Abstract
BACKGROUND
Up to 25% of trauma deaths are related to thoracic injuries.
OBJECTIVE
The primary goal was to analyse the incidence and time distribution of death in adult patients with major thoracic injuries. The secondary goal was to determine if potentially preventable deaths occurred within this time distribution and, if so, identify an associated therapeutic window.
DESIGN
Retrospective observational analysis.
SETTING
TraumaRegister DGU.
PATIENTS
Major thoracic injury was defined as an Abbreviated Injury Scale (AIS) 3 or greater. Patients with severe head injury (AIS ≥ 4) or injuries to other body regions with AIS being greater than the thoracic injury (AIS other >AIS thorax) were excluded to ensure that the most severe injury described was primarily thoracic related.
MAIN OUTCOME MEASURES
Incidence and time distribution of mortality were considered the primary outcome measures. Patient and clinical characteristics and resuscitative interventions were analysed in relation to the time distribution of death.
RESULTS
Among adult major trauma cases with direct admission from the accident scene, 45% had thoracic injuries and overall mortality was 9.3%. In those with major thoracic trauma (n = 24 332) mortality was 5.9% (n = 1437). About 25% of these deaths occurred within the first hour after admission and 48% within the first day. No peak in late mortality was seen. The highest incidences of hypoxia and shock were seen in non-survivors with immediate death within 1 h and early death (1 to 6 h). These groups received the largest number of resuscitative interventions. Haemorrhage was the leading cause of death in these groups, whereas organ failure was the leading cause of death amongst those who survived the first 6 h after admission.
CONCLUSION
About half of adult major trauma cases had thoracic injuries. In non-survivors with primarily major thoracic trauma, most deaths occurred immediately (<1h) or within the first 6 h after injury. Further research should analyse if improvements in trauma resuscitation performed within this time frame will reduce preventable deaths.
TRIAL REGISTRATION
The present study is reported within the publication guidelines of the TraumaRegister DGU® and registered as TR-DGU project ID 2020-022.
KEY POINTS
Most of deaths with primarily major thoracic trauma occur immediately (<1 h) or early (1–6 h) after injury.
Immediate and early deaths are most often related to haemorrhage. Multiple Organ Failure (MOF) is associated with high mortality rates in the days after admission.
In the first hour of hospital trauma resuscitation, life-saving interventions must address reversable causes of immediate and early traumatic death and adequate damage control resuscitation should avoid early onset of MOF.
Introduction
Background
In European studies 15 to 43% of trauma deaths are considered preventable. Recent data from the United States identified 36% of trauma deaths as potentially preventable.1–7 Among the preventable deaths, the thorax is reported as the main injured body region (41%), and blunt trauma the main injury mechanism (80 to 92%).4,7 Delayed treatment, inadequate management, and treatment errors are identified as the most common causes of preventable death.4–6,8
Thoracic injury has been identified as a significant predictor for both early death as well as overall mortality, with approximately 25% of all trauma deaths being directly related to thoracic injuries, emphasising the importance of adequate trauma care in major thoracic trauma.7,9–15
Recently it was demonstrated that severe thoracic trauma is associated with a higher risk of death within the first 24 h in patients with a low estimated risk of death.16 These patients’ likelihood of independent survival improves with early diagnoses and subsequent resuscitative interventions.10,13
Early life-saving interventions (LSI) for thoracic trauma include intubation/ventilation, pleural decompression, haemostatic transfusion and resuscitative thoracotomy.15 These interventions, aimed at addressing reversible causes of traumatic death, have been incorporated in the traumatic cardiac arrest (TCA) guidelines of the European Resuscitation Council and all can be performed in the pre-hospital environment.17,18 Recent data confirm that severe thoracic trauma requires more complex pre-hospital and early clinical management, with a higher rate of early procedural and surgical interventions.15
Therefore, a key question is, to what extent are trauma-related deaths currently preventable when life-threatening conditions are not recognised, or life-saving procedures are omitted or incorrectly performed?
Objectives
The primary goal of this study was to analyse the incidence and time distribution of death in adults with major thoracic injury, amongst over 600 European hospitals contributing to the TraumaRegister DGU. Cases in whom major thoracic injury was the primary injury were studied. The secondary objective was to evaluate the hypothesis that, within this time distribution, a therapeutic window can be identified in which the risk of preventable death can be reduced.
Methods
Study design
A retrospective observational analysis was performed, using data from the TraumaRegister DGU.
Data source
The TraumaRegister DGU of the German Trauma Society (Deutsche Gesellschaft für Unfallchirurgie, DGU) is a multi-centre database of pseudonymised and standardised documentation of severely injured patients. The inclusion criteria were emergency department (ED) admission to hospital and subsequent ICU management or reaching hospital with vital signs but dying before ICU admission.
The participating hospitals are primarily located in Germany (90%), but an increasing number of hospitals from other countries also contribute data (Austria, Belgium, Finland, Luxembourg, Slovenia, Switzerland, The Netherlands, the United Arab Emirates). Currently, over 28 000 cases from almost 700 hospitals are entered into the database per year.
The ‘full dataset’ (standard form) consists of over 100 parameters and is obligatory for all supra-regional trauma centres. Only hospitals certified as a regional or local trauma centre within the TraumaNetzwerk DGU are allowed to complete a ‘reduced dataset’ (QM form) with about 40 parameters. A detailed description of the dataset is included in Appendix 1.
Ethics
Ethical approval for this study was not required because of the retrospective study design. The current study is reported within the publication guidelines of the TraumaRegister DGU and registered as TR-DGU project ID 2020-022. Within the TraumaRegister DGU, prior to every data entry, confirmation of the patients consent is mandatory.
Setting and participants
The study cohort from the TraumaRegister DGU that included documented patients in the ‘2015 version’ of the registry dataset (2015 until 2019) was accessed, and from this an overall group of adult trauma patients (≥16 years) with direct admission to a trauma centre was selected. Only patients directly admitted from the accident scene with vital signs on ED arrival were included. Patients who were transferred out within 48 h to another hospital were excluded.
Severity of injury was described by the Abbreviated Injury Scale (AIS), which is an internationally accepted, anatomically based severity scoring system that classifies each individual injury by body region on a 6-point scale (1 = minor and 6 = maximal).19 AIS and AIS-derived scores are used for injury description, risk adjustment and mortality prediction.20
The TraumaRegister DGU considers a maximum AIS of at least 3 as a serious injury, similar to the definition used by the European Union.21 Therefore, for this analysis major thoracic injury was defined as a maximum AIS thorax 3 or greater. However, to avoid confounding by injuries from other body regions, patients with severe neurotrauma (AIS head ≥4) and patients with region AIS scores that were higher than the AIS thorax were excluded. Moreover, those who died within 6 days because of a patient's clear advanced directive with expressed limitations were excluded.
This left a cohort of patients in whom major thoracic trauma was the primarily injury (major thoracic trauma) (Fig. 1).
Fig. 1.
Selection of study sample.
Statistical methods
Within the overall group the prevalence of thoracic injuries was compared with injuries to other body regions. Mortality rates were compared between the overall group and those with major thoracic trauma. Within this latter group, severity of thoracic injuries (AIS thorax) and associated mortality was documented.
For major thoracic trauma the temporal distribution of death per hour and day was analysed, looking both at absolute and cumulative numbers. Histograms were used to identify patterns.
Subsequently, based on the numbers and histogram patterns, the non-survivors were subdivided into immediate deaths (within 1 h), early deaths (1 to 6 h), delayed deaths (6 h to 1 week), and late deaths (after 1 week). Variables were analysed and compared between these different subgroups, including patient characteristics, clinical findings, management, and associated injuries. Intubation/ventilation, pleural decompression, haemostatic transfusion and resuscitative thoracotomy were considered as LSI in major thoracic trauma. Mortality prediction was calculated using the validated Revised Injury Severity Classification, version 2 (RISC II).22,23
Categorical variables were recorded as numbers with percentages and continuous variables as means ± SD if approximal normally distributed and median [IQR] otherwise.
Chi-squared test was used for categorical variables, and Mann–Whitney U-test for metric and ordinal measures. Statistical significance was defined as P less than 0.05. However, in large database analyses even small differences can become statistically significant without any clinical significance. Therefore, P values are not reported as they do not necessarily reflect clinical significance.
Statistical analysis was conducted with SPSS statistical software (Version 27; IBM Corp., Armonk, New York, USA). Graphs were created with MS Excel (Excel for Mac, Version 16.46. Redmond, WA: Microsoft Corp).
Results
Participants
The study cohort consisted of 175 729 patients registered in the TraumaRegister DGU from 2015 to 2019. After applying the exclusion criteria, 143 503 patients were selected in the overall group. Of these, 43 035 had thoracic injury with AIS at least 3 from which 24 332 were identified with primarily major thoracic trauma.
Prevalence of thoracic injuries
In the overall group, the thorax was the second most injured body region (n = 65 287, 45.5%), after head injury (n = 76 758, 53.5%), and had the highest prevalence of severe injuries (AIS ≥ 3) compared with other body regions (n = 43 035, 30%).
Mortality and temporal distribution of death
The documented all-cause mortality was 9.3% in the overall group. A total of 1437 patients (5.9%) with primarily major thoracic trauma died after hospital arrival. Half of the documented deaths occurred within the first day (n = 693, 48.2%). For 38 (2.6%) cases who died within this first 24 h no exact time of death was recorded, so they were excluded from the hourly distribution analysis. A quarter of the deaths (n = 349, 24.9%) were observed within the first hour and 41.3% (n = 579) within the first 6 h.
After plotting the distribution of death in histograms (Figs. 2 and 3), no late death peak was detected. The initial peak was followed by a rapidly diminishing slope.
Fig. 2.
Time distribution of death for patients with ‘major thoracic trauma’ (days) within the first 30 days of admission. Blue columns: absolute numbers – grey columns: cumulative numbers.
Fig. 3.
Time distribution of death for patients with ‘major thoracic trauma’ (h) within the first day of admission. Blue columns: absolute numbers – grey columns: cumulative numbers.
Characteristics of major thoracic trauma patients
Both demographic and clinical characteristics were analysed and listed in Table 1.
Table 1.
Demographic an clinical characteristics of patients with ‘major thoracic trauma’
| NON-SURVIVORS | ||||||||
| TOTAL | SURVIVORS | Total death | Immediate death | Early death | Delayed death | Late death | MISSING | |
| <1h | 1h to 6h | 6h to 1 week | >1 week | |||||
| Patient characteristics | ||||||||
| Sex, male | 18532 (76.2) | 17489 (76.4) | 1043 (72.4) | 259 (74.2) | 173 (70.0) | 231 (73.3) | 380 (71.8) | 0 (0) |
| Age | 55.0 ± 18.9 | 54.3 ± 18.6 | 65.9 ± 19.8 | 56.5 ± 21.1 | 59.7 ± 20.5 | 67.6 ± 19.0 | 73.8 ± 14.8 | 0 (0) |
| Anticoagulants | 3625 (17.6) | 3258 (17.7) | 367 (41.3) | 27 (16.4) | 43 (35.2) | 87 (34.4) | 210 (52.6) | 5080 (20.9) |
| ASA 3/4 | 3578 (16.1) | 3055 (14.5) | 523 (44.4) | 62 (25.1) | 50 (27.3) | 126 (47.5) | 285 (59.1) | 2065 (8.5) |
| Accident mechanism | ||||||||
| Blunt | 22241 (95.3) | 20972 (95.5) | 1269 (92.0) | 284 (85.5) | 210 (89.0) | 282 (93.4) | 493 (96.7) | 989 (4.1) |
| Car / lorry | 6824 (28.2) | 6430 (28.2) | 394 (27.6) | 117 (33.8) | 65 (26.7) | 85 (27.2) | 127 (24.1) | 140 (0.6) |
| Motorbike | 3895 (16.1) | 3780 (16.6) | 115 (8.1) | 39 (11.3) | 23 (9.5) | 26 (8.3) | 27 (5.1) | 140 (0.6) |
| Bicycle | 2213 (9.1) | 2146 (9.4) | 67 (4.7) | 17 (4.9) | 13 (5.3) | 14 (4.5) | 23 (4.4) | 140 (0.6) |
| Pedestrian | 913 (3.8) | 808 (3.5) | 105 (7.4) | 25 (7.2) | 19 (7.8) | 21 (6.7) | 40 (7.6) | 140 (0.6) |
| High fall, ≥3m | 3454 (14.3) | 3231 (14.2) | 223 (15.6) | 56 (16.2) | 50 (20.6) | 41 (13.1) | 76 (14.4) | 140 (0.6) |
| Low fall, <3m | 4261 (17.6) | 3957 (17.4) | 304 (21.3) | 14 (4.0) | 27 (11.1) | 80 (25.6) | 183 (34.8) | 140 (0.6) |
| Other | 2632 (10.9) | 2413 (10.6) | 219 (15.3) | 78 (22.5) | 46 (18.9) | 45 (14.4) | 50 (9.5) | 140 (0.6) |
| Prehospital vital signs | ||||||||
| GCS ≤8 | 1642 (7.3) | 1010 (4.8) | 632 (48.0) | 276 (85.2) | 132 (57.9) | 113 (39.8) | 111 (23.0) | 1942 (8.0) |
| SBP, mmHg | 134.5 ± 31.9 | 136.5 ± 28.7 | 98.2 ± 56.5 | 42.5 ± 52.8 | 87.1 ± 50.5 | 108.6 ± 46.7 | 125.0 ± 42.8 | 3127 (12.9) |
| SBP ≤90mmHg | 1692 (8.0) | 1246 (6.2) | 446 (40.2) | 184 (81.1) | 93 (50.8) | 78 (30.8) | 91 (20.4) | 3127 (12.9) |
| SpO2 <94% | 3295 (34.0) | 2956 (32.3) | 339 (64.9) | 81 (89.0) | 64 (75.3) | 72 (60.0) | 122 (54.0) | 14655 (60.2) |
| Prehospital management | ||||||||
| Prehospital time, min | 63.9 ± 30.7 | 63.8 ± 30.5 | 66.5 ± 32.4 | 63.2 ± 27.9 | 61.9 ± 26.2 | 73.2 ± 37.9 | 66.7 ± 33.5 | 3716 (15.3) |
| Pleural decompression | 793 (7.1) | 632 (6.0) | 161 (21.4) | 75 (42.6) | 25 (19.1) | 30 (18.0) | 31 (11.1) | 13116 (53.9) |
| Tranexamic acid | 1639 (7.3) | 1444 (6.8) | 195 (14.8) | 64 (20.4) | 33 (14.5) | 46 (16.0) | 52 (10.6) | 1836 (7.5) |
| Intubation | 2097 (18.0) | 1637 (15.0) | 460 (59.4) | 162 (90.5) | 90 (66.7) | 101 (59.4) | 107 (36.8) | 12662 (52.0) |
| Volume, ml | 660.4 ± 519.4 | 643.0 ± 492.4 | 953.5 ± 800.4 | 1231.6 ± 914.9 | 1094.3 ± 802.4 | 930.4 ± 797.5 | 734.6 ± 648.2 | 14876 (61.1) |
| Vital signs on admission | ||||||||
| SBP, mmHg | 136.0 ± 30.8 | 138.2 ± 27.2 | 96.4 ± 54.2 | 35.2 ± 52.7 | 81.1 ± 49.6 | 112.4 ± 37.8 | 124.8 ± 33.9 | 1310 (5.4) |
| SBP ≤90mmHg | 1554 (6.8) | 1069 (4.9) | 485 (39.6) | 208 (84.6) | 120 (56.9) | 75 (26.7) | 82 (16.8) | 1310 (5.4) |
| SpO2 <94% | 2111 (20.1) | 1823 (18.4) | 288 (48.5) | 88 (84.6) | 57 (57.0) | 58 (42.0) | 85 (33.7) | 13827 (56.8) |
| Laboratory values on admission | ||||||||
| Haemoglobin, g dl−1 | 13.5 ± 2.1 | 13.7 ± 1.9 | 11.0 ± 2.9 | 9.6 ± 3.5 | 10.4 ± 3.1 | 11.2 ± 2.8 | 11.9 ± 2.3 | 760 (3.1) |
| Platelets, ×109 l−1 | 230.6 ± 77.6 | 233.5 ± 75.6 | 182.7 ± 93.1 | 111.0 ± 75.5 | 168.2 ± 84.9 | 179.3 ± 85.2 | 210.0 ± 93.0 | 13546 (55.7) |
| PT Quick, % | 90.4 ± 20.8 | 91.6 ± 19.6 | 65.3 ± 27.0 | 46.9 ± 27.5 | 59.1 ± 25.4 | 65.2 ± 26.1 | 72.2 ± 25.3 | 2054 (8.4) |
| INR | 1.1 ± 0.4 | 1.1 ± 0.4 | 1.6 ± 1.1 | 2.4 ± 1.9 | 1.7 ± 1.1 | 1.6 ± 1.1 | 1.4 ± 0.7 | 1554 (6.4) |
| Coagulopathy | 2181 (9.6) | 1682 (7.8) | 499 (45.2) | 90 (73.8) | 112 (58.3) | 130 (45.1) | 167 (33.2) | 1537 (6.3) |
| Base excess, +/– | −1.4 ± 4.5 | −1.0 ± 3.7 | −7.8 ± 9.2 | −15.6 ± 10.2 | −11.9 ± 9.7 | −7.6 ± 8.6 | −3.2 ± 5.4 | 5217 (21.4) |
| Management on admission | ||||||||
| Intubation | 808 (6.9) | 719 (6.6) | 89 (11.5) | 12 (6.7) | 31 (23.0) | 11 (6.5) | 35 (12.0) | 12662 (52.0) |
| Pleural decompression | 3012 (12.4) | 2648 (11.6) | 364 (25.3) | 96 (27.5) | 89 (36.0) | 73 (23.2) | 106 (20.0) | 0 (0) |
| Tranexamic acid | 1301 (12.6) | 1072 (11.1) | 229 (34.4) | 50 (34.2) | 58 (50.9) | 55 (38.7) | 66 (25.0) | 14008 (57.6) |
| Transfusion, PRC | 1345 (5.6) | 907 (4.0) | 438 (32.8) | 124 (45.9) | 137 (59.6) | 89 (29.0) | 88 (16.6) | 104 (0.4) |
| Massive Transfusion >10PRC | 176 (0.7) | 77 (0.3) | 99 (7.4) | 11 (4.1) | 40 (17.4) | 32 (10.4) | 16 (3.0) | 104 (0.4) |
| Volume, ml | 977.2 ± 1266.4 | 933.1 ± 1190.9 | 1615.0 ± 1965.0 | 1250.1 ± 1117.6 | 2453.6 ± 2343.3 | 1915.1 ± 2189.3 | 1293.7 ± 1908.2 | 14876 (61.1) |
| Haemostatic treatment | 1623 (15.6) | 1334 (13.7) | 289 (42.7) | 63 (42.3) | 71 (61.7) | 70 (47.6) | 85 (32.0) | 13931 (57.3) |
| Emergency surgery | 3165 (13.0) | 2776 (12.1) | 389 (27.0) | 77 (22.1) | 99 (40.1) | 95 (30.2) | 118 (22.3) | 0 (0) |
| Thoracotomy | 855 (3.9) | 661 (3.2) | 194 (15.1) | 70 (22.7) | 63 (28.4) | 33 (12.4) | 28 (5.8) | 2256 (9.3) |
| Complications | ||||||||
| Mulitple organ failure | 1301 (13.5) | 955 (10.4) | 346 (74.9) | 1 (50.0) | 22 (61.1) | 116 (79.5) | 207 (74.5) | 14694 (60.4) |
| Sepsis | 532 (5.6) | 402 (4.4) | 130 (29.5) | 0 (0) | 0 (0) | 12 (9.0) | 118 (43.7) | 14791 (60.8) |
| Injury severity | ||||||||
| ISS | 18.5 ± 9.1 | 17.8 ± 8.0 | 30.0 ± 16.1 | 34.9 ± 18.3 | 35.8 ± 18.7 | 29.5 ± 14.7 | 24.5 ± 11.3 | 0 (0) |
| AIS thorax = 3 | 14783 (60.8) | 14492 (63.3) | 291 (20.2) | 20 (5.7) | 22 (8.9) | 72 (22.9) | 177 (33.5) | 0 (0) |
| AIS thorax = 4 | 6461 (26.6) | 6010 (26.3) | 451 (31.3) | 95 (27.2) | 72 (29.1) | 96 (30.5) | 188 (35.5) | 0 (0) |
| AIS thorax = 5 | 2995 (12.3) | 2387 (10.4) | 608 (42.2) | 190 (54.4) | 120 (48.6) | 136 (43.2) | 162 (30.6) | 0 (0) |
| AIS thorax = 6 | 93 (0.4) | 3 (<0.1) | 90 (6.3) | 44 (12.6) | 33 (13.4) | 11 (3.5) | 2 (0.4) | 0 (0) |
| Prognosis | ||||||||
| RISC II, % | 6.4 ± 16.4 | 3.7 ± 9.0 | 48.5 ± 37.3 | 84.4 ± 22.2 | 61.3 ± 34.4 | 40.6 ± 33.6 | 23.6 ± 25.5 | 0 (0) |
| Cause of death | ||||||||
| Traumatic brain injury | 83 (6.2) | 20 (6.9) | 10 (4.4) | 25 (8.3) | 28 (5.5) | 108 (0.4) | ||
| Haemorrhage | 297 (22.3) | 135 (46.7) | 107 (46.7) | 46 (15.2) | 9 (1.8) | 108 (0.4) | ||
| Organ failure | 671 (50.4) | 70 (24.2) | 69 (30.1) | 168 (55.4) | 364 (71.2) | 108 (0.4) | ||
| Other | 281 (21.1) | 64 (22.1) | 43 (18.8) | 64 (21.1) | 110 (21.5) | 108 (0.4) | ||
Data are mean ± SD and number (%).
ASA, American Society of Anesthesiologist Physical Status Classification; GCS, Glasgow coma scale; SBP, systolic blood pressure; PT, prothromin time; INR, international normalised ratio; PRC, packed red cells; ISS, injury severity score; AIS, abbreviated injury scale; RISC-II, revised injury severity classification, version II.
Patient characteristics
The majority of both survivors and non-survivors were male (overall 76.2%), with a mean age of 55 ± 18.9 years. Delayed and late death was associated with an increasing mean age, higher prevalence of co-morbidities (American Society of Anesthesiologist, ASA score 3 and 4) and use of anti-coagulants.
Blunt trauma was most common in all subgroups, but immediate and early deaths had a higher portion of penetrating trauma compared with late deaths.
Road crashes were the leading trauma mechanism for survivors and immediate/early deaths, whereas low falls (<3 m) were the leading trauma mechanism for the late deaths (34.8%).
Injury severity and outcome
The number of patients admitted decreased with increasing severity of thoracic trauma. The majority had an AIS 3 (60.8%), whereas a very small portion had AIS 6 (0.4%). Increasing AIS was associated with increased mortality.
Injury Severity Score was higher in all non-survivor subgroups compared with survivors, with highest values being observed in early deaths.
The observed hospital mortality (5.9%) was lower than the predicted mortality based on RISC II (6.4%). Predicted mortality was higher in earlier deaths compared with late deaths. For immediate and early deaths, haemorrhage was recorded as the leading cause of death (both 46.7%). Organ failure was the leading cause of delayed and late death (55.4% and 71.2%).
Clinical findings
Pre-hospital Glasgow Coma Scale 8 or less, hypotension (systolic blood pressure (SBP) ≤ 90 mmHg), and hypoxia (SpO2 < 94%) were more common in non-survivors, with the highest incidence of these conditions and lowest values for SBP in immediate and early deaths.
Similar findings were seen on admission, along with higher incidences of coagulopathy (prothrombin time (PT) ≤60 and/or activated partial thromboplastin time (aPTT) ≥ 40 and/or International Normalised Ratio (INR) ≥ 1.4) and acidosis (base excess (BE) ≤ −6). In absolute numbers, the mean INR was 2.4 and BE was −15.6 for immediate deaths, compared with 1.1 and −1.0, respectively, in the survivor group.
Management
The mean time from the incident until ED admission was 63.9 min for all patients with major thoracic trauma. No relevant differences were seen between the different subgroups.
Pleural decompression was performed pre-hospital in 7.1% of patients and in 12.4% after hospital admission. Highest rates were seen in immediate deaths (42.6%) for pre-hospital decompression and in early deaths (36.0%) when performed in the ED.
Non-survivors were more likely to have undergone endotracheal intubation. In this group, pre-hospital intubation (59.4%) was more frequent compared with intubation in ED (11.5%). Immediate deaths had the highest intubation rate in the pre-hospital phase (90.5%).
A quarter of the non- survivors (27.0%) underwent emergency surgery prior to ICU admission or death, compared with 12.1% of survivors with an emergency surgical procedure.
For patients with ICU admission, the median length-of-stay (LOS) in ICU for non-survivors was 6 [2 to 14] days with a median of 3 [1 to 10] ventilator days. For survivors 2 [1 to 6] days LOS in ICU was recorded, with a median of 0 [0 to 1] ventilator days.
Early deaths had the highest incidence of blood transfusion (59.6%), massive transfusion with at least 10 units of packed red blood cells (17.4%) and administration of haemostatic agents (61.7%). They also received the largest amounts of intravenous fluids (crystalloids and colloids) in the ED.
Discussion
The current study confirmed the primacy of thoracic trauma in adult major trauma cases surviving to hospital arrival. For those in whom thoracic trauma was the leading injury, mortality was slightly lower (5.9%) compared with the overall trauma cohort (9.3%) due to the exclusion of severe traumatic brain injury (TBI) which is known to be the leading cause of death after trauma.24
Most of deaths occurred shortly after hospital arrival, with a quarter of deaths already in the first hour. In contrast with Trunkey's (1983) sentinel historical description, there was no trimodal distribution of death for patients with major thoracic trauma.25 A late peak of deaths could not be detected. This finding is similar to that described by Bardes et al.24 and Rauf et al.26
The disappearance of the late peak in the weeks after admission may be related to a decreasing incidence of sepsis and multiple organ failure (MOF) in thoracic trauma patients, as documented by Horst et al.27 over a 10-year period. Nevertheless, thoracic trauma is a well known risk factor for the development of MOF.28 In this analysis, organ failure remained the leading cause of death in those who survived the first 6 h. These delayed and late deaths were associated with low-energy trauma, older age, and co-morbidities. Previous research has repeatedly shown that this group of patients have a higher risk of developing single or MOF.28,29 However, in recent analyses by Minei et al.29 and Cole et al.30 the historically described second peak in the onset of MOF between 7 and 14 days had disappeared, likely due to improved intensive care management. Nearly all MOF patients developed organ dysfunction within the first 3 days, with respiratory and cardiovascular dysfunction being the greatest contributors. Different strategies in resuscitation management are associated with the early development of MOF, which is predictive of mortality.
Recently, Brohi et al.31 highlighted the change in the patterns of organ failure. Fulminant cardiac and vascular failure can develop early after initial resuscitation and this group might benefit from early cardiac support and ischaemic protection. A second group will develop a persistent inflammation, immunosuppression and catabolism syndrome, which can cause delayed mortality despite prolonged ICU management. If these patterns are present in thoracic trauma patients, and if early recognition and treatment can further increase the number of ICU survivors needs profound research.
Nevertheless, despite the significantly decreasing incidence of sepsis and MOF in major thoracic trauma patients, mortality rates remained unchanged.27 Moreover, it is known that a high incidence of potentially preventable trauma related mortality remains.32 By excluding severe TBI from our analysis and only including patients with thoracic injuries as the primary diagnosis who arrived with life signs in the ED, we examined a patient group for whom early appropriate in-hospital resuscitation had an increased likelihood of success and where other injuries were unlikely to be confounding. The high portion of immediate and early deaths, a quarter within the first hour after arrival, emphasises the critical role of pre-hospital and primary trauma resuscitation. This patient group had a higher prevalence of hypotension, hypoxia, and coagulopathy, which are potentially reversible with LSI.7 The TCA algorithm by Lockey et al.17, adopted by the European Resuscitation Council, emphasises the importance of simultaneously addressing reversible causes of traumatic death.18
Essential LSI considered key to the management of the (peri-) arresting trauma patient include: control of external haemorrhage, maintaining airway patency, pleural decompression, application of a pelvic splint and massive transfusion. If resources are available and life signs present during the prior 15 min, resuscitative thoracotomy is recommended to relieve cardiac tamponade, control haemorrhage and to gain proximal vascular control.17,18 However, the limited performance of pleural decompression (42.6% pre-hospital, 27.5% in-hospital), and resuscitative thoracotomy (22.7%) in patients who died within the first hour after arrival, seems to be in contrast to the current guidelines and evidence.
Pre-hospital pleural decompression was shown to be associated with lower 24 h mortality compared with decompression in ED and appears to be associated with the probability of survival in patients with TCA.33–35 The implementation of teaching programmes for pleural decompression was found to be associated with a reduced incidence of untreated tension pneumothorax and reduced error and complication rates.36,37 Both finger thoracostomy and needle decompression appear to be feasible and safe to be performed by paramedics.38
The highest success rates of resuscitative thoracotomy are reported in penetrating trauma with witnessed signs of life (61.2%), but survival rates in blunt trauma with witnessed cardiac arrest range up to 26%.39,40
Blood transfusion (45.9%), massive transfusion (4.1%), and administration of tranexamic acid (TXA) (20.4% pre-hospital, 34.2% in-hospital) were limited. However, haemorrhage was identified as cause of death in about half of the patients who died immediately or early after admission to the ED (46.7%) in which also a higher incidence of coagulopathy (73.8%) was recorded. Imach et al.41 demonstrated TXA administration was associated with a significant lower mortality risk at 6 and 12 h after admission. Massive transfusion protocols repeatedly demonstrated their effectiveness in reducing mortality, with better outcomes when transfusion was initiated as early as possible.42
These findings need to be interpreted with caution but may indicate that a portion of deaths could be prevented by rigorous implementation of available guidelines and structured trauma systems with the ability to perform LSI immediately on arrival in the ED. A recent editorial by Cameron and Andrews43 highlighted the need for adequate clinical governance and appropriate training and education. The existence of this window for improvement needs further research.
Limitations
The current study has several limitations. First, a retrospective observational study design comes with its inherent limitations. Associations can be determined, but causality cannot be proven. Although the online portal of the TraumaRegister DGU does not accept data considered impossible for several parameters, errors in data entry could not be ruled out. However, the large sample size of our study helped minimise the above limitation.
Although an analysis of initial management was performed for the entire study cohort, no evaluation of management errors was undertaken for individual cases.
The TraumaRegister DGU only includes patients admitted to hospital. No data on pre-hospital deaths were available. Further evaluation is warranted as previous research demonstrates a high number of potentially preventable pre-hospital deaths.2,7
No data were available on the reason for resuscitation discontinuation. However, as signs of life are necessary to be included in the TraumaRegister DGU, no patient was dead on ED arrival.
Finally, although we aimed to focus on the contribution of thoracic injury to mortality by defining the major thoracic injury group, it cannot be completely excluded that the cause of death was related to another non-thoracic injury or a combination of several injuries from various body regions.
Conclusion
Thoracic injuries have the highest prevalence of severe injuries in adult major trauma patients. In this analysis, approximately 25% of non-survivors with primarily major thoracic injuries, admitted to the ED with life signs on arrival, died within the first hour after admission. Most of these deaths were related to haemorrhage with associated shock and coagulopathy.
LSI can address reversible causes of immediate and early traumatic death and adequate damage control resuscitation can avoid early onset of MOF, which is associated with high mortality rates in the days after admission.
Clearly, the first hour of hospital trauma reception and resuscitation is critical for adult major thoracic trauma patients. Further research should analyse if this window provides an opportunity for improvement in trauma resuscitation outcomes.
Supplementary Material
Acknowledgements relating to this article
Assistance with the study: none.
Financial support and sponsorship: the authors received no specific funding for this work.
Conflicts of interest: RF declares that his institute has an on-going service agreement with AUC GmbH, the holder of TraumaRegister DGU, which includes statistical support in analysis of registry data. All other authors declared no conflicts of interest.
Presentation: preliminary data for this study were presented as a poster presentation at the European Emergency Medicine Congress, 27 to 31 October 2021, Lisbon.
This manuscript was handled by Claire Roger.
Footnotes
Supplemental digital content is available for this article.
References
- 1.Montmany S, Pallisera A, Rebasa P, et al. Preventable deaths and potentially preventable deaths. What are our errors? Injury 2016; 47:669–673. [DOI] [PubMed] [Google Scholar]
- 2.Oliver GJ, Walter DP, Redmond AD. Are prehospital deaths from trauma and accidental injury preventable? A direct historical comparison to assess what has changed in two decades. Injury 2017; 48:978–984. [DOI] [PubMed] [Google Scholar]
- 3.Oliver GJ, Walter DP, Redmond AD. Prehospital deaths from trauma: are injuries survivable and do bystanders help? Injury 2017; 48:985–991. [DOI] [PubMed] [Google Scholar]
- 4.Girard E, Jegousso Q, Boussat B, et al. Preventable deaths in a French regional trauma system: a six-year analysis of severe trauma mortality. J Visc Surg 2019; 156:10–16. [DOI] [PubMed] [Google Scholar]
- 5.Kleber C, Giesecke MT, Tsokos M, et al. Trauma-related preventable deaths in Berlin 2010: need to change prehospital management strategies and trauma management education. World J Surg 2013; 37:1154–1161. [DOI] [PubMed] [Google Scholar]
- 6.Pfeifer R, Halvachizadeh S, Schick S, et al. Are prehospital trauma deaths preventable? A systematic literature review. World J Surg 2019; 43:2438–2446. [DOI] [PubMed] [Google Scholar]
- 7.Drake SA, Holcomb JB, Yang Y, et al. Establishing a regional trauma preventable/potentially preventable death rate. Ann Surg 2020; 271:375–382. [DOI] [PubMed] [Google Scholar]
- 8.Beck B, Smith K, Mercier E, et al. Potentially preventable trauma deaths: a retrospective review. Injury 2019; 50:1009–1016. [DOI] [PubMed] [Google Scholar]
- 9.Veysi VT, Nikolaou VS, Paliobeis C, et al. Prevalence of chest trauma, associated injuries and mortality: a level i trauma centre experience. Int Orthop 2009; 33:1425–1433. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Huber S, Biberthaler P, Delhey P, et al. Predictors of poor outcomes after significant chest trauma in multiply injured patients: a retrospective analysis from the German Trauma Registry (Trauma Register DGU®). Scand J Trauma Resusc Emerg Med 2014; 22:1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Battle C, Hutchings HA, Driscoll T, et al. A multicentre randomised feasibility STUdy evaluating the impact of a prognostic model for Management of BLunt chest wall trauma patients: STUMBL Trial. BMJ Open 2019; 9:e029187. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Pape H-C, Peitzman AB, Schwab CW, et al. Damage control management in the polytrauma patient. New York, NY: Springer; 2010. [Google Scholar]
- 13.Blyth A. Thoracic trauma. BMJ 2014; 348:bmjg.1137–bmjg.11137. [DOI] [PubMed] [Google Scholar]
- 14.Valdez C, Sarani B, Young H, et al. Timing of death after traumatic injury—a contemporary assessment of the temporal distribution of death. J Surg Res 2016; 200:604–609. [DOI] [PubMed] [Google Scholar]
- 15.Bayer J, Lefering R, Reinhardt S, et al. Severity-dependent differences in early management of thoracic trauma in severely injured patients – analysis based on the TraumaRegister DGU®. Scand J Trauma Resusc Emerg Med 2017; 25:1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Nolte PC, Häske D, Lefering R, et al. Training to identify red flags in the acute care of trauma: who are the patients at risk for early death despite a relatively good prognosis? An analysis from the TraumaRegister DGU®. World J Emerg Surg 2020; 15:1–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Lockey DJ, Lyon RM, Davies GE. Development of a simple algorithm to guide the effective management of traumatic cardiac arrest. Resuscitation 2013; 84:738–742. [DOI] [PubMed] [Google Scholar]
- 18.Lott C, Truhlar A, Alfonzo A, et al. European Resuscitation Council Guidelines 2021: cardiac arrest in special circumstances. Resuscitation 2021; 161:152–219. [DOI] [PubMed] [Google Scholar]
- 19.Committee on Medical Aspects of Automotive Safety. Rating the severity of tissue damage. I. The abbreviated scale. JAMA 1971; 215:277–280. [DOI] [PubMed] [Google Scholar]
- 20.Van Ditshuizen JC, Sewalt CA, Palmer CS, et al. The definition of major trauma using different revisions of the abbreviated injury scale. Scand J Trauma Resusc Emerg Med 2021; 29:1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Weijermars W, Bos N, Schoeters A, et al. Serious road traffic injuries in Europe, lessons from the EU research project safetycube. Transp Res Rec 2018; 2672:1–9. [Google Scholar]
- 22.Lefering R, Huber-wagner S, Nienaber U, et al. Update of the trauma risk adjustment model of the TraumaRegister DGUTM: the Revised Injury Severity Classification, version II. Crit Care 2014; 18:476. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Lefering R, Huber-Wagner S, Bouillon B, et al. Cross-validation of two prognostic trauma scores in severely injured patients. Eur J Trauma Emerg Surg 2021; 47:1837–1845. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Rauf R, von Matthey F, Croenlein M, et al. Changes in the temporal distribution of in-hospital mortality in severely injured patients—an analysis of the TraumaRegister DGU. PLoS One 2019; 14:1–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Trunkey DD. Trauma. Accidental and intentional injuries account for more years of life lost. Sci Am 1983; 249:28–35. [PubMed] [Google Scholar]
- 26.Bardes JM, Inaba K, Schellenberg M, et al. The contemporary timing of trauma deaths. J Trauma Acute Care Surg 2018; 84:893–899. [DOI] [PubMed] [Google Scholar]
- 27.Horst K, Andruszkow H, Weber CD, et al. Thoracic trauma now and then: a 10 year experience from 16,773 severely injured patients. PLoS One 2017; 12:1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Fröhlich M, Lefering R, Probst C, et al. Epidemiology and risk factors of multiple-organ failure after multiple trauma: an analysis of 31,154 patients from the TraumaRegister DGU. J Trauma Acute Care Surg 2014; 76:921–927. [DOI] [PubMed] [Google Scholar]
- 29.Minei JP, Cuschieri J, Sperry J, et al. The changing pattern and implications of multiple organ failure after blunt injury with hemorrhagic shock. Crit Care Med 2012; 40:1129–1135. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Cole E, Gillespie S, Vulliamy P, et al. Multiple organ dysfunction after trauma. Br J Surg 2020; 107:402–412. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Brohi K, Gruen RL, Holcomb JB. Why are bleeding trauma patients still dying? Intensive Care Med 2019; 45:709–711. [DOI] [PubMed] [Google Scholar]
- 32.Committee on Military Trauma Care's Learning Health System and Its Translation to the Civilian Sector; Board on Health Sciences Policy; Board on the Health of Select Populations; Health and Medicine Division; National Academies of Sciences, Engineering, and Medicine. A National Trauma Care System: Integrating Military and Civilian Trauma Systems to Achieve Zero Preventable Deaths After Injury. Berwick D, Downey A, Cornett E, editors. Washington (DC): National Academies Press (US); 2016. [PubMed] [Google Scholar]
- 33.Bushby N, Fitzgerald M, Cameron P, et al. Prehospital intubation and chest decompression is associated with unexpected survival in major thoracic blunt trauma. EMA – Emerg Med Australas 2005; 17:443–449. [DOI] [PubMed] [Google Scholar]
- 34.Huber-Wagner S, Lefering R, Qvick M, et al. Outcome in 757 severely injured patients with traumatic cardiorespiratory arrest. Resuscitation 2007; 75:276–285. [DOI] [PubMed] [Google Scholar]
- 35.Muchnok D, Vargo A, Deeb AP, et al. Association of prehospital needle decompression with mortality among injured patients requiring emergency chest decompression. JAMA Surg 2022; 157:934–940. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Cantwell K, Burgess S, Patrick I, et al. Improvement in the prehospital recognition of tension pneumothorax: the effect of a change to paramedic guidelines and education. Injury 2014; 45:71–76. [DOI] [PubMed] [Google Scholar]
- 37.Anderson M, Fitzgerald M, Martin K, et al. A procedural check list for pleural decompression and intercostal catheter insertion for adult major trauma. Injury 2015; 46:42–44. [DOI] [PubMed] [Google Scholar]
- 38.Sharrock MK, Shannon B, Garcia Gonzalez C, et al. Prehospital paramedic pleural decompression: a systematic review. Injury 2021; 52:2778–2786. [DOI] [PubMed] [Google Scholar]
- 39.Narvestad JK, Meskinfamfard M, Søreide K. Emergency resuscitative thoracotomy performed in European civilian trauma patients with blunt or penetrating injuries: a systematic review. Eur J Trauma Emerg Surg 2016; 42:677–685. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Fitzgerald MC, Yong MS, Martin K, et al. Emergency department resuscitative thoracotomy at an adult major trauma centre: Outcomes following a training programme with standardised indications. EMA – Emerg Med Australas 2020; 32:657–662. [DOI] [PubMed] [Google Scholar]
- 41.Imach S, Wafaisade A, Lefering R, et al. The impact of prehospital tranexamic acid on mortality and transfusion requirements: match-pair analysis from the nationwide German TraumaRegister DGU®. Crit Care 2021; 25:1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Meneses E, Boneva D, McKenney M, et al. Massive transfusion protocol in adult trauma population. Am J Emerg Med 2020; 38:2661–2666. [DOI] [PubMed] [Google Scholar]
- 43.Cameron P, Andrews T. Life saving interventions in resuscitation of the injured patient. Injury 2022; 53:2387–2388. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.