- Email: [email protected]
Injury patterns associated with mortality following motorcycle crashes
Injury, Int. J. Care Injured 33 (2002) 473–477
Injury patterns associated with mortality following motorcycle crashes S. Ankarath, P.V. Giannoudis∗ , I. Barlow, M.C. Bellamy, S.J. Matthews, R.M. Smith Department of Trauma and Orthopaedics, St. James’s University Hospital, Leeds LS9 7TF, UK Accepted 12 February 2002
Abstract All patients involved in motorcycle crashes admitted to various hospitals in the Yorkshire region of UK between January 1993 and December 1999 were retrospectively reviewed to identify the factors that are likely to predict a reduced survival. Of the 1239 patients requiring hospital admission, 74 died. The probability of reduced survival was estimated by a logistic regression model using independent variables such as head injury, thoracic trauma, abdominal injury, spinal injury and pelvic fracture and a compound variable of pelvic fracture combined with a long bone fracture. The odds ratio for head injury was 0.349, chest injury 0.39, abdominal injury 0.42, and the compound variable (pelvis plus a long bone fracture) 0.576. The mean injury severity score (ISS) in the fatal group was 35.96 compared to 12.2 in the group that survived (P < 0.01). There was a significant difference in the Glasgow coma scale (GCS) between patients wearing a helmet and those that did not wear any protective headgear (P = 0.0007). Head injury followed by chest and abdominal trauma were found to predict a reduced survival rate. Use of helmets should continue to be compulsory. Chest and abdominal injuries should be diagnosed and treated early to reduce mortality. © 2002 Elsevier Science Ltd. All rights reserved.
1. Introduction Cumulative death rates following motorcycle injuries have shown a decline in UK and Wales between 1985 and 1995 [1]. However, the road accident statistics from the Department of Environment, Transport and the Regions shows that the motorcycle riders and passengers had a casualty rate nearly fifteen times that of car occupants in the year 1997 and they account for 15% of those killed or seriously injured in 1998 [2]. It has been shown that the risk of fatal injuries in motorcycle crashes are related to the engine capacity of the motorcycle, the size of the vehicle collided against and the direction of collision [3]. This study examined the patterns of injury following motorcycle crashes to identify the factors that are likely to predict a reduced survival.
2. Materials and methods Data was collected retrospectively on all motorcycle accident victims who presented to various trauma units in the Yorkshire region of UK during the period from January 1993 to December 1999. The definition used by the Department of the Environment, Transport and the Regions [4] were used to identify fatal and serious injuries. Details of both drivers and ∗ Corresponding author. Tel.: +44-113-2065298; fax: +44-113-2065156. E-mail address: [email protected] (P.V. Giannoudis).
pillion riders involved in each crash were recorded. Severity of the injuries was scored using the abbreviated injury score (AIS98) [5]. Data collection included demographic details of patients, use of a protective device (helmet), clinical details, injury severity score (ISS), Glasgow coma scale (GCS) at the time of presentation in the emergency department, therapeutic interventions, resuscitation requirements, grade of medical personnel involved in the initial management, duration of hospital stay and mortality. A logistic regression model was used to assess the impact of independent variables on the survival rates. The variables included were head injury, chest injury, abdominal injury, spinal injury, pelvic injury and long bone fractures. A compound variable (pelvic injury associated with a long bone fracture) was also included in the model. Variables not predictive of reduced survival were excluded from the final analysis. The GCS were compared between the survivors and non-survivors with respect to the use of protective headgear, using Chi-square test. A P-value of <0.05 was considered significant. 3. Results Between January 1993 and December 1999, 1239 motorcycle crash victims were admitted to various trauma units in the Yorkshire region. There were 1120 males and 119 females (ratio 9:1), with a mean age of 30.7 years (range, 22–36 years). Fifty-one (4.1%) patients were pillion riders.
0020-1383/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 0 - 1 3 8 3 ( 0 2 ) 0 0 0 4 8 - 7
474
S. Ankarath et al. / Injury, Int. J. Care Injured 33 (2002) 473–477
Table 1 Injuries sustained by the motorcycle crash victimsa Injury sustained
Total number of injuries (n = 1239) (%)
Total number of injuries in the fatal group (n = 74) (%)
Percentage of injuries in the fatal group in relation to total number of injuries
Head and facial injuries Intracranial haemorrhage Cerebral contusion Cerebral laceration Fracture skull Fracture of facial bones
146 37 39 6 37 19
(11.8) (25.3) (26.7) (4.1) (25.3) (13)
42 21 2 6 12 6
(56.8) (50) (4.8) (14.2) (28.6) (14.3)
28.8 56.8 5.1 100 32.4 31.6
Chest injuries Lung contusion Lung lacerations Fracture ribs Haemothorax Pneumothorax Flail chest Myocardial contusion Myocardial tear Pericardial tear/haematoma Rupture of thoracic aorta Fracture of sternum
216 56 16 81 24 38 11 6 2 6 6 10
(17.4) (25.9) (7.4) (37.5) (11.1) (17.6) (5.1) (2.8) (0.9) (2.8) (2.8) (4.6)
24 24 8 20 19 7 5 3 2 3 6 2
(32.4) (100) (33.3) (83.3) (79.2) (29.2) (20.9) (12.5) (8.3) (12.5) (25) (8.3)
11.1 42.9 50 24.7 79.2 18.4 45.5 50 100 50 100 20
Abdominal injuries Liver laceration Splenic rupture Mesenteric tear Intestinal perforation Kidney rupture Renal contusion Injury to renal vessels Diaphragmatic tear Tear of inferior Vena cava Superior mesenteric artery tear Bladder rupture Urethral rupture
158 45 21 12 34 8 12 8 3 2 1 18 14
(12.8) (28.5) 13.3) (7.6) (21.5) (5.1) (7.6) (5.1) (1.9) (1.3) (0.6) (11.4) (8.9)
22 13 4 8 3 4 3 6 2 2 1 4 2
(29.8) (59.1) (18.2) (36.4) (13.6) (18.2) (13.6) (27.3) (9.1) (9.1) (4.5) (18.2) (9.1)
13.9 28.9 19 66.6 8.8 50 25 75 66.7 100 100 22.2 14.3
Spinal injuryb Cervical spine fracture Thoracic spine fracture Lumbar spine fracture
129 15 87 17
(10.4) (11.6) (67.4) (13.2)
21 8 13 5
(28.4) (38.1) (61.9) (23.8)
16.3 53.3 14.9 29.4
1168 154 198 4 348 128 43 63 89 23 87 31
(94.3) (13.2) (17) (0.3) (29.8) (11) (3.7) (5.4) (7.6) (2) (7.4) (2.7)
72 20 17 2 3 6 12 10 8 10 5 5
(97.3) (27.8) (23.6) (2.8) (4.2) (8.3) (16.7) (13.9) (11.1) (13.9) (6.9) (6.9)
6.2 13 8.6 50 0.9 4.7 27.9 15.9 9 43.5 5.7 16.1
Axial skeletal injuries Pelvic injuries Femoral shaft Neck of femur Tibial fracture Ankle fracture Forearm Humerus Clavicle Scapula Wrist Hands Crush injury Leg Arm
4 (0.3) 2 (50) 2 (50)
3 (4.1) 1 (33.3) 2 (66.7)
75 50 100
Peripheral arterial injury Subclavian artery External carotid
3 (0.2) 2 (66.7) 1 (33.3)
3 (4.1) 2 (66.7) 1 (33.3)
100 100 100
a Many patients had more than one injury to any one system and in some cases more than one system was involved. Hence, the total number of different injuries sustained by all the patients will be more than the total number of patients. b Two patients had complete paraplegia at initial presentation.
S. Ankarath et al. / Injury, Int. J. Care Injured 33 (2002) 473–477
All patients were initially assessed and resuscitated in the emergency department. The mean volume of crystalloid fluids used for resuscitation was 688 ml (range 0–2000 ml), while the mean blood transfusion required was 0.26 units per patient (95% confidence interval (CI) 0.18–0.34). One hundred and twenty-four (10%) patients did not wear a protective headgear. Seventy-four (6%) patients died from various injuries. The majority of the fatal patients had more than one system involved. The mean ISS in the fatal group was 35.96 (range, 9–75) compared to 12.2 (range, 4–66) in the group that survived (P = 0.0001). 3.1. Head and facial injuries Of the 146 (11.8%) patients with head and facial injuries, 37 (25%) had fracture of the skull and 39 (27%) had cerebral contusion (Table 1). There was no association between the presence of a skull fracture and intracranial bleeding (P = 0.81). The mean GCS of patients at the time of arrival in the emergency department was 13.7 (range, 3–15). Forty-two patients (56.8%) in this group died. This comprised 28.8% of all motorcycle accident victims (Table 1). Eight (19%) patients in the fatal group were not wearing a helmet. The mean GCS in the group not wearing protective headgear was 6 compared to 13 in the group that did (P = 0.0007). Twenty-one of the 37 (56.8%) patients with intracranial bleeding were fatal as also all patients with cerebral laceration. One of them with an unsurvivable ISS of 75 had an open fracture of skull extending to the base with brain herniation. 3.2. Chest injuries Chest trauma accounted for 17.4% (216 patients) of all injuries. Fifty-six (26%) patients in this group had lung contusion. There were 81 patients with fracture of one or more ribs. This included 24 patients with haemothorax (Table 1). The most common combination of injuries in the chest was
475
lung contusion and haemothorax (42%). Although haemothorax was seen only in 11% (24 patients) of all motorcycle accident victims, it was associated with 79% (19 patients) of fatal chest injuries. Out of the three fatal injuries with an ISS of 75, two had rupture of the thoracic aorta. All patients with myocardial tear (two patients) and 50% (three out of six) of patients with myocardial contusion died (Table 1). 3.3. Abdominal injuries Major abdominal trauma was seen in 158 patients (12.8%). Out of this 45 (28%) patients had liver injuries and 21 (13.3%) patients had a splenic tear. Renal laceration was present in eight patients out of which four (50%) died. There were 18 (11%) patients with bladder injuries and 14 patients (9%) with rupture of urethra. Renal tract injuries (21 patients) accounted for a large propotion of intra-abdominal injuries in the fatal group (40%). Hepatic laceration were seen in 59% (13 patients) of fatal abdominal trauma compared to 19% of splenic injuries (Table 1). Both patients with tear of the inferior vena cava and one patient with superior mesneteric artery tear died. 3.4. Spinal injuries Of the 129 (10.4%) patients with spinal injuries, 87 (67.4%) involved the lower thoracic and 17 (13.2%) upper lumbar levels. Two patients had complete paraplegia on initial presentation, both following fracture dislocation of the thoracic spine. Mortality was seen only in 15% (13 out of 87) of thoracic spine fractures when compared to 53.3% (8 out of 15 patients) in cervical spine injuries (Table 1). 3.5. Axial skeletal injuries The most common musculoskeletal injury was fracture of tibia (348 patients (29.8%)), followed by femoral fracture (198 patients (17%)). There were 125 (36%) open tibial
Fig. 1. Odds ratio for factors related to mortality.
476
S. Ankarath et al. / Injury, Int. J. Care Injured 33 (2002) 473–477
fractures. However, femoral fracture was the commonest long bone injury in the fatal group (17 patients) whereas only three patients with tibial fractures died. There were 154 (13.2%) patients with a pelvic ring injury out of which 20 (13%) were fatal. Fifteen of these fatal pelvic injuries were antero-posterior compression type. 3.6. Other injuries Three had injury to neck leading to vascular damage. There were no survivors in the two patients with injury to the subclavian artery and in one patient with external carotid artery tear. 3.7. Odds ratio Odds ratio calculated for variables that were likely predict reduced survival were as follows: head injury 0.349 (95% CI, 0.28–0.43), chest injury 0.39 (95% CI, 0.28–0.56), abdominal injury 0.42 (95% CI, 0.29–0.6) and the compound variable (pelvis plus a long bone fracture) 0.576 (95% CI, 0.41–0.80) (Fig. 1).
4. Discussion Motorcycle crashes continue to be a major cause of fatal road traffic injuries. In this study, an average of 170 motorcycle crash victims required hospital admission per year in the Yorkshire region alone. Head and facial trauma followed by thoracic cage trauma were the largest group of injuries in the fatal group in this series. This is in agreement with the findings of Wyatt et al. [6] who noted that injuries to the head, neck and chest were responsible for most severe injuries. Patients with unsurvivable ISS score of 75 in their series, were found to have significant trauma to brain stem, cervical spinal cord and thoracic aorta, and the majority of the patients were dead at the accident scene. This led the authors to conclude that measures to prevent injury are more important than improved treatment. Intracranial haemorrhage was the single most important predictor of mortality in patients with head injury in this series, other than a cerebral laceration. All three patients with an ISS score of 75 had significant trauma to the head and thoracic aorta. Doyle et al. [7] looking at the results of 96 motorcycle injuries over a period of 1 year reported that although the legs, arms, head and thorax were the commonest regions affected, injury to head and thorax were more likely to be associated with fatal outcomes. Unhelmeted rides are more likely to be critically injured and have a head injury compared to their helmeted counterparts [8,9]. The number of rides not wearing helmets in this series was too small to show any significant difference in the incidence of head injury in the two groups. This probably reflects the compliance of safety helmet usage among motorcyclists in the UK. The significant difference in the GCS
between the patients wearing helmets and the unhelmeted group supports the continued use of a protective headgear. We believe that the high incidence of lung contusion in fatal chest injuries is only reflection of the severity of the injury to the thoracic region. A very high index of suspicion has to be maintained to rule out associated trauma to the myocardium and thoracic aorta, both of which are associated with significant mortality rate. Interestingly, our study also shows a high mortality among patients who had a pericardial tear or haematoma. This may be an incidental association, as the degree of external force applied to the thoracic cage to cause these injuries has to of such a magnitude that it invariably will be associated with injury to other systems. We found that intra-abdominal injury was common in fatal motorcycle crashes. However, it was weaker than head and chest injuries as a predict of reduced survival. Most patients had more than one injury in the abdomen, making it difficult to identify the main cause of death. Almost all deaths from intra-abdominal trauma were secondary to bleeding. Liver injuries were noted in the majority of fatal accidents with abdominal trauma, but only a quarter of these patients died. In contrast, patients with renal tract injuries were found to have a high mortality reaching upto 50% with renal rupture and 75% with injury to the renal vessals (Table 1). This observation may be biased as the increased mortality can secondary to other associated more serious intra-abdominal injuries. The degree of trauma required to produce a retroperitoneal injury can be efficiently high to cause damage to adjacent organs. This is a weakness of this study. Many patients had more than one injury to each system and more than one system involved. Our study did not look specifically at the different combination of injuries in each fatal crash. Hence, we were unable to identify the primary cause of death in many patients. The lower thoracic and upper lumbar area of the spine was the most common site of involvement. Thoracic spine fractures are more likely to be associated with neurological damage, but mortality is more highly associated with cervical spine involvement. Long bone fractures were the commonest injury in the whole group. This is in agreement with a previous study by Bradbury and Robertson [10]. As an independent variable, this was not found to have a significant influence on mortality. Although there were long bone fractures in the fatal group that would have compounded the effect of the injury, none of them were considered severe enough to affect the mortality rate on their own. When long bone fracture was combined with pelvic injury, the odds ratio in predicting the probability of survival was less than head, chest or abdominal injury. Motorcycle crashes continue to be a source of severe injuries. This study shows that the most important variable affecting mortality in motorcycle crashes is head injury. Wearing a helmet should continue to be compulsory, as this single measure can help to reduce the mortality in motorcyclists. Thoracic and abdominal trauma as well as pelvic
S. Ankarath et al. / Injury, Int. J. Care Injured 33 (2002) 473–477
ring fractures associated with long bone injuries appears to be the secondary factors contributing to reduced survival. These should be looked for and excluded early in every motorcycle crash victim. A delay in diagnosis and treatment can influence the final outcome significantly. Acknowledgements S. Ankarath was the principal author of the paper and wrote the first and final draft of the paper and analysed the data. P.V. Giannoudis designed and supervised the study and helped to perform the analysis. I. Barlow and M.C. Bellamy helped in data collection and analysis. S.J. Matthews and R.M. Smith contributed to the interpretation and drafting of the paper. P.V. Giannoudis will act as overall guarantor for the paper. References [1] DiGuiseppi C, Leah L, Roberts I. Influence of travel patterns on mortality from injury among teenagers in UK and Wales, 1985–1995: trend analysis. BMJ 1998;316:904–5.
477
[2] Department of Environment, Transport and the Regions. Transport Statistics: Motorcycle Accidents: Great Britain, 1998; London, 2000. [3] Pang TY, Umar RS, Azhar AA, Ahmad MM, Nasir MT, Harwant S. Accident characteristics of injured motorcyclists in Malaysia. Med J Malaysia 2000;55(1):45–50. [4] Department of the Environment, Transport and the Regions; Royal Ulster Constabulary—Fatal and Serious Road Accidents: Regional Trends Dataset RT351009. [5] American Association for Automotive Medicine. The Abbreviated Injury Scale (AIS), 1998 Revision. Des Plaines Illinois AAAM, 1990. [6] Wyatt JP, O’Donnell J, Beard D, Busuttil A. Injury analyses of fatal motorcycle collisions in south-east Scotland. Forensic Sci Int 1999;104:127–32. [7] Doyle D, Muir M, Chinn B. Motorcycle accidents in strathclyde region, Scotland during 1992: a study of the injuries sustained. Health Bull (Edinburgh) 1995;53(6):386–94. [8] Rowland J, Rivara F, Salzberg R, Maier R, Koepsell T. Motorcycle helmet use and injury outcome and hospitalisation costs from crashes in Washington state. Am J Public Health 1996;86(1):41–5. [9] Sarkar S, Peck C, Kraus JF. Fatal injuries in motorcycle riders according to helmet use. J Trauma 1995;38(2):242–5. [10] Bradbury A, Robertson C. Pattern and severity of injury sustained by motorcyclists in road traffic accidents in Edinburgh, Scotland. Health Bull (Edinburgh) 1993;51(2):86–91.
Injury patterns associated with mortality following motorcycle crashes S. Ankarath, P.V. Giannoudis∗ , I. Barlow, M.C. Bellamy, S.J. Matthews, R.M. Smith Department of Trauma and Orthopaedics, St. James’s University Hospital, Leeds LS9 7TF, UK Accepted 12 February 2002
Abstract All patients involved in motorcycle crashes admitted to various hospitals in the Yorkshire region of UK between January 1993 and December 1999 were retrospectively reviewed to identify the factors that are likely to predict a reduced survival. Of the 1239 patients requiring hospital admission, 74 died. The probability of reduced survival was estimated by a logistic regression model using independent variables such as head injury, thoracic trauma, abdominal injury, spinal injury and pelvic fracture and a compound variable of pelvic fracture combined with a long bone fracture. The odds ratio for head injury was 0.349, chest injury 0.39, abdominal injury 0.42, and the compound variable (pelvis plus a long bone fracture) 0.576. The mean injury severity score (ISS) in the fatal group was 35.96 compared to 12.2 in the group that survived (P < 0.01). There was a significant difference in the Glasgow coma scale (GCS) between patients wearing a helmet and those that did not wear any protective headgear (P = 0.0007). Head injury followed by chest and abdominal trauma were found to predict a reduced survival rate. Use of helmets should continue to be compulsory. Chest and abdominal injuries should be diagnosed and treated early to reduce mortality. © 2002 Elsevier Science Ltd. All rights reserved.
1. Introduction Cumulative death rates following motorcycle injuries have shown a decline in UK and Wales between 1985 and 1995 [1]. However, the road accident statistics from the Department of Environment, Transport and the Regions shows that the motorcycle riders and passengers had a casualty rate nearly fifteen times that of car occupants in the year 1997 and they account for 15% of those killed or seriously injured in 1998 [2]. It has been shown that the risk of fatal injuries in motorcycle crashes are related to the engine capacity of the motorcycle, the size of the vehicle collided against and the direction of collision [3]. This study examined the patterns of injury following motorcycle crashes to identify the factors that are likely to predict a reduced survival.
2. Materials and methods Data was collected retrospectively on all motorcycle accident victims who presented to various trauma units in the Yorkshire region of UK during the period from January 1993 to December 1999. The definition used by the Department of the Environment, Transport and the Regions [4] were used to identify fatal and serious injuries. Details of both drivers and ∗ Corresponding author. Tel.: +44-113-2065298; fax: +44-113-2065156. E-mail address: [email protected] (P.V. Giannoudis).
pillion riders involved in each crash were recorded. Severity of the injuries was scored using the abbreviated injury score (AIS98) [5]. Data collection included demographic details of patients, use of a protective device (helmet), clinical details, injury severity score (ISS), Glasgow coma scale (GCS) at the time of presentation in the emergency department, therapeutic interventions, resuscitation requirements, grade of medical personnel involved in the initial management, duration of hospital stay and mortality. A logistic regression model was used to assess the impact of independent variables on the survival rates. The variables included were head injury, chest injury, abdominal injury, spinal injury, pelvic injury and long bone fractures. A compound variable (pelvic injury associated with a long bone fracture) was also included in the model. Variables not predictive of reduced survival were excluded from the final analysis. The GCS were compared between the survivors and non-survivors with respect to the use of protective headgear, using Chi-square test. A P-value of <0.05 was considered significant. 3. Results Between January 1993 and December 1999, 1239 motorcycle crash victims were admitted to various trauma units in the Yorkshire region. There were 1120 males and 119 females (ratio 9:1), with a mean age of 30.7 years (range, 22–36 years). Fifty-one (4.1%) patients were pillion riders.
0020-1383/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 0 - 1 3 8 3 ( 0 2 ) 0 0 0 4 8 - 7
474
S. Ankarath et al. / Injury, Int. J. Care Injured 33 (2002) 473–477
Table 1 Injuries sustained by the motorcycle crash victimsa Injury sustained
Total number of injuries (n = 1239) (%)
Total number of injuries in the fatal group (n = 74) (%)
Percentage of injuries in the fatal group in relation to total number of injuries
Head and facial injuries Intracranial haemorrhage Cerebral contusion Cerebral laceration Fracture skull Fracture of facial bones
146 37 39 6 37 19
(11.8) (25.3) (26.7) (4.1) (25.3) (13)
42 21 2 6 12 6
(56.8) (50) (4.8) (14.2) (28.6) (14.3)
28.8 56.8 5.1 100 32.4 31.6
Chest injuries Lung contusion Lung lacerations Fracture ribs Haemothorax Pneumothorax Flail chest Myocardial contusion Myocardial tear Pericardial tear/haematoma Rupture of thoracic aorta Fracture of sternum
216 56 16 81 24 38 11 6 2 6 6 10
(17.4) (25.9) (7.4) (37.5) (11.1) (17.6) (5.1) (2.8) (0.9) (2.8) (2.8) (4.6)
24 24 8 20 19 7 5 3 2 3 6 2
(32.4) (100) (33.3) (83.3) (79.2) (29.2) (20.9) (12.5) (8.3) (12.5) (25) (8.3)
11.1 42.9 50 24.7 79.2 18.4 45.5 50 100 50 100 20
Abdominal injuries Liver laceration Splenic rupture Mesenteric tear Intestinal perforation Kidney rupture Renal contusion Injury to renal vessels Diaphragmatic tear Tear of inferior Vena cava Superior mesenteric artery tear Bladder rupture Urethral rupture
158 45 21 12 34 8 12 8 3 2 1 18 14
(12.8) (28.5) 13.3) (7.6) (21.5) (5.1) (7.6) (5.1) (1.9) (1.3) (0.6) (11.4) (8.9)
22 13 4 8 3 4 3 6 2 2 1 4 2
(29.8) (59.1) (18.2) (36.4) (13.6) (18.2) (13.6) (27.3) (9.1) (9.1) (4.5) (18.2) (9.1)
13.9 28.9 19 66.6 8.8 50 25 75 66.7 100 100 22.2 14.3
Spinal injuryb Cervical spine fracture Thoracic spine fracture Lumbar spine fracture
129 15 87 17
(10.4) (11.6) (67.4) (13.2)
21 8 13 5
(28.4) (38.1) (61.9) (23.8)
16.3 53.3 14.9 29.4
1168 154 198 4 348 128 43 63 89 23 87 31
(94.3) (13.2) (17) (0.3) (29.8) (11) (3.7) (5.4) (7.6) (2) (7.4) (2.7)
72 20 17 2 3 6 12 10 8 10 5 5
(97.3) (27.8) (23.6) (2.8) (4.2) (8.3) (16.7) (13.9) (11.1) (13.9) (6.9) (6.9)
6.2 13 8.6 50 0.9 4.7 27.9 15.9 9 43.5 5.7 16.1
Axial skeletal injuries Pelvic injuries Femoral shaft Neck of femur Tibial fracture Ankle fracture Forearm Humerus Clavicle Scapula Wrist Hands Crush injury Leg Arm
4 (0.3) 2 (50) 2 (50)
3 (4.1) 1 (33.3) 2 (66.7)
75 50 100
Peripheral arterial injury Subclavian artery External carotid
3 (0.2) 2 (66.7) 1 (33.3)
3 (4.1) 2 (66.7) 1 (33.3)
100 100 100
a Many patients had more than one injury to any one system and in some cases more than one system was involved. Hence, the total number of different injuries sustained by all the patients will be more than the total number of patients. b Two patients had complete paraplegia at initial presentation.
S. Ankarath et al. / Injury, Int. J. Care Injured 33 (2002) 473–477
All patients were initially assessed and resuscitated in the emergency department. The mean volume of crystalloid fluids used for resuscitation was 688 ml (range 0–2000 ml), while the mean blood transfusion required was 0.26 units per patient (95% confidence interval (CI) 0.18–0.34). One hundred and twenty-four (10%) patients did not wear a protective headgear. Seventy-four (6%) patients died from various injuries. The majority of the fatal patients had more than one system involved. The mean ISS in the fatal group was 35.96 (range, 9–75) compared to 12.2 (range, 4–66) in the group that survived (P = 0.0001). 3.1. Head and facial injuries Of the 146 (11.8%) patients with head and facial injuries, 37 (25%) had fracture of the skull and 39 (27%) had cerebral contusion (Table 1). There was no association between the presence of a skull fracture and intracranial bleeding (P = 0.81). The mean GCS of patients at the time of arrival in the emergency department was 13.7 (range, 3–15). Forty-two patients (56.8%) in this group died. This comprised 28.8% of all motorcycle accident victims (Table 1). Eight (19%) patients in the fatal group were not wearing a helmet. The mean GCS in the group not wearing protective headgear was 6 compared to 13 in the group that did (P = 0.0007). Twenty-one of the 37 (56.8%) patients with intracranial bleeding were fatal as also all patients with cerebral laceration. One of them with an unsurvivable ISS of 75 had an open fracture of skull extending to the base with brain herniation. 3.2. Chest injuries Chest trauma accounted for 17.4% (216 patients) of all injuries. Fifty-six (26%) patients in this group had lung contusion. There were 81 patients with fracture of one or more ribs. This included 24 patients with haemothorax (Table 1). The most common combination of injuries in the chest was
475
lung contusion and haemothorax (42%). Although haemothorax was seen only in 11% (24 patients) of all motorcycle accident victims, it was associated with 79% (19 patients) of fatal chest injuries. Out of the three fatal injuries with an ISS of 75, two had rupture of the thoracic aorta. All patients with myocardial tear (two patients) and 50% (three out of six) of patients with myocardial contusion died (Table 1). 3.3. Abdominal injuries Major abdominal trauma was seen in 158 patients (12.8%). Out of this 45 (28%) patients had liver injuries and 21 (13.3%) patients had a splenic tear. Renal laceration was present in eight patients out of which four (50%) died. There were 18 (11%) patients with bladder injuries and 14 patients (9%) with rupture of urethra. Renal tract injuries (21 patients) accounted for a large propotion of intra-abdominal injuries in the fatal group (40%). Hepatic laceration were seen in 59% (13 patients) of fatal abdominal trauma compared to 19% of splenic injuries (Table 1). Both patients with tear of the inferior vena cava and one patient with superior mesneteric artery tear died. 3.4. Spinal injuries Of the 129 (10.4%) patients with spinal injuries, 87 (67.4%) involved the lower thoracic and 17 (13.2%) upper lumbar levels. Two patients had complete paraplegia on initial presentation, both following fracture dislocation of the thoracic spine. Mortality was seen only in 15% (13 out of 87) of thoracic spine fractures when compared to 53.3% (8 out of 15 patients) in cervical spine injuries (Table 1). 3.5. Axial skeletal injuries The most common musculoskeletal injury was fracture of tibia (348 patients (29.8%)), followed by femoral fracture (198 patients (17%)). There were 125 (36%) open tibial
Fig. 1. Odds ratio for factors related to mortality.
476
S. Ankarath et al. / Injury, Int. J. Care Injured 33 (2002) 473–477
fractures. However, femoral fracture was the commonest long bone injury in the fatal group (17 patients) whereas only three patients with tibial fractures died. There were 154 (13.2%) patients with a pelvic ring injury out of which 20 (13%) were fatal. Fifteen of these fatal pelvic injuries were antero-posterior compression type. 3.6. Other injuries Three had injury to neck leading to vascular damage. There were no survivors in the two patients with injury to the subclavian artery and in one patient with external carotid artery tear. 3.7. Odds ratio Odds ratio calculated for variables that were likely predict reduced survival were as follows: head injury 0.349 (95% CI, 0.28–0.43), chest injury 0.39 (95% CI, 0.28–0.56), abdominal injury 0.42 (95% CI, 0.29–0.6) and the compound variable (pelvis plus a long bone fracture) 0.576 (95% CI, 0.41–0.80) (Fig. 1).
4. Discussion Motorcycle crashes continue to be a major cause of fatal road traffic injuries. In this study, an average of 170 motorcycle crash victims required hospital admission per year in the Yorkshire region alone. Head and facial trauma followed by thoracic cage trauma were the largest group of injuries in the fatal group in this series. This is in agreement with the findings of Wyatt et al. [6] who noted that injuries to the head, neck and chest were responsible for most severe injuries. Patients with unsurvivable ISS score of 75 in their series, were found to have significant trauma to brain stem, cervical spinal cord and thoracic aorta, and the majority of the patients were dead at the accident scene. This led the authors to conclude that measures to prevent injury are more important than improved treatment. Intracranial haemorrhage was the single most important predictor of mortality in patients with head injury in this series, other than a cerebral laceration. All three patients with an ISS score of 75 had significant trauma to the head and thoracic aorta. Doyle et al. [7] looking at the results of 96 motorcycle injuries over a period of 1 year reported that although the legs, arms, head and thorax were the commonest regions affected, injury to head and thorax were more likely to be associated with fatal outcomes. Unhelmeted rides are more likely to be critically injured and have a head injury compared to their helmeted counterparts [8,9]. The number of rides not wearing helmets in this series was too small to show any significant difference in the incidence of head injury in the two groups. This probably reflects the compliance of safety helmet usage among motorcyclists in the UK. The significant difference in the GCS
between the patients wearing helmets and the unhelmeted group supports the continued use of a protective headgear. We believe that the high incidence of lung contusion in fatal chest injuries is only reflection of the severity of the injury to the thoracic region. A very high index of suspicion has to be maintained to rule out associated trauma to the myocardium and thoracic aorta, both of which are associated with significant mortality rate. Interestingly, our study also shows a high mortality among patients who had a pericardial tear or haematoma. This may be an incidental association, as the degree of external force applied to the thoracic cage to cause these injuries has to of such a magnitude that it invariably will be associated with injury to other systems. We found that intra-abdominal injury was common in fatal motorcycle crashes. However, it was weaker than head and chest injuries as a predict of reduced survival. Most patients had more than one injury in the abdomen, making it difficult to identify the main cause of death. Almost all deaths from intra-abdominal trauma were secondary to bleeding. Liver injuries were noted in the majority of fatal accidents with abdominal trauma, but only a quarter of these patients died. In contrast, patients with renal tract injuries were found to have a high mortality reaching upto 50% with renal rupture and 75% with injury to the renal vessals (Table 1). This observation may be biased as the increased mortality can secondary to other associated more serious intra-abdominal injuries. The degree of trauma required to produce a retroperitoneal injury can be efficiently high to cause damage to adjacent organs. This is a weakness of this study. Many patients had more than one injury to each system and more than one system involved. Our study did not look specifically at the different combination of injuries in each fatal crash. Hence, we were unable to identify the primary cause of death in many patients. The lower thoracic and upper lumbar area of the spine was the most common site of involvement. Thoracic spine fractures are more likely to be associated with neurological damage, but mortality is more highly associated with cervical spine involvement. Long bone fractures were the commonest injury in the whole group. This is in agreement with a previous study by Bradbury and Robertson [10]. As an independent variable, this was not found to have a significant influence on mortality. Although there were long bone fractures in the fatal group that would have compounded the effect of the injury, none of them were considered severe enough to affect the mortality rate on their own. When long bone fracture was combined with pelvic injury, the odds ratio in predicting the probability of survival was less than head, chest or abdominal injury. Motorcycle crashes continue to be a source of severe injuries. This study shows that the most important variable affecting mortality in motorcycle crashes is head injury. Wearing a helmet should continue to be compulsory, as this single measure can help to reduce the mortality in motorcyclists. Thoracic and abdominal trauma as well as pelvic
S. Ankarath et al. / Injury, Int. J. Care Injured 33 (2002) 473–477
ring fractures associated with long bone injuries appears to be the secondary factors contributing to reduced survival. These should be looked for and excluded early in every motorcycle crash victim. A delay in diagnosis and treatment can influence the final outcome significantly. Acknowledgements S. Ankarath was the principal author of the paper and wrote the first and final draft of the paper and analysed the data. P.V. Giannoudis designed and supervised the study and helped to perform the analysis. I. Barlow and M.C. Bellamy helped in data collection and analysis. S.J. Matthews and R.M. Smith contributed to the interpretation and drafting of the paper. P.V. Giannoudis will act as overall guarantor for the paper. References [1] DiGuiseppi C, Leah L, Roberts I. Influence of travel patterns on mortality from injury among teenagers in UK and Wales, 1985–1995: trend analysis. BMJ 1998;316:904–5.
477
[2] Department of Environment, Transport and the Regions. Transport Statistics: Motorcycle Accidents: Great Britain, 1998; London, 2000. [3] Pang TY, Umar RS, Azhar AA, Ahmad MM, Nasir MT, Harwant S. Accident characteristics of injured motorcyclists in Malaysia. Med J Malaysia 2000;55(1):45–50. [4] Department of the Environment, Transport and the Regions; Royal Ulster Constabulary—Fatal and Serious Road Accidents: Regional Trends Dataset RT351009. [5] American Association for Automotive Medicine. The Abbreviated Injury Scale (AIS), 1998 Revision. Des Plaines Illinois AAAM, 1990. [6] Wyatt JP, O’Donnell J, Beard D, Busuttil A. Injury analyses of fatal motorcycle collisions in south-east Scotland. Forensic Sci Int 1999;104:127–32. [7] Doyle D, Muir M, Chinn B. Motorcycle accidents in strathclyde region, Scotland during 1992: a study of the injuries sustained. Health Bull (Edinburgh) 1995;53(6):386–94. [8] Rowland J, Rivara F, Salzberg R, Maier R, Koepsell T. Motorcycle helmet use and injury outcome and hospitalisation costs from crashes in Washington state. Am J Public Health 1996;86(1):41–5. [9] Sarkar S, Peck C, Kraus JF. Fatal injuries in motorcycle riders according to helmet use. J Trauma 1995;38(2):242–5. [10] Bradbury A, Robertson C. Pattern and severity of injury sustained by motorcyclists in road traffic accidents in Edinburgh, Scotland. Health Bull (Edinburgh) 1993;51(2):86–91.