indicators were proposed.
2 Methods
2.1 Study approval
This study was approved, and the need for informed consent was waived owing to the retrospective study design and the use of anonymized registry data by the Institutional Review Board of Seoul National University Hospital (H-2303-129-1414). This article was prepared in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.
2.2 Data collection and preprocessing
Among video-assisted thoracoscopic surgeries (VATS) performed at a university-affiliated hospital between October 2004 and May 2023, only adult patients with all required data were included in the analysis. VATS were performed by a team of 5 to 8 board-certified thoracic surgeons, supported by 2 to 6 thoracic anesthesiologists with expertise in one-lung ventilation. Ventilator settings during OLV were determined by the attending anesthesiologists based on individual patient characteristics and intraoperative conditions. Although lung-protective ventilation strategies were generally encouraged, no uniform institutional protocol was applied during the study period. The institution operated 2 to 3 dedicated thoracic surgery operating rooms on a daily basis. Baseline parameters such as patient demographic information (age, sex, body mass index [BMI]), existing comorbidities (chronic obstructive pulmonary disease; COPD, asthma, interstitial lung disease; ILD, heart failure; HF), smoking history, and pulmonary function test results (forced expiratory volume in one second; FEV 1, forced vital capacity; FVC) were collected from electronic medical records (EMR) using SUPREME 2.0, a clinical data warehouse of Seoul National University Hospital, which automatically loads EMR data on a daily basis [ 13, 14]. Preoperative parameters including laboratory test results (hemoglobin [Hb], blood urea nitrogen [BUN], and albumin), the existence of a nasogastric tube (Levin tube), and saturation of partial pressure oxygen (SpO 2), and intraoperative parameters, such as duration of surgery (OP duration), packed red blood cell (pRBC) transfusion, assess respiratory risk in surgical patients in catalonia (ARISCAT) score and lung resection score, were collected. Intraoperative ventilation parameters - Pplat, PIP, PEEP, tidal volume (V T), and respiratory rate – were collected and analyzed exclusively during the OLV phase. The ventilation parameters were systematically recorded at intervals of either one (from 2013 onwards) or five minutes (before 2013). To ensure consistency, data recorded at five-minute intervals were interpolated by carrying forward the most recent value to align with the one-minute interval data. For reliable results, point-of-care laboratory test results were excluded, and only tests conducted within three months prior to surgery were included in the analysis. Serum albumin level was excluded from the analysis because of a high rate of missing values. Erroneous data, such as SpO 2 values exceeding 100 % or PEEP values exceeding PIP values in the breath, were also removed. Information regarding the length of hospital stay was also collected. The clinical data utilized in this study were retrieved directly from the EMR, with all vital signs and ventilator settings automatically recorded in the system. This approach minimized variability in data collection and reduced the potential for inter-investigator discrepancies.
2.3 Exposure
DP in mechanical ventilation is the difference between Pplat and PEEP. In this study, Pplat values automatically recorded during volume-controlled ventilation was used. For cases managed with pressure-controlled ventilation, in which Pplat was unavailable, PIP was used as a surrogate. Accordingly, the MP was calculated using the same methods. These two factors were calculated every minute. Cumulative time-weighted median DP and MP were calculated to analyze the association between each factor and PPC [
15]. The cumulative time-weighted median was calculated by dividing the area under the time curve by the number of minutes of exposure. The formulas for calculating the DP and MP are as follows [
8,
16–18]:
2.4 Outcome in interest
PPC was defined as a composite of various pulmonary complications that occurred within seven days after surgery [ 1]. Given the absence of a universally accepted definition for PPCs, we adopted a widely referenced composite definition from previous literature [ 1] to ensure comprehensive inclusion of clinically relevant spectrum of pulmonary complications. For patients discharged before the seventh postoperative day, only complications that developed during their inpatient stay were considered. Thus, PPCs occurring after discharge were excluded from the analysis. These included acute respiratory distress syndrome, characterized by respiration failure lasting 24 h; mild and severe respiratory failure, defined as SpO 2 < 90 % or partial pressure of oxygen in arterial blood (PaO 2) < 60 mmHg for 10 min in room air. Other PPCs included pleural effusion requiring thoracentesis, pneumonia, and prolonged air leakage, defined as the need for chest tube insertion >7 days post-surgery. Detailed definitions of PPCs and their incidences are provided in Supplementary 1.
2.5 Statistical analysis
Baseline parameters, preoperative parameters, and intraoperative parameters are presented using descriptive statistics for the PPC and non-PPC groups. Separate descriptive statistics were reported according to the two exposure cohorts (DP-Pplat and DP-PIP). Continuous variables were expressed as medians and interquartile ranges, while categorical variables were expressed as n (%). Continuous variables were tested for normality using the Shapiro-Wilk test. Depending on the distribution, the independent Student's t-test or the Mann-Whitney U test was used to compare groups. The chi-squared test was used for categorical variables.
Multivariate logistic regression analysis was performed to assess the association between intraoperative DP, MP, and PPC. Because DP is used in the MP calculation, this might lead to multicollinearity and distortion. Therefore, the logistic regression models for DP and MP were considered independently. All collected clinical variables and parameters were considered covariates to be adjusted in all models and included in the logistic regression models. To examine potential multicollinearity among candidate covariates, Pearson correlation coefficients were computed, and variance inflation factors (VIFs) were calculated for each variable. Variables with VIF ≥ 2.5 [ 19] were excluded from the final models. In particular, FVC was excluded due to strong correlation and multicollinearity with FEV1 (see Supplementary 2). Although clinically relevant, the ARISCAT score was excluded from the multivariate models, as it incorporates variables already included in our models as individual covariates (e.g., hemoglobin, SpO2, duration of surgery). For both the DP and MP models, the dependent variable was the occurrence of PPC. The following covariates were considered in the development of the models: age, sex, BMI, OP duration, Hb, BUN, the existence of Levin tube, SpO 2, the existence of COPD, ILD, asthma, HF, smoking history, FEV 1, pRBC transfusion, and lung resection score. Surgeries with missing data in any of the considered covariates were excluded from the analysis. Multivariable logistic regression analysis was performed by simultaneously including all clinically relevant covariates, without applying automated variable selection procedures. Odds ratios (ORs) with 95 % confidence intervals (CIs) for the associations between the DP or MP and the occurrence of PPC have been reported.
All statistical analyses were performed using Python 3.8.12 (Python Software Foundation, Wilmington, DE, USA) with the TableOne 0.8.0 and the Scipy 1.11.3 packages. p < 0.05 was considered significant.
2.6 Risk threshold analysis
To propose risk thresholds for DP and MP that increase the occurrence of PPC, we conducted a multivariate regression analysis using threshold-based categorization. Each variable was dichotomized at multiple candidate cutoff points to compare outcomes above and below the respective thresholds.
2.7 Subgroup analysis
Reduced FEV 1 and FVC are known risk factors of PPCs [ 1]. To analyze the association between intraoperative DP or MP and PPCs, patients were divided into two groups: those with normal and abnormal pulmonary function test (PFT) results.
3 Results
For the calculation of DP-Pplat, 3386 surgeries from 3219 patients were included, whereas for the DP-PIP calculation, 4951 surgeries from 4701 patients were included (
Fig. 1
3.1 Outcomes
Among included cases, PPCs were observed in 654 cases (19.31 %) in the Pplat group and in 863 cases (17.43 %) in the PIP group.
Table 1
3.2 Associations of outcomes with the occurrence of PPC
In the multivariate model with DP, increases of 1 cmH
2O in DP-Pplat or DP-PIP were associated with the occurrence of PPC after adjusting for all confounders (odds ratio [OR], 1.047 [95 % confidence interval [CI] 1.019–1.075];
p < 0.05; and 1.036 [95 % CI 1.013–1.059], p < 0.05, respectively). The E − value for the odds ratio of DP-Pplat and DP-PIP was 1.268 [95 % CI 1.158–1.359] and 1.229 [95 % CI 1.128–1.309], respectively. In contrast, an increase of 1 J/min in MP was not associated with the occurrence of PPCs (OR 1.033 [95 % CI 0.980–1.089],
p = 0.226, and 1.048 [95 % CI 0.992–1.106,
p = 0.092] for MP-Pplat and MP-PIP, respectively) (
Fig. 2
3.3 Risk threshold analysis
An association with the occurrence of PPCs was observed at a DP-Pplat of 15 cmH
2O or higher, whereas the threshold was 18 cmH
2O for DP-PIP. In both cases, there was a gradual increase in the risk of PPC as the threshold increased (
Fig. 3
3.4 Subgroup analysis of patients with normal and abnormal spirometry results
Significant differences in DP and MP were observed between patients with and without PPC, regardless of the spirometry results. DP was higher in normal spirometry cases with PPC compared to those without PPC (14.0 [12.0, 17.0] vs 14.0 [12.0, 16.0] cmH2O, p < 0.05 for Pplat; 15.0 [12.0, 17.0] vs 14.0 [12.0, 16.0] cmH2O, p < 0.05 for PIP). Similarly, MP was higher in normal spirometry cases with PPC (6.8 [5.8, 8.3] vs 6.7 [5.6, 8.1] J/min, p < 0.05 for Pplat; 7.1 [6.0, 8.5] vs 6.8 [5.6, 8.1] J/min, p < 0.05 for PIP). In the abnormal spirometry group, DP was also higher in PPC cases compared to non-PPC cases (15.0 [12.0, 19.0] vs 14.0 [12.0, 17.0] cmH 2O, p < 0.05 for Pplat; 15.0 [12.0, 18.0] vs 14.0 [12.0, 16.0] cmH 2O, p < 0.05 for PIP), as was MP (7.4 [6.1, 8.7] vs 7.1 [5.8, 8.3] J/min, p < 0.05 for Pplat; 7.1 [6.0, 8.5] vs 6.8 [5.6, 8.1] J/min, p < 0.05 for PIP).
Furthermore, the length of hospital stay was longer in the PPC group than in the non-PPC group. In the normal spirometry group, the length of hospital stay was significantly longer in the PPC group compared to the non-PPC group for both Pplat (5.55 [3.53, 9.19] vs 2.50 [1.59, 3.54] days, p < 0.05) and PIP (6.18 [3.55, 9.35] vs 2.42 [1.48, 3.53] days, p < 0.05). Similarly, in the abnormal spirometry group, the PPC group had a longer hospital stay compared to those without PPC with Pplat (8.28 [5.35, 12.49] vs 3.39 [2.27, 4.55] days, p < 0.05) and PIP (8.31 [5.34, 12.82] vs 3.27 [1.62, 4.49] days, p < 0.05).
In the normal spirometry group, there was no significant association between unit increases in DP and MP or the occurrence of PPC in any scenario. However, in the abnormal spirometry group, a 1 cmH 2O increase in DP-Pplat and DP-PIP increased the odds of PPC by 6.5 % and 5.1 %, respectively. A 1 J/min increase in MP-PIP increased the odds of PPC by 10.8 %. However, there was also no significant association between MP-Pplat and the occurrence of PPC in the abnormal spirometry group ( Fig. 2).
In the abnormal spirometry group, the risk of PPC significantly increased when the DP-Pplat increased to ≥16 cmH 2O and the DP-PIP increased to ≥19 cmH 2O. Additionally, the risk of PPC significantly increased when the MP-PIP was ≥7 J/min ( Fig. 3).
4 Discussion
We analyzed 4956 surgical cases of OLV collected over an 18-year period. These findings consistently indicate that a higher DP during OLV is associated with an increased incidence of PPC. A 1 cmH 2O increase in DP-Pplat led to a 4.7 % increase in the risk of PPC occurrence, whereas a 1 cmH 2O increase in DP-PIP yielded a 3.6 % increase in risk. In risk threshold analysis, a DP-Pplat of 15 cmH 2O or a DP-PIP of ≥18 cmH 2O significantly increases the risk of PPCs. This effect was more pronounced in the patients with abnormal preoperative spirometry results. In these cases, the risk of PPC increased by 6.5 % per 1 cmH 2O increase in DP-Pplat and 5.1 % per 1 cmH 2O increase in DP-PIP. The identified risk thresholds for DP-Pplat and DP-PIP were 16 cmH 2O and 19 cmH 2O, respectively. However, a slightly weaker association was observed between MP and PPCs. Although higher MP values were associated with an increased risk of PPCs, no specific threshold for that was associated with PPC occurrence. In the abnormal spirometry group, a 1 J/min increase in MP-PIP increased the odds of PPC by 10.8 %, and the risk of PPC significantly increased when the MP-PIP exceeded 7 J/min.
This discrepancy can be attributed to several factors. DP, which represents the distending forces exerted on the alveoli, is closely related to barotrauma and volutrauma, both of which are key contributors to PPCs. By contrast, MP, a broader measure of energy transfer to the lungs, may not fully capture the specific alveolar distension stress that directly leads to PPCs. Because MP includes components such as V T and respiratory rate in addition to DP, if these other components remain within safe ranges, the overall MP might not reach a harmful threshold. Additionally, MP, as an aggregate measure, may obscure the short-term detrimental effects of surgery that contribute to PPCs. Therefore, while MP is a valuable parameter, DP may serve as a more direct indicator of the mechanical stress applied to the lungs, making it more strongly associated with PPCs in the context of OLV.
Of the many interoperative factors that can affect the risk of PPC, DP or MP are possible modifiable factors [ 12]. Our study is the first to investigate the association of DP and MP with PPC in OLV surgery and to establish a threshold of DP or MP that significantly increases the risk of PPC. Furthermore, subgroup thresholds based on spirometry results were presented separately and showed that the thresholds were lower in the abnormal spirometry group.
The clinical potential of DP and MP as digital biomarkers for predicting outcomes in mechanically ventilated patients has increased the importance of intraoperative monitoring. Several studies have aimed to understand the relationship between PPC and the DP or MP [ 18, 20–27]. Randomized controlled trials have demonstrated that high DP is associated with an increased risk of PPC in laparoscopic surgery [ 11], robotic surgery, and lung resection surgery [ 7]. A retrospective study of 1191 patients [ 15] also found that high DP was independently associated with PPC in open abdominal surgery. Another study [ 12] demonstrated that patients with restrictive spirometry results who underwent extrathoracic surgery were at a higher risk of PPC with increased intraoperative DP; however, this association was not observed in patients with normal spirometry results.
Compared to DP, studies focusing on the relationship between intraoperative MP and PPCs are limited and mixed [ 28–30]. Our study showed that the MP was not independently associated with PPC. However, two studies [ 28, 29] have reported different results. Possible explanations for this discrepancy include differences in the duration of possible lung injuries. Yoon et al. showed that a higher MP was associated with an increased risk of PPC in cases where the OLV exceeded 150 min. However, the OLV duration in our study was notably shorter—87.55 min in the PIP group and 92.05 min in the Pplat group. The second possible explanation is the differences in the study population and outcomes. Santer's study [ 29] included noncardiac surgical outpatients and inpatients, and the outcome was respiratory failure requiring reintubation within 7 days.
Most previous studies have calculated DP and MP using Pplat. In this study, we calculated DP and MP using both Pplat and PIP, presenting results from both methods, and identifying distinct thresholds for each method. This approach may be particularly useful in broader intraoperative settings, where certain ventilator modes do not routinely provide Pplat measurements. Furthermore, our study included a substantially larger cohort ( N = 4951) compared to previous studies ( N = 609 [ 30], 878 [ 31], and 1055 [ 28], respectively), thereby increasing statistical power and enhancing the generalizability of our findings.
This study had several limitations. First, owing to the inherent limitations of the retrospective design, residual confounding factors may influence the results. However, this study represents one of the largest datasets in thoracic surgery (N = 4951), and while retrospective studies often face challenges related to data quality and consistency, we mitigated these issues by ensuring that data were automatically recorded every minute, thereby enhancing the reliability of the study. Because Pplat was not available for all patients in this retrospective dataset, PIP was used as a surrogate only in cases where Pplat was missing. However, this substitution may lead to overestimation of DP and MP in patients with increased airway resistance, such as those with obstructive lung disease. Also, auto-PEEP was not routinely measured and was assumed to be zero, which may have led to overestimation of DP and MP, particularly in patients with obstructive lung disease. Although this was a single-center study, which may limit the generalizability of the findings, the structured intraoperative management and real-time electronic data collection at this high-volume tertiary center helped minimize variability and improve internal validity. This context should be considered when interpreting the applicability of the findings to other settings. Furthermore, since the data analysis was limited to records up to May 2023, the findings may not fully reflect the most recent changes in ventilation strategies implemented thereafter. Finally, it could be argued that using a composite definition of PPC is a limitation, as there are inherent challenges in retrospectively identifying complications, particularly when dealing with composite outcomes. However, this study closely adheres to the methodology used in previous retrospective studies on PPCs [ 11, 12, 32]. The fact that the abnormal PFT group had a higher incidence of PPCs and that the PPC group exhibited higher DP and MP values along with longer hospital stays suggests that our definitions were applied appropriately.
5 Conclusion
Higher intraoperative DP was independently associated with an increased risk of PPCs in patients undergoing OLV surgery, particularly those with abnormal preoperative pulmonary function. Threshold analysis identified 15 cmH 2O for DP-Pplat and 18 cmH 2O for DP-PIP as clinically relevant cutoffs.
CRediT authorship contribution statement
Hyun-Lim Yang: Writing – review & editing, Writing – original draft, Project administration, Methodology, Investigation, Formal analysis, Data curation. Seong-A Park: Writing – original draft, Project administration, Methodology, Investigation, Formal analysis, Data curation. Hong Yeul Lee: Project administration, Investigation, Conceptualization. Hyeonhoon Lee: Formal analysis, Data curation. Ho-Geol Ryu: Project administration, Investigation. Hyung-Chul Lee: Supervision, Resources, Methodology. Sang-Min Lee: Supervision, Investigation, Funding acquisition. Jinwoo Lee: Writing – review & editing, Writing – original draft, Project administration, Methodology, Investigation, Conceptualization.
Ethics statement
This study was approved, and the need for informed consent was waived owing to the retrospective study design and the use of anonymized registry data by the Institutional Review Board of Seoul National University Hospital (H-2303-129-1414).
Funding
This research was supported by a grant of the
Declaration of competing interest
All authors declared no competing interests.
Appendix A Supplementary data
Supplementary material
Appendix A Supplementary data
Supplementary data to this article can be found online at
Table 1
Time-weighted median of driving pressure (cmH 2O).
Time-weighted median of mechanical power (J/min).
| Pplat | PIP | |||||||
| Overall
( n = 3386) |
Without PPC
( n = 2732) |
PPC
( n = 654) |
P-Value | Overall
(n = 4951) |
Without PPC
( n = 4088) |
With PPC
( n = 863) |
P-Value | |
| Driving pressure † (cmH 2O) | 14.0 [12.0,17.0] | 14.0 [12.0,16.0] | 15.0 [12.0,18.0] | <0.05 | 17.0 [14.0,20.0] | 17.0 [14.0,19.0] | 18.0 [15.0,21.0] | <0.05 |
| Mechanical power ‡ (J/min) | 6.8 [5.7,8.2] | 6.8 [5.6,8.1] | 7.1 [6.0,8.5] | <0.05 | 6.2 [5.1,7.4] | 6.1 [5.1,7.4] | 6.5 [5.5,7.7] | <0.05 |
| Age (year) | 61.0 [53.0,69.0] | 60.0 [52.0,69.0] | 65.0 [57.0,72.0] | <0.05 | 61.0 [53.0,69.0] | 60.0 [52.0,69.0] | 65.0 [57.0,71.5] | <0.05 |
| Sex, male (%) | 1841 (54.4) | 1389 (50.8) | 452 (69.1) | <0.05 | 2641 (53.3) | 2047 (50.1) | 594 (68.8) | <0.05 |
| BMI (kg/m 2) | 23.7 [21.6,25.8] | 23.8 [21.6,25.8] | 23.5 [21.5,26.0] | 0.780 | 23.7 [21.6,25.9] | 23.8 [21.6,25.9] | 23.4 [21.4,25.9] | 0.080 |
| OP duration (hours) | 1.9 [1.2,2.6] | 1.8 [1.1,2.4] | 2.5 [1.8,3.4] | <0.05 | 1.8 [1.0,2.5] | 1.6 [0.9,2.3] | 2.4 [1.7,3.3] | <0.05 |
| Hb (g/dL) | 12.9 [11.9,14.0] | 13.0 [12.0,14.0] | 12.9 [11.8,14.0] | 0.158 | 13.0 [12.0,14.0] | 13.0 [12.0,14.0] | 12.9 [11.7,14.0] | <0.05 |
| BUN (g/dL) | 14.0 [11.0,17.0] | 14.0 [11.0,17.0] | 14.0 [11.0,18.0] | <0.05 | 14.0 [11.0,18.0] | 14.0 [11.0,17.0] | 15.0 [11.0,18.0] | <0.05 |
| Albumin (g/dL) | 3.9 [3.6,4.3] | 4.0 [3.7,4.3] | 3.8 [3.4,4.2] | <0.05 | 4.1 [3.7,4.4] | 4.1 [3.7,4.4] | 3.8 [3.5,4.2] | <0.05 |
| Levin tube (%) | 47 (1.4) | 24 (0.9) | 23 (3.5) | <0.05 | 59 (1.2) | 32 (0.8) | 27 (3.1) | <0.05 |
| SpO 2 (%) | 97.0 [96.0,98.0] | 97.0 [96.0,98.0] | 97.0 [96.0,98.0] | <0.05 | 97.0 [96.0,98.0] | 97.0 [96.0,98.0] | 97.0 [96.0,98.0] | <0.05 |
| COPD (%) | 267 (7.9) | 167 (6.1) | 100 (15.3) | <0.05 | 317 (6.4) | 205 (5.0) | 112 (13.0) | <0.05 |
| ILD (%) | 124 (3.7) | 83 (3.0) | 41 (6.3) | <0.05 | 184 (3.7) | 128 (3.1) | 56 (6.5) | <0.05 |
| Asthma (%) | 141 (4.2) | 116 (4.2) | 25 (3.8) | 0.706 | 225 (4.5) | 189 (4.6) | 36 (4.2) | 0.625 |
| HF (%) | 42 (1.2) | 29 (1.1) | 13 (2.0) | 0.084 | 67 (1.4) | 46 (1.1) | 21 (2.4) | <0.05 |
| Smoking (%) | 848 (25.0) | 625 (22.9) | 223 (34.1) | <0.05 | 1194 (24.1) | 908 (22.2) | 286 (33.1) | <0.05 |
| FEV1 (%) | 101.0 [89.0,112.0] | 102.0 [91.0,113.0] | 94.5 [81.0,108.0] | <0.05 | 102.0 [90.0,113.0] | 103.0 [92.0,114.0] | 95.0 [82.0,109.0] | <0.05 |
| FVC (%) | 98.0 [88.0,108.0] | 98.0 [89.0,108.0] | 96.0 [83.0,107.0] | <0.05 | 99.0 [88.0,108.0] | 99.0 [89.0,109.0] | 96.0 [83.0,107.0] | <0.05 |
| pRBC (%) | 35 (1.0) | 15 (0.5) | 20 (3.1) | <0.05 | 54 (1.1) | 23 (0.6) | 31 (3.6) | <0.05 |
| Lung resection score (%)
1 |
776 (22.9) | 636 (23.3) | 140 (21.4) | <0.05 | 1481 (29.9) | 1270 (31.1) | 211 (24.4) | <0.05 |
| 2 | 115 (3.4) | 92 (3.4) | 23 (3.5) | 191 (3.9) | 157 (3.8) | 34 (3.9) | ||
| 3 | 1012 (29.9) | 897 (32.8) | 115 (17.6) | 1476 (29.8) | 1312 (32.1) | 164 (19.0) | ||
| 4 | 1428 (42.2) | 1070 (39.2) | 358 (54.7) | 1742 (35.2) | 1307 (32.0) | 435 (50.4) | ||
| 5 | 34 (1.0) | 20 (0.7) | 14 (2.1) | 39 (0.8) | 24 (0.6) | 15 (1.7) | ||
| 6 | 21 (0.6) | 17 (0.6) | 4 (0.6) | 22 (0.4) | 18 (0.4) | 4 (0.5) | ||
| ARISCAT Score | 35.0 [27.0,43.0] | 32 [27.0,43.0] | 43.0 [35.0,50.0] | <0.05 | 35.0 [27.0,43.0] | 27.0 [27.0,43.0] | 43.0 [29.5,50.0] | <0.05 |
| Hospital LOS | 3.19 [1.60,4.55] | 2.57 [1.60,4.22] | 7.31 [4.40,10.58] | <0.05 | 3.19 [1.60,4.55] | 2.52 [1.53,3.60] | 7.35 [4.40,11.15] | <0.05 |
Table 2
Time-weighted median of driving pressure (cmH2O).
Time-weighted median of mechanical power (J/min).
| Normal spirometry results | Abnormal spirometry results | |||||||||||||||
| Pplat | PIP | Pplat | PIP | |||||||||||||
| Overall
( n = 2237) |
Without PPC
( n = 1920) |
With
PPC ( n = 317) |
P-
Value |
Overall
( n = 3384) |
Without PPC
( n = 2944) |
With
PPC ( n = 440) |
P-Value | Overall
( n = 1149) |
Without PPC
( n = 812) |
With
PPC ( n = 337) |
P-Value | Overall
( n = 1567) |
Without PPC
( n = 1144) |
With
PPC ( n = 423) |
P-Value | |
| Driving pressure † (cmH 2O) | 14.0 [12.0,16.0] | 14.0 [12.0,16.0] | 14.0 [12.0,17.0] | <0.05 | 14.0 [12.0,17.0] | 14.0 [12.0,16.0] | 15.0 [12.0,18.0] | <0.05 | 14.0 [12.0,17.0] | 14.0 [12.0,17.0] | 15.0 [12.0,19.0] | <0.05 | 14.0 [12.0,17.0] | 14.0 [12.0,16.0] | 15.0 [12.0,18.0] | <0.05 |
| Mechanical power ‡ (J/min) | 6.7 [5.6,8.1] | 6.7 [5.6,8.1] | 6.8 [5.8,8.3] | <0.05 | 6.8 [5.7,8.2] | 6.8 [5.6,8.1] | 7.1 [6.0,8.5] | <0.05 | 7.2 [5.9,8.4] | 7.1 [5.8,8.3] | 7.4 [6.1,8.7] | <0.05 | 6.8 [5.7,8.2] | 6.8 [5.6,8.1] | 7.1 [6.0,8.5] | <0.05 |
| Age (year) | 59.0 [51.0,67.0] | 59.0 [51.0,66.0] | 61.0 [54.0,70.0] | <0.05 | 14.0 [12.0,17.0] | 14.0 [12.0,16.0] | 15.0 [12.0,18.0] | <0.05 | 67.0 [58.0,72.0] | 66.0 [57.0,72.0] | 67.0 [60.0,73.0] | <0.05 | 61.0 [53.0,69.0] | 60.0 [52.0,69.0] | 65.0 [57.0,72.0] | <0.05 |
| Sex, male (%) | 995 (44.5) | 809 (42.1) | 186 (58.7) | <0.05 | 1497 (44.2) | 1231 (41.8) | 266 (60.5) | <0.05 | 846 (73.6) | 580 (71.4) | 266 (78.9) | <0.05 | 1144 (73.0) | 816 (71.3) | 328 (77.5) | <0.05 |
| BMI (kg/m 2) | 23.8 [21.6,25.9] | 23.8 [21.6,25.9] | 23.7 [21.7,26.2] | 0.345 | 23.7 [21.6,25.8] | 23.8 [21.6,25.8] | 23.5 [21.5,26.0] | 0.780 | 23.6 (3.2) | 23.6 (3.1) | 23.5 (3.5) | 0.570 | 23.7 [21.6,25.8] | 23.8 [21.6,25.8] | 23.5 [21.5,26.0] | 0.780 |
| OP duration (hours) | 1.8 [1.1,2.4] | 1.7 [1.0,2.3] | 2.3 [1.8,3.1] | <0.05 | 1.9 [1.2,2.6] | 1.8 [1.1,2.4] | 2.5 [1.8,3.4] | <0.05 | 2.1 [1.2,3.0] | 1.9 [1.2,2.7] | 2.7 [1.8,3.6] | <0.05 | 1.9 [1.2,2.6] | 1.8 [1.1,2.4] | 2.5 [1.8,3.4] | <0.05 |
| Hb (g/dL) | 12.9 [12.0,14.0] | 12.9 [12.0,14.0] | 13.0 [12.0,14.0] | 0.797 | 12.9 [11.9,14.0] | 13.0 [12.0,14.0] | 12.9 [11.8,14.0] | 0.158 | 12.9 [11.8,14.0] | 13.0 [11.9,14.0] | 12.8 [11.5,14.0] | <0.05 | 12.9 [11.9,14.0] | 13.0 [12.0,14.0] | 12.9 [11.8,14.0] | 0.158 |
| BUN (g/dL) | 13.0 [11.0,16.0] | 13.0 [10.0,16.0] | 14.0 [11.0,17.0] | <0.05 | 14.0 [11.0,17.0] | 14.0 [11.0,17.0] | 14.0 [11.0,18.0] | <0.05 | 14.0 [11.0,18.0] | 14.0 [12.0,18.0] | 15.0 [11.0,18.0] | 0.935 | 14.0 [11.0,17.0] | 14.0 [11.0,17.0] | 14.0 [11.0,18.0] | <0.05 |
| ALB (g/dL) | 4.0 [3.7,4.3] | 4.0 [3.7,4.3] | 3.8 [3.6,4.2] | <0.05 | 3.9 [3.6,4.3] | 4.0 [3.7,4.3] | 3.8 [3.4,4.2] | <0.05 | 3.8 [3.5,4.2] | 3.9 [3.5,4.2] | 3.7 [3.3,4.1] | <0.05 | 3.9 [3.6,4.3] | 4.0 [3.7,4.3] | 3.8 [3.4,4.2] | <0.05 |
| Levin tube (%) | 25 (1.1) | 16 (0.8) | 9 (2.8) | 0.005 | 31 (0.9) | 21 (0.7) | 10 (2.3) | 0.004 | 22 (1.9) | 8 (1.0) | 14 (4.2) | <0.05 | 28 (1.8) | 11 (1.0) | 17 (4.0) | <0.05 |
| SpO 2 (%) | 97.0 [96.0,98.0] | 97.0 [96.0,98.0] | 97.0 [96.0,98.0] | 0.584 | 97.0 [96.0,98.0] | 97.0 [96.0,98.0] | 97.0 [96.0,98.0] | 0.147 | 97.0 [96.0,98.0] | 97.0 [96.0,98.0] | 97.0 [96.0,98.0] | <0.05 | 97.0 [96.0,98.0] | 97.0 [96.0,98.0] | 97.0 [96.0,98.0] | <0.05 |
| COPD (%) | 35 (1.6) | 28 (1.5) | 7 (2.2) | 0.326 | 43 (1.3) | 34 (1.2) | 9 (2.0) | 0.184 | 232 (20.2) | 139 (17.1) | 93 (27.6) | <0.05 | 274 (17.5) | 171 (14.9) | 103 (24.3) | <0.05 |
| ILD (%) | 45 (2.0) | 38 (2.0) | 7 (2.2) | 0.958 | 75 (2.2) | 64 (2.2) | 11 (2.5) | 0.795 | 79 (6.9) | 45 (5.5) | 34 (10.1) | <0.05 | 109 (7.0) | 64 (5.6) | 45 (10.6) | 0.001 |
| Asthma (%) | 71 (3.2) | 63 (3.3) | 8 (2.5) | 0.589 | 131 (3.9) | 119 (4.0) | 12 (2.7) | 0.230 | 70 (6.1) | 53 (6.5) | 17 (5.0) | 0.412 | 94 (6.0) | 70 (6.1) | 24 (5.7) | 0.834 |
| HF (%) | 19 (0.8) | 15 (0.8) | 4 (1.3) | 0.332 | 21 (0.6) | 17 (0.6) | 4 (0.9) | 0.341 | 23 (2.0) | 14 (1.7) | 9 (2.7) | 0.417 | 46 (2.9) | 29 (2.5) | 17 (4.0) | 0.169 |
| Smoking (%) | 424 (19.0) | 338 (17.6) | 86 (27.1) | <0.05 | 636 (18.8) | 511 (17.4) | 125 (28.4) | <0.05 | 424 (36.9) | 287 (35.3) | 137 (40.7) | 0.103 | 558 (35.6) | 397 (34.7) | 161 (38.1) | 0.241 |
| FEV 1 (%) | 106.0 [97.0,116.0] | 106.0 [98.0,116.0] | 106.0 [96.0,116.0] | 0.236 | 101.0 [89.0,112.0] | 102.0 [91.0,113.0] | 94.5 [81.0,108.0] | <0.05 | 84.0 [75.0,96.0] | 86.0 [76.0,98.0] | 82.0 [71.0,93.0] | <0.05 | 101.0 [89.0,112.0] | 102.0 [91.0,113.0] | 94.5 [81.0,108.0] | <0.05 |
| FVC (%) | 100.0 [91.0,108.0] | 100.0 [92.0,108.0] | 98.0 [89.0,108.0] | 0.043 | 98.0 [88.0,108.0] | 98.0 [89.0,108.0] | 96.0 [83.0,107.0] | <0.05 | 93.0 [76.0,106.0] | 94.0 [77.0,106.2] | 92.0 [73.0,106.0] | 0.071 | 98.0 [88.0,108.0] | 98.0 [89.0,108.0] | 96.0 [83.0,107.0] | <0.05 |
| pRBC (%) | 12 (0.5) | 10 (0.5) | 2 (0.6) | 0.683 | 22 (0.7) | 14 (0.5) | 8 (1.8) | <0.05 | 23 (2.0) | 5 (0.6) | 18 (5.3) | <0.05 | 32 (2.0) | 9 (0.8) | 23 (5.4) | <0.05 |
| Lung resection
score (%) 1 |
506 (22.6) | 451 (23.5) | 55 (17.4) | <0.05 | 1030 (30.4) | 936 (31.8) | 94 (21.4) | <0.05 | 270 (23.5) | 185 (22.8) | 85 (25.2) | <0.05 | 451 (28.8) | 334 (29.2) | 117 (27.7) | <0.05 |
| 2 | 37 (1.7) | 32 (1.7) | 5 (1.6) | 75 (2.2) | 69 (2.3) | 6 (1.4) | 78 (6.8) | 60 (7.4) | 18 (5.3) | 116 (7.4) | 88 (7.7) | 28 (6.6) | ||||
| 3 | 722 (32.3) | 665 (34.6) | 57 (18.0) | 1058 (31.3) | 975 (33.1) | 83 (18.9) | 290 (25.2) | 232 (28.6) | 58 (17.2) | 418 (26.7) | 337 (29.5) | 81 (19.1) | ||||
| 4 | 951 (42.5) | 757 (39.4) | 194 (61.2) | 1197 (35.4) | 946 (32.1) | 251 (57.0) | 477 (41.5) | 313 (38.5) | 164 (48.7) | 545 (34.8) | 361 (31.6) | 184 (43.5) | ||||
| 5 | 13 (0.6) | 10 (0.5) | 3 (0.9) | 16 (0.5) | 13 (0.4) | 3 (0.7) | 21 (1.8) | 10 (1.2) | 11 (3.3) | 23 (1.5) | 11 (1.0) | 12 (2.8) | ||||
| 6 | 8 (0.4) | 5 (0.3) | 3 (0.9) | 8 (0.2) | 5 (0.2) | 3 (0.7) | 13 (1.1) | 12 (1.5) | 1 (0.3) | 14 (0.9) | 13 (1.1) | 1 (0.2) | ||||
| ARISCAT Score | 32.0 [27.0,43.0] | 27.0 [27.0,43.0] | 43.0 [27.0,50.0] | <0.05 | 35.0 [27.0,43.0] | 32.0 [27.0,43.0] | 43.0 [35.0,50.0] | <0.05 | 43.0 [27.0,50.0] | 38.0 [27.0,43.0] | 43.0 [35.0,50.0] | <0.05 | 35.0 [27.0,43.0] | 32.0 [27.0,43.0] | 43.0 [35.0,50.0] | <0.05 |
| Hospital LOS | 2.58 [1.61,4.32] | 2.50 [1.59,3.54] | 5.55 [3.53,9.19] | <0.05 | 2.55 [1.55,4.28] | 2.42 [1.48,3.53] | 6.18 [3.55,9.35] | <0.05 | 4.29 [2.48,6.52] | 3.39 [2.27,4.55] | 8.28 [5.35,12.49] | <0.05 | 3.62 [2.34,6.38] | 3.27 [1.62,4.49] | 8.31 [5.34,12.82] | <0.05 |
© 2025 Elsevier Inc.