- Email: [email protected]
Searching for ski-lift injury: An uphill struggle?
Available online at www.sciencedirect.com
Journal of Science and Medicine in Sport 13 (2010) 205–209
Original paper
Searching for ski-lift injury: An uphill struggle? Pam Smartt ∗ , David Chalmers Injury Prevention Research Unit (IPRU), University of Otago, New Zealand Received 20 October 2008; received in revised form 12 December 2008; accepted 11 January 2009
Abstract Injuries arising from ski-lift malfunction are rare. Most arise from skier error when embarking or disembarking, or from improper lift operation. A search of the literature failed to uncover any studies focusing specifically on ski-lift injuries. The purpose of this study was to identify and characterise ski-lift injury resulting in hospitalisation and comment on barriers to reporting and reporting omissions. New Zealand hospitalised injury discharges 2000–2005 formed the primary dataset. To aid case identification these data were linked to ACC compensated claims for the same period and the data searched for all hospitalised cases of injury arising from ski-lifts. 44 cases were identified representing 2% of snow-skiing/snowboarding cases. 28 cases (64%) were male and 16 (36%) female, the average age was 32 yrs (range 5–73 yrs). The majority of cases were snow-skiers (35 cases, 80%). Most of the injuries were serious, or potentially so, with 1 case of traumatic pneumothorax, one of pulmonary embolism (after jumping from a ski-lift) and 28 cases sustaining fractures (six to the neck-of-femur, one to the lumbar spine and one to the pubis). ICISS scores for all cases ranged from 1.00 to 0.8182 (probability of dying in hospital 0–18.18%). Only 14 (32%) cases could be easily identified from ICD-10-AM e-codes and activity codes in the discharge summary. The ICD-10-AM external cause code for ski-lift injury V98 (“other specified transport accidents”) was only assigned to 39% of cases. The type of ski-lift could only be determined in 24 cases (55%). © 2009 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved. Keywords: MESH terms snow sports; Wounds and injuries; Accident prevention; Skiing
1. Introduction The term “ski-lift” includes a large number of different types of machinery used to transport snow-skiers and snowboarders on a ski slope. They are known variously as skitow, handle tow, T-bar, chair-lift, chain-lift, rope-tow, magic carpet, poma, nutcracker and gondola. Injury arising from ski-lift malfunctions are rare.1 Most injuries arise from skier error when embarking or disembarking, or from improper lift operation.2 Prospective surveys suggest that injury associated with ski-lifts comprises 7–8% of snow-skiing and snowboarding injury.3–5 The injury often requires in-patient hospital treatment. A study of skiers and snowboarders admitted to the Vancouver hospital acute spinal cord injury unit over two ski seasons,6 reported that for 5 of 56 cases (9% of cases, ∗
Corresponding author. E-mail address: [email protected] (P. Smartt).
15% of skiers) the cause of injury was a chair-lift. More recently, a study of serious winter sports injury to children indicated that 13 of 101 cases (13%) received their injuries from falling from a ski-lift.7 The absolute number of cases reported in these studies is small in relation to all reported skiing and snowboarding injury, however, the nature and severity of injuries sustained in these incidents (particularly in children) is a cause for concern. It is therefore surprising that a search of the peer-reviewed scientific literature failed to uncover any studies focusing specifically on ski-lift injuries. New Zealand is a mountainous country with 15 commercial and 10 club ski areas; in 2001 there were 1.25 million skier visitors to these areas.8 Ski-lifts of various types are in operation in most ski areas and although minor injury involving ski-lifts are treated on-site or in local medical centres, the most serious injuries are hospitalised. The purpose of this study was to search New Zealand’s national collection of public and private hospital discharge information (NMDS)
1440-2440/$ – see front matter © 2009 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.jsams.2009.01.004
206
P. Smartt, D. Chalmers / Journal of Science and Medicine in Sport 13 (2010) 205–209
for ski-lift injury cases with a view to describing their characteristics and commenting on any barriers to reporting that may help explain the paucity of information relating to these injuries.
2. Methods This study was undertaken as a part of a larger study of hospitalised injury across 187 sporting and recreational activities. Cases were eligible for inclusion if they were (a) first time admissions for that injury, (b) discharged between 1 January, 2000 and 31 December, 2005, (c) hospitalised for at least one night for an injury arising from snowboarding or snow-skiing activities (ICD-10-AM 3rd edition activity codes U 55.20-U 55.29, and U 55.4) and (d) were determined by e-code (V98) and/or relevant keyword in a case narrative to involve machinery used to transport skiers or snowboarders. Injuries involving snowmobiles were excluded. To assist case ascertainment and maximise injury incident information, hospital discharge records were linked to a national dataset of sport and recreation reimbursed entitlement claims for the same period. These data were provided by the Accident Compensation Corporation (ACC) which provides personal injury cover for all New Zealanders and visitors. Records were matched by National Health Index (NHI) number, surname, first name, injury day and second initial using the probabilistic linkage programme AUTOMATCH.9 Only ACC claims that could be linked to a hospital discharge record were retained. This linked dataset was searched using three separate procedures (implemented sequentially) to identify (a) cases with ICD-10-AM external cause codes which were likely to be associated with sporting/recreational activities or have an ICD-10-AM activity code indicating “sport” or “recreation”, (b) cases with key-words in the NMDS case narratives that were likely to indicate sporting or recreational activities and (c) linked NMDS/ACC claim cases identified through an ACC sport code, sport occupation or key-words in the ACC narrative that were likely to indicate sport/recreational activity. Cases identified in this way constituted a “master” dataset of 56,144 sport and recreation injury in-patient hospital cases. Each case was allocated to one of 231 possible sporting or recreational activities identified in the ICD-10-AM-3 classification of sports and recreational activities.10 Snow-skiing and snowboarding cases were identified using ICD-10-AM external cause codes, ACC snow sport codes 82 (snow-skiing) and 101 (snowboarding), and a keyword search of the national minimum dataset for hospital discharges (NMDS) and ACC narrative fields for snow-skiing and snowboarding related text. However, only ICD-10-AM-3 contained specific activity and location codes for snowskiing and snowboarding (Table 1), these were used to search the records coded to this version of ICD-10-AM (1 July 2004–31 December 2005). 1934 cases of injury attributable to snow-skiing and snowboarding were identified using these
Table 1 ICD-10-AM-3 external cause codes associated with snow-skiing and snowboarding injury. ICD-10-AM-3a external cause codes Snow-skiing activity Snowboarding activity Fall involving snow ski Fall involving snow board Skiing, snowboarding area a
U55.20–U55.29 U55.4 W02.3 W02.4 Y92.36
Cases discharged between 1 July 2004–31 December 2005.
procedures. This dataset (hereafter referred to as the “snowsport” dataset) was the starting point for this study. The ICD-10 external cause code for injury arising from the use of a ski-lift is V98 (“other specified transport accidents”) which includes incidents involving a “ski chair-lift” or “ski-lift with gondola”. The snow-sport dataset was searched for all cases coded to V98, or with key-words associated with ski-lifts or similar ski-field transport e.g., T-bar, skitow, chair-lift. As the initial search of the master dataset (N = 56,144) for snow-skiing and snowboarding injury cases did not specifically look for ski-lift injury cases, this dataset was searched again using the ski-lift search procedure outlined above. A further three cases, that had been assigned to the generic group “snow sport” i.e. cases where it was not clear if they were snow-skiers or snowboarders, were identified in this way. As a measure of injury severity, the ICD-derived injury severity score (ICISS)11 was determined for each case. This was computed as the product of the survival risk ratio (using survival probabilities based on 1999–2001 mortality data) for each of their injuries. Serious non-fatal injury cases have been defined by Cryer and Langley as cases with an ICISS score of less than or equal to 0.941, i.e. cases whose injuries at admission give them a survival probability of 94.1% or worse (probability of death in hospital of at least 5.90%).12
3. Results A total of 44 cases were identified (7 cases per year, range 3–8 cases) where the mechanism of injury was a ski-lift or similar ski-field transport. 23 (52%) discharges were matched to an ACC claim; 7 of these cases (16% of 44 cases) were identified primarily through information in the ACC narrative. Ski-lifts were responsible for 2% of all snow-skiing/snowboarding cases hospitalised in the period. 28 (64%) cases were male and 16 (36%) female, with an average age of 32 yrs (range 5–73 yrs). The majority of cases were skiers (35 cases, 80%) rather than snowboarders (6 cases, 14%). In 3 cases it was not clear which snow sport activity was involved. An analysis of the mechanism of injury ascribed to each incident was illuminating from a coding perspective, Table 2. Only 17 of 44 cases (39%) were assigned to the appropriate e-code i.e. V98 “Other specified transport accidents”. The
P. Smartt, D. Chalmers / Journal of Science and Medicine in Sport 13 (2010) 205–209 Table 2 External cause codes (mechanism of injury) for ski-lift injury cases identified from the New Zealand National Minimum Data Set of hospital discharges for the period 2000–2005. e-code V98 W02b W03 W17 W19 W21 W31 W51 X50 Total
Description
N accidentsa
Other specified transport Fall involving ice-skates, skis, roller-skates or skateboards Fall on other level due to collision with, or pushing by, another person Fall from one level to another Unspecified fall Striking against or struck by sports equipment Contact with other and unspecified machinery Striking against or bumped into by another person Overexertion and strenuous or repetitive movements
17 7 1 8 1 4 3 1 2 44
a
Includes cable-car not on rails, ski-chair lift, ski-lift with gondola. ICD-10-AM-1-2 W02 codes are non-specific, ICD-10-AM-3 subdivides the code into W023 “fall involving a snow ski” and W024 “fall involving a snowboard”. b
remaining 27 cases (61%) were assigned to eight other external cause codes, including 17 cases in which the injury was ascribed to a fall of one sort or another (e-codes: W02, W03, W17, W19), 5 cases in which the injury was attributed to being “Struck by or against an object” (e-codes: W21, W51), 3 cases in which the injury was attributed to “Contact with other and unspecified machinery” and 2 cases in which the injury was attributed to “overexertion and strenuous or repetitive movements”. The type of ski-lift involved could only be determined from narrative information in the NMDS or ACC record. In 14 cases (32%) the lift was reported to be a “tow” (e.g. T-bar, pomma, ski-tow), in 10 cases (23%) the lift was reported as a “chair-lift”; in the remaining 20 cases (45%) it was not possible to determine what type of ski-lift was involved. No single ICD-10-AM code reliably identified injury arising from/associated with ski-lifts as the e-code V98 may also be used for injury arising from other transport such as domestic chair-lifts. Other codes, defining the activity at the time of injury and/or the place of injury occurrence, are also required (see Table 1). Only the third revision of ICD-10-AM provides a combination of relevant activity (snow-skiing U 55.20-U 55.29, snowboarding U 55.4), and location (Skiing/snowboarding area Y 92.36) codes, that, when used in conjunction with the e-code V98, will reliably identify a skilift injury case. In the current study, only 5 cases (11%) were identified in this way. For the remaining 39 cases, key-words identified in the narratives recorded in the NMDS (30 cases) or ACC dataset (9 cases) were the main means of identifying cases. An examination of the primary diagnosis accompanying the discharge summary revealed that a number of the injuries were serious (probability of death in hospital of at least 5.90%) e.g. pulmonary embolism and fracture of the subcapital section of the femur, or potentially so (probability of death 2.93–5.54%) e.g. traumatic pneumothorax, other fractures of
207
the femur, fracture of the pubis, fracture of the lumbar vertebra, Table 3. For these cases, the anatomical injury severity score ranged from 0.8182 to 0.9707, with the corresponding probability of death (in hospital) arising from these injuries ranging from 2.93% to 18.18%. The ICISS score for all cases ranged from 1.00 to 0.8182, median score 0.9940. The number of bed-days spent in hospital for all cases ranged from 1 to 8 days (median 2 days, mean 3 days).
4. Discussion Ski-lift injuries are not common and this in part may explain the paucity of epidemiological data available to inform injury prevention strategies. Nevertheless, it is surprising that despite the very large number of studies reporting on injury in snow-skiers and snowboarders, little is known about how ski-lift injuries occur, which ski-lift types are involved, who is most likely to be injured, the severity of the injury and the outcome following injury. One potentially useful source of information in New Zealand is the NMDS which contains coded information on the diagnosis, treatment, length of stay and the external cause (mechanism) of injury. This dataset also includes a free text field (narrative) which the NZHIS advises should “always be used with external cause codes” and is intended to be used to describe the circumstances of injury.13 Despite the potential of this data source to inform research and action, the retrieval of ski-lift injury information from the NMDS was problematic. The ICD-10-AM external cause code for ski-lift injury is coded to V98 “other specified transport accidents” which includes incidents on, or involving a “ski chair-lift” or “skilift with gondola”. This code is operational in all versions of ICD-10-AM, however, by itself this code is not enough to identify ski-lift injury. An activity code or key-word associating the e-code with an appropriate activity (e.g. snow-skiing or snowboarding) is also required. Only 14 of 44 cases (32%) identified in the current study could be easily identified using ICD-10-AM e-codes and activity codes; 12 for snow-skiing and 2 for snowboarding. Identification of the majority of cases was dependent upon key-words in the NMDS and ACC narratives. A wide range of external cause codes, other than V98, were used to describe the mechanism of injury. This suggests that this type of injury either poses particular difficulties for coders and/or that the coding system for this type of injury is inadequate. In this context it may be noted that the indexing of this code in the ICD-10-AM code book is not intuitive i.e. the term “ski-lift” is not listed in the index and neither is “fall from ski-lift”. It was not possible to estimate the cost of ski-lift injury. ACC entitlement claims records for the period July 2007–June 2008 indicate that “serious” (not defined) snowskiing and snowboarding injury cost ACC on average NZ $ 57,000–NZ $ 68,000 per case; these are likely to include skilift injury cases. It was also not possible (with any degree of certainty) to determine which was the most common type of
208
P. Smartt, D. Chalmers / Journal of Science and Medicine in Sport 13 (2010) 205–209
Table 3 Primary diagnosis and injury severity score for ski-lift injury cases. Free text describing the injury event
Diag code
Primary diagnosis code description
Injury severity
Days
Aa
B
Stay
Superficial injury (N = 1) Fall from T-bar
S0085
Superficial injury of head, contusion
0.9872
1.28%
1
Open wound (N = 1) Contact with ski-tow pulley
S611
Open wound of finger with damage to nail
1.0000
0.00%
1
Fracture (N = 28) Fall involving ice-skates, skis, roller-skates, etc.b Other fall from ski-lift Fall from ski-lift Accident involving chair-lift Other specified transport accidentsb Fall involving snow board Overexertion/strenuous or repetitive movementsb Fell whilst disembarking from ski chair-lift Fell off t bar on ski field Fall from ski-lift Fall from ski-lift Fall from ski-lift Struck by chair-lift at ski field Struck by ski-lift tow bar Other fall same level collision/push by person Dragged by T-bar Struck by ski board waiting for ski-lift Fall involving snow ski Fall getting off ski-lift on mountain Fall involving ice-skates, skis, roller-skates, etc.b Fell off ski-lift Other specified transport accidents – snowboard Other specified transport accidentsb Fell from ski-lift while engaged in sports activity Fall while alighting from ski-lift Fell off ski-lift whilst snowboarding Fall while skiing Fall while attempting to alight from ski-lift
S0265 S3204 S325 S5200 S5250 S526 S6261 S7203 S7204 S722 S722 S723 S723 S8218 S8221 S8221 S8228 S8228 S8231 S8231 S8238 S825 S826 S826 S826 S826 S826 S8282
Fracture of angle of jaw Fracture of lumbar vertebra, L4 level Fracture of pubis Fracture of proximal ulna, NOS Fracture of lower end of radius NOS Fracture of lower end of radius/ulnar styloid Fracture of 4th and 5th proximal phalanx Fracture of subcapital section of femur Fracture of neck of femur Subtrochanteric fracture of neck of femur Subtrochanteric fracture of neck of femur Fracture of shaft of femur Fracture of shaft of femur Fracture of upper end of tibia Fracture of shaft of tibia with fracture of fibula Fracture of shaft of tibia with fracture of fibula Fracture of shaft of tibia Fracture of shaft of tibia Fracture of lower end of tibia with fracture of fibula Fracture of distal tibia with fracture of fibula Fracture of lower end of tibia Fracture of medial malleolus Fracture of lateral malleolus Fracture of lateral malleolus Fracture of lateral malleolus Fracture of lateral malleolus Weber C fracture of fibula Trimalleolar fracture
0.9952 0.9563 0.9542 0.9740 0.9983 0.9985 0.9978 0.9254 0.9582 0.9446 0.9446 0.9707 0.9707 0.9896 0.9841 0.9841 0.9768 0.9973 0.9758 0.9963 0.9965 0.9965 0.9969 0.9969 0.9969 0.9969 0.9969 0.9985
0.48% 4.37% 4.58% 2.60% 0.17% 0.15% 0.22% 7.46% 4.18% 5.54% 5.54% 2.93% 2.93% 1.04% 1.59% 1.59% 2.32% 0.27% 2.42% 0.37% 0.35% 0.35% 0.31% 0.31% 0.31% 0.31% 0.31% 0.15%
1 1 1 3 2 1 1 6 6 8 6 5 8 3 4 7 2 2 8 5 3 1 4 3 2 8 3 5
Strain of cervical spine Rupture of anterior cruciate ligament Rupture of posterior cruciate ligament
1.0000 1.0000 1.0000
0.00% 0.00% 0.00%
2 2 1
S764
Laceration of muscle, thigh
0.9941
0.59%
1
S270
Traumatic pneumothorax
0.9485
5.15%
5
Other and unspecified injury (N = 2) Striking against or struck by other sports equipment Hit in left side by tip of ski/coming off ski-tow
S099 S398
Unspecified injury of head Blunt abdominal trauma
0.9952 0.9903
0.48% 0.97%
1 2
Concussive injury (N = 4) Fall involving skis T-bar Fell backwards off chair-lift Fell from ski-lift Fall involving skisb
S0602 S0602 S0602 S0602
Loss of consciousness (<30 mins) Concussion/loss of consciousness (15 mins) Loss of consciousness (<30 mins) Loss of consciousness (<30 mins)
0.9769 0.9940 0.9940 0.9940
2.31% 0.60% 0.60% 0.60%
1 1 1 1
Pain at hip, pelvic region and thigh Disorders of excessive somnolence [hypersomnias] Pulmonary embolism
nc nc nc
Dislocation, strain or sprain of joints and ligaments (N = 3) Fell backwards from chair-lift S134 Tangled with other skiers coming off chair-lift S8353 S8354 Overexertion/strenuous or repetitive movements b Injury of muscle and tendon (N = 1) Struck by snow boardb Crushing injury (N = 1) Fall involving skisb
Not an “injury” diagnosis – injury severity score not calculated (N−3) Specified transport accidents = on T-bar M2555 Jumping off a ski-lift G471 Jumped from ski-lift I269
Note: Specific locations have been removed from the free text field. a ICISS score is calculated across each individual injury diagnosis code not just the primary code. b Indentified via the ACC free text. A = ICISS score. B = probability of death in hospital due to injury. nc = ICISS not calculated.
1 1 7
P. Smartt, D. Chalmers / Journal of Science and Medicine in Sport 13 (2010) 205–209
ski-lift involved in the injuries as 45% of the records simply indicated that a “ski-lift” was involved in the incident. For those cases that did contain information relating to the type of ski-lift, ski-tow-lifts were the most often cited (32% of all cases). This study has a number of limitations. Only injury cases hospitalised for at least one night and ACC registered claims cases that could be linked to a hospital discharge, were eligible for review. Moreover, the probabilistic matching technique used meant that false-positive and false-negative matches could not be ruled out. The definition of “serious” injury used here complies with that of Cryer and Langley12 ; this is not necessarily the most appropriate definition for sport related injury. Because of the difficulty in identifying skilift injury from hospital discharge records, it is possible that there were more cases than we have reported. Also while one may be reasonably confident that the most severe ski-lift injuries will be captured in hospital in-patient records, less severe and disabling injury cases presenting to other facilities will not.
5. Conclusion The results from this study suggest that identifying ski-lift injury from hospital discharge records is an “uphill struggle” and that this may in part explain the paucity of information about this cause of injury.
Practical implications It is suggested that; • the indexing of ski-lift injury in the ICD-10 code books is improved, and • a fourth character subdivision of the V98 code be introduced to distinguish incidents on or involving ski-lifts from those involving domestic chair-lifts, cable-cars (not on rails), ice-yachts and land-yachts. The new on-line public submission process for suggesting modifications, announced by The National Centre for Classification in Health in June 2008, may be an appropriate vehicle for achieving such changes.14
209
Acknowledgements The New Zealand Accident Compensation Corporation (ACC) funded this research. Data for the study was supplied by the New Zealand Health Information Service (NZHIS) and the New Zealand Accident Compensation Corporation (ACC). Helpful comments on an earlier draft of this paper were received from J Langley.
References 1. Nicholas RA, Oberheide JE. EMS response to a ski lift disaster in the Colorado mountains. J Trauma 1988;28(5):672–5. 2. Chalat Hatten, Koupal PC. Practice areas ski accidents. Lift cases. 2008 [cited] 2008; Available from: http://www.chalathatten.com/CM/ OtherSeriousPersonalInjuries/OtherSeriousPersonalInjuries13.asp]. 3. Davidson TM, Laliotis AT. Snowboarding injuries: a four-year study with comparison with alpine ski injuries. West J Med 1996;164(3):231–7. 4. Langran M, Jachacy GB, MacNeill A. Ski injuries in Scotland. A review of statistics from Cairngorm ski area Winter 1993/94. Scott Med J 1996;41(6):169–72. 5. Sulheim S, et al. Helmet use and risk of head injuries in alpine skiers and snowboarders. J Am Med Assoc 2006;295(8):919–24. 6. Tarazi F, Dvorak MFS, Wing PC. Spinal injuries in skiers and snowboarders. Am J Sports Med 1999;27(2):177–80. 7. Skokan EG, Junkins Jr EP, Kadish H. Serious winter sport injuries in children and adolescents requiring hospitalization. Am J Emerg Med 2003;21(2):95–9. 8. New Zealand Snow Sports Council. Ski and Snowboard statistics:Skier Visits/Open Days and Turnover; 2008. Available from: http://www.snow.co.nz/snowsports/stats.htm. 9. Match Ware Technologies Inc. AUTOMATCH generalized record linkage system. MD: Silver Spring; 1995. 10. National Centre for Classification in Health. The International Statistical Classification of Diseases and Related Health Problems. 10th Revision, Australian Modification (ICD-10-CM), third ed. Sydney: National Centre for Classification in Health; 2002. 11. Stephenson SC, et al. Comparing measures of injury severity for use with large databases. J Trauma 2002;53(2):326–32. 12. Cryer C, Langley JD. Developing valid indicators of injury incidence for “all injury”. Inj Prev 2006;12(3):202–7. 13. New Zealand Health Information Service. NMDS (Hospital Events) Data Dictionary. In: Diagnosis Procedure Table. Wellington: New Zealand Health Information Service; 2007. 14. National Centre for Classification in Health. Coding Matters. University of Sydney; 2008.
Journal of Science and Medicine in Sport 13 (2010) 205–209
Original paper
Searching for ski-lift injury: An uphill struggle? Pam Smartt ∗ , David Chalmers Injury Prevention Research Unit (IPRU), University of Otago, New Zealand Received 20 October 2008; received in revised form 12 December 2008; accepted 11 January 2009
Abstract Injuries arising from ski-lift malfunction are rare. Most arise from skier error when embarking or disembarking, or from improper lift operation. A search of the literature failed to uncover any studies focusing specifically on ski-lift injuries. The purpose of this study was to identify and characterise ski-lift injury resulting in hospitalisation and comment on barriers to reporting and reporting omissions. New Zealand hospitalised injury discharges 2000–2005 formed the primary dataset. To aid case identification these data were linked to ACC compensated claims for the same period and the data searched for all hospitalised cases of injury arising from ski-lifts. 44 cases were identified representing 2% of snow-skiing/snowboarding cases. 28 cases (64%) were male and 16 (36%) female, the average age was 32 yrs (range 5–73 yrs). The majority of cases were snow-skiers (35 cases, 80%). Most of the injuries were serious, or potentially so, with 1 case of traumatic pneumothorax, one of pulmonary embolism (after jumping from a ski-lift) and 28 cases sustaining fractures (six to the neck-of-femur, one to the lumbar spine and one to the pubis). ICISS scores for all cases ranged from 1.00 to 0.8182 (probability of dying in hospital 0–18.18%). Only 14 (32%) cases could be easily identified from ICD-10-AM e-codes and activity codes in the discharge summary. The ICD-10-AM external cause code for ski-lift injury V98 (“other specified transport accidents”) was only assigned to 39% of cases. The type of ski-lift could only be determined in 24 cases (55%). © 2009 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved. Keywords: MESH terms snow sports; Wounds and injuries; Accident prevention; Skiing
1. Introduction The term “ski-lift” includes a large number of different types of machinery used to transport snow-skiers and snowboarders on a ski slope. They are known variously as skitow, handle tow, T-bar, chair-lift, chain-lift, rope-tow, magic carpet, poma, nutcracker and gondola. Injury arising from ski-lift malfunctions are rare.1 Most injuries arise from skier error when embarking or disembarking, or from improper lift operation.2 Prospective surveys suggest that injury associated with ski-lifts comprises 7–8% of snow-skiing and snowboarding injury.3–5 The injury often requires in-patient hospital treatment. A study of skiers and snowboarders admitted to the Vancouver hospital acute spinal cord injury unit over two ski seasons,6 reported that for 5 of 56 cases (9% of cases, ∗
Corresponding author. E-mail address: [email protected] (P. Smartt).
15% of skiers) the cause of injury was a chair-lift. More recently, a study of serious winter sports injury to children indicated that 13 of 101 cases (13%) received their injuries from falling from a ski-lift.7 The absolute number of cases reported in these studies is small in relation to all reported skiing and snowboarding injury, however, the nature and severity of injuries sustained in these incidents (particularly in children) is a cause for concern. It is therefore surprising that a search of the peer-reviewed scientific literature failed to uncover any studies focusing specifically on ski-lift injuries. New Zealand is a mountainous country with 15 commercial and 10 club ski areas; in 2001 there were 1.25 million skier visitors to these areas.8 Ski-lifts of various types are in operation in most ski areas and although minor injury involving ski-lifts are treated on-site or in local medical centres, the most serious injuries are hospitalised. The purpose of this study was to search New Zealand’s national collection of public and private hospital discharge information (NMDS)
1440-2440/$ – see front matter © 2009 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.jsams.2009.01.004
206
P. Smartt, D. Chalmers / Journal of Science and Medicine in Sport 13 (2010) 205–209
for ski-lift injury cases with a view to describing their characteristics and commenting on any barriers to reporting that may help explain the paucity of information relating to these injuries.
2. Methods This study was undertaken as a part of a larger study of hospitalised injury across 187 sporting and recreational activities. Cases were eligible for inclusion if they were (a) first time admissions for that injury, (b) discharged between 1 January, 2000 and 31 December, 2005, (c) hospitalised for at least one night for an injury arising from snowboarding or snow-skiing activities (ICD-10-AM 3rd edition activity codes U 55.20-U 55.29, and U 55.4) and (d) were determined by e-code (V98) and/or relevant keyword in a case narrative to involve machinery used to transport skiers or snowboarders. Injuries involving snowmobiles were excluded. To assist case ascertainment and maximise injury incident information, hospital discharge records were linked to a national dataset of sport and recreation reimbursed entitlement claims for the same period. These data were provided by the Accident Compensation Corporation (ACC) which provides personal injury cover for all New Zealanders and visitors. Records were matched by National Health Index (NHI) number, surname, first name, injury day and second initial using the probabilistic linkage programme AUTOMATCH.9 Only ACC claims that could be linked to a hospital discharge record were retained. This linked dataset was searched using three separate procedures (implemented sequentially) to identify (a) cases with ICD-10-AM external cause codes which were likely to be associated with sporting/recreational activities or have an ICD-10-AM activity code indicating “sport” or “recreation”, (b) cases with key-words in the NMDS case narratives that were likely to indicate sporting or recreational activities and (c) linked NMDS/ACC claim cases identified through an ACC sport code, sport occupation or key-words in the ACC narrative that were likely to indicate sport/recreational activity. Cases identified in this way constituted a “master” dataset of 56,144 sport and recreation injury in-patient hospital cases. Each case was allocated to one of 231 possible sporting or recreational activities identified in the ICD-10-AM-3 classification of sports and recreational activities.10 Snow-skiing and snowboarding cases were identified using ICD-10-AM external cause codes, ACC snow sport codes 82 (snow-skiing) and 101 (snowboarding), and a keyword search of the national minimum dataset for hospital discharges (NMDS) and ACC narrative fields for snow-skiing and snowboarding related text. However, only ICD-10-AM-3 contained specific activity and location codes for snowskiing and snowboarding (Table 1), these were used to search the records coded to this version of ICD-10-AM (1 July 2004–31 December 2005). 1934 cases of injury attributable to snow-skiing and snowboarding were identified using these
Table 1 ICD-10-AM-3 external cause codes associated with snow-skiing and snowboarding injury. ICD-10-AM-3a external cause codes Snow-skiing activity Snowboarding activity Fall involving snow ski Fall involving snow board Skiing, snowboarding area a
U55.20–U55.29 U55.4 W02.3 W02.4 Y92.36
Cases discharged between 1 July 2004–31 December 2005.
procedures. This dataset (hereafter referred to as the “snowsport” dataset) was the starting point for this study. The ICD-10 external cause code for injury arising from the use of a ski-lift is V98 (“other specified transport accidents”) which includes incidents involving a “ski chair-lift” or “ski-lift with gondola”. The snow-sport dataset was searched for all cases coded to V98, or with key-words associated with ski-lifts or similar ski-field transport e.g., T-bar, skitow, chair-lift. As the initial search of the master dataset (N = 56,144) for snow-skiing and snowboarding injury cases did not specifically look for ski-lift injury cases, this dataset was searched again using the ski-lift search procedure outlined above. A further three cases, that had been assigned to the generic group “snow sport” i.e. cases where it was not clear if they were snow-skiers or snowboarders, were identified in this way. As a measure of injury severity, the ICD-derived injury severity score (ICISS)11 was determined for each case. This was computed as the product of the survival risk ratio (using survival probabilities based on 1999–2001 mortality data) for each of their injuries. Serious non-fatal injury cases have been defined by Cryer and Langley as cases with an ICISS score of less than or equal to 0.941, i.e. cases whose injuries at admission give them a survival probability of 94.1% or worse (probability of death in hospital of at least 5.90%).12
3. Results A total of 44 cases were identified (7 cases per year, range 3–8 cases) where the mechanism of injury was a ski-lift or similar ski-field transport. 23 (52%) discharges were matched to an ACC claim; 7 of these cases (16% of 44 cases) were identified primarily through information in the ACC narrative. Ski-lifts were responsible for 2% of all snow-skiing/snowboarding cases hospitalised in the period. 28 (64%) cases were male and 16 (36%) female, with an average age of 32 yrs (range 5–73 yrs). The majority of cases were skiers (35 cases, 80%) rather than snowboarders (6 cases, 14%). In 3 cases it was not clear which snow sport activity was involved. An analysis of the mechanism of injury ascribed to each incident was illuminating from a coding perspective, Table 2. Only 17 of 44 cases (39%) were assigned to the appropriate e-code i.e. V98 “Other specified transport accidents”. The
P. Smartt, D. Chalmers / Journal of Science and Medicine in Sport 13 (2010) 205–209 Table 2 External cause codes (mechanism of injury) for ski-lift injury cases identified from the New Zealand National Minimum Data Set of hospital discharges for the period 2000–2005. e-code V98 W02b W03 W17 W19 W21 W31 W51 X50 Total
Description
N accidentsa
Other specified transport Fall involving ice-skates, skis, roller-skates or skateboards Fall on other level due to collision with, or pushing by, another person Fall from one level to another Unspecified fall Striking against or struck by sports equipment Contact with other and unspecified machinery Striking against or bumped into by another person Overexertion and strenuous or repetitive movements
17 7 1 8 1 4 3 1 2 44
a
Includes cable-car not on rails, ski-chair lift, ski-lift with gondola. ICD-10-AM-1-2 W02 codes are non-specific, ICD-10-AM-3 subdivides the code into W023 “fall involving a snow ski” and W024 “fall involving a snowboard”. b
remaining 27 cases (61%) were assigned to eight other external cause codes, including 17 cases in which the injury was ascribed to a fall of one sort or another (e-codes: W02, W03, W17, W19), 5 cases in which the injury was attributed to being “Struck by or against an object” (e-codes: W21, W51), 3 cases in which the injury was attributed to “Contact with other and unspecified machinery” and 2 cases in which the injury was attributed to “overexertion and strenuous or repetitive movements”. The type of ski-lift involved could only be determined from narrative information in the NMDS or ACC record. In 14 cases (32%) the lift was reported to be a “tow” (e.g. T-bar, pomma, ski-tow), in 10 cases (23%) the lift was reported as a “chair-lift”; in the remaining 20 cases (45%) it was not possible to determine what type of ski-lift was involved. No single ICD-10-AM code reliably identified injury arising from/associated with ski-lifts as the e-code V98 may also be used for injury arising from other transport such as domestic chair-lifts. Other codes, defining the activity at the time of injury and/or the place of injury occurrence, are also required (see Table 1). Only the third revision of ICD-10-AM provides a combination of relevant activity (snow-skiing U 55.20-U 55.29, snowboarding U 55.4), and location (Skiing/snowboarding area Y 92.36) codes, that, when used in conjunction with the e-code V98, will reliably identify a skilift injury case. In the current study, only 5 cases (11%) were identified in this way. For the remaining 39 cases, key-words identified in the narratives recorded in the NMDS (30 cases) or ACC dataset (9 cases) were the main means of identifying cases. An examination of the primary diagnosis accompanying the discharge summary revealed that a number of the injuries were serious (probability of death in hospital of at least 5.90%) e.g. pulmonary embolism and fracture of the subcapital section of the femur, or potentially so (probability of death 2.93–5.54%) e.g. traumatic pneumothorax, other fractures of
207
the femur, fracture of the pubis, fracture of the lumbar vertebra, Table 3. For these cases, the anatomical injury severity score ranged from 0.8182 to 0.9707, with the corresponding probability of death (in hospital) arising from these injuries ranging from 2.93% to 18.18%. The ICISS score for all cases ranged from 1.00 to 0.8182, median score 0.9940. The number of bed-days spent in hospital for all cases ranged from 1 to 8 days (median 2 days, mean 3 days).
4. Discussion Ski-lift injuries are not common and this in part may explain the paucity of epidemiological data available to inform injury prevention strategies. Nevertheless, it is surprising that despite the very large number of studies reporting on injury in snow-skiers and snowboarders, little is known about how ski-lift injuries occur, which ski-lift types are involved, who is most likely to be injured, the severity of the injury and the outcome following injury. One potentially useful source of information in New Zealand is the NMDS which contains coded information on the diagnosis, treatment, length of stay and the external cause (mechanism) of injury. This dataset also includes a free text field (narrative) which the NZHIS advises should “always be used with external cause codes” and is intended to be used to describe the circumstances of injury.13 Despite the potential of this data source to inform research and action, the retrieval of ski-lift injury information from the NMDS was problematic. The ICD-10-AM external cause code for ski-lift injury is coded to V98 “other specified transport accidents” which includes incidents on, or involving a “ski chair-lift” or “skilift with gondola”. This code is operational in all versions of ICD-10-AM, however, by itself this code is not enough to identify ski-lift injury. An activity code or key-word associating the e-code with an appropriate activity (e.g. snow-skiing or snowboarding) is also required. Only 14 of 44 cases (32%) identified in the current study could be easily identified using ICD-10-AM e-codes and activity codes; 12 for snow-skiing and 2 for snowboarding. Identification of the majority of cases was dependent upon key-words in the NMDS and ACC narratives. A wide range of external cause codes, other than V98, were used to describe the mechanism of injury. This suggests that this type of injury either poses particular difficulties for coders and/or that the coding system for this type of injury is inadequate. In this context it may be noted that the indexing of this code in the ICD-10-AM code book is not intuitive i.e. the term “ski-lift” is not listed in the index and neither is “fall from ski-lift”. It was not possible to estimate the cost of ski-lift injury. ACC entitlement claims records for the period July 2007–June 2008 indicate that “serious” (not defined) snowskiing and snowboarding injury cost ACC on average NZ $ 57,000–NZ $ 68,000 per case; these are likely to include skilift injury cases. It was also not possible (with any degree of certainty) to determine which was the most common type of
208
P. Smartt, D. Chalmers / Journal of Science and Medicine in Sport 13 (2010) 205–209
Table 3 Primary diagnosis and injury severity score for ski-lift injury cases. Free text describing the injury event
Diag code
Primary diagnosis code description
Injury severity
Days
Aa
B
Stay
Superficial injury (N = 1) Fall from T-bar
S0085
Superficial injury of head, contusion
0.9872
1.28%
1
Open wound (N = 1) Contact with ski-tow pulley
S611
Open wound of finger with damage to nail
1.0000
0.00%
1
Fracture (N = 28) Fall involving ice-skates, skis, roller-skates, etc.b Other fall from ski-lift Fall from ski-lift Accident involving chair-lift Other specified transport accidentsb Fall involving snow board Overexertion/strenuous or repetitive movementsb Fell whilst disembarking from ski chair-lift Fell off t bar on ski field Fall from ski-lift Fall from ski-lift Fall from ski-lift Struck by chair-lift at ski field Struck by ski-lift tow bar Other fall same level collision/push by person Dragged by T-bar Struck by ski board waiting for ski-lift Fall involving snow ski Fall getting off ski-lift on mountain Fall involving ice-skates, skis, roller-skates, etc.b Fell off ski-lift Other specified transport accidents – snowboard Other specified transport accidentsb Fell from ski-lift while engaged in sports activity Fall while alighting from ski-lift Fell off ski-lift whilst snowboarding Fall while skiing Fall while attempting to alight from ski-lift
S0265 S3204 S325 S5200 S5250 S526 S6261 S7203 S7204 S722 S722 S723 S723 S8218 S8221 S8221 S8228 S8228 S8231 S8231 S8238 S825 S826 S826 S826 S826 S826 S8282
Fracture of angle of jaw Fracture of lumbar vertebra, L4 level Fracture of pubis Fracture of proximal ulna, NOS Fracture of lower end of radius NOS Fracture of lower end of radius/ulnar styloid Fracture of 4th and 5th proximal phalanx Fracture of subcapital section of femur Fracture of neck of femur Subtrochanteric fracture of neck of femur Subtrochanteric fracture of neck of femur Fracture of shaft of femur Fracture of shaft of femur Fracture of upper end of tibia Fracture of shaft of tibia with fracture of fibula Fracture of shaft of tibia with fracture of fibula Fracture of shaft of tibia Fracture of shaft of tibia Fracture of lower end of tibia with fracture of fibula Fracture of distal tibia with fracture of fibula Fracture of lower end of tibia Fracture of medial malleolus Fracture of lateral malleolus Fracture of lateral malleolus Fracture of lateral malleolus Fracture of lateral malleolus Weber C fracture of fibula Trimalleolar fracture
0.9952 0.9563 0.9542 0.9740 0.9983 0.9985 0.9978 0.9254 0.9582 0.9446 0.9446 0.9707 0.9707 0.9896 0.9841 0.9841 0.9768 0.9973 0.9758 0.9963 0.9965 0.9965 0.9969 0.9969 0.9969 0.9969 0.9969 0.9985
0.48% 4.37% 4.58% 2.60% 0.17% 0.15% 0.22% 7.46% 4.18% 5.54% 5.54% 2.93% 2.93% 1.04% 1.59% 1.59% 2.32% 0.27% 2.42% 0.37% 0.35% 0.35% 0.31% 0.31% 0.31% 0.31% 0.31% 0.15%
1 1 1 3 2 1 1 6 6 8 6 5 8 3 4 7 2 2 8 5 3 1 4 3 2 8 3 5
Strain of cervical spine Rupture of anterior cruciate ligament Rupture of posterior cruciate ligament
1.0000 1.0000 1.0000
0.00% 0.00% 0.00%
2 2 1
S764
Laceration of muscle, thigh
0.9941
0.59%
1
S270
Traumatic pneumothorax
0.9485
5.15%
5
Other and unspecified injury (N = 2) Striking against or struck by other sports equipment Hit in left side by tip of ski/coming off ski-tow
S099 S398
Unspecified injury of head Blunt abdominal trauma
0.9952 0.9903
0.48% 0.97%
1 2
Concussive injury (N = 4) Fall involving skis T-bar Fell backwards off chair-lift Fell from ski-lift Fall involving skisb
S0602 S0602 S0602 S0602
Loss of consciousness (<30 mins) Concussion/loss of consciousness (15 mins) Loss of consciousness (<30 mins) Loss of consciousness (<30 mins)
0.9769 0.9940 0.9940 0.9940
2.31% 0.60% 0.60% 0.60%
1 1 1 1
Pain at hip, pelvic region and thigh Disorders of excessive somnolence [hypersomnias] Pulmonary embolism
nc nc nc
Dislocation, strain or sprain of joints and ligaments (N = 3) Fell backwards from chair-lift S134 Tangled with other skiers coming off chair-lift S8353 S8354 Overexertion/strenuous or repetitive movements b Injury of muscle and tendon (N = 1) Struck by snow boardb Crushing injury (N = 1) Fall involving skisb
Not an “injury” diagnosis – injury severity score not calculated (N−3) Specified transport accidents = on T-bar M2555 Jumping off a ski-lift G471 Jumped from ski-lift I269
Note: Specific locations have been removed from the free text field. a ICISS score is calculated across each individual injury diagnosis code not just the primary code. b Indentified via the ACC free text. A = ICISS score. B = probability of death in hospital due to injury. nc = ICISS not calculated.
1 1 7
P. Smartt, D. Chalmers / Journal of Science and Medicine in Sport 13 (2010) 205–209
ski-lift involved in the injuries as 45% of the records simply indicated that a “ski-lift” was involved in the incident. For those cases that did contain information relating to the type of ski-lift, ski-tow-lifts were the most often cited (32% of all cases). This study has a number of limitations. Only injury cases hospitalised for at least one night and ACC registered claims cases that could be linked to a hospital discharge, were eligible for review. Moreover, the probabilistic matching technique used meant that false-positive and false-negative matches could not be ruled out. The definition of “serious” injury used here complies with that of Cryer and Langley12 ; this is not necessarily the most appropriate definition for sport related injury. Because of the difficulty in identifying skilift injury from hospital discharge records, it is possible that there were more cases than we have reported. Also while one may be reasonably confident that the most severe ski-lift injuries will be captured in hospital in-patient records, less severe and disabling injury cases presenting to other facilities will not.
5. Conclusion The results from this study suggest that identifying ski-lift injury from hospital discharge records is an “uphill struggle” and that this may in part explain the paucity of information about this cause of injury.
Practical implications It is suggested that; • the indexing of ski-lift injury in the ICD-10 code books is improved, and • a fourth character subdivision of the V98 code be introduced to distinguish incidents on or involving ski-lifts from those involving domestic chair-lifts, cable-cars (not on rails), ice-yachts and land-yachts. The new on-line public submission process for suggesting modifications, announced by The National Centre for Classification in Health in June 2008, may be an appropriate vehicle for achieving such changes.14
209
Acknowledgements The New Zealand Accident Compensation Corporation (ACC) funded this research. Data for the study was supplied by the New Zealand Health Information Service (NZHIS) and the New Zealand Accident Compensation Corporation (ACC). Helpful comments on an earlier draft of this paper were received from J Langley.
References 1. Nicholas RA, Oberheide JE. EMS response to a ski lift disaster in the Colorado mountains. J Trauma 1988;28(5):672–5. 2. Chalat Hatten, Koupal PC. Practice areas ski accidents. Lift cases. 2008 [cited] 2008; Available from: http://www.chalathatten.com/CM/ OtherSeriousPersonalInjuries/OtherSeriousPersonalInjuries13.asp]. 3. Davidson TM, Laliotis AT. Snowboarding injuries: a four-year study with comparison with alpine ski injuries. West J Med 1996;164(3):231–7. 4. Langran M, Jachacy GB, MacNeill A. Ski injuries in Scotland. A review of statistics from Cairngorm ski area Winter 1993/94. Scott Med J 1996;41(6):169–72. 5. Sulheim S, et al. Helmet use and risk of head injuries in alpine skiers and snowboarders. J Am Med Assoc 2006;295(8):919–24. 6. Tarazi F, Dvorak MFS, Wing PC. Spinal injuries in skiers and snowboarders. Am J Sports Med 1999;27(2):177–80. 7. Skokan EG, Junkins Jr EP, Kadish H. Serious winter sport injuries in children and adolescents requiring hospitalization. Am J Emerg Med 2003;21(2):95–9. 8. New Zealand Snow Sports Council. Ski and Snowboard statistics:Skier Visits/Open Days and Turnover; 2008. Available from: http://www.snow.co.nz/snowsports/stats.htm. 9. Match Ware Technologies Inc. AUTOMATCH generalized record linkage system. MD: Silver Spring; 1995. 10. National Centre for Classification in Health. The International Statistical Classification of Diseases and Related Health Problems. 10th Revision, Australian Modification (ICD-10-CM), third ed. Sydney: National Centre for Classification in Health; 2002. 11. Stephenson SC, et al. Comparing measures of injury severity for use with large databases. J Trauma 2002;53(2):326–32. 12. Cryer C, Langley JD. Developing valid indicators of injury incidence for “all injury”. Inj Prev 2006;12(3):202–7. 13. New Zealand Health Information Service. NMDS (Hospital Events) Data Dictionary. In: Diagnosis Procedure Table. Wellington: New Zealand Health Information Service; 2007. 14. National Centre for Classification in Health. Coding Matters. University of Sydney; 2008.