Original Research
Cardiothoracic Imaging
September 1, 2021

Preoperative Thoracic CT Findings Associated With Postoperative Mechanical Ventilation in Patients Undergoing Major Abdominal or Pelvic Surgery: A Matched Case-Control Study

Abstract

Please see the Editorial Comment by Babina Gosangi discussing this article.
BACKGROUND. Postoperative prolonged mechanical ventilation is associated with increased morbidity and mortality. Reliable predictors of the need for postoperative mechanical ventilation after abdominal or pelvic surgeries are lacking.
OBJECTIVE. The purpose of this study was to explore associations between preoperative thoracic CT findings and the need for postoperative mechanical ventilation after major abdominal or pelvic surgeries.
METHODS. This retrospective case-control study included patients who underwent abdominal or pelvic surgeries during the period from January 1, 2014, through December 31, 2018, and had undergone preoperative thoracic CT. Case patients were patients who required postoperative mechanical ventilation. Control patients and case patients were matched at a 3:1 ratio on the basis of age, sex, body mass index, chronic obstructive pulmonary disease, smoking status, and surgery type. Two radiologists (readers 1 and 2) reviewed the CT images. Findings were compared between groups.
RESULTS. The study included 165 patients (70 women, 95 men; mean age, 67.0 ± 9.7 [SD] years; 42 case patients and 123 matched control patients). Bronchial wall thickening and pericardial effusion were more frequent in case patients than control patients for reader 2 (10% vs 2%, p = .03; 17% vs 5%, p = .01) but not for reader 1. Pulmonary artery diameter (mean ± SD) was greater in case patients than control patients for reader 2 (2.9 ± 0.5 cm vs 2.8 ± 0.5 cm, p = .045) but not reader 1. Right lung height was lower in case patients than control patients for reader 1 (18.4 ± 2.9 cm vs 19.9 ± 2.7 cm, p = .01) and reader 2 (18.3 ± 2.9 cm vs 19.8 ± 2.7 cm, p = .01). Left lung height was lower in case patients than control patients for reader 1 (19.5 ± 3.1 cm vs 21.1 ± 2.6 cm, p = .01) and reader 2 (19.6 ± 2.4 cm vs 20.9 ± 2.6 cm, p = .01). Anteroposterior (AP) chest diameter was greater for case patients than control patients for reader 1 (14.0 ± 2.3 cm vs 12.9 ± 3.7 cm, p = .02) and reader 2 (14.2 ± 2.2 cm vs 13.2 ± 3.6 cm, p = .04). In a multivariable regression model using pooled reader data, bronchial wall thickening exhibited an odds ratio (OR) of 4.6 (95% CI, 1.3–16.5; p = .02); pericardial effusion, an OR of 5.1 (95% CI, 1.7–15.5; p = .004); pulmonary artery diameter, an OR of 1.4 per 1-cm increase (95% CI, 0.7–3.0; p = .32); mean lung height, an OR of 0.8 per 1-cm increase (95% CI, 0.7–1.001; p = .05); and AP chest diameter, an OR of 1.2 per 1-cm increase (95% CI, 1.013–1.4; p = .03).
CONCLUSION. CT features are associated with the need for postoperative mechanical ventilation after abdominal or pelvic surgery.
CLINICAL IMPACT. Many patients undergo thoracic CT before abdominal or pelvic surgery; the CT findings may complement preoperative clinical risk factors.

HIGHLIGHTS

Key Finding
In a matched case-control study of 165 patients undergoing major abdominal or pelvic surgery, independent predictors (p < .05) of postoperative mechanical ventilation using preoperative thoracic CT included bronchial wall thickening, pericardial effusion, shorter lung height, and greater AP chest diameter (OR = 4.8, 5.3, 0.8 per 1-cm increase, and 1.2 per 1-cm increase, respectively).
Importance
Preoperative thoracic CT findings can be used to improve risk assessment for need for postoperative mechanical ventilation in patients undergoing major abdominal or pelvic surgery.
Advances in surgical techniques and perioperative management strategies have resulted in an increasing number of surgeries being performed in patients who have multiple significant chronic comorbidities [1]. Patients with chronic health conditions are thought to be at higher risk for requiring prolonged mechanical ventilation after leaving the operating room [2]. The incidence of postoperative mechanical ventilation varies significantly depending on the definition of postoperative mechanical ventilation and the type of surgery performed (cardiac vs noncardiac). Patients who undergo cardiac surgery commonly remain intubated and receive postoperative mechanical ventilation for several hours after surgery. Thus, the cardiac surgery literature focuses on prolonged postoperative mechanical ventilation (e.g., mechanical ventilation lasting > 24 [3, 4], 48 [5–7], 72 [8], or 96 [9] hours). Prolonged postoperative mechanical ventilation is needed in approximately 3–23% of patients who undergo cardiac surgery [310].
Patients who undergo noncardiac surgery are typically extubated in the operating room and thus generally do not receive postoperative mechanical ventilation. Postoperative mechanical ventilation is needed in approximately 2–7% of patients who undergo noncardiac surgery [1114]. Regardless of the definition, mechanical ventilation after noncardiac surgeries and the need for extended mechanical ventilation after cardiac surgeries have been associated with significantly increased mortality and morbidity, longer hospital and ICU stays, higher health care costs, and increased utilization of health care resources [7, 9, 10, 12, 14].
Several prior studies have investigated the preoperative and intraoperative risk factors for the need for postoperative mechanical ventilation, especially with regard to their association with prolonged intubation after cardiac or abdominal surgeries [1, 4, 6, 7, 1016]. Although some studies have found positive correlations with patient age, smoking status, and presence of congestive heart failure, currently no preoperative patient physical characteristics, laboratory results, or lung function tests have been shown to reliably identify patients who cannot be successfully extubated at the end of surgery [4, 1619]. This observation is particularly true for patients undergoing noncardiac surgeries [1, 1113]. However, preoperative chest CT is performed in many patients scheduled to undergo major abdominal or pelvic noncardiac surgeries as part of the workup for potential metastatic cancer or as part of preoperative surgical clearance. Findings on chest CT images may help predict which patients undergoing major abdominal and pelvic surgery will require postoperative mechanical ventilation. We therefore conducted this study to explore associations between findings on preoperative chest CT and the need for postoperative mechanical ventilation after major abdominal or pelvic surgery.

Methods

Study Population and Design

This study was a retrospective single-center case-control study of patients who underwent major nonemergency abdominal or pelvic surgery at University Hospitals at The University of Texas (UT) Southwestern Medical Center in Dallas. The study was approved by UT Southwestern Medical Center's institutional review board with a waiver of the need for individual patient informed consent.
The anesthesia registry in the electronic medical record (EMR) system was searched for patients who underwent major abdominal or pelvic surgery requiring general endotracheal tube anesthesia between January 1, 2014, and December 31, 2018, which yielded 54,557 patients. The search start date was selected as January 1, 2014, because the EMR did not contain discrete data for intraoperative events before that date. Among the patients retrieved by the initial search, a review of the radiology registry showed that 1479 patients had undergone thoracic CT (conventional chest CT with or without IV contrast material, pulmonary CTA, or high-resolution chest CT) within the 12 months preceding the surgery. All of these thoracic CT examinations were performed according to a standard institutional protocol with the patient using inspiratory breath-holding under technologist instruction, after having been coached by the technologist before the acquisition.
A cardiothoracic radiologist (A.C., with 7 years of experience) manually reviewed the EMR, resulting in patient exclusions for the following reasons: other abdominal procedures (e.g., esophagogastroduodenoscopy or colonoscopy) (n = 348), emergency surgery (n = 302), solid organ transplant surgery (n = 36), preoperatively planned postoperative transfer to the ICU (n = 40), postoperative transfer to the ICU for nonrespiratory reasons (e.g., bleeding control or hemodynamic stability) (n = 33), or the available thoracic CT performed within 12 months before surgery was insufficient for analysis (e.g., only limited imaging of chest for biopsy or radiation therapy planning) (n = 334) (Fig. 1). The patients with planned postoperative ICU transfer were excluded because the postoperative mechanical ventilation may have been planned preoperatively. For the remaining 386 patients, the previously noted investigator (A.C.) reviewed the EMR to extract clinical data including age, sex, body mass index (BMI), smoking history (former, current, or never smoker), clinical diagnosis of chronic obstructive pulmonary disease (COPD) (present or absent), and type of surgery (genitourinary [GU], gastrointestinal [GI], or gynecologic [GYN]).
Fig. 1 —Schematic description of patient selection. EMR = electronic medical record, BMI = body mass index (weight in kilograms divided by square of height in meters), GI = gastrointestinal, GU = genitourinary, GYN = gynecologic.
Patients were categorized into two groups. The case group was defined by the need for postoperative mechanical ventilation via endotracheal tube after leaving the operating room at the end of primary surgery and unplanned postoperative admission to the ICU. Forty-two patients met these criteria; all were included in the case group. To be eligible for the control group, patients needed to be extubated in the operating room at the end of the procedure and then leave the operating room after surgery breathing spontaneously, without transfer from the operating room to the ICU. A total of 344 patients were successfully extubated in the operating room at the end of the primary surgery without subsequent ICU transfer and were categorized as potential control patients. Potential control patients were matched to the case patients in a 3:1 ratio (as described in the Statistical Methods section) on the basis of age in decile, sex (female vs male), BMI ([weight in kilograms divided by square of height in meters] < 30 vs ≥ 30), clinical diagnosis of COPD (present vs absent), smoking status (former, current, or never smoker), and type of surgery (GI, GU, or GYN), providing initially 126 control patients. During the image analysis that occurred after matching of case patients and control patients, three control patients were excluded because the thoracic CT study was found to provide incomplete anatomic coverage. The final study cohort for analysis thus included a total of 165 patients: 42 case patients and 123 matched control patients.

Image Analysis

At least 3 months after the earlier EMR review, the previously noted cardiothoracic radiologist (A.C.) and two additional cardiothoracic radiologists (J.L.L., with 14 years of experience, and A.K., with 17 years of experience) performed an initial joint review of 20 thoracic CT examinations that were not from patients included in the study sample to develop consensus in the approach to measuring the CT parameters. After this joint review of 20 CT examinations, A.C. and J.L.L. independently reviewed a separate group of 16 CT examinations from unmatched control patients to assess interobserver agreement and identify potential sources of reader discrepancy. After that preliminary review, the two investigators independently reviewed the thoracic CT examinations of the 165 study patients using a dedicated PACS station while blinded to each other's CT readings, patient clinical information, and case-versus-control group assignment.

Qualitative Findings

The axial CT images were evaluated for the following qualitative findings: respiratory motion, atelectasis (assessed separately for right and left sides), pleural effusion (assessed separately for right and left sides), pneumothorax, Kerley B lines, emphysema, bullous disease, bronchial wall thickening, bronchiectasis, pulmonary nodules, mediastinal or hilar lymphadenopathy, cardiomegaly, pericardial effusion, CT evidence of prior lung surgery, CT evidence of pulmonary fibrosis, and CT evidence of interstitial lung disease (ILD). Most features were assessed in a binary (present vs absent) fashion. Respiratory motion was categorized as none, mild, or severe, and emphysema was classified as none, mild, moderate, or severe. Mild respiratory motion was defined as motion resulting in only blurring of lung markings. Severe respiratory motion was defined as motion resulting in a double appearance of the diaphragm that compromised assessment of the quantitative parameters. Bronchial wall thickening was assessed visually relative to the bronchial luminal diameter; if the bronchial wall thickness was more than 20% of the bronchial luminal diameter on visual assessment, then the wall was considered to be thickened [20]. Table S1, which can be viewed in the AJR electronic supplement to this article at https://doi.org/10.2214/AJR.21.26411, provides additional descriptions of the definitions of the qualitative findings for purposes of this study, reflecting the Glossary of Terms for Thoracic Imaging published by the Fleischner Society [21].

Quantitative Parameters

The following quantitative parameters were recorded: number of lobes affected by any type of pathology (from one to five lobes), anteroposterior (AP) diameter of trachea, transverse diameter of trachea, pulmonary artery diameter, right lung height, left lung height, AP diameter of chest, transverse diameter of chest, and subcutaneous fat thickness. The two tracheal diameters were measured orthogonally on axial images. The height of each lung was measured from apex to the mid diaphragm craniocaudally at the hilar level on sagittal images (Figs. 2A and 2B). The AP diameter of the chest (from inner border of the sternum to the anterior border of the vertebral body) (Fig. 2C) and transverse diameter of the chest (from inner borders of the right and left costal surfaces, orthogonal to the AP diameter), along with the subcutaneous fat tissue thickness (from skin to outer border of the sternum), were obtained at the level of the xiphoid process on axial images.
Fig. 2A —Preoperative chest CT images of two patients who underwent major surgery and did not require postoperative mechanical ventilation.
A, Sagittal chest CT images in 62-year-old woman who underwent hysterectomy for uterine cancer show measurement of right (A) and left (B) lung length (arrow) at hilar level from apex to diaphragmatic dome. Right lung length was 20.1 cm for reader 1 and 20.0 cm for reader 2; left lung length was 21.7 cm for reader 1 and 21.3 cm for reader 2. Patient did not require postoperative mechanical ventilation.
Fig. 2B —Preoperative chest CT images of two patients who underwent major surgery and did not require postoperative mechanical ventilation.
B, Sagittal chest CT images in 62-year-old woman who underwent hysterectomy for uterine cancer show measurement of right (A) and left (B) lung length (arrow) at hilar level from apex to diaphragmatic dome. Right lung length was 20.1 cm for reader 1 and 20.0 cm for reader 2; left lung length was 21.7 cm for reader 1 and 21.3 cm for reader 2. Patient did not require postoperative mechanical ventilation.
Fig. 2C —Preoperative chest CT images of two patients who underwent major surgery and did not require postoperative mechanical ventilation.
C, Axial chest CT image in 65-year-old woman who underwent nephrectomy for renal cell carcinoma shows measurement of anteroposterior (AP) diameter of chest (arrow) at level of xiphoid junction, from posterior border of sternum to anterior border of vertebral column. AP chest diameter was 11.5 cm for reader 1 and 11.4 cm for reader 2. Patient did not require postoperative mechanical ventilation.

Statistical Methods

Potential control patients were matched to case patients using a stratification score. The stratification score represents the probability of the outcome of postoperative mechanical ventilation modeled as a function of potential confounders including sex, smoking status, COPD, type of surgery, age at encounter, and BMI [22]. Three control patients were matched to each case patient in a manner designed to minimize overall dissimilarity using the optmatch package in R (version R 4.0.2, R Foundation for Statistical Computing) [23].
Reader agreement for categoric measures was assessed by simple and weighted kappa coefficients. Prevalence-adjusted bias-adjusted kappa (PABAK) coefficients were also determined given the low prevalence of many features. For continuous measurements, agreement was assessed by intraclass correlation coefficients (ICCs) (absolute agreement with two-way mixed model). Reader agreement was interpreted as follows: 0.76–1.00, excellent agreement; 0.60–0.75, good; 0.40–0.60, fair; and less than 0.40, poor. The 95% CIs were reported for all estimates.
Descriptive statistics were reported as a mean and SD for continuous measurements and as a frequency and percentage for categoric measurements. The qualitative and quantitative parameters were compared between case and control patients for each reader using conditional logistic regression accounting for the paired design.
Further analysis was performed of selected variables that showed a significant difference between the two groups for at least one of the two readers. This further analysis was conducted using pooled data from the two readers for the selected variables via generalized estimating equations (GEEs), adjusting for the matching of patients and the assessment of each was assessed by two independent readers. In addition, before the further analysis of the selected variables, the variables' collinearity was assessed through Pearson correlation coefficients; initial model adjustment was performed for any identified collinearity. AUC was computed for quantitative measures. Sensitivity and specificity were computed for qualitative and quantitative measures. Considering the large impact of an unexpected need for postoperative mechanical ventilation (i.e., of false-negatives), thresholds were determined for the quantitative measurements to ensure a sensitivity of at least 90%.
GEEs were used to compare the selected variables between case and control patients, adjusting for the fact that patients were matched and that each patient's CT examination was assessed by two independent readers. From the GEE, odds ratios (ORs) and 95% CIs were derived for the selected variables. Multivariable GEE regression was applied for the selected variables to identify independent predictors. The ORs were adjusted for patient height as a covariate in both the univariable and multivariable models.
A mixed-effects model was used to model the association between lung height and patient height for case and control patients for each lung for each reader. The slopes of the relationships between lung height and patient height were compared between case and control patients via interaction terms using F tests.
Statistical analyses were performed in SAS software (version 9.4, SAS Institute).

Sample Size Calculation

Due to the low incidence rate of the imaging findings in the control group, we assumed a probability of exposure among sampled control patients of 5% and a correlation coefficient for exposure between matched case and control patients of 0.2 [24]. A sample of 42 case patients with a 3:1 matching sample from the control group achieves 80% power to detect an OR of 5.89 or above with a .05 significance level. Sample size calculation was performed using PASS software (version 14, NCSS).

Results

Patient Characteristics

The 165 patients included 70 (42.4%) women and 95 (57.6%) men. The mean patient age was 67.0 ± 9.7 (SD) years. The mean BMI (weight in kilograms divided by square of height in meters) was 29.3 ± 7.7. Eighteen (10.9%) patients had a known preoperative clinical diagnosis of COPD. A total of 79 (47.9%) patients were never smokers, 64 (38.8%) were former smokers, and 22 (13.3%) were current smokers. A total of 45 (27.3%) patients had GI surgery, 113 (68.5%) had GU surgery, and seven (4.2%) had GYN surgery. Table 1 compares these characteristics between the case and control patients. The preoperative thoracic CT examinations in the 165 patients included 115 contrast-enhanced chest CT examinations, 47 unenhanced chest CT examinations, and three pulmonary CTA examinations.
TABLE 1: Characteristics of Case and Matched Control Patients
Matching VariableCase Patients (n = 42)Matched Control Patients (n = 123)
Age (y), mean ± SD66.1 ± 10.367.3 ± 9.5
Body mass indexa, mean ± SD29.1 ± 6.629.4 ± 8.1
Chronic obstructive pulmonary disease4 (9.5)14 (11.4)
Sex  
 Female18 (42.9)52 (42.3)
 Male24 (57.1)71 (57.7)
Smoking status  
 Never smoker20 (47.6)59 (48.0)
 Current smoker6 (14.3)16 (13.0)
 Former smoker16 (38.1)48 (39.0)
Type of surgery  
 Gastrointestinal12 (28.6)33 (26.8)
 Genitourinary28 (66.7)85 (69.1)
 Gynecologic2 (4.8)5 (4.1)

Note—Unless otherwise indicated, values represent number of patients with percentage in parentheses. Some percentages do not total 100 because of rounding.

a
Weight in kilograms divided by the square of height in meters.

Interreader Agreement

Table S2, which can be viewed in the AJR electronic supplement to this article at https://doi.org/10.2214/AJR.21.26411, provides interobserver agreement for qualitative measures. Based on the PABAK coefficients, agreement between the readers was almost perfect for all qualitative parameters (0.88–1.00). Table S3, which can be viewed in the AJR electronic supplement to this article at https://doi.org/10.2214/AJR.21.26411, provides interobserver agreement for quantitative measures. There was no agreement for the number of lobes with any pathology (ICC = –0.09), but there was substantial agreement for pulmonary artery diameter (ICC = 0.70), transverse tracheal diameter (ICC = 0.71), and AP tracheal diameter (ICC = 0.86). Agreement was almost perfect for the remaining measures (ICC = 0.91–0.98).

Comparison of Case and Control Patients

Table 2 compares the qualitative and quantitative parameters between the case and control patients for the two readers. Mild respiratory motion was present in 3–7% of cases across patient groups and readers and was not significantly different in frequency between groups for either reader (both, p > .05). No patient exhibited severe respiratory motion. Bronchial wall thickening (Fig. 3) was significantly more frequent in case patients than control patients for reader 2 (10% vs 2%, respectively; p = .03) but not for reader 1 (12% vs 4%, p = .13). Pericardial effusion (Fig. 4) was significantly more frequent in case than control patients for reader 2 (17% vs 5%, p = .01) but not for reader 1 (7% vs 0%, p = .99). The remaining qualitative features atelectasis, pleural effusion, pneumothorax, Kerley B lines, emphysema, bullous disease, bronchiectasis, pulmonary nodules, lymphadenopathy, cardiomegaly, and CT evidence of prior lung surgery, pulmonary fibrosis, or ILD) were not significantly different between case and control patients for either reader (all, p > .05).
TABLE 2: Comparison of Qualitative and Quantitative CT Parameters Between Case and Control Patients
CT ParameterReader 1Reader 2
Case (n = 42)Control (n = 123)paCase (n = 42)Control (n = 123)pa
Qualitative parameters      
 Respiratory motion  .72  .99
  None40 (95)119 (97) 39 (93)115 (93) 
  Mild2 (5)4 (3) 3 (7)8 (7) 
  Severe0 (0)0 (0) 0 (0)0 (0) 
 Right-sided atelectasis  .40  .30
  None15 (36)55 (45) 16 (38)60 (49) 
  Subsegmental27 (64)68 (55) 26 (62)63 (51) 
 Left-sided atelectasis  .72  .27
  None15 (36)50 (41) 12 (29)49 (40) 
  Subsegmental27 (64)73 (59) 30 (71)74 (60) 
 Right-sided pleural effusion  .21  .27
  None40 (95)120 (98) 40 (95)120 (98) 
  Small2 (5)3 (2) 2 (5)3 (2) 
 Left-sided pleural effusion  .81  .76
  None41 (98)118 (96) 41 (98)118 (96) 
  Small1 (2)5 (4) 1 (2)5 (4) 
 Pneumothorax  NA  NA
  No42 (100)123 (100) 42 (100)123 (100) 
  Yes0 (0)0 (0) 0 (0)0 (0) 
 Kerley B lines  .19  .46
  No36 (86)115 (93) 40 (95)120 (98) 
  Yes6 (14)8 (7) 2 (5)3 (2) 
 Emphysema  .85  .83
  None33 (79)99 (80) 33 (79)98 (80) 
  Mild3 (7)12 (10) 6 (14)16 (13) 
  Moderate6 (14)9 (7) 3 (7)7 (6) 
  Severe0 (0)3 (2) 0 (0)2 (2) 
 Bullous disease  .81  .11
  No38 (90)113 (92) 40 (95)122 (99) 
  Yes4 (10)10 (8) 2 (5)1 (1) 
 Bronchial wall thickening  .13  .03
  No37 (88)118 (96) 38 (90)121 (98) 
  Yes5 (12)5 (4) 4 (10)2 (2) 
 Bronchiectasis  .30  .18
  No38 (90)117 (95) 37 (88)116 (94) 
  Yes4 (10)6 (5) 5 (12)7 (6) 
 Pulmonary nodules  .73  .81
  No9 (21)30 (24) 8 (19)27 (22) 
  Yes33 (79)93 (76) 34 (81)96 (78) 
 Lymphadenopathy  .12  .21
  No36 (86)115 (93) 36 (86)114 (93) 
  Yes6 (14)8 (7) 6 (14)9 (7) 
 Cardiomegaly  .90  .49
  No36 (86)106 (86) 35 (83)110 (89) 
  Yes6 (14)17 (14) 7 (17)13 (11) 
 Pericardial effusion  .99  .01
  No39 (93)123 (100) 35 (83)117 (95) 
  Yes3 (7)0 (0) 7 (17)6 (5) 
 CT evidence of prior lung surgery  .99  .99
  No42 (100)122 (99) 42 (100)122 (99) 
  Yes0 (0)1 (1) 0 (0)1 (1) 
 CT evidence of pulmonary fibrosis  NA  NA
  No42 (100)123 (100) 42 (100)123 (100) 
  Yes0 (0)0 (0) 0 (0)0 (0) 
 CT evidence of interstitial lung disease  .24  NA
  No40 (95)122 (99) 42 (100)123 (100) 
  Yes2 (5)1 (1) 0 (0)0 (0) 
Quantitative parameters      
 No. of lung lobes involved, mean ± SD2.9 ± 1.92.7 ± 1.8.601.9 ± 1.51.7 ± 1.5.48
 Tracheal AP diameter (cm), mean ± SD2.3 ± 0.52.4 ± 0.5.672.2 ± 0.42.2 ± 0.5.70
 Tracheal transverse diameter (cm), mean ± SD1.9 ± 0.31.9 ± 0.3.441.8 ± 0.31.7 ± 0.3.38
 Pulmonary artery diameter (cm), mean ± SD2.9 ± 0.42.8 ± 0.5.162.9 ± 0.52.8 ± 0.5.045
 Right lung height (cm), mean ± SD18.4 ± 2.919.9 ± 2.7.0118.3 ± 2.919.8 ± 2.7.01
 Left lung height (cm), mean ± SD19.5 ± 3.121.1 ± 2.6.0119.6 ± 2.420.9 ± 2.6.01
 AP diameter of chest (cm), mean ± SD14.0 ± 2.312.9 ± 3.7.0214.2 ± 2.213.2 ± 3.6.04
 Transverse diameter of chest (cm), mean ± SD27.3 ± 2.926.9 ± 4.0.1726.3 ± 2.725.9 ± 4.0.16
 Subcutaneous fat thickness (cm), mean ± SD2.1 ± 0.92.2 ± 0.9.622.0 ± 0.72.1 ± 0.8.61

Note—Unless otherwise indicated, values indicate number of patients with percentage in parentheses. Some percentages do not total 100 because of rounding. NA = not applicable, AP = anteroposterior.

a
Listed in boldface when statistically significant at p < .05.
Fig. 3A —Preoperative chest CT images of two patients who underwent major surgery.
A, Preoperative unenhanced axial chest CT in 82-year-old man who underwent cystectomy for bladder cancer shows mild diffuse bronchial wall thickening (arrows). Patient required postoperative mechanical ventilation.
Fig. 3B —Preoperative chest CT images of two patients who underwent major surgery.
B, Preoperative unenhanced axial chest CT in 49-year-old man who underwent cystectomy shows no bronchial wall thickening (arrows). Patient did not require postoperative mechanical ventilation.
Fig. 4 —Unenhanced axial chest CT in 55-year-old man who underwent nephrectomy for renal cell carcinoma shows mild pericardial effusion (arrows). Patient required postoperative mechanical ventilation.
Pulmonary artery diameter was significantly greater in case than control patients for reader 2 (2.9 ± 0.5 cm vs 2.8 ± 0.5 cm, respectively; p = .045) but not for reader 1 (2.9 ± 0.4 cm vs 2.8 ± 0.5 cm, p = .16). Right lung height was significantly lower in case than control patients for reader 1 (18.4 ± 2.9 cm vs 19.9 ± 2.7 cm, p = .01) and reader 2 (18.3 ± 2.9 cm vs 19.8 ± 2.7 cm, p = .01). Left lung height was significantly lower in case than control patients for reader 1 (19.5 ± 3.1 cm vs 21.1 ± 2.6 cm, p = .01) and reader 2 (19.6 ± 2.4 cm vs 20.9 ± 2.6 cm, p = .01). AP chest diameter was significantly greater for case than control patients for reader 1 (14.0 ± 2.3 cm vs 12.9 ± 3.7 cm, p = .02) and reader 2 (14.2 ± 2.2 cm vs 13.2 ± 3.6 cm, p = .04). The remaining quantitative features (number of lung lobes involved, AP and transverse tracheal diameters, transverse chest diameter, subcutaneous fat thickness) were not significantly different between case and control patients for either reader (all, p > .05).
Based on the comparison of findings between case and control patients, bronchial wall thickening, pericardial effusion, right lung height, left lung height, AP chest diameter, and pulmonary artery diameter were selected for further performance analysis using GEEs. Table S4, which can be viewed in the AJR electronic supplement to this article at https://doi.org/10.2214/AJR.21.26411, shows the pairwise Pearson correlation coefficients among these selected study variables. Substantial correlation was observed only between right and left lung height (r = 0.86); the remaining pairwise combinations did not suggest collinearity (all, r < 0.01). To account for the observed collinearity, the mean of right and left lung height was used for further analyses.
Table 3 shows the results of the pooled assessment of diagnostic performance of the selected variables. Pulmonary artery diameter, mean lung height, and AP chest diameter exhibited AUCs of 0.57, 0.66, and 0.55, respectively. Pulmonary artery diameter, mean lung height, and AP chest diameter were subsequently assessed using threshold values of 2.4 cm, 22.7 cm, and 11.5 cm, respectively. For predicting postoperative mechanical ventilation, bronchial wall thickening achieved a sensitivity of 11% and specificity of 97%, pericardial effusion achieved a sensitivity of 12% and specificity of 98%, pulmonary artery diameter at the given threshold achieved a sensitivity of 94% and specificity of 12%, mean lung height at the given threshold achieved a sensitivity of 90% and specificity of 19%, and AP chest diameter at the selected threshold achieved a sensitivity of 92% and specificity of 24%.
TABLE 3: Diagnostic Performance of Selected CT Parameters
MeasureAUCThreshold (cm)Sensitivity (%)Specificity (%)
Bronchial wall thickeningNANA11 (5–21)97 (93–99)
Pericardial effusionNANA12 (6–23)98 (95–99)
Pulmonary artery diameter0.57 (0.50–0.64)2.494 (85–98)12 (8–18)
Mean lung height0.66 (0.59–0.72)22.790 (77–96)19 (13–26)
AP diameter of chest0.55 (0.48–0.62)11.592 (81–97)24 (17–32)

Note—Values in parentheses indicate 95% CIs. NA = not applicable, AP = anteroposterior.

Table 4 shows results of the univariable GEE analysis and multivariable regression model. All five measures were significant predictors in the univariable analysis that adjusted for patient height (all, p < .05) and were also significant independent predictors in the multivariable regression model. In the multivariable regression model for predicting postoperative mechanical ventilation, bronchial wall thickening exhibited an OR of 4.6 (95% CI, 1.3–16.5; p = .02), pericardial effusion exhibited an OR of 5.1 (95% CI, 1.7–15.5; p = .004), pulmonary artery diameter exhibited an OR of 1.4 per 1-cm increase (95% CI, 0.7–3.0; p = .32), mean lung height exhibited an OR of 0.8 per 1-cm increase (95% CI, 0.7–1.001; p = .05), and AP chest diameter exhibited an OR of 1.2 per 1-cm increase (95% CI, 1.013–1.4; p = .03).
TABLE 4: Univariable and Multivariable Regression Analysis of Significant Features, Adjusted for Patient Height
CT ParameterUnivariableMultivariable
Odds RatiopOdds Ratiop
Bronchial wall thickening3.9 (1.2–12.8).024.6 (1.3–16.5).02
Pericardial effusion5.8 (1.9–17.9).0025.1 (1.7–15.5).004
Pulmonary artery diameter (per 1-cm increase)1.8 (0.9–3.7).091.4 (0.7–3.0).32
Lung height (mean of right and left lungs) (per 1-cm increase)0.8 (0.7–0.96).010.8 (0.7–1.001).05
AP diameter of chest (per 1-cm increase)1.2 (1.04–1.4).011.2 (1.013–1.4).03

Note—Values in parentheses indicate 95% CIs. AP = anteroposterior.

Figure 5 shows the relationship between patient height and lung height, separately for left and right sides for each reader, for case and control patients. A positive association was observed between lung height and patient height for both sides for both patient groups for both readers (slope ranging from x to y). However, the incremental change in lung height to patient height, as indicated by the slope, was not significantly different between case and control patients for any comparison (all, p > .05) (Table S5, which can be viewed in the AJR electronic supplement to this article at https://doi.org/10.2214/AJR.21.26411).
Fig. 5 —Scatterplots show relationship between patient height and lung height for case and control patients separately for each lung and each reader. Dashed and solid lines indicate relationship between patient height and lung height based on mixed-effects model.

Discussion

In this exploratory single-center matched case-control study, patients who unexpectedly required postoperative mechanical ventilation after major abdominal or pelvic surgery exhibited more frequent presence of bronchial wall thickening and pericardial ef-fusion, shorter lung height (measured from apex to the middle of the diaphragm), and increased AP chest diameter. These thoracic CT findings were positively associated with a need for postoperative mechanical ventilation even after matching for preoperative clinical COPD, BMI, age, sex, and type of surgery. These findings have potential clinical relevance for preoperatively anticipating which patients undergoing abdominal or pelvic surgery are at risk of requiring postoperative mechanical ventilation.
Although there was an association between lung height and patient height, shorter lung height was significantly associated with a need for postoperative mechanical ventilation after adjusting for body height. Despite the significantly shorter lung height in the case group, the patients in the case group also had significantly larger AP chest diameters. Increased AP chest diameter coupled with increased lung height (i.e., barrel chest) on imaging (both radiography and CT) has been reported to predict clinical COPD with high sensitivity and specificity, because this combination indicates pulmonary hyperinflation [25, 26]. Hightower et al. [25] reported that increased lung height (≥ 19.5 cm) measured on sagittal CT is significantly associated with obstructive airway disease and is an accurate marker of COPD with a sensitivity of 67%, specificity of 78%, and accuracy of 74% and that increased AP chest diameter (> 21 cm) was significantly associated with obstructive airway disease, but with less accuracy (64%). The presence of clinically diagnosed COPD independently predicts postoperative mechanical ventilation, as do greater age and female sex, after cardiac surgery [4, 16].
Although increased lung height on CT is suggestive of COPD, case and control groups in this study were matched on the basis of preoperative clinical diagnosis of COPD, age, sex, BMI, and smoking status. This approach allowed identification of unique CT parameters, other than these recognized risk factors, to potentially aid in the prediction of which patients will need postoperative mechanical ventilation. Shorter lung heights are seen in patients with restrictive lung disease. Robbie et al. [27] found that decreased CT-derived lung heights measured on sagittal images are associated with lung volume loss in patients with idiopathic pulmonary fibrosis. They also reported that CT-derived lung height has the strongest relationship with percentage force vital capacity. The mechanism of the association between shorter lung heights and/or larger chest AP diameter with the need for postoperative mechanical ventilation remains unknown from our study. We speculate that altered thoracic geometry may reduce the efficiency of diaphragmatic contraction, which in turn may require increased respiratory effort.
Bronchial wall thickening was the only qualitative parameter involving the lung parenchyma that differed between the case and control groups for both readers, with significantly increased need for postoperative mechanical ventilation in patients with bronchial wall thickening. Bronchial wall thickness can result from many entities and is associated with COPD [28, 29]. In one study, bronchial wall thickening exhibited an independent association with COPD and had 90.6% specificity and 42.3% sensitivity for its early detection [29]. Because the groups were matched on the basis of a clinical diagnosis of COPD, the greater frequency of bronchial wall thickening in the case group may not relate to COPD. The association may indicate a relation between small airways disease and difficulty in weaning from mechanical ventilation.
Our study has limitations. First, this was a retrospective study. Spirometry monitoring was not available to assess for differences in respiratory efforts during preoperative thoracic CT acquisition. Despite the institution's standard inspiratory breath-hold acquisition protocol, variation in inspiratory effort and the degree of inspiration between individuals is inevitable and could affect lung height. However, respiratory spirometry guidance for CT is not available in routine clinical practice. Also, preoperative pulmonary function test results were not available in the EMR, because most patients undergoing abdominal and pelvic surgery do not routinely undergo preoperative pulmonary function tests. This limitation indicates the potential utility of using preoperative thoracic CT findings for assessing risk of the need for postoperative mechanical ventilation, especially because these patients commonly undergo preoperative CT. Although the severity of emphysema on CT was not significantly different between the two groups, other quantitative variables reflecting the severity of COPD, such as the Global Initiative for Obstructive Lung Disease criteria [30], were not included in the analysis. In addition, we did not assess the presence of additional comorbidities or causes of pericardial effusion, such as congestive heart failure, renal failure, and immunosuppression. Finally, tracheal shape and dimensions were not evaluated at levels other than the thoracic inlet.
In conclusion, this case-control study indicates that the presence of bronchial wall thickening, presence of pericardial effusion, smaller lung height, and increased AP chest diameter on preoperative thoracic CT are significantly associated with the need for postoperative mechanical ventilation after elective major abdominal or pelvic surgery. Because many patients undergo thoracic CT before abdominal or pelvic surgery, these findings may complement traditional clinical risk factors when assessing such patients' postoperative risk.

Supplemental Content

File (02_21_26411_suppl.pdf)

References

1.
Smetana GW, Lawrence VA, Cornell JE. Preoperative pulmonary risk stratification for noncardiothoracic surgery: systematic review for the American College of Physicians. Ann Intern Med 2006; 144:581–595
2.
Yang HX, Ling L, Zhang X, Lin P, Rong TH, Fu JH. Outcome of elderly patients with oesophageal squamous cell carcinoma after surgery. Br J Surg 2010; 97:862–867
3.
Suarez-Pierre A, Fraser CD, Zhou X, et al. Predictors of operative mortality among cardiac surgery patients with prolonged ventilation. J Card Surg 2019; 34:759–766
4.
Légaré JF, Hirsch GM, Buth KJ, MacDougall C, Sullivan JA. Preoperative prediction of prolonged mechanical ventilation following coronary artery bypass grafting. Eur J Cardiothorac Surg 2001; 20:930–936
5.
Prapas SN, Panagiotopoulos IA, Hamed Abdelsalam A, et al. Predictors of prolonged mechanical ventilation following aorta no-touch off-pump coronary artery bypass surgery. Eur J Cardiothorac Surg 2007; 32:488–492
6.
Kollef MH, Wragge T, Pasque C. Determinants of mortality and multiorgan dysfunction in cardiac surgery patients requiring prolonged mechanical ventilation. Chest 1995; 107:1395–1401
7.
Kern H, Redlich U, Hotz H, et al. Risk factors for prolonged ventilation after cardiac surgery using APACHE II, SAPS II, and TISS: comparison of three different models. Intensive Care Med 2001; 27:407–415
8.
Tabib A, Abrishami SE, Mahdavi M, Mortezaeian H, Totonchi Z. Predictors of prolonged mechanical ventilation in pediatric patients after cardiac surgery for congenital heart disease. Res Cardiovasc Med 2016; 5:e30391
9.
Rajakaruna C, Rogers CA, Angelini GD, Ascione R. Risk factors for and economic implications of prolonged ventilation after cardiac surgery. J Thorac Cardiovasc Surg 2005; 130:1270–1277
10.
Branca P, McGaw P, Light R. Factors associated with prolonged mechanical ventilation following coronary artery bypass surgery. Chest 2001; 119:537–546
11.
Arozullah AM, Daley J, Henderson WG, Khuri SF. Multifactorial risk index for predicting postoperative respiratory failure in men after major noncardiac surgery. The National Veterans Administration Surgical Quality Improvement Program. Ann Surg 2000; 232:242–253
12.
Yuan H, Tuttle-Newhall JE, Chawa V, et al. Prognostic impact of mechanical ventilation after liver transplantation: a national database study. Am J Surg 2014; 208:582–590
13.
Yang CK, Teng A, Lee DY, Rose K. Pulmonary complications after major abdominal surgery: National Surgical Quality Improvement Program analysis. J Surg Res 2015; 198:441–449
14.
Fernandez-Bustamante A, Frendl G, Sprung J, et al. Postoperative pulmonary complications, early mortality, and hospital stay following noncardiothoracic surgery: a multicenter study by the perioperative research network investigators. JAMA Surg 2017; 152:157–166
15.
Spivack SD, Shinozaki T, Albertini JJ, Deane R. Preoperative prediction of postoperative respiratory outcome: coronary artery bypass grafting. Chest 1996; 109:1222–1230
16.
Cislaghi F, Condemi AM, Corona A. Predictors of prolonged mechanical ventilation in a cohort of 5123 cardiac surgical patients. Eur J Anaesthesiol 2009; 26:396–403
17.
Butler J, Chong GL, Pillai R, Westaby S, Rocker GM. Early extubation after coronary artery bypass surgery: effects on oxygen flux and haemodynamic variables. J Cardiovasc Surg (Torino) 1992; 33:276–280
18.
Gall SA Jr, Olsen CO, Reves JG, et al. Beneficial effects of endotracheal extubation on ventricular performance: implications for early extubation after cardiac operations. J Thorac Cardiovasc Surg 1988; 95:819–827
19.
Gass GD, Olsen GN. Preoperative pulmonary function testing to predict postoperative morbidity and mortality. Chest 1986; 89:127–135
20.
Theriault MM, Eddy K, Borgaonkar JN, Babar JL, Manos D. Diseases involving the central bronchi: multidetector CT for detection, characterization, and differential diagnosis. RadioGraphics 2018; 38:58–59
21.
Hansell DM, Bankier AA, MacMahon H, McLoud TC, Müller NL, Remy J. Fleischner Society: glossary of terms for thoracic imaging. Radiology 2008; 246:697–722
22.
Allen AS, Satten GA. Control for confounding in case-control studies using the stratification score, a retrospective balancing score. Am J Epidemiol 2011; 173:752–760
23.
Hansen BB, Klopfer SO. Optimal full matching and related designs via network flows. J Comput Graph Stat 2006; 15:609–627
24.
Dupont WD. Power calculations for matched case-control studies. Biometrics 1988; 44:1157–1168
25.
Hightower JS, Amadi C, Den E, Schmitt JE, Shah RM, Miller WT Jr. Back to the future: sagittal CT in the evaluation of COPD. Eur Radiol 2016; 26:2730–2739
26.
Pierce JA, Ebert RV. The barrel deformity of the chest, the senile lung and obstructive pulmonary emphysema. Am J Med 1958; 25:13–22
27.
Robbie H, Wells AU, Jacob J, et al. Visual and automated CT measurements of lung volume loss in idiopathic pulmonary fibrosis. AJR 2019; 213:318–324
28.
Webb WR. Radiology of obstructive pulmonary disease. AJR 1997; 169:637–647
29.
Mets OM, Schmidt M, Buckens CF, et al. Diagnosis of chronic obstructive pulmonary disease in lung cancer screening computed tomography scans: independent contribution of emphysema, air trapping and bronchial wall thickening. Respir Res 2013; 14:59
30.
Rabe KF, Hurd S, Anzueto A, et al.; Global Initiative for Chronic Obstructive Lung Disease. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: GOLD executive summary. Am J Respir Crit Care Med 2007; 176:532–555

Information & Authors

Information

Published In

American Journal of Roentgenology
Pages: 279 - 288
PubMed: 34467781

History

Submitted: June 12, 2021
Revision requested: June 30, 2021
Revision received: July 30, 2021
Accepted: August 23, 2021
Version of record online: September 1, 2021

Keywords

  1. CT
  2. postoperative mechanical ventilation
  3. surgery

Authors

Affiliations

Arzu Canan, MD
Department of Radiology, UT Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX 75390
Asha Kandathil, MD
Department of Radiology, UT Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX 75390
Jing-Lei Li, MD
Department of Radiology, Guangdong Provincial People's Hospital, Guangdong Academy of Medical Sciences, Guangzhou, China
Yin Xi, PhD
Department of Radiology, UT Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX 75390
Lauren Wehrmann, MS
Department of Anesthesiology and Pain Management, UT Southwestern Medical Center, Dallas, TX
Trenton Bryson, MD
Department of Anesthesiology and Pain Management, UT Southwestern Medical Center, Dallas, TX
Travis Browning, MD
Department of Radiology, UT Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX 75390
Suhny Abbara, MD
Department of Radiology, UT Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX 75390
Amanda A. Fox, MD, MPH
Department of Anesthesiology and Pain Management, UT Southwestern Medical Center, Dallas, TX
McDermott Center for Human Growth and Development, UT Southwestern Medical Center, Dallas, TX

Notes

Address correspondence to A. Canan ([email protected]).
The authors declare that they have no disclosures relevant to the subject matter of this article.

Metrics & Citations

Metrics

Citations

Export Citations

To download the citation to this article, select your reference manager software.

Articles citing this article

Media

Figures

Other

Tables

Share

Share

Copy the content Link

Share on social media