Trauma Pneumonectomy: Has Survival Improved Over Two Decades?

Abstract

Introduction

The hypothesis of this study is that recent advances in mechanical cardiopulmonary support and operative management have improved survival in patients requiring a trauma pneumonectomy.

Methods

Retrospective, single center study from January 2003 to December 2023 of all patients who underwent a pneumonectomy for trauma. Data collected included demographics, admission physiology, use of venovenous extracorporeal membrane oxygenation (VV-ECMO), and mortality, defined as early (within 24 hours of surgery) and late (>24 hours after surgery). Outcomes were compared between decades, the first decade (2003-2010) and second decade (2011-2023).

Results

Twenty patients met inclusion criteria, 9 in the first decade and 11 in the second decade. There were no differences in Injury Severity Score (ISS) (26.4 vs 34.3, P = 0.23). However, those in the second decade had significantly lower mean admission pH (6.89 vs 7.14, P = 0.01) and higher admission base deficit (19.3 vs 9.8, P = 0.003). The use of thoracic damage control surgery increased from 33% in the first decade to 100% in the second decade (P = 0.002). VV-ECMO with lung rest ventilation increased from 22% to 64% (P = 0.06). Overall and early mortality did not change (55.6% vs 45.5%, p-0.65 and 11% vs 36.3%, P = 0.09, respectively). However, late mortality was dramatically lower in the second decade compared to the first (9% vs 50%, P = 0.06).

Conclusion

Early mortality remains high; however, the combination of thoracic damage control and early initiation of VV-ECMO may contribute to the dramatic decrease in late mortality in the second decade.

Keywords

trauma thoracic surgery pneumonectomy

Introduction

Traumatic chest injuries necessitating pneumonectomy occur in <1% of patients undergoing thoracic exploration.^1,2 Despite the fact that this procedure is rarely performed, it has a prodigious mortality, ranging from 47% to 100%, which has remained unchanged over decades.^1,3,4 Intraoperative mortality results from exsanguination and acute right ventricular failure. Right ventricular failure is also the predominate cause of death in the immediate postoperative period.^1,4-9 Acute right ventricular failure results from the precipitous increase in pulmonary vascular resistance immediately after pneumonectomy in a patient in hemorrhagic shock.¹⁰

Thoracic damage control (TDC) is a proven operative approach in the physiologically depleted patient with severe chest trauma, including traumatic pneumonectomy.^11-13 In addition, veno-venous extracorporeal membrane oxygenation (VV-ECMO) allows for the rapid correction of hypoxemia and hypercarbia, which both result in increased pulmonary vascular resistance. Lung rest ventilation, with minimal driving pressures, lowers intrathoracic pressure and is protective for the failing right ventricle.^9,13,14

The hypothesis of this study is that with recent advances in operative management and cardiopulmonary support, the mortality rates of trauma pneumonectomy have improved over time.

Methods

Study Population

This was a retrospective single-center study at a Level 1 trauma center of patients >18 years of age who underwent a pneumonectomy for trauma between January 1, 2003 and December 31, 2023. Patients were selected from a trauma registry if they had an operative intervention described as a pneumonectomy at the initial operation.

Variables and Outcomes

Data collected included demographics, mechanism of injury, injury severity score (ISS), admission physiology, and hemodynamics. Intraoperative data included operative details such as thoracic damage control, estimated blood loss, intraoperative hemodynamics, and time to chest closure. Thoracic damage control was defined as delayed closure of the chest following the initial operation. ECMO specific variables included time to cannulation, duration of ECMO, VV vs veno-arterial (VA) ECMO, and pre-ECMO physiology. Post-operative data included complications such as the development of acute respiratory distress syndrome (ARDS), hospital length of stay (LOS), intensive care unit (ICU) LOS, number of ventilator days, and use of continuous renal replacement therapy (CRRT). Outcomes included mortality, including cause and timing of death. Mortality was divided into early mortality (within 24 hours of surgery) and late mortality (>24 hours after surgery).

Statistical Analysis

Outcomes were compared between the first decade (2003-2010) and the second decade (2011-2023) using t-tests, ANOVA, and chi-squared tests. Data points that were not available were excluded in the analysis. Data were reported as mean with standard deviation (SD) or median with interquartile range (IQR) as appropriate. This study was approved by our institutional review board and waiver of consent was obtained.

Results

Study Population and Demographics

Twenty patients were included in the analysis, with 9 patients in the first decade and 11 patients in the second decade. Three patients in the first decade had minimal data available except for operative intervention and outcomes.

Comparing patients in the first decade and second decade, there were no significant differences in age (32.2 vs 25.8, P = 0.16), sex (male 88.9% male vs 90.9%, P = 0.88), ISS (26.4 vs 34.3, P = 0.23), or admission Glasgow Coma Score (GCS) (10.2 vs 9.0, P = 0.34). Overall, 15 (75%) sustained a penetrating injury (4 stab wounds, 11 gunshot wounds), 2 (10%) sustained a blunt injury from motor vehicle collisions, and 3 (15%) had incomplete data regarding mechanism of injury.

Preoperative characteristics

Comparing patients in the second decade to patients in the first decade, patients in the second decade had a lower admission systolic blood pressure (88.6 mmHg vs 114.4 mmHg, P = 0.06), lower admission pH (6.89 vs 7.14, P = 0.01), and higher admission base deficit (19.3 vs 9.8, P = 0.003). Fewer patients were intubated in the field (5 patients vs 1 patient, P = 0.001). In addition, 36% (4/11) of patients in the second decade received Tranexamic Acid (TXA). Due to incomplete data in the first decade, the use of TXA could not be compared between decades Table 1.

Table 1.

Preoperative Characteristics of 20 Patients Who Underwent a Pneumonectomy for Trauma From 2003-2023. P-Value Represents the Comparison Between the Early Decade (2003-2010) and the Late Decade (2011-2023).

	Overall cohort (n = 20)	2003-2010 (n = 9)	2011-2023 (n = 11)	P-value
Age (years), mean (SD)	28.7 (9.1)	32.2 (11.7)	25.8 (5.1)	0.16
Male sex, n (%)	18 (90.0)	8 (88.9)	10 (90.9)	0.88
Race, n (%)				0.02
White	4 (20.0)	4 (44.4)	0 (0.0)
Black	15 (75)	4 (44.4)	11 (100.0)
Other	1 (5.0)	1 (11.1)	0 (0.0)
Mechanism of injury, n (%)				0.10^a
Stab		3 (33.3)	1 (9.1)
GSW		2 (22.2)	9 (81.8)
Motor vehicle or cycle		1 (11.1)	1 (9.1)
Collision		3 (33.3)	0 (0.0)
Unknown
Admission GCS, mean (SD)	9.4 (5.1)	10.2 (5.9)	9.0 (4.9)	0.34
Injury Severity Score, mean (SD)	31.7 (10.6)	26.4 (11.7)	34.3 (9.5)	0.23
Lowest admission systolic blood pressure (mmHg), mean (SD)	99.3 (27.4)	114.4 (25.8)	88.6 (24.6)	0.06
Admission pH, mean (SD)	6.99 (0.21)	7.14 (0.13)	6.89 (0.20)	0.01
Admission lactate, mean (SD)	11.3 (5.1)	8.8 (5.0)	12.5 (4.9)	0.22
Admission base deficit, mean (SD)	15.7 (7.4)	9.8 (4.0)	19.3 (6.6)	0.003
Crystalloid (liters), mean (SD)	1.0 (1.6)	3.5 (0.71)	0.2 (0.49)	0.03
Preoperative packed red blood cells, mean (SD)	7.8 (6.4)	12.7 (4.8)	3.7 (4.3)	0.005
Preoperative fresh frozen plasma, mean (SD)	5.7 (5.0)	9.4 (4.7)	2.7 (2.7)	0.014
Preoperative platelets, mean (SD)	2.0 (2.8)	4.2 (2.9)	0.2 (0.41)	0.02
Tranexamic acid, n (%)	Incomplete data	Incomplete data	4 (36.3)
Intubated in the field, n (%)	6 (30.0)	5 (55.6)	1 (9.1)	0.001

^aP-value comparing penetrating vs blunt mechanism between early vs late decades.

Operative Management

Of the patients with intraoperative data, 5/12 (41.7%) underwent an anterolateral thoracotomy, 6/12 (50.0%) a clamshell thoracotomy, and 1/12 (8.3%) a posterolateral thoracotomy. Patients with an anterolateral or clamshell thoracotomy appeared to be more physiologically deranged than the patient with a posterolateral thoracotomy, with a lower SBP (92mmHg vs 110mmHg), lower pH (7.12 vs 6.93), higher lactate (11.6 vs 9.5), and higher base deficit (18 vs 7.8). Of the patients with the available data, 5/11 (45.5%) patients had the entire hilum stapled en masse and 6/11 (54.5%) patients had the hilar structures individually ligated. Patients with the hilum stapled en masse also appeared to be more, although not statically significant, physiologically deranged than patients with the hilum individually ligated, with a lower SBP (84mmHg vs 112mmHg, P = 0.19), lower pH (6.87 vs 6.97, P = 0.47), higher lactate (13.8 vs 10.1, P = 0.21), and higher base deficit (20.4 vs 15.5, P = 0.27).

There was a higher rate of TDC in the second decade compared to the first (100% vs 22%, P = 0.002). Of the patients who survived to discharge, 66.7% (6/9) underwent TDC at the initial operation, 0% in the first decade, and 100% in the second decade. Of the patients who survived the first 24 hours after surgery, 69.2% (9/13) underwent TDC at the initial operation (75.0% survived to discharge), 33.3% in the first decade (0% survived to discharge), and 100% in the second decade (89% survived to discharge). For patients who survived the first 24 hours, in univariate analysis, TDC was not significantly associated with mortality (P = 0.72). Average time to chest closure was 3.7 (SD 3.8) days. Of the patient with intraoperative data available and who survived to chest closure, 6/8 (75.0%) had the bronchial stump buttressed with a muscle or pericardial flap.

There was a higher rate of VV-ECMO cannulation in the second decade compared to the first (64% vs 22%, P = 0.06). Of the patients who survived to discharge, 50% (5/10) were cannulated for VV-ECMO, 25% in the first decade and 66.7% in the second decade. Overall, there was no significant difference in mortality for patients on ECMO and patients not on ECMO (50% vs 55%, P = 0.65) Of the patients who survived the first 24 hours after surgery, survival rate for patients canulated for VV-ECMO was 71.4% (5/7) (50% in the first decade and 80% in the second decade). For patients who survived the first 24 hours, in univariate analysis, VV-ECMO was not significantly associated with mortality (P = 0.26). Overall, average time to cannulation was 1.6 days (±2.2). Average number of days on VV-ECMO was 24.8 days (±36.9). One patient was converted from VV to VA-ECMO due to persistent hemodynamic instability and significant cardiac biventricular dysfunction.

Outcomes

Most common postoperative complications were sepsis (30%), cardiac arrest (25%), bronchopleural fistula (20%), and acute kidney injury requiring renal replacement therapy (40%). Average ICU length of stay was 37.4 days (±45.4) (45.1 days in the first decade vs 34.0 days in the second decade, P = 0.29), average ventilator days were 35.2 days (±45.1) (56.2 days in the first decade vs 29.2 days in the second decade, P = 0.17), and average hospital length of stay was 46.5 days (±55.6) (60.0 days in the first decade vs 40.2 days in the second decade, P = 0.22). Of the patients who survived to discharge in the second decade, 1 (16.7%) was ventilator dependent and none were on dialysis at the time of discharge.

The overall mortality was 10/20 (50%), with no statistical difference between the first decade (55.6%) and the second decade (45.5%), P = 0.65. Early mortality was 25% (5/20), with no significant difference between the first decade (11%) and the second decade (36.3%), P = 0.09. All patients who suffered early mortality died from hemorrhagic shock, with one patient who had evidence of both hemorrhagic shock and right heart failure. Late mortality was 25% (5/20), with a dramatic reduction between the first decade (50%) and the second decade (9%), P = 0.06 Table 2.

Table 2.

Intraoperative and Post-operative Characteristics of 20 Patients Who Underwent a Pneumonectomy for Trauma From 2003-2023. P-Value Represents the Comparison Between the Early Decade (2003-2010) and the Late Decade (2011-2023).

	Overall cohort (n = 20)	2003-2010 (n = 9)	2011-2023 (n = 11)	P-value
Damage control thoracotomy, n (%)	13 (65.0)	2 (22.2)	11 (100.0)	0.001
Veno-venous extracorporeal membrane oxygenation, n (%)	8 (40.0)	2 (22.0)	7 (63.6)	0.06
Veno-arterial extracorporeal membrane oxygenation, n (%)	1 (5.0)	0 (0.0)	1 (9.1)	0.35
Days on ECMO, mean (SD)	24.8 (36.9)	54 (50.9)	18.3 (33.5)	0.50
Bronchial stump leak, n (%)	4 (20.0)	2 (22.2)	2 (18.2)	0.82
Hospital length of stay (days), mean (SD)	46.5 (55.6)	60.0 (53.5)	40.2 (57.8)	0.22
Intensive care unit length of stay (days), mean (SD)	37.4 (45.4)	45.1 (38.6)	34.0 (50.6)	0.29
Ventilator days, mean (SD)	35.2 (45.1)	56.2 (51.4)	29.2 (41.3)	0.17
Mortality, n (%)
Overall	10 (50.0)	5 (55.6)	5 (45.5)	0.65
Early (<24 hours after surgery)	5 (25.0)	1 (11.0)	4 (36.3)	0.19
Late (>24 hours after surgery)	5 (25.0)	4 (44.4)	1 (9.0)	0.06

Eliminating the patients with missing data, on univariate regression analysis, the preoperative and intraoperative variables that were independently associated with mortality were injury severity score (P = 0.01), admission pH (P = 0.007), and admission lactate (P = 0.01).

Discussion

Although infrequently performed, trauma pneumonectomy is a formidable procedure, with a staggering mortality between 50% and 100%.^1,3,4,9 Exsanguination and acute right heart failure are the most common causes of mortality.^1,4-7,9 Two reports from the National Trauma Data Bank (NTDB), over a decade apart, both reported a 60% mortality.^4,15 Aside from these large studies from the NTDB, most reports consist of smaller series of trauma pneumonectomies, underscoring their rarity.^1-3,5-7,9

Patients undergoing a trauma pneumonectomy almost always present in shock, requiring rapid evaluation and operative intervention.^2-4,13 Expeditious control of the hilum is essential to preventing fatal intraoperative hemorrhage. A posterolateral thoracotomy gives wide exposure to the entire thorax. In unstable patients, an anterolateral approach is more rapid, more versatile, and often provides adequate exposure to perform a pneumonectomy. If additional exposure is needed, extension to a clamshell should provide superb exposure. This would seem to be our practice as patients treated with anterolateral or clamshell thoracotomies were less stable than those treated with a posterolateral thoracotomy.

While control with individual ligation of the pulmonary artery, pulmonary vein, and main stem bronchus is ideal, that may be time consuming and technically quite challenging, particularly in those with very proximal injuries. Mass ligation is quicker, especially in those with injuries proximal in the hilum. In our series, patients who had their hilum stapled en masse were more physiologically deranged, with lower SBPs, lower pH, higher base deficit, and higher serum lactate levels. When using this technique, we prefer to ligate the hilar structures using a TA^TM stapler and reinforce the staple line with a running 3-0 prolene suture. We always cover these staple lines with a muscle or pericardial flap, usually at the time of takeback as TDC is usually used in these desperately ill patients.

Another disastrous complication is acute right heart failure, which contributes to both intraoperative and early postoperative mortality. A landmark study by Cryer and colleagues¹⁰ demonstrated the impact of a pneumonectomy on pulmonary vascular resistance.¹⁰ A porcine model compared three groups: hemorrhagic shock, pneumonectomy, and hemorrhagic shock with pneumonectomy. In the first two groups, there was a modest increase in the pulmonary vascular resistance compared to baseline values. However, in the hemorrhagic shock with pneumonectomy group, the pulmonary vascular resistance increased 500% over baseline. This precipitous rise in pulmonary vascular resistance is poorly tolerated by the right ventricle and the effect is rapidly progressive right ventricular dysfunction, which if not immediately treated, results in further ventricular dilation and dysfunction. In patients with impeding right heart failure, we sometimes widely open the pericardium in the initial operation to give the right ventricle room to swell, using the same rationale as when doing a prophylactic fasciotomy in the extremity. If possible, we close the pericardium later in those treated with TDC, though the heart is often still swollen to allow for pericardial closure. In those cases, we simply leave the pericardium open and have not seen any problems with that strategy.

In this current study, the mortality was 50% which is comparable to multiple published reports.^1-5,7,9,16 Similar to other studies, there was no significant difference in mortality between the first and second decades.^1,3,4 However, in the second decade, if a patient survived 24 hours after surgery, almost all of them survived to discharge. This is particularly noteworthy as patients in the second decade had a significantly lower pH and higher base deficit, indicating profound shock.

Several factors may contribute to these results. TDC is a well described strategy to manage physiologically depleted patients with severe chest injuries.^11-13 It involves initial hemorrhage control, temporary chest closure, resuscitation in the ICU to achieve normal physiology and then planned delayed definitive closure. Studies report a mortality of 24%, with approximately 75% of deaths occurring prior to chest closure.¹¹ Complications occur in 70% of patients but only the need for CRRT is associated with late mortality.¹³

The second important advancement is the increasing use of early VV-ECMO following trauma pneumonectomy. Several reports demonstrate the benefit of VV-ECMO in trauma patients. A study from a high-volume trauma center had a 78% survival to discharge rate of trauma patients on VV-ECMO.¹⁷ In a propensity-matched cohort study, VV-ECMO improved survival in patients with ARDS after trauma from 24% to 65%.¹⁸ In another single-center retrospective study, the authors found that early VV-ECMO (<48 hours from admission) did not increase mortality and also allowed for easier procedural treatment of injuries.¹⁹ Early initiation of VV-ECMO has several benefits. It allows for minimal respiratory support (lung rest), thereby keeping the intrathoracic pressures low and reducing right ventricular afterload and improving cardiac function. ECMO also allows for tight control of pO₂ and pCO₂ avoiding hypoxemia and hypercarbia, both of which cause pulmonary vasoconstriction and elevated pulmonary vascular resistance. An added benefit of VV-ECMO over VA-ECMO is the ability to avoid systemic anticoagulation, particularly in trauma patients who are already coagulopathic and may have other critical non-thoracic injuries.²⁰ The literature, however, on the use of VV-ECMO in trauma pneumonectomy patients is extremely limited. Two small retrospective studies have demonstrated a small, but non-significant, survival advantage.^2,9

It is likely that the small numbers of patients in our dataset limited the ability to show statistical significance on whether the use of TDC or VV-ECMO impacted mortality. It seems reasonable to think that those patients in profound shock who develop severe right heart failure early simply do not survive. It is, however, important to note that those treated with TDC had over a 75% survival, impressive in a group that had a 50% mortality. In those who survived 24 hours and were treated with VV-ECMO, survival was over 70%. We have clearly embraced both TDC and the use of mechanical circulatory support in these desperately ill patients and the results are encouraging. Larger numbers will be needed to demonstrate whether these techniques improve survival.

Our data also demonstrates the need for institutional commitment as these patients require long-term support. In our series, 40% of patients required CRRT and average time on VV-ECMO was 25 days. ICU length of stay was 37 days, with an average of 35 days on the ventilator. Overall hospital length of stay was 46 days. It would seem that investment in these patients is worthwhile as survival, if a patient makes it beyond the immediate pneumonectomy, is quite good.

Recent advances in trauma care such as the administration of TXA, which has been shown to have a survival benefit,²¹ could not be analyzed in this current study due to incomplete data. Likewise, the small numbers in this study precluded the evaluation of balanced transfusion and the use of whole blood, which have also been shown to have demonstrable survival benefits compared to crystalloid resuscitation.^22-25

The results of this study must be taken in the context of its limitations. This is a retrospective cohort study, over two decades, at an extremely busy single level 1 trauma center with experience in TDC surgery and the use of ECMO in trauma patients. Even in our center, we perform about one pneumonectomy for trauma per year. Therefore, the results of this study may not be generalizable to other institutions. In addition, given the dates of this study, which began prior to the universal use of electronic medical records, there were several missing data points, particularly in the first decade. This limits the ability to identify other factors that may have contributed to changes in outcomes such as the use of pulmonary vasodilators, TXA, and balanced transfusion. We were also not able to account for any advances in ICU care that may have also contributed to improved mortalities over two decades. Finally, the comparative population in the two decades were relatively small, making it more difficult to detect statistically significant differences between patient populations.

Conclusion

This study demonstrates that the overall mortality of trauma pneumonectomy has not changed in the last 20 years. The decrease in mortality from 55% to 45%, from the first to second decade, while not statistically significant, seems to be clinically relevant. Injury severity and depth of shock are independently associated with mortality. Interestingly, in the second decade, most patients who survived the initial operation survived to discharge, implying that the inability to rapidly correct profound acidosis results in early mortality. Compared to the first decade, the patients in the second decade were significantly more physiologically depleted, all had thoracic damage control and almost two-thirds had early initiation of VV-ECMO. This suggests that utilizing thoracic damage control and VV-ECMO in patients who survive the index operation may well help improve survival.

Footnotes

Author Contribution

Research plan and idea developed by TS, JOC, and SA, data was collected by AD and SA, statistical analysis was done by SA and EP, writing was completed by SA, TS and JOC with critical review by AD, EP, TS and JOC.

Declaration of conflicting interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: JTACS disclosure forms have been supplied and are provided as supplemental digitial content.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Meeting

AAST, Las Vegas, September 12, 2024 – poster presentation

ORCID iD

Seema P. Anandalwar

References

Martin

McDonald

Mullenix

Steele

Demetriades

. Operative management and outcomes of traumatic lung resection. J Am Coll Surg. 2006;203(3):336-344. doi:10.1016/j.jamcollsurg.2006.05.009

O’Connor

Moran

Galvagno

Deane

Feliciano

Scalea

. Admission physiology vs blood pressure: predicting the need for operating room thoracotomy after penetrating thoracic trauma. J Am Coll Surg. 2020;230(4):494-500. doi:10.1016/j.jamcollsurg.2019.12.019

Phillips

Turco

Mirzaie

Fernandez

. Trauma pneumonectomy: a narrative review. Int J Surg. 2017;46:71-74. doi:10.1016/j.ijsu.2017.08.570

Matsushima

Aiolfi

Park

, et al. Surgical outcomes after trauma pneumonectomy: revisited. J Trauma Acute Care Surg. 2017;82(5):927-932. doi:10.1097/TA.0000000000001416

Bowling

Mavroudis

Richardson

Flint

Howe

Gray

. Emergency pneumonectomy for penetrating and blunt trauma. Am Surg. 1985;51(3):136-139.

Baumgartner

Omari

Lee

, et al. Survival after trauma pneumonectomy: the pathophysiologic balance of shock resuscitation with right heart failure. Am Surg. 1996;62(11):967-972.

Carrillo

Block

Zeppa

Sosa

. Urgent lobectomy and pneumonectomy. Eur J Emerg Med. 1994;1(3):126-130.

Alfici

Ashkenazi

Kounavsky

Kessel

. Total pulmonectomy in trauma: a still unresolved problem--our experience and review of the literature. Am Surg. 2007;73(4):381-384.

Halonen-Watras

O’Connor

Scalea

. Traumatic pneumonectomy: a viable option for patients in extremis. Am Surg. 2011;77(4):493-497.

10.

Cryer

Mavroudis

, et al. Shock, transfusion, and pneumonectomy. Death is due to right heart failure and increased pulmonary vascular resistance. Ann Surg. 1990;212(2):197-201. doi:10.1097/00000658-199008000-00014

11.

O’Connor

DuBose

Scalea

. Damage-control thoracic surgery: management and outcomes. J Trauma Acute Care Surg. 2014;77(5):660-665. doi:10.1097/TA.0000000000000451

12.

Garcia

Martinez

Rodriguez

, et al. Damage-control techniques in the management of severe lung trauma. J Trauma Acute Care Surg. 2015;78(1):45-50. doi:10.1097/TA.0000000000000482

13.

Deane

Galvagno

Moran

Stein

Scalea

O’Connor

. Shock, not blood pressure or shock index, determines the need for thoracic damage control following penetrating trauma. Shock. 2020;54(1):4-8. doi:10.1097/SHK.0000000000001472

14.

Combes

Hajage

Capellier

, et al. Extracorporeal membrane oxygenation for severe acute respiratory distress syndrome. N Engl J Med. 2018;378(21):1965-1975. doi:10.1056/NEJMoa1800385

15.

Aiolfi

Inaba

Martin

, et al. Lung resection for trauma: a propensity score adjusted analysis comparing wedge resection, lobectomy, and pneumonectomy. Am Surg. 2020;86(3):261-265.

16.

Karmy-Jones

Jurkovich

Shatz

, et al. Management of traumatic lung injury: a western trauma association multicenter review. J Trauma. 2001;51(6):1049-1053. doi:10.1097/00005373-200112000-00004

17.

Menaker

Tesoriero

Tabatabai

, et al. Veno-venous extracorporeal membrane oxygenation (VV ECMO) for acute respiratory failure following injury: outcomes in a high-volume adult trauma center with a dedicated unit for VV ECMO. World J Surg. 2018;42(8):2398-2403. doi:10.1007/s00268-018-4480-6

18.

Guirand

Okoye

Schmidt

, et al. Venovenous extracorporeal life support improves survival in adult trauma patients with acute hypoxemic respiratory failure: a multicenter retrospective cohort study. J Trauma Acute Care Surg. 2014;76(5):1275-1281. doi:10.1097/TA.0000000000000213

19.

Powell

Reynolds

Webb

, et al. Early veno-venous extracorporeal membrane oxygenation is an effective strategy for traumatically injured patients presenting with refractory respiratory failure. J Trauma Acute Care Surg. 2023;95(2S Suppl 1):S50-S59. doi:10.1097/TA.0000000000004057

20.

Muellenbach

Kredel

Kunze

, et al. Prolonged heparin-free extracorporeal membrane oxygenation in multiple injured acute respiratory distress syndrome patients with traumatic brain injury. J Trauma Acute Care Surg. 2012;72(5):1444-1447. doi:10.1097/TA.0b013e31824d68e3

21.

Roberts

Shakur

Coats

, et al. The CRASH-2 trial: a randomised controlled trial and economic evaluation of the effects of tranexamic acid on death, vascular occlusive events and transfusion requirement in bleeding trauma patients. Health Technol Assess. 2013;17(10):1-79. doi:10.3310/hta17100

22.

Sheppard

Schaub

Cap

, et al. Whole blood mitigates the acute coagulopathy of trauma and avoids the coagulopathy of crystalloid resuscitation. J Trauma Acute Care Surg. 2018;85(6):1055-1062. doi:10.1097/TA.0000000000002046

23.

Hazelton

Ssentongo

, et al. Use of cold-stored whole blood is associated with improved mortality in hemostatic resuscitation of major bleeding: a multicenter study. Ann Surg. 2022;276(4):579-588. doi:10.1097/SLA.0000000000005603

24.

Brill

Tang

Hatton

, et al. Impact of incorporating whole blood into hemorrhagic shock resuscitation: analysis of 1,377 consecutive trauma patients receiving emergency-release uncrossmatched blood products. J Am Coll Surg. 2022;234(4):408-418. doi:10.1097/XCS.0000000000000086

25.

Holcomb

Tilley

Baraniuk

, et al. Transfusion of plasma, platelets, and red blood cells in a 1:1:1 vs a 1:1:2 ratio and mortality in patients with severe trauma: the PROPPR randomized clinical trial. JAMA. 2015;313(5):471-482. doi:10.1001/jama.2015.12