ORIGINAL ARTICLE
Does chest tube location matter? An analysis of chest tube position and the need for secondary interventions Matthew V. Benns, MD, Michael E. Egger, MD, Brian G. Harbrecht, MD, Glen A. Franklin, MD, Jason W. Smith, MD, PhD, Keith R. Miller, MD, Nicholas A. Nash, MD, and J. David Richardson, MD, Louisville, Kentucky
Tube thoracostomy is a common procedure used in the management of thoracic trauma. Traditional teaching suggests that chest tubes should be directed in specific locations to improve function. Common examples include anterior and superior placement for pneumothorax, inferior and posterior placement for hemothorax, and avoidance of the pulmonary fissure. The purpose of this study was to examine the effect of specific chest tube position on subsequent chest tube function. METHODS: A retrospective review of all patients undergoing tube thoracostomy for trauma from January 1, 2010, to September 30, 2012, was performed. Only patients undergoing computed tomography scans following chest tube insertion were included so that positioning could be accurately determined. Rib space insertion level and positioning of the tube relative to the lung parenchyma were recorded. The duration of chest tube drainage and the need for secondary interventions were determined and compared for tubes in different rib spaces and locations. For purposes of comparison, tubes placed above the sixth rib space were considered ‘‘high,’’ and those at or below it were considered ‘‘low.’’ RESULTS: A total of 291 patients met criteria for inclusion. Forty-eight patients (16.5%) required secondary intervention. Neither high chest tube placement nor chest tube location relative to lung parenchyma was associated with an increased need for secondary interventions. On multivariate analysis, only chest Abbreviated Injury Scale (AIS) scores, mechanism, and volume of hemothorax were found to be significant risk factors for the need for secondary interventions. CONCLUSION: Chest tube location does not influence the need for secondary interventions as long as the tube resides in the pleural space. The severity of chest injury is the most important factor influencing outcome in patients undergoing tube thoracostomy for trauma. Tube thoracostomy technique should focus on safe insertion within the pleural space and not on achieving a specific tube location. (J Trauma Acute Care Surg. 2015;78: 386Y390. Copyright * 2015 Wolters Kluwer Health, lnc. All rights reserved.) LEVEL OF EVIDENCE: Therapeutic study, level IV. KEY WORDS: Chest tubes; thoracic trauma; tube thoracostomy. BACKGROUND:
T
ube thoracostomy is a common procedure used in the management of thoracic trauma. Current teaching in tube thoracostomy technique emphasizes placement within the ‘‘triangle of safety,’’ an area bordered inferiorly by the nipple line, posteriorly by the lateral border of the latissimus dorsi, and anteriorly by the lateral border of the pectoralis major.1 Because there are no major neurovascular structures and a paucity of muscle tissue in this area, safe and efficient access to the pleural space can be readily obtained. Adherence to these guidelines should allow placement of most chest tubes at the fifth intercostal space or above. These principles are often balanced with traditional teaching by thoracic surgeons for tube thoracostomy that emphasizes ‘‘directed’’ placement of chest tubes with three primary tenets as follows: superior and anterior placement for pneumothorax, inferior and posterior placement
for hemothorax, and the avoidance of fissure placement.2Y4 These latter principles of ideal chest tube placement have persisted despite limited evidence to the contrary.5Y7 In our study, we attempted to provide a more comprehensive evaluation of the effects of chest tube positioning using the multidimensional visualization afforded by current computed tomography (CT) scans. The purpose of our study was to examine the effect of specific chest tube position on subsequent chest tube function, specifically rib interspace placement and location relative to lung parenchyma. We also evaluated the effect of location of chest tube placement on the development of complications and patient outcome. Because the thorax should be a closed system, we hypothesized that chest tube position should not matter provided that the tube was in the pleural space.
PATIENTS AND METHODS Submitted: May 18, 2014, Revised: August 19, 2014, Accepted: August 21, 2014. From the Hiram C. Polk, Jr., M.D. Department of Surgery, University of Louisville, Louisville, Kentucky. This study was presented as a poster at the 72nd annual meeting of the American Association for the Surgery of Trauma, September 18Y21, 2013, in San Francisco, California. Address for reprints: Matthew Benns, MD, Hiram C. Polk, Jr., M.D. Department of Surgery, University of Louisville, 550 South Jackson St, Louisville, KY 40202; email: [email protected]. DOI: 10.1097/TA.0000000000000479
With institutional review board approval, charts of all patients undergoing tube thoracostomy for trauma from January 1, 2010, to September 30, 2012, were reviewed. All patients included in the study had chest tubes placed in the emergency department by the trauma team as part of their initial evaluation and management. Only patients who had a CT scan following chest tube insertion were included so that positioning could be accurately determined. With the use of the scan images, J Trauma Acute Care Surg Volume 78, Number 2
386
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
J Trauma Acute Care Surg Volume 78, Number 2
Benns et al.
TABLE 1. Demographics and Outcome Variables Age, y ISS Chest AIS score Chest tube duration, d ICU LOS, d Total LOS, d Ventilator days
Mean
SD
Median
IQR
40.3 20.2 3 6 9.3 16.2 6.4
17.9 11.1 0.8 4 11.1 14.3 8.9
39 18 3 5 4 12 2
25Y53 11Y27 3Y3 3Y8 1Y14 6Y22 0Y11
IQR, interquartile range; LOS, length of stay.
rib interspaces were counted, and the level of chest tube insertion was recorded. Position of the tube within the thorax relative to the lung parenchyma was determined based on the terminal end of the chest tube. The hemithorax was divided into 60-degree ranges in the transverse plane with tubes recorded as anterior, lateral, or posterior based on this scheme. Tubes terminating in the pulmonary fissure or within the lung parenchyma itself were labeled accordingly. A minority of chest tubes terminated below the pulmonary parenchyma and were thus labeled as ‘‘inferior.’’ The duration of chest tube drainage and the need for secondary interventions were determined and compared for tubes placed in different rib spaces and locations. Placement in the right or left hemithorax was also included. Tubes placed above the sixth rib space were compared with those tubes placed below it (‘‘high’’ vs. ‘‘low’’ placement). Criteria for chest tube removal was determined by the clinical judgment of the treating physicians and based on local clinical practice guidelines that included the absence of an air leak, radiographic resolution of pneumothorax or hemothorax, and a recorded chest tube output of less than 200 mL in a 24-hour period. Patients who died or had an early (G24 hours from admission) thoracotomy were excluded. Other variables included in the analysis were patient age, Injury Severity Score (ISS), ventilator days, trauma mechanism, chest Abbreviated Injury Scale (AIS) score, chest tube size, and initial chest tube output. Length of stay and ventilator days were also included in the predictive models on a univariate basis and were used in the multivariate models if significant on univariate analyses. These factors are included as potential risk factors to adjust the models for any confounding factors associated with duration of chest tube. Length of stay may capture the degree of complexity of patient care not included in injury severity adjustments such as social disposition issues and the coordination of care among multiple specialties. This level of complexity may
influence chest tube duration and need for secondary intervention. The number ventilator days is often a confounding factor to chest tube removal, as chest tubes may be left in for longer times because of concerns related to positivepressure ventilation or repeat trips to the operating room. The presence of coagulopathy as defined by preexisting warfarin use, antiplatelet agent use, or cirrhosis was also included in the analysis. Continuous and categorical variables were compared using Wilcoxon rank-sum or Kruskal-Wallis test and W2 or Fisher’s exact test, as appropriate. Logistic regression modeling was used to identify risk factors associated with binary outcomes; Wald confidence limits are reported. Risk factors identified on univariate analysis with a p G 0.05 were included in the multivariate models. Linear regression modeling was used to identify risk factors associated with duration of chest tube drainage. Risk factors identified on univariate analysis with a p G 0.05 were included in the multivariate models. All tests were two sided, and a p G 0.05 was considered statistically significant. Analyses were performed using SAS version 9.3 (SAS, Cary, NC).
RESULTS A total of 8,186 trauma patients were screened during the study period, with 862 patients undergoing tube thoracostomy. There were 571 patients who had chest tubes placed who died, underwent emergency thoracotomy, or did not have a CT scan of the chest performed and were excluded from further analysis. There were 291 patients who had tube thoracostomy as part of their initial management with subsequent CT scans of the chest who were able to be analyzed. Of these patients, 236 (81.1%) were male. A total of 196 patients (67.3%) experienced blunt trauma, while 95 patients (32.7%) had a penetrating mechanism of injury. Table 1 shows the demographics and main outcome variables of patients included in the study. Tables 2 and 3 illustrate the anatomic distribution of chest tubes. The sixth intercostal space was the most common craniocaudal location for chest tube placement (36.4%). The posterior position was the most common location (38.8%) relative to the lung parenchyma. Three chest tubes were placed in an extrathoracic position, 1 intra-abdominally and 2 within the soft tissue of the axilla. Thirty-two French was the most common sized chest tube used (52.9%), followed by 36 French (35.7%), 28 French (6.5%), and 40 French (3.8%).
TABLE 2. Anatomic Distribution of Chest Tubes by Rib Space
TABLE 3. Anatomic Distribution of Chest Tubes by Location Relative to Lung Parenchyma
Rib Space
Location
n (%)
3 4 5 6 7 8 9
2 (0.6) 29 (10.0) 75 (25.8) 106 (36.4) 53 (18.2) 26 (8.6) 1 (0.3)
Anterior Fissure Inferior Lateral Posterior Intraparenchymal Extrathoracic
* 2015 Wolters Kluwer Health, lnc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
n (%) 42 (14.4) 98 (33.7) 4 (1.4) 21 (7.2) 113 (38.8) 10 (3.4) 3 (1.0)
387
J Trauma Acute Care Surg Volume 78, Number 2
Benns et al.
TABLE 4. Risk Factors for the Need for Secondary Intervention Factor High placement AIS score Age ICU duration ISS LOS Ventilator days Kinked tube Fissure tube Male sex Right side Penetrating injury Ideal placement Chest tube size Initial output, mL Anticoagulation Antiplatelet therapy Any coagulopathy
Univariate OR (95% CI)
Univariate p
1.77 (0.95Y3.31) 1.16 (1.05Y1.29) 1.01 (0.99Y1.03) 1.01 (0.98Y1.03) 1.01 (0.98Y1.04) 1.01 (0.99Y1.03) 1.00 (0.96Y1.03) 2.71 (0.88Y8.32) 0.98 (0.51Y1.89) 1.01 (0.46Y2.24) 0.69 (0.37Y1.30) 0.42 (0.20Y0.91) 0.45 (0.17Y1.20) 1.01 (0.91Y1.13) 1.001 (1.000Y1.003) 5.24 (0.72Y38.15) 2.60 (0.46Y14.60) 2.67 (0.77Y9.25)
0.073 0.006 0.34 0.67 0.53 0.64 0.82 0.082 0.96 0.98 0.25 0.028 0.11 0.84 0.016 0.10 0.28 0.12
Multivariate OR (95% CI)
Multivariate p
1.13 (1.01Y1.26)
0.03
0.36 (0.16Y0.81)
0.014
1.002 (1.000Y1.003)
0.010
p G 0.05 included in multivariate model. LOS, length of stay; OR, odds ratio.
Forty-eight patients (16.5%) required a secondary intervention. The most common secondary intervention was an additional tube thoracostomy (58.5%), followed by videoassisted thoracoscopic surgery (15.4%), thoracotomy (13.8%), and percutaneous drainage by interventional radiology (12.3%). Among the 10 patients undergoing video-assisted thoracoscopic surgery, 7 were performed for retained hemothorax and 3 for decortication of empyema. Nine patients underwent open thoracotomy, seven for decortication of empyema, one for persistent air leak that required a lung resection, and one for retained hemothorax. Table 4 summarizes the results of our analysis of the need for secondary interventions. On univariate analysis, significant variables associated with the need for secondary intervention included chest AIS score, chest tube duration, penetrating mechanism, initial chest tube output, and increased length of hospital stay. ‘‘High’’ chest tube placement (third to fifth interspace) was not a significant risk factor for the need for secondary interventions. When excluding extrathoracic placement, chest tube location relative to the lung parenchyma was also not significant. On multivariate analysis, chest AIS score, penetrating mechanism, and initial chest tube output were found to be significant risk factors for the need for secondary interventions. The multivariate model c statistic (area under the curve) was 0.69 (95% confidence interval [CI], 0.61Y0.77). Hosmer-Lemeshow goodness-of-fit test was appropriate (p = 0.22). Diagnostics for the performance of the logistic model were appropriate. Table 5 summarizes the results of our analysis of chest tube duration. Univariate analysis of chest tube duration demonstrated an increasing duration associated with increased chest AIS score, total ISS, penetrating mechanism, initial chest tube output, intensive care unit (ICU) length of stay, total length of stay, and ventilator days. All of these variables remained 388
significant on multivariate analysis. ‘‘High’’ chest tube placement or chest tube location relative to the lung parenchyma was not significant with regard to chest tube duration. The overall multivariate model was statistically significant (p G 0.0001), with an adjusted r2 of 0.31. Linear model assumptions were evaluated. Standard evaluation of residuals did not reveal any evidence for nonnormal distribution of error or heteroscedasticity in any of the model parameters. Lack-offit testing was appropriate (p = 0.23). Polynomial modeling of factors did not improve model performance; thus, simple linear associations of factors were used.
DISCUSSION Chest tube insertion is the most common invasive procedure used in the management of thoracic trauma and is the only intervention required in 85% of cases of chest trauma that require an intervention.8 While safe placement is stressed, persistent teachings exist regarding the ideal placement of a chest tube and the ability of tubes to adequately function relative to their position in the pleural space. The results of this study suggest that outcomes and the need for secondary thoracic interventions associated with tube thoracostomy are not related to tube position, provided the tube resides within the pleural space. This concept has been previously suggested by Hegarty,6 who noted that blood rarely clotted within the pleural space and was therefore amenable to drainage from any location. Our study, however, is the first to use the multidimensional capabilities of CT imaging to determine and examine specific chest tube locations. With respect to tube location relative to pulmonary parenchyma, we saw no difference in complications or tube duration related to anterior or posterior placement, fissure * 2015 Wolters Kluwer Health, lnc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
J Trauma Acute Care Surg Volume 78, Number 2
Benns et al.
TABLE 5. Risk Factors for Chest Tube Duration Factor High placement AIS score Age ICU duration ISS LOS Ventilator days Kinked tube Fissure tube Male sex Right side Penetrating injury Ideal placement Chest tube size Initial output, mL Anticoagulation Antiplatelet therapy Any coagulopathy
Univariate Effect (95% CI)
Univariate p
j0.06 (j1.02 to 0.91) 0.29 (0.14 to 0.45) 0.01 (j0.02 to 0.04) 0.16 (0.12 to 0.20) 0.13 (0.10 to 0.17) 0.13 (0.11 to 0.16) 0.21 (0.17 to 0.26) j0.66 (j2.76 to 1.43) 0.17 (j0.81 to 1.15) j0.50 (j1.69 to 0.68) 0.49 (j0.44 to 1.42) j1.99 (j2.96 to j1.03) 0.48 (j0.70 to 1.66) j0.04 (j0.21 to 0.12) 0.002 (G0.001 to 0.004) 0.23 (j3.76 to 4.21) 0.31 (j2.95 to 3.58) j0.03 (j2.36 to 2.30)
0.91 0.0003 0.41 G0.0001 G0.0001 G0.0001 G0.0001 0.53 0.73 0.40 0.30 G0.0001 0.42 0.60 0.049 0.91 0.85 0.98
Multivariate Effect (95% CI)
Multivariate p
0.15 (j0.01 to 0.29)
0.052
j0.25 (j0.39 to j0.11) 0.02 (j0.02 to 0.07) 0.14 (0.07 to 0.21) 0.29 (0.17 to 0.42)
0.0004 0.30 G0.0001 G0.0001
j1.05 (j1.97 to j0.13)
0.026
0.003 (0.001 to 0.004)
0.006
p G 0.05 included in multivariate model. LOS, length of stay.
placement, or even intraparenchymal placement. These results provide some evidence against traditional teaching regarding the benefits of ‘‘directed’’ tube placement within the thorax. The effectiveness of fissure tubes has been previously studied with similar results.7 The performance of the intraparenchymal tubes was surprising and seems counterintuitive. However, in all cases, only the terminal end of the tube was intraparenchymal, with a portion residing within the pleural space. Landay et al.9 previously used CT imaging to identify intraparenchymal placement of chest tubes. They concluded that intraparenchymal positioning was likely more common than previously thought. In their series, 26% had prolonged air leaks or recurrent pneumothorax, but nearly half of patients had no apparent complications. While we would certainly recommend avoidance of intraparenchymal chest tubes, we could detect no difference in the need for secondary interventions with intraparenchymal tubes compared with other locations. Intercostal level of insertion was not associated with an increased need for secondary interventions, with the exception of very low placement. There was one chest tube placed within the ninth intercostal space that terminated in an intraabdominal location and required secondary intervention. This complication could clearly have been avoided by adherence to the ‘‘triangle of safety’’ method of chest tube insertion. Because there is no apparent benefit to chest tube insertion below these landmarks, these guidelines seem prudent in the training of safe chest tube insertion. The only significant variables associated with complications and the need for secondary intervention were chest AIS score, penetrating mechanism, and initial volume of hemothorax. Similarly, the duration of chest tube drainage was only significantly affected by the severity of chest injury and other variables directly associated with overall injury burden. A recent study by Menger et al.10 also demonstrated
increased chest AIS score to be the only significant factor in the development of chest tube complications, although chest tube location was not considered. There are several limitations to this study. First, as a retrospective review, we can only demonstrate associations and not causality. Second, the design of our study introduces an inherent selection bias. All of the chest tubes placed in this study were placed before CT imaging. The need for chest tube placement was therefore determined by physical examination findings or portable chest x-ray findings. This effectively excludes patients with smaller pneumothoraces, particularly in the anterior location. Because this is a group that would benefit from a directionally placed chest tube by traditional teaching, its exclusion is a weakness of this study. It is also important to note that this study only includes patients receiving tube thoracostomy for initial management of thoracic injuries. Tube thoracostomy is also commonly used in the treatment of patients well beyond their initial presentation. Many of these patients may develop intrapleural adhesions or loculations as a result of localized injury and inflammation or previous interventions. It would stand to reason that specific chest tube location may be of greater importance in these patients, but this was not studied in the current analysis. Finally, although CT imaging was used to document the initial chest tube location, we did not routinely use follow-up CT imaging to document resolution of pneumothorax or hemothorax. Our group practice pattern is to aggressively look for undrained collections that might require intervention,11 but follow-up CT imaging was only obtained if sufficient suspicion was generated by clinical examination or chest x-ray findings. We therefore cannot exclude greater residual hemothorax or pneumothorax associated with a particular tube location. However, as there was no difference in the need for secondary interventions, this was not a clinically relevant outcome.
* 2015 Wolters Kluwer Health, lnc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
389
J Trauma Acute Care Surg Volume 78, Number 2
Benns et al.
Our study provides evidence that chest tube location within the thorax is not associated with the need for secondary interventions. The severity and mechanism of chest injury is the most important factor influencing outcome in patients undergoing tube thoracostomy for chest trauma. Inserting a tube in any location and applying suction should provide adequate drainage of hemothorax and allow for reexpansion of pneumothorax in patients without significant pleural adhesions. Practitioners performing tube thoracostomy should focus their efforts on safe placement within the pleural space and avoid excessive manipulations to achieve a specific tube location. AUTHORSHIP M.V.B. and J.D.R. provided the study design. M.V.B. and M.E.E. performed the data collection. M.V.B., M.E.E., and K.R.M. performed the data analysis. M.V.B. and M.E.E. drafted the manuscript. M.V.B., M.E.E., B.G.H., N.A.N., J.W.S., G.A.F., K.R.M., and J.D.R. provided critical revisions.
DISCLOSURE The authors declare no conflicts of interest.
REFERENCES 1. Laws D, Neville E, Duffy J; Pleural Diseases Group, Standards of Care Committee, British Thoracic Society. BTS guidelines for the insertion of a chest drain. Thorax. 2003;58(Suppl 2):53Y59.
390
2. Batchelder TL, Morris KA. Critical factors in determining adequate pleural drainage in both the operated and nonoperated chest. Am Surg. 1962; 28:296Y302. 3. Drummond DS. Traumatic hemothorax: complications and management. Am Surg. 1967;33(5):403Y408. 4. Miller KS, Sahn SA. Chest tubes. indications, technique, management and complications. Chest. 1987;91(2):258Y264. 5. Duponselle EF. The level of the intercostal drain and other determinant factors in the conservative approach to penetrating chest injuries. Cent Afr J Med. 1980;26:52Y55. 6. Hegarty MM. A conservative approach to penetrating injuries of the chest. Experience with 131 successive cases. Injury. 1976;8(1): 53Y59. 7. Curtin JJ, Goodman LR, Quebbeman EJ, Haasler GB. Thoracostomy tubes after acute chest injury: relationship between location in a pleural fissure and function. AJR Am J Roentgenol. 1994;163(6):1339Y1342. 8. Luchette FA, Barrie PS, Oswanski MF, et al. Practice management guidelines for prophylactic antibiotic use in tube thoracostomy for traumatic hemopneumothorax: the EAST Practice Management Guidelines Work Group. Eastern Association For Trauma. J Trauma. 2000;48(4): 753Y757. 9. Landay M, Oliver Q, Estrera A, Friese R, Boonswang N, DiMaio JM. Lung penetration by thoracostomy tubes: imaging findings on CT. J Thorac Imaging. 2006;21(3):197Y204. 10. Menger R, Telford G, Kim P, et al. Complications following thoracic trauma managed with tube thoracostomy. Injury. 2012; 43(1):46Y50. 11. Smith JW, Franklin GA, Harbrecht BG, Richardson JD. Early VATS for blunt chest trauma: a management technique underutilized by acute care surgeons. J Trauma. 2011;71(1):102Y105.
* 2015 Wolters Kluwer Health, lnc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Does chest tube location matter? An analysis of chest tube position and the need for secondary interventions Matthew V. Benns, MD, Michael E. Egger, MD, Brian G. Harbrecht, MD, Glen A. Franklin, MD, Jason W. Smith, MD, PhD, Keith R. Miller, MD, Nicholas A. Nash, MD, and J. David Richardson, MD, Louisville, Kentucky
Tube thoracostomy is a common procedure used in the management of thoracic trauma. Traditional teaching suggests that chest tubes should be directed in specific locations to improve function. Common examples include anterior and superior placement for pneumothorax, inferior and posterior placement for hemothorax, and avoidance of the pulmonary fissure. The purpose of this study was to examine the effect of specific chest tube position on subsequent chest tube function. METHODS: A retrospective review of all patients undergoing tube thoracostomy for trauma from January 1, 2010, to September 30, 2012, was performed. Only patients undergoing computed tomography scans following chest tube insertion were included so that positioning could be accurately determined. Rib space insertion level and positioning of the tube relative to the lung parenchyma were recorded. The duration of chest tube drainage and the need for secondary interventions were determined and compared for tubes in different rib spaces and locations. For purposes of comparison, tubes placed above the sixth rib space were considered ‘‘high,’’ and those at or below it were considered ‘‘low.’’ RESULTS: A total of 291 patients met criteria for inclusion. Forty-eight patients (16.5%) required secondary intervention. Neither high chest tube placement nor chest tube location relative to lung parenchyma was associated with an increased need for secondary interventions. On multivariate analysis, only chest Abbreviated Injury Scale (AIS) scores, mechanism, and volume of hemothorax were found to be significant risk factors for the need for secondary interventions. CONCLUSION: Chest tube location does not influence the need for secondary interventions as long as the tube resides in the pleural space. The severity of chest injury is the most important factor influencing outcome in patients undergoing tube thoracostomy for trauma. Tube thoracostomy technique should focus on safe insertion within the pleural space and not on achieving a specific tube location. (J Trauma Acute Care Surg. 2015;78: 386Y390. Copyright * 2015 Wolters Kluwer Health, lnc. All rights reserved.) LEVEL OF EVIDENCE: Therapeutic study, level IV. KEY WORDS: Chest tubes; thoracic trauma; tube thoracostomy. BACKGROUND:
T
ube thoracostomy is a common procedure used in the management of thoracic trauma. Current teaching in tube thoracostomy technique emphasizes placement within the ‘‘triangle of safety,’’ an area bordered inferiorly by the nipple line, posteriorly by the lateral border of the latissimus dorsi, and anteriorly by the lateral border of the pectoralis major.1 Because there are no major neurovascular structures and a paucity of muscle tissue in this area, safe and efficient access to the pleural space can be readily obtained. Adherence to these guidelines should allow placement of most chest tubes at the fifth intercostal space or above. These principles are often balanced with traditional teaching by thoracic surgeons for tube thoracostomy that emphasizes ‘‘directed’’ placement of chest tubes with three primary tenets as follows: superior and anterior placement for pneumothorax, inferior and posterior placement
for hemothorax, and the avoidance of fissure placement.2Y4 These latter principles of ideal chest tube placement have persisted despite limited evidence to the contrary.5Y7 In our study, we attempted to provide a more comprehensive evaluation of the effects of chest tube positioning using the multidimensional visualization afforded by current computed tomography (CT) scans. The purpose of our study was to examine the effect of specific chest tube position on subsequent chest tube function, specifically rib interspace placement and location relative to lung parenchyma. We also evaluated the effect of location of chest tube placement on the development of complications and patient outcome. Because the thorax should be a closed system, we hypothesized that chest tube position should not matter provided that the tube was in the pleural space.
PATIENTS AND METHODS Submitted: May 18, 2014, Revised: August 19, 2014, Accepted: August 21, 2014. From the Hiram C. Polk, Jr., M.D. Department of Surgery, University of Louisville, Louisville, Kentucky. This study was presented as a poster at the 72nd annual meeting of the American Association for the Surgery of Trauma, September 18Y21, 2013, in San Francisco, California. Address for reprints: Matthew Benns, MD, Hiram C. Polk, Jr., M.D. Department of Surgery, University of Louisville, 550 South Jackson St, Louisville, KY 40202; email: [email protected]. DOI: 10.1097/TA.0000000000000479
With institutional review board approval, charts of all patients undergoing tube thoracostomy for trauma from January 1, 2010, to September 30, 2012, were reviewed. All patients included in the study had chest tubes placed in the emergency department by the trauma team as part of their initial evaluation and management. Only patients who had a CT scan following chest tube insertion were included so that positioning could be accurately determined. With the use of the scan images, J Trauma Acute Care Surg Volume 78, Number 2
386
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
J Trauma Acute Care Surg Volume 78, Number 2
Benns et al.
TABLE 1. Demographics and Outcome Variables Age, y ISS Chest AIS score Chest tube duration, d ICU LOS, d Total LOS, d Ventilator days
Mean
SD
Median
IQR
40.3 20.2 3 6 9.3 16.2 6.4
17.9 11.1 0.8 4 11.1 14.3 8.9
39 18 3 5 4 12 2
25Y53 11Y27 3Y3 3Y8 1Y14 6Y22 0Y11
IQR, interquartile range; LOS, length of stay.
rib interspaces were counted, and the level of chest tube insertion was recorded. Position of the tube within the thorax relative to the lung parenchyma was determined based on the terminal end of the chest tube. The hemithorax was divided into 60-degree ranges in the transverse plane with tubes recorded as anterior, lateral, or posterior based on this scheme. Tubes terminating in the pulmonary fissure or within the lung parenchyma itself were labeled accordingly. A minority of chest tubes terminated below the pulmonary parenchyma and were thus labeled as ‘‘inferior.’’ The duration of chest tube drainage and the need for secondary interventions were determined and compared for tubes placed in different rib spaces and locations. Placement in the right or left hemithorax was also included. Tubes placed above the sixth rib space were compared with those tubes placed below it (‘‘high’’ vs. ‘‘low’’ placement). Criteria for chest tube removal was determined by the clinical judgment of the treating physicians and based on local clinical practice guidelines that included the absence of an air leak, radiographic resolution of pneumothorax or hemothorax, and a recorded chest tube output of less than 200 mL in a 24-hour period. Patients who died or had an early (G24 hours from admission) thoracotomy were excluded. Other variables included in the analysis were patient age, Injury Severity Score (ISS), ventilator days, trauma mechanism, chest Abbreviated Injury Scale (AIS) score, chest tube size, and initial chest tube output. Length of stay and ventilator days were also included in the predictive models on a univariate basis and were used in the multivariate models if significant on univariate analyses. These factors are included as potential risk factors to adjust the models for any confounding factors associated with duration of chest tube. Length of stay may capture the degree of complexity of patient care not included in injury severity adjustments such as social disposition issues and the coordination of care among multiple specialties. This level of complexity may
influence chest tube duration and need for secondary intervention. The number ventilator days is often a confounding factor to chest tube removal, as chest tubes may be left in for longer times because of concerns related to positivepressure ventilation or repeat trips to the operating room. The presence of coagulopathy as defined by preexisting warfarin use, antiplatelet agent use, or cirrhosis was also included in the analysis. Continuous and categorical variables were compared using Wilcoxon rank-sum or Kruskal-Wallis test and W2 or Fisher’s exact test, as appropriate. Logistic regression modeling was used to identify risk factors associated with binary outcomes; Wald confidence limits are reported. Risk factors identified on univariate analysis with a p G 0.05 were included in the multivariate models. Linear regression modeling was used to identify risk factors associated with duration of chest tube drainage. Risk factors identified on univariate analysis with a p G 0.05 were included in the multivariate models. All tests were two sided, and a p G 0.05 was considered statistically significant. Analyses were performed using SAS version 9.3 (SAS, Cary, NC).
RESULTS A total of 8,186 trauma patients were screened during the study period, with 862 patients undergoing tube thoracostomy. There were 571 patients who had chest tubes placed who died, underwent emergency thoracotomy, or did not have a CT scan of the chest performed and were excluded from further analysis. There were 291 patients who had tube thoracostomy as part of their initial management with subsequent CT scans of the chest who were able to be analyzed. Of these patients, 236 (81.1%) were male. A total of 196 patients (67.3%) experienced blunt trauma, while 95 patients (32.7%) had a penetrating mechanism of injury. Table 1 shows the demographics and main outcome variables of patients included in the study. Tables 2 and 3 illustrate the anatomic distribution of chest tubes. The sixth intercostal space was the most common craniocaudal location for chest tube placement (36.4%). The posterior position was the most common location (38.8%) relative to the lung parenchyma. Three chest tubes were placed in an extrathoracic position, 1 intra-abdominally and 2 within the soft tissue of the axilla. Thirty-two French was the most common sized chest tube used (52.9%), followed by 36 French (35.7%), 28 French (6.5%), and 40 French (3.8%).
TABLE 2. Anatomic Distribution of Chest Tubes by Rib Space
TABLE 3. Anatomic Distribution of Chest Tubes by Location Relative to Lung Parenchyma
Rib Space
Location
n (%)
3 4 5 6 7 8 9
2 (0.6) 29 (10.0) 75 (25.8) 106 (36.4) 53 (18.2) 26 (8.6) 1 (0.3)
Anterior Fissure Inferior Lateral Posterior Intraparenchymal Extrathoracic
* 2015 Wolters Kluwer Health, lnc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
n (%) 42 (14.4) 98 (33.7) 4 (1.4) 21 (7.2) 113 (38.8) 10 (3.4) 3 (1.0)
387
J Trauma Acute Care Surg Volume 78, Number 2
Benns et al.
TABLE 4. Risk Factors for the Need for Secondary Intervention Factor High placement AIS score Age ICU duration ISS LOS Ventilator days Kinked tube Fissure tube Male sex Right side Penetrating injury Ideal placement Chest tube size Initial output, mL Anticoagulation Antiplatelet therapy Any coagulopathy
Univariate OR (95% CI)
Univariate p
1.77 (0.95Y3.31) 1.16 (1.05Y1.29) 1.01 (0.99Y1.03) 1.01 (0.98Y1.03) 1.01 (0.98Y1.04) 1.01 (0.99Y1.03) 1.00 (0.96Y1.03) 2.71 (0.88Y8.32) 0.98 (0.51Y1.89) 1.01 (0.46Y2.24) 0.69 (0.37Y1.30) 0.42 (0.20Y0.91) 0.45 (0.17Y1.20) 1.01 (0.91Y1.13) 1.001 (1.000Y1.003) 5.24 (0.72Y38.15) 2.60 (0.46Y14.60) 2.67 (0.77Y9.25)
0.073 0.006 0.34 0.67 0.53 0.64 0.82 0.082 0.96 0.98 0.25 0.028 0.11 0.84 0.016 0.10 0.28 0.12
Multivariate OR (95% CI)
Multivariate p
1.13 (1.01Y1.26)
0.03
0.36 (0.16Y0.81)
0.014
1.002 (1.000Y1.003)
0.010
p G 0.05 included in multivariate model. LOS, length of stay; OR, odds ratio.
Forty-eight patients (16.5%) required a secondary intervention. The most common secondary intervention was an additional tube thoracostomy (58.5%), followed by videoassisted thoracoscopic surgery (15.4%), thoracotomy (13.8%), and percutaneous drainage by interventional radiology (12.3%). Among the 10 patients undergoing video-assisted thoracoscopic surgery, 7 were performed for retained hemothorax and 3 for decortication of empyema. Nine patients underwent open thoracotomy, seven for decortication of empyema, one for persistent air leak that required a lung resection, and one for retained hemothorax. Table 4 summarizes the results of our analysis of the need for secondary interventions. On univariate analysis, significant variables associated with the need for secondary intervention included chest AIS score, chest tube duration, penetrating mechanism, initial chest tube output, and increased length of hospital stay. ‘‘High’’ chest tube placement (third to fifth interspace) was not a significant risk factor for the need for secondary interventions. When excluding extrathoracic placement, chest tube location relative to the lung parenchyma was also not significant. On multivariate analysis, chest AIS score, penetrating mechanism, and initial chest tube output were found to be significant risk factors for the need for secondary interventions. The multivariate model c statistic (area under the curve) was 0.69 (95% confidence interval [CI], 0.61Y0.77). Hosmer-Lemeshow goodness-of-fit test was appropriate (p = 0.22). Diagnostics for the performance of the logistic model were appropriate. Table 5 summarizes the results of our analysis of chest tube duration. Univariate analysis of chest tube duration demonstrated an increasing duration associated with increased chest AIS score, total ISS, penetrating mechanism, initial chest tube output, intensive care unit (ICU) length of stay, total length of stay, and ventilator days. All of these variables remained 388
significant on multivariate analysis. ‘‘High’’ chest tube placement or chest tube location relative to the lung parenchyma was not significant with regard to chest tube duration. The overall multivariate model was statistically significant (p G 0.0001), with an adjusted r2 of 0.31. Linear model assumptions were evaluated. Standard evaluation of residuals did not reveal any evidence for nonnormal distribution of error or heteroscedasticity in any of the model parameters. Lack-offit testing was appropriate (p = 0.23). Polynomial modeling of factors did not improve model performance; thus, simple linear associations of factors were used.
DISCUSSION Chest tube insertion is the most common invasive procedure used in the management of thoracic trauma and is the only intervention required in 85% of cases of chest trauma that require an intervention.8 While safe placement is stressed, persistent teachings exist regarding the ideal placement of a chest tube and the ability of tubes to adequately function relative to their position in the pleural space. The results of this study suggest that outcomes and the need for secondary thoracic interventions associated with tube thoracostomy are not related to tube position, provided the tube resides within the pleural space. This concept has been previously suggested by Hegarty,6 who noted that blood rarely clotted within the pleural space and was therefore amenable to drainage from any location. Our study, however, is the first to use the multidimensional capabilities of CT imaging to determine and examine specific chest tube locations. With respect to tube location relative to pulmonary parenchyma, we saw no difference in complications or tube duration related to anterior or posterior placement, fissure * 2015 Wolters Kluwer Health, lnc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
J Trauma Acute Care Surg Volume 78, Number 2
Benns et al.
TABLE 5. Risk Factors for Chest Tube Duration Factor High placement AIS score Age ICU duration ISS LOS Ventilator days Kinked tube Fissure tube Male sex Right side Penetrating injury Ideal placement Chest tube size Initial output, mL Anticoagulation Antiplatelet therapy Any coagulopathy
Univariate Effect (95% CI)
Univariate p
j0.06 (j1.02 to 0.91) 0.29 (0.14 to 0.45) 0.01 (j0.02 to 0.04) 0.16 (0.12 to 0.20) 0.13 (0.10 to 0.17) 0.13 (0.11 to 0.16) 0.21 (0.17 to 0.26) j0.66 (j2.76 to 1.43) 0.17 (j0.81 to 1.15) j0.50 (j1.69 to 0.68) 0.49 (j0.44 to 1.42) j1.99 (j2.96 to j1.03) 0.48 (j0.70 to 1.66) j0.04 (j0.21 to 0.12) 0.002 (G0.001 to 0.004) 0.23 (j3.76 to 4.21) 0.31 (j2.95 to 3.58) j0.03 (j2.36 to 2.30)
0.91 0.0003 0.41 G0.0001 G0.0001 G0.0001 G0.0001 0.53 0.73 0.40 0.30 G0.0001 0.42 0.60 0.049 0.91 0.85 0.98
Multivariate Effect (95% CI)
Multivariate p
0.15 (j0.01 to 0.29)
0.052
j0.25 (j0.39 to j0.11) 0.02 (j0.02 to 0.07) 0.14 (0.07 to 0.21) 0.29 (0.17 to 0.42)
0.0004 0.30 G0.0001 G0.0001
j1.05 (j1.97 to j0.13)
0.026
0.003 (0.001 to 0.004)
0.006
p G 0.05 included in multivariate model. LOS, length of stay.
placement, or even intraparenchymal placement. These results provide some evidence against traditional teaching regarding the benefits of ‘‘directed’’ tube placement within the thorax. The effectiveness of fissure tubes has been previously studied with similar results.7 The performance of the intraparenchymal tubes was surprising and seems counterintuitive. However, in all cases, only the terminal end of the tube was intraparenchymal, with a portion residing within the pleural space. Landay et al.9 previously used CT imaging to identify intraparenchymal placement of chest tubes. They concluded that intraparenchymal positioning was likely more common than previously thought. In their series, 26% had prolonged air leaks or recurrent pneumothorax, but nearly half of patients had no apparent complications. While we would certainly recommend avoidance of intraparenchymal chest tubes, we could detect no difference in the need for secondary interventions with intraparenchymal tubes compared with other locations. Intercostal level of insertion was not associated with an increased need for secondary interventions, with the exception of very low placement. There was one chest tube placed within the ninth intercostal space that terminated in an intraabdominal location and required secondary intervention. This complication could clearly have been avoided by adherence to the ‘‘triangle of safety’’ method of chest tube insertion. Because there is no apparent benefit to chest tube insertion below these landmarks, these guidelines seem prudent in the training of safe chest tube insertion. The only significant variables associated with complications and the need for secondary intervention were chest AIS score, penetrating mechanism, and initial volume of hemothorax. Similarly, the duration of chest tube drainage was only significantly affected by the severity of chest injury and other variables directly associated with overall injury burden. A recent study by Menger et al.10 also demonstrated
increased chest AIS score to be the only significant factor in the development of chest tube complications, although chest tube location was not considered. There are several limitations to this study. First, as a retrospective review, we can only demonstrate associations and not causality. Second, the design of our study introduces an inherent selection bias. All of the chest tubes placed in this study were placed before CT imaging. The need for chest tube placement was therefore determined by physical examination findings or portable chest x-ray findings. This effectively excludes patients with smaller pneumothoraces, particularly in the anterior location. Because this is a group that would benefit from a directionally placed chest tube by traditional teaching, its exclusion is a weakness of this study. It is also important to note that this study only includes patients receiving tube thoracostomy for initial management of thoracic injuries. Tube thoracostomy is also commonly used in the treatment of patients well beyond their initial presentation. Many of these patients may develop intrapleural adhesions or loculations as a result of localized injury and inflammation or previous interventions. It would stand to reason that specific chest tube location may be of greater importance in these patients, but this was not studied in the current analysis. Finally, although CT imaging was used to document the initial chest tube location, we did not routinely use follow-up CT imaging to document resolution of pneumothorax or hemothorax. Our group practice pattern is to aggressively look for undrained collections that might require intervention,11 but follow-up CT imaging was only obtained if sufficient suspicion was generated by clinical examination or chest x-ray findings. We therefore cannot exclude greater residual hemothorax or pneumothorax associated with a particular tube location. However, as there was no difference in the need for secondary interventions, this was not a clinically relevant outcome.
* 2015 Wolters Kluwer Health, lnc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
389
J Trauma Acute Care Surg Volume 78, Number 2
Benns et al.
Our study provides evidence that chest tube location within the thorax is not associated with the need for secondary interventions. The severity and mechanism of chest injury is the most important factor influencing outcome in patients undergoing tube thoracostomy for chest trauma. Inserting a tube in any location and applying suction should provide adequate drainage of hemothorax and allow for reexpansion of pneumothorax in patients without significant pleural adhesions. Practitioners performing tube thoracostomy should focus their efforts on safe placement within the pleural space and avoid excessive manipulations to achieve a specific tube location. AUTHORSHIP M.V.B. and J.D.R. provided the study design. M.V.B. and M.E.E. performed the data collection. M.V.B., M.E.E., and K.R.M. performed the data analysis. M.V.B. and M.E.E. drafted the manuscript. M.V.B., M.E.E., B.G.H., N.A.N., J.W.S., G.A.F., K.R.M., and J.D.R. provided critical revisions.
DISCLOSURE The authors declare no conflicts of interest.
REFERENCES 1. Laws D, Neville E, Duffy J; Pleural Diseases Group, Standards of Care Committee, British Thoracic Society. BTS guidelines for the insertion of a chest drain. Thorax. 2003;58(Suppl 2):53Y59.
390
2. Batchelder TL, Morris KA. Critical factors in determining adequate pleural drainage in both the operated and nonoperated chest. Am Surg. 1962; 28:296Y302. 3. Drummond DS. Traumatic hemothorax: complications and management. Am Surg. 1967;33(5):403Y408. 4. Miller KS, Sahn SA. Chest tubes. indications, technique, management and complications. Chest. 1987;91(2):258Y264. 5. Duponselle EF. The level of the intercostal drain and other determinant factors in the conservative approach to penetrating chest injuries. Cent Afr J Med. 1980;26:52Y55. 6. Hegarty MM. A conservative approach to penetrating injuries of the chest. Experience with 131 successive cases. Injury. 1976;8(1): 53Y59. 7. Curtin JJ, Goodman LR, Quebbeman EJ, Haasler GB. Thoracostomy tubes after acute chest injury: relationship between location in a pleural fissure and function. AJR Am J Roentgenol. 1994;163(6):1339Y1342. 8. Luchette FA, Barrie PS, Oswanski MF, et al. Practice management guidelines for prophylactic antibiotic use in tube thoracostomy for traumatic hemopneumothorax: the EAST Practice Management Guidelines Work Group. Eastern Association For Trauma. J Trauma. 2000;48(4): 753Y757. 9. Landay M, Oliver Q, Estrera A, Friese R, Boonswang N, DiMaio JM. Lung penetration by thoracostomy tubes: imaging findings on CT. J Thorac Imaging. 2006;21(3):197Y204. 10. Menger R, Telford G, Kim P, et al. Complications following thoracic trauma managed with tube thoracostomy. Injury. 2012; 43(1):46Y50. 11. Smith JW, Franklin GA, Harbrecht BG, Richardson JD. Early VATS for blunt chest trauma: a management technique underutilized by acute care surgeons. J Trauma. 2011;71(1):102Y105.
* 2015 Wolters Kluwer Health, lnc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
