Secondary Logo

Journal Logo

Utilizing Forced Vital Capacity to Predict Low Lung Compliance and Select Intraoperative Tidal Volume During Thoracic Surgery

Hoftman, Nir MD*; Eikermann, Eric MD*; Shin, John MD*; Buckley, Jack MD*; Navab, Kaveh MD*; Abtin, Fereidoun MD; Grogan, Tristan MS; Cannesson, Maxime MD, PhD*; Mahajan, Aman MD, PhD*

doi: 10.1213/ANE.0000000000001885
Patient Safety: Original Clinical Research Report
Free

BACKGROUND: Tidal volume selection during mechanical ventilation utilizes dogmatic formulas that only consider a patient’s predicted body weight (PBW). In this study, we investigate whether forced vital capacity (FVC) (1) correlates better to total lung capacity (TLC) than PBW, (2) predicts low pulmonary compliance, and (3) provides an alternative method for tidal volume selection.

METHODS: One hundred thirty thoracic surgery patients had their preoperative TLC calculated via 2 methods: (1) pulmonary function test (PFT; TLCPFT) and (2) computed tomography 3D reconstruction (TLCCT). We compared the correlation between TLC and PBW with the correlation between TLC and FVC to determine which was stronger. Dynamic pulmonary compliance was then calculated from intraoperative ventilator data and logistic regression models constructed to determine which clinical measure best predicted low compliance. Ratios of tidal volume/FVC plotted against peak inspiratory pressure were utilized to construct a new model for tidal volume selection. Calculated tidal volumes generated by this model were then compared with those generated by the standard lung-protective formula Vt = 7 cc/kg.

RESULTS: The correlation between FVC and TLC (0.82 for TLCPFT and 0.76 for TLCCT) was stronger than the correlation between PBW and TLC (0.65 for TLCPFT and 0.58 for TLCCT). Patients with very low compliance had significantly smaller lung volumes (forced expiratory volume at 1 second, FVC, TLC) and lower diffusion capacity of the lungs for carbon monoxide when compared with patients with normal compliance. An FVC cutoff of 3470 cc was 100% sensitive and 51% specific for predicting low compliance. The proposed equation Vt = FVC/8 significantly reduced calculated tidal volume by a mean of 22.5% in patients with low pulmonary compliance without affecting the mean tidal volume in patients with normal compliance (mean difference 0.9%).

CONCLUSIONS: FVC is more strongly correlated to TLC than PBW and a cutoff of about 3.5 L can be utilized to predict low pulmonary compliance. The equation Vt = FVC/8 reduced mean calculated tidal volume in patients with low pulmonary compliance and/or small lungs.

Published ahead of print March 8, 2017.

From the Departments of *Anesthesiology and Perioperative Medicine

Radiology

Medicine Statistics Core, University of California, Los Angeles, California.

Published ahead of print March 8, 2017.

Accepted for publication December 12, 2016.

Funding: All support for this project came from Departmental Research Funds from the UCLA Department of Anesthesiology and Perioperative Medicine.

The authors declare no conflicts of interest.

Reprints will not be available from the authors.

Address correspondence to Nir Hoftman, MD, Department of Anesthesiology and Perioperative Medicine, University of California, 757 Westwood Blvd, Los Angeles, CA 90095. Address e-mail to nhoftman@mednet.ucla.edu.

Proper selection of tidal volume is crucial in mechanically ventilated patients. Numerous animal and clinical trials have demonstrated reduced lung injury, morbidity, and even mortality when smaller tidal volumes are utilized.1–4 The currently accepted method for determining tidal volume is based only on patients’ predicted body weight (PBW). Such a strategy for tidal volume selection assumes a robust correlation between PBW and lung volume. This assumption, however, is not supported in the scientific literature. In fact, normal variability of thoracic anatomy within a population, combined with changes that occur as a result of aging and lung disease, raise doubts whether a simple weight- or height-based equation can be expected to fit all individuals.5–7 Patients with reduced lung volume and/or pulmonary compliance may be particularly at risk for ventilator-induced barotrauma if the accepted PBW-based equation is used to determine tidal volume size.

Thoracic surgery patients are a subpopulation at high risk for postoperative acute lung injury given their (1) high incidence of preoperative pulmonary disease and (2) intraoperative trauma of their surgical lung.8–12 Unfortunately, the standard preoperative history, physical, and laboratory survey is unlikely to reliably identify low lung volumes and poor pulmonary compliance. Modern perioperative medicine is currently transitioning from standardized patient care to patient-centered precision medicine.13,14 Pulmonary function tests (PFTs) that measure (1) respiratory spirometry, (2) lung volumes, and/or (3) diffusion of carbon monoxide are uniquely positioned to deliver lung-focused data. Personalizing tidal volume selection based on measured lung parameters, rather than relying on standardized monograms such as weight and/or height could be the natural progression of this evolution. In particular, the PFT measure forced vital capacity (FVC) encompasses both patients’ pulmonary-specific performance as well as their physical stature. Such pulmonary measures may be well suited in helping to determine optimal tidal volume.

In this study involving patients undergoing thoracic surgery, we (1) determine if several PFT-derived measures are more strongly correlated to total lung capacity (TLC) than PBW, and (2) if these measures can be utilized to identify patients at high risk for poor pulmonary compliance. We hypothesize that FVC would more strongly correlate to TLC compared to PBW, and could be a useful tool for identifying patients at risk for low pulmonary compliance. Finally, (3) we propose an alternative formula for tidal volume selection that may better fit patients with low pulmonary compliance.

Back to Top | Article Outline

METHODS

Patient Selection and Data Collection

Figure 1.

Figure 1.

After obtaining institutional review board approval and permission to waive consent, the electronic medical records of 200 consecutive patients who underwent thoracic surgery between May 2013 and February 2014 at our institution were selected for review. The presentation of this article follows the STROBE checklist for observational studies and all data sheets have been deidentified. Eligibility for enrollment required the following inclusion criteria: (1) recent (<6 months) preoperative chest computed tomography (CT) scan and pulmonary function tests, (2) general anesthesia with controlled mechanical ventilation utilizing a large-bore endotracheal tube (at least 8.0-mm single lumen or 35F dual lumen), (3) stable period of 2 lung ventilation with neuromuscular blockade in the supine position before surgery. Patients with the following findings were excluded from enrollment: (1) age < 18, (2) tracheal or bronchial mass of any size, (3) previous lung lobectomy or pneumonectomy, (4) previous lung transplantation, (5) large (≥6.0 cm diameter) intrathoracic tumor, and (6) indwelling pleural catheter at the time of lung volume determination. One hundred thirty patients satisfied inclusion/exclusion criteria, and their data were analyzed in the study (Figure 1).

Back to Top | Article Outline

Lung Volume Investigation

Figure 2.

Figure 2.

Preoperative TLC was independently calculated by both (1) pulmonary function tests (primarily via plethysmography technique, rarely via helium dilution) and (2) CT 3D reconstruction. Utilizing 2 “Gold Standards” for measuring TLC added a layer of quality control by ensuring that errors in measurement due to limitations of 1 technique would not confound the measured results of the other. Having 2 independently measured TLC values for each patient and comparing the correlation between these measurements to a previously published reference enabled us to assess the accuracy of our measurements.15 Furthermore, the PFT-derived TLC (TLCPFT) data enabled analysis of lung function (gas diffusion, flow spirometry), whereas the CT-derived TLC (TLCCT) in turn allowed detailed assessment of pulmonary/thoracic anatomy including pathology that PFTs could not reveal. Vitrea (V6.4, Vital Images, Minnetonka, MN), an automated postimage acquisition software package, was utilized to generate the 3D reconstruction of the patients’ lungs from the chest CT data (Figure 2).16 Aerated lung and trachea were segmented from CT scan data, and the software allowed for separation of lung segmentation from the tracheobronchial tree. The lung volume from each individual CT slice was calculated and subsequently added together to generate a complete 3D lung image with corresponding TLC volume reported in milliliters. Not all patients had CT scans that were considered eligible for volumetric measurement and quantification. Sixteen patients whose CT scans exhibited motion artifact or were obtained in the expiratory phase of respiration were excluded from the study. A further 15 patients who did not have TLC measured in their PFTs were also excluded. Thus, we analyzed data from a total of 99 subjects with suitable CT and PFT studies in the lung volume investigation. PBW was calculated for each patient based on the ARDSnet definition3 (men: PBW [kg] = 50 + 2.3 [height {in}-60]; women: PBW [kg] = 45.5 + 2.3 [height {in}-60]).

Back to Top | Article Outline

Lung Compliance Investigation

The electronic intraoperative anesthesia record of every patient (n = 130) was manually reviewed by an anesthesiologist with expertise in thoracic anesthesia. Mechanical ventilation data (tidal volume, respiratory rate, peak inspiratory pressure, positive end expiratory pressure) was extracted from the time period immediately after bronchoscope-assisted tracheal tube positioning but before patient positioning for surgery. At the time, all patients were undergoing 2-lung volume control mechanical ventilation in the supine position under deep anesthesia and neuromuscular blockade. This highly controlled setting promoted low ventilator pressures; peak inspiratory pressures ≥ 25 cm H2O were unusual (seen in only 6% of patients) and were defined as “elevated” based on our clinical experience. Pulmonary static compliance could not be accurately determined due to the retrospective nature of the data collection and lack of a verifiable inspiratory pause (with true cessation of gas-flow and pressure equilibration) during the collection period of plateau pressure. Therefore, pulmonary dynamic compliance was calculated for each patient using the following equation: compliance = tidal volume / (peak inspiratory pressure – positive end expiratory pressure).17,18 Patients whose dynamic compliance was reduced by at least 50% from the lower limit of normal (compliance ≤ 25 mL/cm H2O) were identified as having markedly reduced pulmonary compliance.19,20 This subgroup of patients was then compared with the remaining patients (compliance > 25 mL/cm H2O) to identify any differences in their baseline characteristics.

Back to Top | Article Outline

Analysis and Statistics

Total lung capacity values (TLCPFT and TLCCT) were individually plotted against PBW and then FVC to determine each of those correlations, respectively. Patient characteristics and outcomes were summarized using means/SDs for continuous variables and frequencies/percentages for categorical variables. Pearson correlation coefficients were computed to assess the associations between lung volume measures (TLCCT and TLCPFT) and PBW or FVC. We compared the correlation coefficients using Williams Test.21

Dynamic pulmonary compliance was separately plotted against all baseline variables which differed statistically between the 2 compliance groups (compliance ≤ 25 mL/cm H2O versus compliance > 25 mL/cm H2O). Fisher exact test was utilized to compare incidence of elevated peak inspiratory pressure between the 2 compliance groups. Logistic regression models were constructed for the lung compliance analyses. Receiver operating characteristic curves were created and optimal cutoff points that maximized sensitivity/specificity were identified using the Youden index.22 The respective areas under the curve were then compared utilizing the Delong test.

Ratios of tidal volume/FVC were plotted against peak inspiratory pressure to characterize this relationship. We hypothesized that as tidal volume/FVC ratio decreased, so would the probability of seeing high peak pressures due to excessive tidal volume. We then set out to identify the ideal tidal volume/FVC ratio which would achieve the clinical needs of ventilation while adequately protecting the lungs from barotrauma. These data were utilized to construct a model for tidal volume selection that would better fit patients with wide ranges of pulmonary compliance. The tidal volume generated by this novel formula was then calculated for each patient and compared to the tidal volume determined from the accepted lung-protective equation (tidal volume = 7 mL/kg PBW).

Lung volume analyses were performed using R V3.1.2 (Vienna, Austria). Lung compliance and tidal volume analyses were performed using IBM SPSS V23 (Armonk, NY). P < .05 were considered statistically significant.

Before data collection, we hypothesized that the correlation between PBW and TLC would be around 0.5 (Pearson correlation coefficient) and between FVC and TLC would be about 0.7. This estimate was based on informal clinical observation and analysis over the past decade. An a priori power analysis determined that at least 80 patients would be needed to show such a difference with 80% power (at α = .05). Based on our initial observation that only around 50% of patients met all the inclusion and exclusion criteria, we decided to review 200 patient charts. Given our secondary goal of developing a model for selecting tidal volume in patients with low pulmonary compliance, we chose to enroll all eligible patients from the original patient cohort to enhance that analysis.

Back to Top | Article Outline

RESULTS

Table 1.

Table 1.

Baseline patient demographics and pulmonary characteristics are summarized in Table 1.

Back to Top | Article Outline

Lung Volume Investigation

Table 2.

Table 2.

Figure 3.

Figure 3.

The correlation between FVC and TLC (0.82 for TLCPFT and 0.76 for TLCCT) was stronger when compared with the correlation between PBW and TLC (0.65 for TLCPFT and 0.58 for TLCCT). This finding remained consistent irrespective of whether TLCCT or TLCPFT was used in the comparisons, as graphically illustrated in Figure 3. Such consistency was the result of the very strong correlation between TLCCT and TLCPFT, equal to that published in the historical reference. All the correlations coefficients and statistical comparisons are summarized in Table 2.

Back to Top | Article Outline

Lung Compliance Investigation

Table 3.

Table 3.

Figure 4.

Figure 4.

A comparison of baseline patient demographics and pulmonary characteristics between patients with markedly low pulmonary compliance and the rest of the population is summarized in Table 3. The low compliance subgroup included 21 patients (16% of total study population), 8 of whom experienced elevated intraoperative peak inspiratory pressures (≥25 cm H2O), while no patients from the normal compliance subgroup experienced elevated peak pressures (P < .001). Patients with markedly low compliance had significantly smaller lung volumes (forced expiratory volume in 1 second [FEV1], FVC, TLC) and lower carbon monoxide diffusing capacity (DLCO) when compared with patients with compliance >25 mL/cm H2O. Receiver operator curves generated for the comparisons between pulmonary compliance and PBW, FEV1, TLC, FVC, and DLCO, respectively, demonstrated that DLCO and FVC yielded the highest area under the curve (0.79 and 0.78, respectively). The area under the curve for FVC was significantly greater than that of PBW (0.78 vs 0.58, P = .002). These results are summarized in Figure 4. As a clinical test, an FVC cutoff of 3470 mL was 100% sensitive (95% CI 0.81–1.00) and 51% specific (95% CI 0.41–0.61) for predicting low pulmonary compliance.

Back to Top | Article Outline

Proposed Formula for Tidal Volume Selection

Figure 5.

Figure 5.

Figure 6.

Figure 6.

We propose a new “lung-centric formula” for selecting tidal volume: Vt = FVC/8. Tidal volumes calculated for each patient by this formula differed by an average of 4.4% from those calculated by the standard PBW-based lung-protective equation (tidal volume = 7 mL/kg PBW). This proposed lung-centric formula reduced calculated tidal volume in patients with markedly low compliance by an average of 22.5%; 8 of these patients who also experienced high intraoperative peak airway pressures saw a 30% mean reduction in calculated tidal volume compared to the PBW-based equation. Those with compliance >25 mL/cm H2O demonstrated an average difference of 0.9% between the 2 methods. When compared with the standard equation, the proposed formula reduced calculated tidal volume in patients with small lungs (smallest quartile) and enlarged calculated tidal volume in all patients with large lungs (largest quartile). These results are summarized in Figures 5 and 6.

Back to Top | Article Outline

DISCUSSION

This study demonstrated that (1) FVC is more strongly correlated to TLC than is PBW, (2) low FVC can identify low lung compliance, and (3) FVC can be utilized to determine tidal volume. The weaker correlation between PBW and TLC was not surprising given that the PBW calculation uses height as its major determinant, and human height is much more heavily influenced by femur length than thoracic girth.23 Clinically, this weaker correlation translated to a greater than 2-fold difference between the smallest and largest TLC values measured among patients with identical PBW. Given the natural variability in thoracic proportions within a population, a tidal volume equation based solely on PBW could over- or underestimate required tidal volumes, especially in patients with small or large lungs.5

An FVC value less than 3.5 L was found to be 100% sensitive and about 50% specific for detecting low compliance. Although other variables including TLCCT and DLCO had equivalent areas under their respective receiver operator curves, FVC was better suited for clinical care given its high sensitivity and ease of acquisition at the bedside. We deliberately chose a cutoff point that maximized sensitivity to ensure that all patients with low compliance would be identified to receive lung protection. The low specificity of the cutoff was not surprising, because patients with low FVC likely represented two different subpopulations: (1) small patients with normal, compliant lungs, and (2) average sized patients with diseased, small noncompliant lungs. Differentiating which subpopulation a patient with low FVC belonged to is beyond the capabilities of a single cutoff value, because it would require calculation of pulmonary compliance. Nevertheless, having a simple tool to identify patients potentially at risk for barotrauma could still aid clinicians in everyday ventilator management.

Precisely to aid in these challenging clinical scenarios, we developed an experimental model that may better select tidal volumes for patients with small lung volumes and low pulmonary compliance. Compared with the standard lung-protective formula, the proposed equation reduced mean calculated tidal volume in patients with small, noncompliant lungs, increased it in patients with large lungs, and left it largely unchanged in the remainder of the population. The mechanism for ventilator-associated lung injury is complex and not fully understood, but evidence suggests that excessive alveolar stretch, combined with repetitive alveolar open/close cycles, contribute to the injury.24 For this reason, low lung volumes and positive end expiratory pressure have been widely adopted with the goal of reducing barotrauma/volutrauma and atelectotrauma, respectively. However, ventilator settings that reduce tidal volume often increase atelectasis, requiring increasing levels of positive end expiratory pressure, which can be detrimental in some settings.25,26 A tidal volume formula based on lung dimensions rather than weight or height could possibly aid in selection of safe ventilator settings while balancing these often opposing requirements. This pilot study was not designed or powered to answer such questions, nor were its results validated on a secondary, nonthoracic surgery patient cohort. Future studies would need to be conducted to determine whether the tidal volume reductions suggested by our proposed equation actually reduce lung inflammation and injury, and whether tidal volume increases in patients with large lungs are indeed safe. Nevertheless, the time has come to look past the dogma of weight- and height-based formulas; new parameters must be studied in the quest for individualized, precision medical care delivery.

There were several limitations due to study design that we wish to address. First, the TLCCT was generated from images acquired in a CT scanner, and thus patients were not coached to the same degree as they were when they had their TLCPFT measured. It is therefore possible that the TLCCT values could have been artificially low due to reduced patient inspiratory effort. However, CT technicians at our facility routinely instruct patients to take a maximally deep breath and hold it during scanning. Furthermore, any images with motion artifact or evidence of submaximal inspiratory effort were eliminated from data analysis. Finally, the correlation between TLCCT and TLCPFT was very strong and identical to a study that compared prospectively collected CT volumes with PFT volumes, thus validating the quality of the CT-generated data.15 Second, our decision to enroll thoracic surgery patients could confound the relationship between lung volumes and PBW given the higher incidence of lung pathology in this patient population. However, to see an effect difference in a small pilot trial we needed to study a population with an expected high incidence of pulmonary disease and reduced compliance that would likely benefit most from the intervention. Detailed patient history review and analysis of 3D lung images enabled us to identify and exclude patients with anatomical thoracic abnormalities (such as previous lung lobectomy, large intrathoracic mass) that could confound the expected relationship between TLC and PBW. Third, the retrospective nature of ventilator data collection introduced several potential confounders. Because of the lack of a consistent inspiratory pause, we could not reliably measure plateau pressure and thus could not calculate static airway compliance. Instead, we measured dynamic airway compliance, which could be influenced by endotracheal tube diameter, inspiration to expiration (I:E) ratio, and respiratory rate.27,28 To mitigate the effects of these and other confounders, we excluded patient with tracheal tubes smaller than 8.0-mm single lumen or 35F dual lumen, and collected ventilator data during deep anesthesia and neuromuscular blockade. We are thus confident that endotracheal tube gas flow resistance and patient extrinsic muscle tone did not significantly affect our measured compliance value. Ventilator settings such as respiratory rate and I:E ratio were not different among the 2 compliance subgroups, making it unlikely that these variables had an appreciable effect on our results. Fourth, given that a significant portion of our initial sample population was excluded from analysis due to missing data, we could not exclude the possibility of unforeseen biases skewing the results. For example, only patients with both TLCCT and TLCPFT were included in the lung volume investigation, which further reduced the sample size of that analysis to n = 99. We chose this design in order to eliminate comparisons between 2 overlapping but different (and numerically uneven) patient cohorts, thus simplifying the methods without significantly impacting the results. Finally, the perioperative surgical population studied did not include any critically ill patients with acute lung injury, a cohort that could perhaps benefit most from precise ventilator settings. FVC measurement, being highly effort dependent, may not be well suited for this debilitated patient population.

In conclusion, (1) FVC was more strongly correlated to TLC than was PBW. (2) An FVC value below 3.5 L was 100% sensitive and about 50% specific for detecting patients with low pulmonary compliance. (3) When compared with the PBW-based lung-protective formula (tidal volume = 7 mL/kg PBW), the proposed lung-centric formula Vt = FVC/8 reduced calculated tidal volume in patients with low pulmonary compliance without significantly affecting mean tidal volume in patients with normal compliance. Future research will need to confirm the validity of this model in prospective studies with diverse patient populations.

Back to Top | Article Outline

DISCLOSURES

Name: Nir Hoftman, MD.

Contribution: This author helped conceive and design the study; acquire, analyze and interpret the data; and draft and revise the manuscript.

Name: Eric Eikermann, MD.

Contribution: This author helped acquire, analyze and interpret the data; and draft and revise the manuscript.

Name: John Shin, MD.

Contribution: This author helped acquire, analyze and interpret the data; and draft and revise the manuscript.

Name: Jack Buckley, MD.

Contribution: This author helped acquire, analyze and interpret the data; and draft and revise the manuscript.

Name: Kaveh Navab, MD.

Contribution: This author helped acquire, analyze and interpret the data; and draft and revise the manuscript.

Name: Fereidoun Abtin, MD.

Contribution: This author helped acquire, analyze and interpret the data; and draft and revise the manuscript.

Name: Tristan Grogan, MS.

Contribution: This author helped acquire, analyze and interpret the data; and draft and revise the manuscript.

Name: Maxime Cannesson, MD, PhD.

Contribution: This author helped acquire, analyze and interpret the data; and draft and revise the manuscript.

Name: Aman Mahajan, MD, PhD.

Contribution: This author helped conceive and design the study; acquire, analyze and interpret the data; and draft and revise the manuscript.

This manuscript was handled by: Richard C. Prielipp, MD.

Back to Top | Article Outline

REFERENCES

1. Brégeon F, Roch A, Delpierre S, et al.Conventional mechanical ventilation of healthy lungs induced pro-inflammatory cytokine gene transcription. Respir Physiol Neurobiol. 2002;132:191–203.
2. Licker M, Diaper J, Villiger Y, et al.Impact of intraoperative lung-protective interventions in patients undergoing lung cancer surgery. Crit Care. 2009;13:R41.
3. The Acute Respiratory Distress Syndrome Network. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med. 2000;342:1301–1308.
4. Vaneker M, Halbertsma FJ, van Egmond J, et al.Mechanical ventilation in healthy mice induces reversible pulmonary and systemic cytokine elevation with preserved alveolar integrity: an in vivo model using clinical relevant ventilation settings. Anesthesiology. 2007;107:419–426.
5. Bellemare JF, Cordeau MP, Leblanc P, Bellemare FThoracic dimensions at maximum lung inflation in normal subjects and in patients with obstructive and restrictive lung diseases. Chest. 2001;119:376–386.
6. Lowery EM, Brubaker AL, Kuhlmann E, Kovacs EJThe aging lung. Clin Interv Aging. 2013;8:1489–1496.
7. Sharma G, Goodwin JEffect of aging on respiratory system physiology and immunology. Clin Interv Aging. 2006;1:253–260.
8. Licker M, de Perrot M, Spiliopoulos A, et al.Risk factors for acute lung injury after thoracic surgery for lung cancer. Anesth Analg. 2003;97:1558–1565.
9. Fernández-Pérez ER, Keegan MT, Brown DR, Hubmayr RD, Gajic OIntraoperative tidal volume as a risk factor for respiratory failure after pneumonectomy. Anesthesiology. 2006;105:14–18.
10. Licker M, Fauconnet P, Villiger Y, Tschopp JMAcute lung injury and outcomes after thoracic surgery. Curr Opin Anaesthesiol. 2009;22:61–67.
11. Ruffini E, Parola A, Papalia E, et al.Frequency and mortality of acute lung injury and acute respiratory distress syndrome after pulmonary resection for bronchogenic carcinoma. Eur J Cardiothorac Surg. 2001;20:30–36.
12. Schilling T, Kozian A, Huth C, et al.The pulmonary immune effects of mechanical ventilation in patients undergoing thoracic surgery. Anesth Analg. 2005;101:957–965.
13. Hamburg MA, Collins FSThe path to personalized medicine. N Engl J Med. 2010;363:301–304.
14. Miller TE, Roche AM, Mythen MFluid management and goal-directed therapy as an adjunct to Enhanced Recovery After Surgery (ERAS). Can J Anesth. 2015;62:158–168.
15. Brown MS, Kim HJ, Abtin F, et al.Reproducibility of lung and lobar volume measurements using computed tomography. Acad Radiol. 2010;17:316–322.
16. Saito T, Horie M, Sato M, et al.Low-dose computed tomography volumetry for subtyping chronic lung allograft dysfunction. J Heart Lung Transplant. 2016;35:59–66.
17. Grinnan DC, Truwit JDClinical review: respiratory mechanics in spontaneous and assisted ventilation. Crit Care. 2005;9:472–484.
18. Kress JP, Hall JBApproach to the Patient With Critical Illness. Harrison’s Principles of Internal Medicine. 2015.New York, NY: McGraw-Hill
19. MacIntyre NREvidence-based guidelines for weaning and discontinuing ventilatory support: a collective task force facilitated by the American College of Chest Physicians; the American Association for Respiratory Care; and the American College of Critical Care Medicine. Chest. 2001;120:375S–379S.
20. Singer BD, Corbridge TCBasic invasive mechanical ventilation. South Med J. 2009;102:1238–1245.
21. Williams EJThe comparison of regression variables. J R Stat Soc. 1959;Series B:396–399.
22. Youden WJIndex for rating diagnostic tests. Cancer. 1950;3:32–35.
23. Soranzo N, Rivadeneira F, Chinappen-Horsley U, et al.Meta-analysis of genome-wide scans for human adult stature identifies novel loci and associations with measures of skeletal frame size. PLoS Genet. 2009;5:e1000445.
24. Albaiceta GM, Blanch LBeyond volutrauma in ARDS: the critical role of lung tissue deformation. Crit Care. 2011;15:304.
25. Luecke T, Pelosi PClinical review: positive end-expiratory pressure and cardiac output. Crit Care. 2005;9:607–621.
26. Tuxen DVDetrimental effects of positive end-expiratory pressure during controlled mechanical ventilation of patients with severe airflow obstruction. Am Rev Respir Dis. 1989;140:5–9.
27. Gal TJPulmonary mechanics in normal subjects following endotracheal intubation. Anesthesiology. 1980;52:27–35.
28. Sullivan M, Paliotta J, Saklad MEndotracheal tube as a factor in measurement of respiratory mechanics. J Appl Physiol. 1976;41:590–592.
Copyright © 2017 International Anesthesia Research Society