bs_bs_banner
Emergency Medicine Australasia (2015) 27, 323–327
doi: 10.1111/1742-6723.12407
ORIGINAL RESEARCH
Mechanisms, injuries and helmet use in cyclists presenting to an inner city emergency department Michael M DINH,1,2,3 Christopher KASTELEIN,1,4 Roy HOPKINS,5 Timothy J ROYLE,1 Kendall J BEIN,1 Dane R CHALKLEY1 and Rebecca IVERS3 1 Department of Trauma Services, Royal Prince Alfred Hospital, Sydney, New South Wales, Australia, 2Emergency Department, Royal Prince Alfred Hospital, Sydney, New South Wales, Australia, 3The George Institute for Global Health, Sydney Medical School, Sydney, New South Wales, Australia, 4Sydney Nursing School, The University of Sydney, Sydney, New South Wales, Australia, and 5Division of Surgery, Royal Prince Alfred Hospital, Sydney, New South Wales, Australia
Abstract Objectives: The objectives of the present study were to describe the injury profiles of cyclists presenting to an ED and determine the risk of significant head injury associated with bicycle helmet use. Methods: This was a retrospective single trauma centre study of all adult cyclists presenting to an inner city ED and undergoing a trauma team review between January 2012 and June 2014. The outcome of interest was significant head injury defined as any head injury with an Abbreviated Injury Scale score of two or more. Variables analysed included demographic characteristics, helmet use at time of incident, location, time and the presence of intoxication. Results: The most common body regions were upper limb injuries (57%), followed by head injuries (43%), facial injuries (30%) and lower limb injuries (24%). A lower proportion of people wearing helmets had significant head injury (17% vs 31%, P = 0.018) or facial injury (26% vs 48%, P = 0.0017) compared with nonhelmet users. After adjustment for im-
portant covariates, helmet use was associated with a 70% decrease in the odds of significant head injury (odds ratio 0.34, 95% confidence interval 0.15, 0.76, P = 0.008). Conclusions: Head injuries were common after inner city cycling incidents. The use of helmets was associated with a reduction in significant head injury. Key words: cycling, helmets, road trauma.
Introduction Cycling participation rates have been increasing in many inner city and metropolitan areas in Australia over the past decade1,2 particularly in the inner Sydney area where journeys to work by bicycle have increased around 25% between 2001 and 2011.3 Debate continues around measures designed to improve the safety of cycling in particular the implementation of mandatory helmet laws across Australia in the 1990s.3,4 Several studies have investigated the role of helmets in reducing head injuries in injured cyclists. A Cochrane
Correspondence: Clinical Associate Professor Michael M Dinh, Department of Trauma Services, Royal Prince Alfred Hospital, Missenden Road, Camperdown, NSW 2050, Australia. Email: [email protected] Michael M Dinh, MBBS, MPH, Co-Director of Trauma; Christopher Kastelein, BN, Trauma Research Nurse; Roy Hopkins, MBBS, Senior Registrar; Timothy J Royle, MBBS, Emergency Physician; Kendall J Bein, MBBS, Emergency Physician; Dane R Chalkley, MBBS, BSc (Hons), Emergency Physician; Rebecca Ivers, MPH, PhD, Director of Injury Division. Accepted 7 April 2015
Key findings • Upper limb and head injuries were common in cyclists presenting to EDs. • Helmet use was associated with a substantial reduction in the risk of significant head injuries in cyclists.
Collaboration review published in 2009 reviewed five case–controlled studies and found that helmet use was associated with a 63–88% reduction in risk of head injuries regardless of the mechanism of injury.5 Since the work of McDermott et al.,6 who demonstrated an association between helmet use and reduction in head injuries, further trauma related studies in Australia have been limited. Olivier et al.7 postulated that bicycle helmet laws had reduced the rate of head injuries in NSW by demonstrating that head injuries had decreased relative to arm injuries since the introduction of mandatory helmet laws. A retrospective study from Western Australia found a reduced odds of any head injury associated with helmet use although this was not found to be statistically significant (P = 0.06).8 Other studies have suggested that evidence of intoxication is associated with nonhelmet use and could therefore be a potential confounder in the association between helmet use and head injuries.9 Most trauma studies involving cyclists to date have used trauma registry or administrative databases that have included only admitted
© 2015 Australasian College for Emergency Medicine and Australasian Society for Emergency Medicine
324
MM DINH ET AL.
patients with particular injuries, excluding the majority of hospitalised cyclists who are subsequently discharged from the ED.6,10 It would therefore be useful to describe injury profiles and the association between helmet use and head injuries in a cohort of patients that includes those with minor injuries and those discharged from ED. The specific objectives of the present study were to describe the injury profiles of cyclists presenting to an inner city ED and determine associations between significant head injury and bicycle helmet use.
Study outcome The outcome of interest was significant head injury defined as a coded injury with an Abbreviated Injury Scale (AIS) score in the head region of two or more.12 This included concussion (any loss of consciousness > 1 min) skull fracture or intracranial haemorrhage. Injuries to other body regions (face, chest, abdomen, upper and lower limb regions) were similarly categorised by AIS codes and used to describe the injury profile of the study population.
Variables (definitions)
Methods Design This was a retrospective trauma registry-based study at a single inner city major trauma centre in Australia.
Setting The major trauma service covers most of the inner-west and inner city Sydney. There are around 700 trauma presentations (requiring a trauma team assessment in the ED) to the hospital each year, of which 20% are related to bicycle incidents. The hospital trauma data registry has collected complete demographic, prehospital and hospital clinical data and incident details on all trauma patients presenting to the ED regardless of disposition since 2012.
Patient population (inclusion/ exclusion criteria) All adult (age 15 years and over) patients who received a trauma team assessment on arrival to the ED between January 2012 and June 2014. Patients were assessed by the trauma team in the ED if they met pre-specified and validated trauma triage criteria, which included any cyclists with evidence of severe or multi-region injury or significant mechanism of injury (collision with vehicle or fall at estimated speed >20 km/h).11 We included transfers from all hospitals within the local health district. Transfers from facilities outside the local health district were excluded.
Data on incident characteristics, such as helmet use (yes, no or unknown), intoxication and collision type (any collision with any road vehicle vs other mechanism) were routinely extracted from ambulance case records and structured trauma team assessment notes and formed part of mandatory data collection for the trauma registry. Incident location within the inner city was defined as being within the inner city if the incident postcode was within 5 km of the Central Business District, and daylight hours were defined as 0700–1800 during Australian Eastern Standard time and corrected for daylight saving periods. Injury Severity Scores, hospital utilisation data, such as disposition from the ED (discharge or admission intensive care unit admission) and inpatient length of stay in hospital were abstracted from the trauma data registry. Data variables and injury codes were routinely collected and scored by a trained data manager without prior knowledge of the study. Cases where helmet data were missing were crosschecked by the data manager a second time with paper and electronic medical records.
Statistical analysis Descriptive statistics were used to compare injury profiles and head injuries in the relevant study groups. To ascertain the odds ratio for significant head injury in non-helmet users, we used multivariable logistic regression adjusting for a priori defined variables – age, sex, mechanism of injury
(collision with vehicle versus other) incident location (inner city versus other) and incident time, consistent with previous work in this area.10
Ethics The study was approved by the Sydney Local Health District Human Research Ethics Committee.
Results Study population Over the study period 258 cases were identified. The mean age was 37 years (standard deviation 13) and 80% were male. The median ISS score was 5 (interquartile range 2–6) and 15% of the study group had an ISS of 12 or more. Table 1 summarises the baseline characteristics of the study group. Around two-thirds of incidents occurred during daylight hours and over half of patients were discharged from ED. There was one in-hospital death recorded during the study period.
Injury profile The most common body regions were upper limb injuries (57%), followed by head injuries (43%), facial injuries (30%) and lower limb injuries (24%). Sixteen percent had spine or vertebral column injuries and 3% had abdominal injuries. Mechanisms of injury that did not involve collision with a road vehicle were associated with increased head and chest injuries compared with collisions with road vehicles (Table 2). None of the patients with documented intoxication were involved in collisions with road vehicles.
Helmet use There were no cases where helmet use was unknown at the time of incident. Helmets were worn in 79% of cases. Non-helmet use was more frequent in inner city incidents and in intoxicated patients (Table 3). Helmet use was associated with decreased proportion of significant head injury (17% vs 31% P = 0.018), facial injury (26% vs 48%, P = 0.0017) compared with non-helmet users, respectively (Table 3). After adjusting for the
© 2015 Australasian College for Emergency Medicine and Australasian Society for Emergency Medicine
325
INJURIES AND BICYCLE HELMETS
TABLE 1.
Baseline characteristics of study population n = 258
Variable Age, mean (SD) (years) Male (%) Injury severity score ≥ 12 (%) Intensive care unit admission (%) ED discharge (%) Length of stay for admitted patients (median IQR) In-hospital death (%) Inner city incident location (%) Intoxication (%) Helmet use (%) Daylight hours (%) Collision with road vehicle (%)
37 207 38 14 139 2 1 68 16 200 179 111
(13) (80) (15) (5) (54) (1–4) (0.4) (26) (6) (79) (69) (43)
IQR, interquartile range.
TABLE 2. Injury characteristics of cyclists involved in a collision with road vehicle versus cyclists involved in other mechanisms of injury
Age, mean (SD) (years) Male (%) Body region injured (%) Head Significant head injury Face Chest Abdomen Vertebral column Upper limb Lower limb ICU admission ISS 12 or more Inner city incident location Intoxication
Collision with road vehicle n = 111
No collision n = 147
P value
37 (11) 89 (80)
38 (14) 118 (80)
0.26 0.99
35 15 32 12 2 20 62 34 5 15 33 0
(32) (14) (29) (11) (2) (18) (56) (31) (5) (14) (30) (0)
77 36 46 30 7 9 85 30 6 23 35 16
(52) (24) (31) (20) (5) (6) (58) (20) (6) (16) (24) (11)
0.001 0.028 0.67 0.039 0.20 0.003 0.75 0.06 0.57 0.63 0.29 12 Any head injury Significant head injury Facial injury Chest injury Vertebral column injury Abdominal injury Upper limb injury Lower limb injury Intoxication Inner city incident location
39 158 11 26 79 34 52 32 24 8 123 52 4 45
No helmet n = 54
(13) (79) (6) (13) (40) (17) (26) (16) (12) (4) (62) (26) (2) (23)
31 45 3 12 32 17 26 9 4 1 22 12 11 22
(11) (83) (6) (22) (59) (31) (48) (17) (7) (2) (41) (22) (20) (41)
P value
Emergency Medicine Australasia (2015) 27, 323–327
doi: 10.1111/1742-6723.12407
ORIGINAL RESEARCH
Mechanisms, injuries and helmet use in cyclists presenting to an inner city emergency department Michael M DINH,1,2,3 Christopher KASTELEIN,1,4 Roy HOPKINS,5 Timothy J ROYLE,1 Kendall J BEIN,1 Dane R CHALKLEY1 and Rebecca IVERS3 1 Department of Trauma Services, Royal Prince Alfred Hospital, Sydney, New South Wales, Australia, 2Emergency Department, Royal Prince Alfred Hospital, Sydney, New South Wales, Australia, 3The George Institute for Global Health, Sydney Medical School, Sydney, New South Wales, Australia, 4Sydney Nursing School, The University of Sydney, Sydney, New South Wales, Australia, and 5Division of Surgery, Royal Prince Alfred Hospital, Sydney, New South Wales, Australia
Abstract Objectives: The objectives of the present study were to describe the injury profiles of cyclists presenting to an ED and determine the risk of significant head injury associated with bicycle helmet use. Methods: This was a retrospective single trauma centre study of all adult cyclists presenting to an inner city ED and undergoing a trauma team review between January 2012 and June 2014. The outcome of interest was significant head injury defined as any head injury with an Abbreviated Injury Scale score of two or more. Variables analysed included demographic characteristics, helmet use at time of incident, location, time and the presence of intoxication. Results: The most common body regions were upper limb injuries (57%), followed by head injuries (43%), facial injuries (30%) and lower limb injuries (24%). A lower proportion of people wearing helmets had significant head injury (17% vs 31%, P = 0.018) or facial injury (26% vs 48%, P = 0.0017) compared with nonhelmet users. After adjustment for im-
portant covariates, helmet use was associated with a 70% decrease in the odds of significant head injury (odds ratio 0.34, 95% confidence interval 0.15, 0.76, P = 0.008). Conclusions: Head injuries were common after inner city cycling incidents. The use of helmets was associated with a reduction in significant head injury. Key words: cycling, helmets, road trauma.
Introduction Cycling participation rates have been increasing in many inner city and metropolitan areas in Australia over the past decade1,2 particularly in the inner Sydney area where journeys to work by bicycle have increased around 25% between 2001 and 2011.3 Debate continues around measures designed to improve the safety of cycling in particular the implementation of mandatory helmet laws across Australia in the 1990s.3,4 Several studies have investigated the role of helmets in reducing head injuries in injured cyclists. A Cochrane
Correspondence: Clinical Associate Professor Michael M Dinh, Department of Trauma Services, Royal Prince Alfred Hospital, Missenden Road, Camperdown, NSW 2050, Australia. Email: [email protected] Michael M Dinh, MBBS, MPH, Co-Director of Trauma; Christopher Kastelein, BN, Trauma Research Nurse; Roy Hopkins, MBBS, Senior Registrar; Timothy J Royle, MBBS, Emergency Physician; Kendall J Bein, MBBS, Emergency Physician; Dane R Chalkley, MBBS, BSc (Hons), Emergency Physician; Rebecca Ivers, MPH, PhD, Director of Injury Division. Accepted 7 April 2015
Key findings • Upper limb and head injuries were common in cyclists presenting to EDs. • Helmet use was associated with a substantial reduction in the risk of significant head injuries in cyclists.
Collaboration review published in 2009 reviewed five case–controlled studies and found that helmet use was associated with a 63–88% reduction in risk of head injuries regardless of the mechanism of injury.5 Since the work of McDermott et al.,6 who demonstrated an association between helmet use and reduction in head injuries, further trauma related studies in Australia have been limited. Olivier et al.7 postulated that bicycle helmet laws had reduced the rate of head injuries in NSW by demonstrating that head injuries had decreased relative to arm injuries since the introduction of mandatory helmet laws. A retrospective study from Western Australia found a reduced odds of any head injury associated with helmet use although this was not found to be statistically significant (P = 0.06).8 Other studies have suggested that evidence of intoxication is associated with nonhelmet use and could therefore be a potential confounder in the association between helmet use and head injuries.9 Most trauma studies involving cyclists to date have used trauma registry or administrative databases that have included only admitted
© 2015 Australasian College for Emergency Medicine and Australasian Society for Emergency Medicine
324
MM DINH ET AL.
patients with particular injuries, excluding the majority of hospitalised cyclists who are subsequently discharged from the ED.6,10 It would therefore be useful to describe injury profiles and the association between helmet use and head injuries in a cohort of patients that includes those with minor injuries and those discharged from ED. The specific objectives of the present study were to describe the injury profiles of cyclists presenting to an inner city ED and determine associations between significant head injury and bicycle helmet use.
Study outcome The outcome of interest was significant head injury defined as a coded injury with an Abbreviated Injury Scale (AIS) score in the head region of two or more.12 This included concussion (any loss of consciousness > 1 min) skull fracture or intracranial haemorrhage. Injuries to other body regions (face, chest, abdomen, upper and lower limb regions) were similarly categorised by AIS codes and used to describe the injury profile of the study population.
Variables (definitions)
Methods Design This was a retrospective trauma registry-based study at a single inner city major trauma centre in Australia.
Setting The major trauma service covers most of the inner-west and inner city Sydney. There are around 700 trauma presentations (requiring a trauma team assessment in the ED) to the hospital each year, of which 20% are related to bicycle incidents. The hospital trauma data registry has collected complete demographic, prehospital and hospital clinical data and incident details on all trauma patients presenting to the ED regardless of disposition since 2012.
Patient population (inclusion/ exclusion criteria) All adult (age 15 years and over) patients who received a trauma team assessment on arrival to the ED between January 2012 and June 2014. Patients were assessed by the trauma team in the ED if they met pre-specified and validated trauma triage criteria, which included any cyclists with evidence of severe or multi-region injury or significant mechanism of injury (collision with vehicle or fall at estimated speed >20 km/h).11 We included transfers from all hospitals within the local health district. Transfers from facilities outside the local health district were excluded.
Data on incident characteristics, such as helmet use (yes, no or unknown), intoxication and collision type (any collision with any road vehicle vs other mechanism) were routinely extracted from ambulance case records and structured trauma team assessment notes and formed part of mandatory data collection for the trauma registry. Incident location within the inner city was defined as being within the inner city if the incident postcode was within 5 km of the Central Business District, and daylight hours were defined as 0700–1800 during Australian Eastern Standard time and corrected for daylight saving periods. Injury Severity Scores, hospital utilisation data, such as disposition from the ED (discharge or admission intensive care unit admission) and inpatient length of stay in hospital were abstracted from the trauma data registry. Data variables and injury codes were routinely collected and scored by a trained data manager without prior knowledge of the study. Cases where helmet data were missing were crosschecked by the data manager a second time with paper and electronic medical records.
Statistical analysis Descriptive statistics were used to compare injury profiles and head injuries in the relevant study groups. To ascertain the odds ratio for significant head injury in non-helmet users, we used multivariable logistic regression adjusting for a priori defined variables – age, sex, mechanism of injury
(collision with vehicle versus other) incident location (inner city versus other) and incident time, consistent with previous work in this area.10
Ethics The study was approved by the Sydney Local Health District Human Research Ethics Committee.
Results Study population Over the study period 258 cases were identified. The mean age was 37 years (standard deviation 13) and 80% were male. The median ISS score was 5 (interquartile range 2–6) and 15% of the study group had an ISS of 12 or more. Table 1 summarises the baseline characteristics of the study group. Around two-thirds of incidents occurred during daylight hours and over half of patients were discharged from ED. There was one in-hospital death recorded during the study period.
Injury profile The most common body regions were upper limb injuries (57%), followed by head injuries (43%), facial injuries (30%) and lower limb injuries (24%). Sixteen percent had spine or vertebral column injuries and 3% had abdominal injuries. Mechanisms of injury that did not involve collision with a road vehicle were associated with increased head and chest injuries compared with collisions with road vehicles (Table 2). None of the patients with documented intoxication were involved in collisions with road vehicles.
Helmet use There were no cases where helmet use was unknown at the time of incident. Helmets were worn in 79% of cases. Non-helmet use was more frequent in inner city incidents and in intoxicated patients (Table 3). Helmet use was associated with decreased proportion of significant head injury (17% vs 31% P = 0.018), facial injury (26% vs 48%, P = 0.0017) compared with non-helmet users, respectively (Table 3). After adjusting for the
© 2015 Australasian College for Emergency Medicine and Australasian Society for Emergency Medicine
325
INJURIES AND BICYCLE HELMETS
TABLE 1.
Baseline characteristics of study population n = 258
Variable Age, mean (SD) (years) Male (%) Injury severity score ≥ 12 (%) Intensive care unit admission (%) ED discharge (%) Length of stay for admitted patients (median IQR) In-hospital death (%) Inner city incident location (%) Intoxication (%) Helmet use (%) Daylight hours (%) Collision with road vehicle (%)
37 207 38 14 139 2 1 68 16 200 179 111
(13) (80) (15) (5) (54) (1–4) (0.4) (26) (6) (79) (69) (43)
IQR, interquartile range.
TABLE 2. Injury characteristics of cyclists involved in a collision with road vehicle versus cyclists involved in other mechanisms of injury
Age, mean (SD) (years) Male (%) Body region injured (%) Head Significant head injury Face Chest Abdomen Vertebral column Upper limb Lower limb ICU admission ISS 12 or more Inner city incident location Intoxication
Collision with road vehicle n = 111
No collision n = 147
P value
37 (11) 89 (80)
38 (14) 118 (80)
0.26 0.99
35 15 32 12 2 20 62 34 5 15 33 0
(32) (14) (29) (11) (2) (18) (56) (31) (5) (14) (30) (0)
77 36 46 30 7 9 85 30 6 23 35 16
(52) (24) (31) (20) (5) (6) (58) (20) (6) (16) (24) (11)
0.001 0.028 0.67 0.039 0.20 0.003 0.75 0.06 0.57 0.63 0.29 12 Any head injury Significant head injury Facial injury Chest injury Vertebral column injury Abdominal injury Upper limb injury Lower limb injury Intoxication Inner city incident location
39 158 11 26 79 34 52 32 24 8 123 52 4 45
No helmet n = 54
(13) (79) (6) (13) (40) (17) (26) (16) (12) (4) (62) (26) (2) (23)
31 45 3 12 32 17 26 9 4 1 22 12 11 22
(11) (83) (6) (22) (59) (31) (48) (17) (7) (2) (41) (22) (20) (41)
P value
