Original Article
Cohort Study on the Association Between Helmet Use and Traumatic Brain Injury in Snowboarders From a Swiss Tertiary Trauma Center Rebecca M. Hasler1, Dominik Baschera2, David Taugwalder1, Aristomenis K. Exadaktylos1, Andreas Raabe2
BACKGROUND: Since the introduction of helmets in winter sports there is on-going debate on whether they decrease traumatic brain injuries (TBI).
-
METHODS: This cohort study included 117 adult (‡16 years) snowboarders with TBI admitted to a level I alpine trauma center in Switzerland between 2000/2001 and 2010/ 2011. The primary objective was to examine the association between helmet use and moderate-to-severe TBI. Secondary objectives were to describe the epidemiology of TBI during the past decade in relation to increased helmet use.
TBI despite helmet use and the dangers of riding off-piste should be a focus of future prevention programs.
-
RESULTS: Of 691 injured snowboarders evaluated, 117 (17%) suffered TBI. Sixty-six percent were men (median age, 23 years). Two percent of accidents were fatal. Ninetytwo percent of patients sustained minor, 1% moderate, and 7% severe TBI according to the Glasgow coma scale. Pathologic computed tomography findings were present in 16% of patients, 26% of which required surgery. Eightythree percent of TBIs occurred while riding on-slope. There was no trend in the TBI rate during the studied period, although helmet use increased from 10% to 69%. Comparing patients with and without a helmet showed no significant difference in odds ratios for the severity of TBI. However, of the 5 patients requiring surgery only 1 was wearing a helmet. Off-piste compared with on-slope snowboarders showed an odds ratio of 26.5 (P [ 0.003) for sustaining a moderate-to-severe TBI.
-
CONCLUSIONS: Despite increased helmet use we found no decrease in TBI among snowboarders. The possibility of
-
INTRODUCTION
W
ith 350,000 active participants in Switzerland, snowboarding has become a major winter sport (4); however, with about 8000 injuries per season it has also become a public health concern (16, 34). Comparison of medical costs resulting from sports injuries shows that skiing and snowboarding take the leading positions (16, 31, 34). Lack of preparedness for speed, bad weather, poor visibility, boarding on old snow, or a combination of these factors may increase the risk of a snowboard injury (12). Jumping is reported to be responsible for most snowboarding injuries (10, 15, 21, 28). Because they tend to fall backward, snowboarders mainly injure the upper extremities, most frequently the wrist (9, 15, 21). Although head injuries account only for 6%e22% of all snowboarding injuries—depending on the definition of head injury and the level of the trauma center—they are the main reason for referral to a major trauma center after an accident and are the leading cause of death and morbidity after snowboarding accidents (1, 2, 4, 12, 13, 18, 23, 30, 33, 38). Since the introduction of protective helmets in winter sports there is on-going debate on whether helmets effectively prevent traumatic brain injury (TBI) in skiers and snowboarders (8, 20, 26, 33). Some recent studies suggested a 15%e57% reduction of the odds ratio for head injuries in helmeted skiers and snowboarders compared with those not wearing a helmet (11, 27, 33). However, these studies had major limitations: they used different definitions of head injury, sometimes
Key words Helmet - Snowboarding - TBI - Traumatic brain injury
From the Departments of 1Emergency Medicine and 2Neurosurgery, Inselspital, Bern University Hospital, Bern, Switzerland
Abbreviations and Acronyms AIS: Abbreviated injury scale CT: Computed tomography ED: Emergency department GCS: Glasgow coma scale GOS: Glasgow outcome scale IQR: Interquartile range OR: Odds ratio TBI: Traumatic brain injury
Citation: World Neurosurg. (2015) 84, 3:805-812. http://dx.doi.org/10.1016/j.wneu.2015.05.016
-
WORLD NEUROSURGERY 84 [3]: 805-812, SEPTEMBER 2015
To whom correspondence should be addressed: Rebecca M. Hasler, M.D., M.Sc. [E-mail: [email protected]] Rebecca M. Hasler and Dominik Baschera are equally contributing first authors.
Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2015 Elsevier Inc. All rights reserved.
www.WORLDNEUROSURGERY.org
805
ORIGINAL ARTICLE REBECCA M. HASLER ET AL.
TBI IN SNOWBOARDERS
including face and neck injuries, included a mixed population of skiers and snowboarders, or focused only on patients up to 21 years of age. Furthermore, most studies used ski patrol data instead of hospital discharge diagnoses (11, 19, 26, 27, 33). Rughani et al. (26), who evaluated neurosurgical discharge diagnosis in children and young adults, found a protective effect of helmets for skull fractures, but not for intracranial lesions or need for neurosurgical intervention. In contrast with skiers, TBI in snowboarders generally impacts the occiput due to backward falls (10). Our hospital is a tertiary alpine trauma center and is the only referral center for moderate-to-severe TBI in one of the largest winter sports regions in Switzerland. Although helmet use has steadily increased from 20% in 2002/2003 to above 89% in 2012/ 2013, we hypothesized—based on the impression from daily clinical work—that the number of patients with TBI due to snowboarding accidents, and especially severe TBI, did not change in our institution (36). Furthermore, our hypothesis was that helmet use is not associated with a significant decrease in moderate-tosevere TBI, but riding off-piste might increase the risk.
patients with TBI the discharge diagnosis from the Department of Neurosurgery was used. In addition, all computed tomography (CT) scans of these patients were reviewed by a neurosurgeon to determine the presence and extent of midline shift or intracranial bleeding. For prospective data collection on helmet use and Glasgow outcome score (GOS) patients were contacted by phone by specially trained study staff speaking German, English, and French (14). Only patients who survived the accident, and for whom no recordings of later death were found on the patient information system of the hospital, were contacted. Five contact attempts were made per patient at different times during the day during several weeks. Only patients who gave oral consent to take part in the study, after receiving a standardized study introduction, were interviewed. For the data on skier days per season we used the officially published numbers by the Swiss ski lifts association (37). For the seasons 2000/2001 to 2003/2004 only estimated numbers exist, and these were used for analyses (37).
METHODS
Statistical Analysis Descriptive data of the study population were examined and tabulated. Parametric data are presented as means, together with the corresponding standard deviations, and nonparametric data as medians with interquartile ranges (IQRs). Categorical data are reported as numbers and percentages. The primary outcome, TBI, was classified as mild, moderate, or severe according to commonly used cutoff levels of GCS (mild, 13e15; moderate, 9e12; severe, 3e8) (35). The primary exposure was helmet use at the time of the accident. Secondary exposures were age, gender, and the place of skiing (on-slope vs. off-piste vs. in a park). The main analysis compared the occurrence of moderate and severe TBI versus mild TBI in patients with or without a helmet using logistic regression analysis providing odds ratios (OR) and 95% confidence intervals with corresponding P values. As secondary analysis, the regression model was repeated with off-piste versus park versus on-slope as the exposure. All regression models were adjusted for age, gender, and skier days. Cross-tabulation and identification of missing data were used to check for errors throughout the analyses. With the c2 test and the Wilcoxon rank sum test, including TBI patients with or without data on helmet use, results were compared with regard to age, gender, GCS, and AIS head. Statistical significance was defined as an alpha value of 5%. Analyses were performed using R version 2.15.2 (22).
Study Design and Participants This cohort study compared the association between helmet use and incidence of TBI in adult snowboarders admitted to a university hospital between the seasons 2000/2001 and 2010/2011. Descriptive data of the study population were also evaluated. Setting The hospital is situated in central Switzerland, surrounded by famous Swiss ski resorts such as Adelboden, Grindelwald, and Zermatt. It is a level I trauma center and is the only referral hospital for moderate-to-severe neurotrauma within a catchment area of about 1.8 million inhabitants. It covers ski resorts from the Bernese Oberland to the Valais, which account for approximately 60% of skier days in the country (37). Therefore, the study population is representative for moderate-to-severe TBI resulting from snowboarding or alpine skiing. Inclusion and Exclusion Criteria We included all adult (16 years) patients admitted to the emergency department (ED) of our university hospital between the seasons 2000/2001 and 2010/2011. We only included patients directly injured from snowboarding. Patients with injuries sustained from ski lifts, activities after snowboarding, or handling of snowboard equipment were excluded. We included patients with isolated TBI or TBI combined with injuries to other body regions as defined by the abbreviated injury scale (AIS) for head injuries, with scores ranging from 1e6 (6). We excluded patients with isolated soft tissue injuries to the head. Study Procedures and Data Collection Data regarding age, gender, Glasgow coma scale (GCS) score and injured body regions were acquired from the patient information system of the ED. For GCS, the first recorded value on admission to the ED was used (35). In case of prehospital intubation of the patient the last GCS before intubation was used for analysis. For each patient the injury severity score was calculated (6). For
806
www.SCIENCEDIRECT.com
Ethics and Funding The study received ethical approval from the Internal Review Board of Bern University Hospital (Nr. 12-006). No specific funding was provided for the study. All study participants provided oral informed consent for this analysis. RESULTS Study Population During the seasons 2000/2001 to 2010/2011, 797 patients injured during snowboarding were treated at our university hospital. Of
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2015.05.016
ORIGINAL ARTICLE REBECCA M. HASLER ET AL.
these, 96 (12.0%) suffered nonacute snowboarding injuries (chronic diseases, overuse injuries) and were therefore excluded. Ten patients (1.3%) were excluded due to injuries sustained from nonsnowboarding-related activities (i.e., ski lift, activities after snowboarding). Of the remaining 691 patients, 117 (16.9%) suffered TBI and were included in the analysis (Figure 1) (34). Descriptive Data Of the 117 snowboarders with TBI, 65.8% (n ¼ 77) were men. Median age was 23.0 years (IQR, 18.0e27.6), median GCS 15 (IQR 15e15), and the median AIS head score was 1 (IQR 1e2). TBI occurred while riding on a prepared slope (82.9%; n ¼ 97), in a park (12.0%; n ¼ 14), and off-piste (5.1%; n ¼ 6). Two accidents (1.7%) were fatal. Patients who sustained minor (GCS 13e15; 92.3%, n ¼108), moderate (GCS 9e12; 0.9%, n ¼ 1), and severe (GCS 3e8; 6.8%, n ¼ 8) TBI (34). A complicated minor TBI was sustained by 8.5% of patients (n ¼ 10). Pathologic CT findings were present in 16.2% (n ¼ 19) of patients. Among patients with CT pathology, 26.3% (n ¼ 5) required surgical intervention and 15.8% (n ¼ 3) had skull fractures. On the initial CT scan in the emergency department 2 patients with TBI had a midline shift of 5 mm. Patients With Surgical Interventions The following (main) lesions were found in patients who underwent neurosurgical interventions: contusions with parenchymatous hemorrhage in 2 patients and acute subdural hematoma in 3 patients. The surgery in these patients included 3 interventions with different types of invasive monitoring of intracerebral
TBI IN SNOWBOARDERS
pressure, including 1 with osteoplastic craniotomy and 1 with osteoclastic craniectomy. The severity according to grading by GCS was minor in 1 and severe in 4 patients. The mean GOS of these patients was 4. Three patients had a good outcome with a GOS of 5 (i.e., slight damage with minor neurological and psychologic deficits), 1 died, and 1 was lost to follow-up. Four had initial unconsciousness and 1 had a focal neurological deficit (Table 1). Figure 2 displays CT images from a snowboarder with both parenchymatous hemorrhage and acute subdural hematoma. The number and proportion of patients with snowboardingrelated TBI with low GCS or high AIS head score differed from season to season, with no trend toward increase or decrease of these injuries (Figures 3 and 4). Helmet Use Data on helmet use were available for 74 patients (63.2%). Of these 52.7% (n ¼ 39) were wearing a helmet at the time of the accident. Helmet use steadily increased from 10% in the 2000/2001 and 2001/2002 seasons to 69.2% in the 2010/2011 season (P ¼ 0.002). Among the 5 patients requiring surgical treatment only 1 was wearing a helmet. Riding Off-piste Of the 6 snowboarders riding off-piste, 1 fell 20 m into a crevasse, 2 fell down cliffs of up to 100 m, 1 hit his head on a stone, 1 crashed into an avalanche screen, and 1 was captured in a floe avalanche. The 2 snowboarders who died were both riding offpiste (fall from cliff, hit head on stone). The snowboarder who hit his head on a stone suffered from multiple dislocated skull and base fractures, massive frontotemporal cerebral contusions, diffuse parachymatous hemorrhage, and a midline shift >5 mm. Due to his dire neurological prognosis no neurosurgical intervention was attempted and the patient died shortly after in the critical care unit. Logistic Regression Analyses Helmet wearing. The main analysis, comparing TBI according to GCS score in patients with or without a helmet, showed decreased crude and adjusted ORs of 0.26 and 0.24, respectively, for suffering moderate or severe TBI in helmet wearers. However, both 95% confidence intervals were wide and included 1, the value of no difference between the 2 groups (Table 2). Secondary analyses, with the GOS as the outcome, resulted in a decreased OR of 0.84 for sustaining GOS 5 mm
Outcome (GOS)y
REBECCA M. HASLER ET AL.
Comparison to Literature Seventeen percent of patients with snowboarding accidents admitted to our trauma center were treated for TBI. As expected, this percentage is on the upper end compared with accident statistics (range, 6%e22%) as the data are from a level I trauma center that is the only neurosurgical referral hospital in the region (1, 2, 4, 18, 23, 30, 33, 37). Wasden et al. (38) report a head injury rate at approximately 22% at a level I trauma center in Salt Lake City, Utah, USA. In our study population, 92.3% of patients with TBI sustained mild TBI. This finding is supported by Levy et al. (18) who report that 81.4% of their patients sustained minor TBI with a GCS between 13 ando 15. We found a relatively low number of patients with severe TBI (6.8%). Only 19 patients (16.2%) showed pathologic findings in CT scans, and only 5 of these required neurosurgical intervention. The most frequent intracranial injury pattern requiring neurosurgical intervention was acute subdural
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2015.05.016
ORIGINAL ARTICLE REBECCA M. HASLER ET AL.
TBI IN SNOWBOARDERS
Figure 2. Computed tomography scans from a nonhelmeted snowboarder, who fell on a skiing slope. At admission the Glasgow coma scale score of this patient was 3, with anisocoria. The patient suffered from an acute subdural hematoma causing a midline shift more than 2 cm (A). After urgent decompressive hemicraniectomy the midline shift was almost normalized and the contusions were demarcated (B1, B2).
hemorrhage. This is similar to the findings of Koyama et al. (17), who found subdural hemorrhages to be the most common intracranial lesion in snowboarders. Helmet use has steadily increased and recent studies report that 89% of snowboarders in Switzerland and 70% of skiers and snowboarders in the United States wear helmets (5, 36). Despite this tremendous increase in helmet use, multivariable regression analysis showed only a decreased OR for sustaining moderate-to-severe TBI in helmet wearers, but the difference was not statistically significant. Furthermore, we found no decrease in the number of moderate-to-severe TBI in snowboarders during the study period. This is in contrast to the findings of previous studies that relied on ski patrol data (11, 27, 33), but similar to the findings of Rughani et al. (26), who used pediatric neurosurgical data in
their analysis. Among the 5 patients who required neurosurgical interventions only 1 was wearing a helmet. It is therefore unclear whether some neurosurgical interventions could have been avoided if the other 4 patients had been using helmets. Apart from helmets, other factors, such as riding off-piste, bad slope, or weather conditions, may also play important roles (12). The much higher male-to-female ratio in patients requiring neurosurgical interventions compared with the overall population of snowboarders suggests that a risk-taking behavior may influence the severity of TBI. A recent study found that male gender is associated with a higher risk-taking behavior in skiers and snowboarders (25). This finding, along with the lack of a decrease in severe TBI despite increased use of winter sport helmets, supports the theory of risk compensation, suggesting that those
Figure 3. Distribution of traumatic brain injury (TBI) grades during the seasons. Proportions of patients with mild (dark gray) and moderate-to-severe (light gray) TBI during the seasons 2000/2001 to 2010/2011. The width of the bar indicates the number of snowboarding injuries per season.
WORLD NEUROSURGERY 84 [3]: 805-812, SEPTEMBER 2015
www.WORLDNEUROSURGERY.org
809
ORIGINAL ARTICLE REBECCA M. HASLER ET AL.
TBI IN SNOWBOARDERS
Figure 4. Proportions of patients with different abbreviated injury scale (AIS) scores (2000/2001 to 2010/2011). The dark bars represent an AIS head score of 5 and the gray bars represent AIS head scores of 4e1, with lighter gray bars representing lower scores. The width of the bar indicates the number of snowboarding injuries per season.
were caused by the additional dangers encountered off-piste, such as crevasses, cliffs, stones, avalanche screens, or floe avalanches.
wearing a helmet are more likely to ride at higher speeds (29). Furthermore, it is important to be aware that currently available helmets—depending on the testing standard—are made to protect from an impact speed of 20e23 km/h (3, 7, 32). Doubling the speed produces a fourfold increase in the energy of the impact. Keeping this in mind, along with the average speed of 48 km/h measured in skiers by Ruedl et al. (24), an impact at such speed would involve 4 times higher impact energy than current helmets are designed to withstand. The parameter riding “offpiste” was not part of previous studies. However, in our study snowboarders riding off-piste versus those riding on-slope had a significantly increased adjusted OR of 26.5 (P ¼ 0.003) for sustaining a moderate-to-severe TBI. In our cohort, these injuries
Implications Our results may have implications for future prevention campaigns for snowboarding safety by making riders aware that a helmet protects, but does not eliminate the risk of TBI. Furthermore, these campaigns should not only target the use of protective devices but also the importance of responsible behavior on slopes and the increased risk of riding off-piste. In addition, the results of our study may have implications for physicians and nurses working in emergency departments by raising awareness that severe head trauma is possible even in patients wearing a helmet at
Table 2. Logistic Regression Analyses in Snowboarders with TBI Exposure variables Helmet vs. No Helmet
Off Piste vs. On Slope
Park vs. On Slope
Crude OR (95% CI)
Adj.* OR (95% CI) P Valuey
Crude OR (95% CI)
Adj.* OR (95% CI) P Valuey
Crude OR (95% CI)
Adj.* OR (95% CI) P Valuey
TBI II/III vs. TBI I
0.26 (0.05e1.39)
0.24 (0.04e1.44) 0.120
23.3 (3.52e153)
26.5 (3.06e230) 0.003
3.87 (0.64e23.5)
4.59 (0.54e39.3) 0.164
GOS ¼ 5 vs. GOS
Cohort Study on the Association Between Helmet Use and Traumatic Brain Injury in Snowboarders From a Swiss Tertiary Trauma Center Rebecca M. Hasler1, Dominik Baschera2, David Taugwalder1, Aristomenis K. Exadaktylos1, Andreas Raabe2
BACKGROUND: Since the introduction of helmets in winter sports there is on-going debate on whether they decrease traumatic brain injuries (TBI).
-
METHODS: This cohort study included 117 adult (‡16 years) snowboarders with TBI admitted to a level I alpine trauma center in Switzerland between 2000/2001 and 2010/ 2011. The primary objective was to examine the association between helmet use and moderate-to-severe TBI. Secondary objectives were to describe the epidemiology of TBI during the past decade in relation to increased helmet use.
TBI despite helmet use and the dangers of riding off-piste should be a focus of future prevention programs.
-
RESULTS: Of 691 injured snowboarders evaluated, 117 (17%) suffered TBI. Sixty-six percent were men (median age, 23 years). Two percent of accidents were fatal. Ninetytwo percent of patients sustained minor, 1% moderate, and 7% severe TBI according to the Glasgow coma scale. Pathologic computed tomography findings were present in 16% of patients, 26% of which required surgery. Eightythree percent of TBIs occurred while riding on-slope. There was no trend in the TBI rate during the studied period, although helmet use increased from 10% to 69%. Comparing patients with and without a helmet showed no significant difference in odds ratios for the severity of TBI. However, of the 5 patients requiring surgery only 1 was wearing a helmet. Off-piste compared with on-slope snowboarders showed an odds ratio of 26.5 (P [ 0.003) for sustaining a moderate-to-severe TBI.
-
CONCLUSIONS: Despite increased helmet use we found no decrease in TBI among snowboarders. The possibility of
-
INTRODUCTION
W
ith 350,000 active participants in Switzerland, snowboarding has become a major winter sport (4); however, with about 8000 injuries per season it has also become a public health concern (16, 34). Comparison of medical costs resulting from sports injuries shows that skiing and snowboarding take the leading positions (16, 31, 34). Lack of preparedness for speed, bad weather, poor visibility, boarding on old snow, or a combination of these factors may increase the risk of a snowboard injury (12). Jumping is reported to be responsible for most snowboarding injuries (10, 15, 21, 28). Because they tend to fall backward, snowboarders mainly injure the upper extremities, most frequently the wrist (9, 15, 21). Although head injuries account only for 6%e22% of all snowboarding injuries—depending on the definition of head injury and the level of the trauma center—they are the main reason for referral to a major trauma center after an accident and are the leading cause of death and morbidity after snowboarding accidents (1, 2, 4, 12, 13, 18, 23, 30, 33, 38). Since the introduction of protective helmets in winter sports there is on-going debate on whether helmets effectively prevent traumatic brain injury (TBI) in skiers and snowboarders (8, 20, 26, 33). Some recent studies suggested a 15%e57% reduction of the odds ratio for head injuries in helmeted skiers and snowboarders compared with those not wearing a helmet (11, 27, 33). However, these studies had major limitations: they used different definitions of head injury, sometimes
Key words Helmet - Snowboarding - TBI - Traumatic brain injury
From the Departments of 1Emergency Medicine and 2Neurosurgery, Inselspital, Bern University Hospital, Bern, Switzerland
Abbreviations and Acronyms AIS: Abbreviated injury scale CT: Computed tomography ED: Emergency department GCS: Glasgow coma scale GOS: Glasgow outcome scale IQR: Interquartile range OR: Odds ratio TBI: Traumatic brain injury
Citation: World Neurosurg. (2015) 84, 3:805-812. http://dx.doi.org/10.1016/j.wneu.2015.05.016
-
WORLD NEUROSURGERY 84 [3]: 805-812, SEPTEMBER 2015
To whom correspondence should be addressed: Rebecca M. Hasler, M.D., M.Sc. [E-mail: [email protected]] Rebecca M. Hasler and Dominik Baschera are equally contributing first authors.
Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2015 Elsevier Inc. All rights reserved.
www.WORLDNEUROSURGERY.org
805
ORIGINAL ARTICLE REBECCA M. HASLER ET AL.
TBI IN SNOWBOARDERS
including face and neck injuries, included a mixed population of skiers and snowboarders, or focused only on patients up to 21 years of age. Furthermore, most studies used ski patrol data instead of hospital discharge diagnoses (11, 19, 26, 27, 33). Rughani et al. (26), who evaluated neurosurgical discharge diagnosis in children and young adults, found a protective effect of helmets for skull fractures, but not for intracranial lesions or need for neurosurgical intervention. In contrast with skiers, TBI in snowboarders generally impacts the occiput due to backward falls (10). Our hospital is a tertiary alpine trauma center and is the only referral center for moderate-to-severe TBI in one of the largest winter sports regions in Switzerland. Although helmet use has steadily increased from 20% in 2002/2003 to above 89% in 2012/ 2013, we hypothesized—based on the impression from daily clinical work—that the number of patients with TBI due to snowboarding accidents, and especially severe TBI, did not change in our institution (36). Furthermore, our hypothesis was that helmet use is not associated with a significant decrease in moderate-tosevere TBI, but riding off-piste might increase the risk.
patients with TBI the discharge diagnosis from the Department of Neurosurgery was used. In addition, all computed tomography (CT) scans of these patients were reviewed by a neurosurgeon to determine the presence and extent of midline shift or intracranial bleeding. For prospective data collection on helmet use and Glasgow outcome score (GOS) patients were contacted by phone by specially trained study staff speaking German, English, and French (14). Only patients who survived the accident, and for whom no recordings of later death were found on the patient information system of the hospital, were contacted. Five contact attempts were made per patient at different times during the day during several weeks. Only patients who gave oral consent to take part in the study, after receiving a standardized study introduction, were interviewed. For the data on skier days per season we used the officially published numbers by the Swiss ski lifts association (37). For the seasons 2000/2001 to 2003/2004 only estimated numbers exist, and these were used for analyses (37).
METHODS
Statistical Analysis Descriptive data of the study population were examined and tabulated. Parametric data are presented as means, together with the corresponding standard deviations, and nonparametric data as medians with interquartile ranges (IQRs). Categorical data are reported as numbers and percentages. The primary outcome, TBI, was classified as mild, moderate, or severe according to commonly used cutoff levels of GCS (mild, 13e15; moderate, 9e12; severe, 3e8) (35). The primary exposure was helmet use at the time of the accident. Secondary exposures were age, gender, and the place of skiing (on-slope vs. off-piste vs. in a park). The main analysis compared the occurrence of moderate and severe TBI versus mild TBI in patients with or without a helmet using logistic regression analysis providing odds ratios (OR) and 95% confidence intervals with corresponding P values. As secondary analysis, the regression model was repeated with off-piste versus park versus on-slope as the exposure. All regression models were adjusted for age, gender, and skier days. Cross-tabulation and identification of missing data were used to check for errors throughout the analyses. With the c2 test and the Wilcoxon rank sum test, including TBI patients with or without data on helmet use, results were compared with regard to age, gender, GCS, and AIS head. Statistical significance was defined as an alpha value of 5%. Analyses were performed using R version 2.15.2 (22).
Study Design and Participants This cohort study compared the association between helmet use and incidence of TBI in adult snowboarders admitted to a university hospital between the seasons 2000/2001 and 2010/2011. Descriptive data of the study population were also evaluated. Setting The hospital is situated in central Switzerland, surrounded by famous Swiss ski resorts such as Adelboden, Grindelwald, and Zermatt. It is a level I trauma center and is the only referral hospital for moderate-to-severe neurotrauma within a catchment area of about 1.8 million inhabitants. It covers ski resorts from the Bernese Oberland to the Valais, which account for approximately 60% of skier days in the country (37). Therefore, the study population is representative for moderate-to-severe TBI resulting from snowboarding or alpine skiing. Inclusion and Exclusion Criteria We included all adult (16 years) patients admitted to the emergency department (ED) of our university hospital between the seasons 2000/2001 and 2010/2011. We only included patients directly injured from snowboarding. Patients with injuries sustained from ski lifts, activities after snowboarding, or handling of snowboard equipment were excluded. We included patients with isolated TBI or TBI combined with injuries to other body regions as defined by the abbreviated injury scale (AIS) for head injuries, with scores ranging from 1e6 (6). We excluded patients with isolated soft tissue injuries to the head. Study Procedures and Data Collection Data regarding age, gender, Glasgow coma scale (GCS) score and injured body regions were acquired from the patient information system of the ED. For GCS, the first recorded value on admission to the ED was used (35). In case of prehospital intubation of the patient the last GCS before intubation was used for analysis. For each patient the injury severity score was calculated (6). For
806
www.SCIENCEDIRECT.com
Ethics and Funding The study received ethical approval from the Internal Review Board of Bern University Hospital (Nr. 12-006). No specific funding was provided for the study. All study participants provided oral informed consent for this analysis. RESULTS Study Population During the seasons 2000/2001 to 2010/2011, 797 patients injured during snowboarding were treated at our university hospital. Of
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2015.05.016
ORIGINAL ARTICLE REBECCA M. HASLER ET AL.
these, 96 (12.0%) suffered nonacute snowboarding injuries (chronic diseases, overuse injuries) and were therefore excluded. Ten patients (1.3%) were excluded due to injuries sustained from nonsnowboarding-related activities (i.e., ski lift, activities after snowboarding). Of the remaining 691 patients, 117 (16.9%) suffered TBI and were included in the analysis (Figure 1) (34). Descriptive Data Of the 117 snowboarders with TBI, 65.8% (n ¼ 77) were men. Median age was 23.0 years (IQR, 18.0e27.6), median GCS 15 (IQR 15e15), and the median AIS head score was 1 (IQR 1e2). TBI occurred while riding on a prepared slope (82.9%; n ¼ 97), in a park (12.0%; n ¼ 14), and off-piste (5.1%; n ¼ 6). Two accidents (1.7%) were fatal. Patients who sustained minor (GCS 13e15; 92.3%, n ¼108), moderate (GCS 9e12; 0.9%, n ¼ 1), and severe (GCS 3e8; 6.8%, n ¼ 8) TBI (34). A complicated minor TBI was sustained by 8.5% of patients (n ¼ 10). Pathologic CT findings were present in 16.2% (n ¼ 19) of patients. Among patients with CT pathology, 26.3% (n ¼ 5) required surgical intervention and 15.8% (n ¼ 3) had skull fractures. On the initial CT scan in the emergency department 2 patients with TBI had a midline shift of 5 mm. Patients With Surgical Interventions The following (main) lesions were found in patients who underwent neurosurgical interventions: contusions with parenchymatous hemorrhage in 2 patients and acute subdural hematoma in 3 patients. The surgery in these patients included 3 interventions with different types of invasive monitoring of intracerebral
TBI IN SNOWBOARDERS
pressure, including 1 with osteoplastic craniotomy and 1 with osteoclastic craniectomy. The severity according to grading by GCS was minor in 1 and severe in 4 patients. The mean GOS of these patients was 4. Three patients had a good outcome with a GOS of 5 (i.e., slight damage with minor neurological and psychologic deficits), 1 died, and 1 was lost to follow-up. Four had initial unconsciousness and 1 had a focal neurological deficit (Table 1). Figure 2 displays CT images from a snowboarder with both parenchymatous hemorrhage and acute subdural hematoma. The number and proportion of patients with snowboardingrelated TBI with low GCS or high AIS head score differed from season to season, with no trend toward increase or decrease of these injuries (Figures 3 and 4). Helmet Use Data on helmet use were available for 74 patients (63.2%). Of these 52.7% (n ¼ 39) were wearing a helmet at the time of the accident. Helmet use steadily increased from 10% in the 2000/2001 and 2001/2002 seasons to 69.2% in the 2010/2011 season (P ¼ 0.002). Among the 5 patients requiring surgical treatment only 1 was wearing a helmet. Riding Off-piste Of the 6 snowboarders riding off-piste, 1 fell 20 m into a crevasse, 2 fell down cliffs of up to 100 m, 1 hit his head on a stone, 1 crashed into an avalanche screen, and 1 was captured in a floe avalanche. The 2 snowboarders who died were both riding offpiste (fall from cliff, hit head on stone). The snowboarder who hit his head on a stone suffered from multiple dislocated skull and base fractures, massive frontotemporal cerebral contusions, diffuse parachymatous hemorrhage, and a midline shift >5 mm. Due to his dire neurological prognosis no neurosurgical intervention was attempted and the patient died shortly after in the critical care unit. Logistic Regression Analyses Helmet wearing. The main analysis, comparing TBI according to GCS score in patients with or without a helmet, showed decreased crude and adjusted ORs of 0.26 and 0.24, respectively, for suffering moderate or severe TBI in helmet wearers. However, both 95% confidence intervals were wide and included 1, the value of no difference between the 2 groups (Table 2). Secondary analyses, with the GOS as the outcome, resulted in a decreased OR of 0.84 for sustaining GOS 5 mm
Outcome (GOS)y
REBECCA M. HASLER ET AL.
Comparison to Literature Seventeen percent of patients with snowboarding accidents admitted to our trauma center were treated for TBI. As expected, this percentage is on the upper end compared with accident statistics (range, 6%e22%) as the data are from a level I trauma center that is the only neurosurgical referral hospital in the region (1, 2, 4, 18, 23, 30, 33, 37). Wasden et al. (38) report a head injury rate at approximately 22% at a level I trauma center in Salt Lake City, Utah, USA. In our study population, 92.3% of patients with TBI sustained mild TBI. This finding is supported by Levy et al. (18) who report that 81.4% of their patients sustained minor TBI with a GCS between 13 ando 15. We found a relatively low number of patients with severe TBI (6.8%). Only 19 patients (16.2%) showed pathologic findings in CT scans, and only 5 of these required neurosurgical intervention. The most frequent intracranial injury pattern requiring neurosurgical intervention was acute subdural
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2015.05.016
ORIGINAL ARTICLE REBECCA M. HASLER ET AL.
TBI IN SNOWBOARDERS
Figure 2. Computed tomography scans from a nonhelmeted snowboarder, who fell on a skiing slope. At admission the Glasgow coma scale score of this patient was 3, with anisocoria. The patient suffered from an acute subdural hematoma causing a midline shift more than 2 cm (A). After urgent decompressive hemicraniectomy the midline shift was almost normalized and the contusions were demarcated (B1, B2).
hemorrhage. This is similar to the findings of Koyama et al. (17), who found subdural hemorrhages to be the most common intracranial lesion in snowboarders. Helmet use has steadily increased and recent studies report that 89% of snowboarders in Switzerland and 70% of skiers and snowboarders in the United States wear helmets (5, 36). Despite this tremendous increase in helmet use, multivariable regression analysis showed only a decreased OR for sustaining moderate-to-severe TBI in helmet wearers, but the difference was not statistically significant. Furthermore, we found no decrease in the number of moderate-to-severe TBI in snowboarders during the study period. This is in contrast to the findings of previous studies that relied on ski patrol data (11, 27, 33), but similar to the findings of Rughani et al. (26), who used pediatric neurosurgical data in
their analysis. Among the 5 patients who required neurosurgical interventions only 1 was wearing a helmet. It is therefore unclear whether some neurosurgical interventions could have been avoided if the other 4 patients had been using helmets. Apart from helmets, other factors, such as riding off-piste, bad slope, or weather conditions, may also play important roles (12). The much higher male-to-female ratio in patients requiring neurosurgical interventions compared with the overall population of snowboarders suggests that a risk-taking behavior may influence the severity of TBI. A recent study found that male gender is associated with a higher risk-taking behavior in skiers and snowboarders (25). This finding, along with the lack of a decrease in severe TBI despite increased use of winter sport helmets, supports the theory of risk compensation, suggesting that those
Figure 3. Distribution of traumatic brain injury (TBI) grades during the seasons. Proportions of patients with mild (dark gray) and moderate-to-severe (light gray) TBI during the seasons 2000/2001 to 2010/2011. The width of the bar indicates the number of snowboarding injuries per season.
WORLD NEUROSURGERY 84 [3]: 805-812, SEPTEMBER 2015
www.WORLDNEUROSURGERY.org
809
ORIGINAL ARTICLE REBECCA M. HASLER ET AL.
TBI IN SNOWBOARDERS
Figure 4. Proportions of patients with different abbreviated injury scale (AIS) scores (2000/2001 to 2010/2011). The dark bars represent an AIS head score of 5 and the gray bars represent AIS head scores of 4e1, with lighter gray bars representing lower scores. The width of the bar indicates the number of snowboarding injuries per season.
were caused by the additional dangers encountered off-piste, such as crevasses, cliffs, stones, avalanche screens, or floe avalanches.
wearing a helmet are more likely to ride at higher speeds (29). Furthermore, it is important to be aware that currently available helmets—depending on the testing standard—are made to protect from an impact speed of 20e23 km/h (3, 7, 32). Doubling the speed produces a fourfold increase in the energy of the impact. Keeping this in mind, along with the average speed of 48 km/h measured in skiers by Ruedl et al. (24), an impact at such speed would involve 4 times higher impact energy than current helmets are designed to withstand. The parameter riding “offpiste” was not part of previous studies. However, in our study snowboarders riding off-piste versus those riding on-slope had a significantly increased adjusted OR of 26.5 (P ¼ 0.003) for sustaining a moderate-to-severe TBI. In our cohort, these injuries
Implications Our results may have implications for future prevention campaigns for snowboarding safety by making riders aware that a helmet protects, but does not eliminate the risk of TBI. Furthermore, these campaigns should not only target the use of protective devices but also the importance of responsible behavior on slopes and the increased risk of riding off-piste. In addition, the results of our study may have implications for physicians and nurses working in emergency departments by raising awareness that severe head trauma is possible even in patients wearing a helmet at
Table 2. Logistic Regression Analyses in Snowboarders with TBI Exposure variables Helmet vs. No Helmet
Off Piste vs. On Slope
Park vs. On Slope
Crude OR (95% CI)
Adj.* OR (95% CI) P Valuey
Crude OR (95% CI)
Adj.* OR (95% CI) P Valuey
Crude OR (95% CI)
Adj.* OR (95% CI) P Valuey
TBI II/III vs. TBI I
0.26 (0.05e1.39)
0.24 (0.04e1.44) 0.120
23.3 (3.52e153)
26.5 (3.06e230) 0.003
3.87 (0.64e23.5)
4.59 (0.54e39.3) 0.164
GOS ¼ 5 vs. GOS
