- Email: [email protected]
Sport- or leisure-related injury hospital admissions: Do we need to get more out of being struck?
Journal of Science and Medicine in Sport (2006) 9, 498—505
ORIGINAL PAPER
Sport- or leisure-related injury hospital admissions: Do we need to get more out of being struck? Rebecca Mitchell ∗, Andrew Hayen NSW Injury Risk Management Research Centre, University of New South Wales, Sydney, NSW 2052, Australia Received 11 January 2006 ; received in revised form 28 April 2006; accepted 1 May 2006 KEYWORDS Sport or leisure injuries; Hospitalised injury; Injury classification; Struck by/struck against
Summary The usefulness of New South Wales (NSW) hospitalisation data for the identification of prevention measures for sport- or leisure-related injury hospitalisations for one common injury mechanism, struck by/struck against injuries, is illustrated. Sport- or leisure-related hospitalisations were identified during 1999—2000 to 2003—2004 from the NSW hospitalisation data using activity and place of occurrence information. Struck by/struck against injury hospitalisations were identified using the International Classification of Disease, 10th Revision, Australian Modified (ICD-10-AM) codes W20—W23 and W50—W52. Information regarding the number of hospitalisations for basic demographic descriptors (such as age and sex), the type of injury experienced, the injury mechanism, the activity, and the place of occurrence of the injury event are available from NSW hospitalisation data. Additional information than what is currently available would be required for the identification of targeted injury prevention strategies for sport- or leisure-related struck by/struck against injuries leading to hospitalisation. Assessing the feasibility of collecting information regarding the object or agent of injury, the phase of activity at the time of the injury, the collection of narrative text and the date of injury are all recommended. These recommendations have national and international implications as ICD-10 is widely used to classify hospitalised morbidity data. © 2006 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved.
Background ∗
Corresponding author. Tel.: +61 2 9385 4837; fax: +61 2 9385 6040. E-mail address: [email protected] (R. Mitchell).
In Australia, information regarding the magnitude of sport- or leisure-related injuries can be obtained
1440-2440/$ — see front matter © 2006 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jsams.2006.05.005
Getting more out of being struck from a number of sources, including hospital admissions, sports medicine clinics, general practitioners, insurance records, surveys, and local sportsspecific data collections.1 Many of these data sources use different taxonomies to collect information on sport- or leisure-related injuries. Internationally, the International Classification of Disease, 9th (ICD-9)2 or 10th Revision (ICD10),3 is often used to classify hospital admissions. However, detailed information regarding the type of sport or leisure activity is not available from these classification systems. In New South Wales (NSW) and throughout Australia, injury-related hospital admissions are classified using the International Classification of Disease, 10th Revision, Australian Modified (ICD-10-AM).4—6 Within chapter XIX of the ICD-10-AM on Injury, poisoning and certain other consequences of external causes certain injury types are identified by three character codes, and a fourth and fifth character identified the place of occurrence and the activity of the individual at the time of the incident, respectively, in the first edition4 and in supplementary variables in later editions.5—7 Since 1998, sport- and leisure-related activities have been able to be recorded, but not until more recent editions have particular sports (e.g. soccer, cricket, Australian rules) been able to be reported. Unfortunately, information obtained from hospitalisation data is not always ideal for injury prevention purposes since it often lacks specificity or is absent.8 Evaluation of data quality from hospital collections has been published elsewhere,9 but use of hospitalised data to identify injury prevention strategies has not been examined in NSW following the implementation of ICD-10-AM. Using one of the most common injury mechanisms for sportor leisure-related hospitalisation in NSW, struck by/struck against injuries, (A. Hayen unpublished analysis) an illustration of the type of information available at a state level upon which to base the development of prevention strategies is provided.
NSW inpatient statistics collection (ISC) The NSW Inpatient Statistics Collection is a census of all services for admitted patients to public and private hospitals, private day procedures, and public psychiatric hospitals in NSW.10 Hospitalisation data include information on episodes of care in hospital, which end with the discharge, transfer, or death of the patient, or when the service category for the admitted patient changes. Also included within the ISC are hospitalisations of NSW residents that occurred in another state or territory. How-
499 ever, these data were not available for 2003—2004, but were estimated to include 233 or 3% of all cases of hospitalisation for struck by/struck against injuries based on the average of the past 4 years. Hospitalisations relating to transfers or statistical discharges were excluded using the separation mode variable in order to attempt to partly eliminate ‘multiple counts’, which occur when an individual has more than one episode of care for a given injury. These exclusions refer to transfers between hospitals or changes in the on-going clinical care requirements (e.g. from acute to rehabilitation) for a patient during the one period of stay in a single facility, respectively.10
Identification of sport- or leisure-related struck by/struck against injuries Sport- or leisure-related injuries of NSW residents were identified in the NSW ISC for the 1999—2000 to 2003—2004 financial years using the following criteria: • the hospitalisation was for a patient who was a resident of NSW; • a principal diagnosis was in the ICD-10-AM range S00-T98; and • an activity code in the fifth character subdivision in ICD-10-AM version 1 was equal to 0 or 1 or the activity code in ICD-10-AM versions 2 and 3 indicated the activity of the injured person was sport- or leisure-related (i.e., U50—U72); or • a location code in the fourth character subdivision in ICD-10-AM version 1 was equal to 3 or the location code in ICD-10-AM versions 2 and 3 was equal to ‘sports and athletic areas’ (i.e., Y92.3). Struck by/struck against injuries refer to the contact made between one person and another person(s) or object(s) as the result of an unintentional event.3 For example, struck by injuries can refer to the type of unintentional contact that may occur between players participating in a sporting activity. Struck by/struck against injury hospitalisations were identified using the ICD-10-AM external cause codes in the range W20—W23 for struck by injuries and W50—W52 for struck against injuries.
Changes to ICD-10-AM During the period under study there were three changes to the coding frames used by ICD-10-AM to classify both the activity and location of the injurious incident, largely in the form of creating additional codes. As a result, different cod-
500
R. Mitchell, A. Hayen
ing frames were used to classify the activity and location variables during the three time periods of 1999—2000,4 2000—2001 to 2001—2002,5 and 2002—2003 to 2003—2004.6 For consistency, place of occurrence was classified using ICD-10-AM version one.4 Activity at time of incident was only analysed for the years 2002—2003 to 2003—2004 using ICD-10-AM version 3.6 As appropriate, injury types were grouped by bodily region.
Data analysis Analysis was performed using SAS.11 Age- and sexspecific population estimates as at 31 December of each of the years under study were obtained from the NSW Health Department. These estimates are based on the Australian Bureau of Statistics (ABS) population estimates as at 30 June.12 Directly age standardised rates were calculated using the estimated Australian residential population as at 30 June 2001 as the standard population. Ninety-five per cent confidence intervals (95% CI) were calculated assuming a Poisson distribution.13 To examine the association between age group and body location of injury, a chi-squared test of independence was used.14 Differences between age-standardised rates for males and females were assessed for significance using the method described in Armitage et al.14
Results and discussion While information needs from hospitalised injury data can vary depending on who requires the information and for what purpose, all injury professionals seek information of high quality and of enough detail to allow the identification of injury prevention priorities and/or to assist in evaluation of the effectiveness of injury prevention programs.
Basic demographics The type of information that can be obtained from the NSW ISC to assist with the identification of prevention measures for sport- or leisure-related injuries, includes basic demographic information (such as age and sex), which can be used in conjunction with demographic data to calculate injury incidence rates to illustrate the magnitude of the issue. For example, overall during 1999—2000 to 2003—2004, there were 35,416 hospitalisations for struck by/struck against injuries, at a rate of 107.8 per 100,000 population (95% CI 106.6—108.9). Of these, 8181 (23.1%) were recorded as sport- or leisure-related, at a rate of 25.1 per 100,000 population (95% CI 24.5—25.6). Males had a significantly higher hospitalisation rate of sport- or leisure-related struck by/struck against injuries than females (2 = 4221.9, d.f. = 1, p < 0.0001) with over three-quarters of the struck by/struck against hospitalised injuries of males (85.9%) (Table 1). Males aged 15—19 years had the highest age adjusted hospitalisation rates for struck by/struck against injuries during 1999—2000 to 2003—2004. For females, girls aged 10—14 years had the highest hospitalisation rates during this timeframe (Fig. 1).
Injury mechanism and type of injury Fundamental information regarding the type of injury mechanism and information regarding the type of injury experienced can be obtained from the NSW ISC. For sport- and leisure-related struck by/struck against injuries these included: being hit, struck, kicked, twisted, bitten or scratched by another person (33.9%), striking against or struck by sports equipment (29.4%), and striking against or bumped into by another person (24.9%) were the most common types of struck by/struck against injuries (Table 2). Head injuries (38.3%), wrist and
Table 1 Sport- or leisure-related struck by/struck against injury hospitalisations by year, NSW, number, rate and 95% CI, 1999—2000 to 2003—2004 Year
Persons
Males
N
Rate
1999—2000 2000—2001 2001—2002 2002—2003 2003—2004
1469 1287 1396 2063 1966
22.8 19.9 21.4 31.2 29.8
Total
8181
25.1
a b
a
95% CI
b
Females a
N
Rate
21.7—24.0 18.8—21.0 20.3—22.5 29.8—32.5 28.5—31.2
1290 1125 1202 1741 1671
39.7 34.3 36.4 51.9 50.1
37.5—41.9 32.3—36.3 34.4—38.6 49.4—54.4 47.7—52.5
24.5—25.6
7029
42.5
41.5—43.5
Age-standardised rate per 100,000 population. 95% confidence interval.
95% CI
b
Ratea
95% CIb
179 162 194 322 295
5.6 5.1 5.9 9.8 9.0
4.8—6.5 4.3—5.9 5.1—6.8 8.8—11.0 8.0—10.1
1152
7.1
6.7—7.5
N
Getting more out of being struck
501
Figure 1 Age adjusted rates for sport- or leisure-related struck by/struck against injury hospitalisations by sex, NSW, 1999—2000 to 2003—2004.
hand injuries (19.6%), and injuries to the knee and lower leg (15.8%) were the most common principal diagnoses following admission for a struck by/struck against injury (Fig. 2). Injury type varied significantly by age group (2 = 458.5, d.f. = 90, p < 0.001), with head injuries common in individuals aged less than 24 years. However, for sport- and leisure-related struck by/struck against injuries, the lack of specific information regarding, for example, the type of object that struck the person, or the circumstances surrounding or contributing to each injury mechanism, limits injury prevention activities aimed at reduc-
ing the incidence of injuries due to struck by/struck against injuries. There are several ways in which improvements could be made, such as by expanding the information that is currently recorded in the existing inpatient data collection on injuries. One option to achieve this would be an assessment of the feasibility of collecting relevant additional variables, such as those specified in the International Classification of External Causes of Injuries (ICECI), particularly regarding the type of object that struck the person.15 The ICECI is a multi-axial, modular, hierarchical system for classifying external causes
Table 2 Mechanism of sport- or leisure-related struck by/struck against injury hospitalisations, NSW, number and per cent, 1999—2000 to 2003—2004 Mechanism Struck by thrown, projected or falling object (W20)
Number
Per cent
182
2.2
Striking against or struck by sports equipment (W21) Striking against or struck by bat and racquet (W21.0) Striking against or struck by ball (W21.1) Striking against or struck by object or structure on or near sports area (W21.2) Striking against or struck by other sports equipment (W21.8) Striking against or struck by unspecified sports equipment (W21.9)
2408 117 758 28 243 53
29.4 4.9 31.5 1.2 10.1 2.2
Striking against or struck by other objects (W22) Caught, crushed, jammed or pinched in or between objects (W23) Hit, struck, kicked, twisted, bitten or scratched by another person (W50) Striking against or bumped into by another person (W51) Crushed, pushed or stepped on by crowd or human stampede (W52)
487 197 2775 2036 96
6.0 2.4 33.9 24.9 1.2
Total
8181
100.0
502
R. Mitchell, A. Hayen
Figure 2 Principal body location of injury of sport- or leisure-related struck by/struck injury hospitalisations by sex, NSW, per cent, 1999—2000 to 2003—2004.
of injuries. It was developed by an international collaboration of injury experts to complement the International Classification of Diseases (ICD) coding of injuries and enable further detail on the circumstances of the incident to be recorded, including information on the risk and (lack of) protective factors that led or contributed to the injury event occurring.
Activity and location of the injury event Some information regarding the activity and location of the injury event is available in the NSW ISC. For sport- and leisure-related struck by/struck against injuries, common activities reported to be undertaken at the time of the incident were team ball sports (58.3%) and team bat or stick sports (11.7%) (Table 3). Sports and athletic areas (82.2%), schools, other institution and public administrative areas (2.6%), and the home (1.3%) were the most common specified locations where the injury occurred. Unspecified locations accounted for only 9% of the places where the struck by/struck against injury occurred (Table 4). In recent editions of ICD-10-AM, there has been particular attention to the addition of activity codes to indicate different types of sport- or leisure-related activities that were performed at the time of the injurious incident. For example, at
present there are around 260 types of sport- and leisure-related activities (such as, cricket, rugby union, basketball) that can be classified. Unfortunately, there is no specific detail regarding the type of activity being performed at the time of the incident, which is often necessary to identify effective preventive actions. For example, if the sport- or leisure-related activity was identified as cricket, no information is available regarding whether the player was batting, bowling, or fielding when the struck by/struck against injury occurred, nor what phase of activity was being undertaken, such as training or competition. Nevertheless, having information about the sport played at time of injury is a major advance for sports injury researchers in Australia, but this has not yet been adopted universally internationally in ICD-10.
Sources of additional information Another option to gain further detail regarding the injurious incident may be to combine or link different data sources together to obtain a more comprehensive picture of injury events and outcomes.16,17 Data linkage involves combining information from two or more different data sources either through manual methods, deterministic procedures, or probabilistic linking techniques.18 Combining information from different data sources regarding the
Getting more out of being struck
503
Table 3 Activity of incident for sport- or leisure-related struck by/struck against injury hospitalisations, NSW, number and per cent, 2002—2003 to 2003—2004 Activity
Number
Team ball sports Football (including Australian Rules, Rugby Union, Rugby League, Soccer) Basketball
2350 2146 124
58.3 53.3 3.1
471 57 269 64 34
11.7 1.4 6.7 1.6 0.8
6 35
0.1 0.9
Individual water sports Fishing Surfing and boogie boarding Swimming
238 15 127 61
5.9 0.4 3.2 1.5
Ice and snow sports Individual athletic sports Aesthetic activities (e.g. dancing) Racquet sports (e.g. tennis) Target and precision sports Combat and martial arts Power sports Equestrian activities Adventure sports (e.g. hiking) Wheeled motor sports Non-wheeled motor sports (e.g. cycling, roller balding, scooters) Aero sports (e.g. hang gliding) Other school-related sports Other sport, exercise and leisure activities Other activities Not known/missing
43 29 8 33 58 146 19 12 4 11 42 4 25 403 86 6
1.1 0.7 0.2 0.8 1.4 3.6 0.5 0.3 0.1 0.3 1.0 0.1 0.6 10.0 2.1 0.1
4029
100.0
Team bat or stick sports Baseball Cricket Field hockey Softball Team water sports (e.g. water polo) Boat sports
Total
Per cent
Table 4 Location of incident for sport- or leisure-related struck by/struck against injury hospitalisations, NSW, number and per cent, 1999—2000 to 2003—2004 Location
Number
Home Residential institution School, other institution and public administrative area Sports and athletics area Street and highway Trade and service area Industrial and construction area Farm Other specified places Unspecified place Not known/missing
108 10 216 6723 14 38 2 3 330 736 1
Per cent 1.3 0.1 2.6 82.2 0.2 0.5 0.0 0.0 4.0 9.0 0.0
Total
8181
100.0
504 same injury event is often able to provide an increased understanding of an event and/or outcome as the combined data sources can often provide a temporal ordering to the sequence of events from the causal factors leading to a injury-related incident, to a description of the incident itself and of its outcome. The addition of a narrative text field to complement the coded data may be another alternative that has been investigated elsewhere with some success.19,20 The use of data text mining of narrative text fields is not a new concept to the NSW Health system as this technique has been used to examine records of presentations to a number of emergency departments and has potential use for injury prevention efforts.21 Additionally, the NSW ISC does not capture information relating to date of injury. If such a variable were introduced, it could be used with probabilistic data linkage methods (or a unique patient identifier), to determine whether a given episode of care for a patient was the first admission to hospital for a particular injury. This would enable better estimates of the incidence of sports- and leisurerelated activity.
Limitations This paper only reviewed one type of injury, sportor leisure-related injuries, and one type of injury mechanism, struck by/struck against injuries. It is possible that analyses of other injury mechanisms will reveal additional issues and outline further recommendations to improve hospitalisation data for injury prevention purposes. This paper also only reviewed a number of variables from the NSW ISC and does not represent the total information that is available for analysis. One limitation of the analysis conducted is the use of the whole population estimates to calculate injury incidence rates as these are likely to result in an underestimate of the true incidence rate. In fact, only 28.4% of NSW residents reported taking part in some form of organised sporting activity in the 12 months prior to 1999—2000.22
Conclusion This examination of sport- or leisure-related struck by/struck against injury hospitalisation data sought to provide an indication of several areas for improvement both in the collection of hospitalisation data in NSW and in the ICD-10-AM external causes classification system. In NSW, during
R. Mitchell, A. Hayen 1999—2000 to 2003—2004 injury was the sixth leading cause of hospitalisation with 606,954 hospitalisations, representing 6.3% of all hospitalisations of NSW residents (A Hayen and R Mitchell unpublished analysis). Injury is thus an important area for the identification of appropriate prevention strategies. Although, these recommendations for improvement have been illustrated using NSW hospitalisation data, they have national and international implications as ICD is one of the most widely used classification system for hospital morbidity records.23
Practical applications • To develop targeted injury prevention strategies for hospitalised sport- or leisure-related struck by/struck against injuries additional information is required. • Both the activity and location codes in the existing ICD-10-AM classification system require further enhancement to allow for additional detail to be collected. • Use of the International Classification of External Causes of Injuries would augment hospital injury-related morbidity data currently using ICD-10-AM. • Injury-related hospitalisation data would benefit from the addition of narrative text fields and from data linkage with other data collections that include information on the circumstances of injury.
Acknowledgements The authors are supported by the NSW Injury Risk Management Research Centre, with core funding provided by the NSW Health Department, the NSW Roads and Traffic Authority and the Motor Accidents Authority. Rebecca Mitchell is also supported by a Ph.D. scholarship from Injury Prevention and Control Australia. The authors wish to thank the Centre for Epidemiology and Research at the NSW Health Department for providing the data from the Health Outcomes and Information Statistical Toolkit (HOIST) analysed in this study.
References 1. Australian Government Department of Health and Ageing. Sports safety in Australia. Canberra: Commonwealth of Australia; 2003.
Getting more out of being struck 2. World Health Organization. ICD-9 International Classification of Diseases, 9th revision. Geneva: World Health Organization; 1977. 3. World Health Organization. ICD-10 International Classification of Diseases, 10th revision. Geneva: World Health Organization; 1992. 4. National Centre for Classification in Health. ICD-10-AM. 1st ed. Sydney: National Centre for Classification in Health; 1998. 5. National Centre for Classification in Health. ICD-10-AM. 2nd ed. Sydney: National Centre for Classification in Health; 2000. 6. National Centre for Classification in Health. ICD-10-AM. 3rd ed. Sydney: National Centre for Classification in Health; 2002. 7. National Centre for Classification in Health. ICD-10-AM. 4th ed. Sydney: Commonwealth Department of Health; 2004. 8. Langley J, Chalmers D. Coding the circumstances of injury: ICD-10 a step forwards or backwards? Inj Prev 1999;5:247—53. 9. MacIntyre C, Ackland M, Chandraraj E. Accuracy of injury coding in Victorian hospital morbidity data. Aust N Z J Public Health 1997;21(7):779—83. 10. Population Health Division. The Health of the People of NSW. Report of the Chief Health Officer. Sydney: NSW Health Department; 2004. 11. SAS Institute. SAS: statistical software, version 8.02. Cary, North Carolina: SAS Institute; 2000. 12. Centre for Epidemiology and Research. Health Outcomes and Information Statistical Toolkit (HOIST). NSW Department of Health; 2006. 13. Dobson A, Kuulasmaa K, Eberle E, Schere J. Confidence intervals for weighted sums of Poisson parameters. Stat Med 1991;10:457—762.
505 14. Armitage P, Berry G, Matthews J. Statistical methods in medical research. 4th ed. Cornwell: Blackwell Science; 2002. 15. WHO Collaborating Center on Injury Surveillance. International Classification of External Causes of Injury World Health Organization; 2005; Available from: URL:http:// www.iceci.org/. [cited 29 June 2005]. 16. Graitcer PL. Injury surveillance. In: Halperin W, Baker E, editors. Public health surveillance. New York: John Wiley & Sons Inc.; 1992. p. 142—56. 17. Horan JM, Mallonee S. Injury surveillance. Epidemiol Rev 2003;25:24—42. 18. Clark DE. Practical introduction to record linkage for injury research. Inj Prev 2004;10(3):186—91. 19. Lincoln AE, Sorock GS, Courtney TK, Wellman HM, Smith GS, Amoroso PJ. Using narrative text and coded data to develop hazard scenarios for occupational injury interventions. Inj Prev 2004;10(4):249—54. 20. Williamson A, Feyer A-M, Stout N, Driscoll T, Usher H. Use of narrative analysis for comparisons of the causes of fatal accidents in three countries: New Zealand, Australia, and the United States. Inj Prev 2001;7(Suppl. I):i15— 20. 21. Muscatello D, Churches T, Kaldor J, Zheng W, Chiu C, Correll P, et al. An automated, broad-based, near real-time public health surveillance system using presentations to hospital Emergency Departments in New South Wales, Australia. BMC Public Health 2005;4(141). 22. Australian Bureau of Statistics. Participation in Sport and Physical Activities. (Cat. No. 4177.0) Canberra: ABS; 2000. 23. Harrison J. Injury classification: balancing continuity and utility. Inj Control Saf Promot 2000;7(1):51—63.
ORIGINAL PAPER
Sport- or leisure-related injury hospital admissions: Do we need to get more out of being struck? Rebecca Mitchell ∗, Andrew Hayen NSW Injury Risk Management Research Centre, University of New South Wales, Sydney, NSW 2052, Australia Received 11 January 2006 ; received in revised form 28 April 2006; accepted 1 May 2006 KEYWORDS Sport or leisure injuries; Hospitalised injury; Injury classification; Struck by/struck against
Summary The usefulness of New South Wales (NSW) hospitalisation data for the identification of prevention measures for sport- or leisure-related injury hospitalisations for one common injury mechanism, struck by/struck against injuries, is illustrated. Sport- or leisure-related hospitalisations were identified during 1999—2000 to 2003—2004 from the NSW hospitalisation data using activity and place of occurrence information. Struck by/struck against injury hospitalisations were identified using the International Classification of Disease, 10th Revision, Australian Modified (ICD-10-AM) codes W20—W23 and W50—W52. Information regarding the number of hospitalisations for basic demographic descriptors (such as age and sex), the type of injury experienced, the injury mechanism, the activity, and the place of occurrence of the injury event are available from NSW hospitalisation data. Additional information than what is currently available would be required for the identification of targeted injury prevention strategies for sport- or leisure-related struck by/struck against injuries leading to hospitalisation. Assessing the feasibility of collecting information regarding the object or agent of injury, the phase of activity at the time of the injury, the collection of narrative text and the date of injury are all recommended. These recommendations have national and international implications as ICD-10 is widely used to classify hospitalised morbidity data. © 2006 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved.
Background ∗
Corresponding author. Tel.: +61 2 9385 4837; fax: +61 2 9385 6040. E-mail address: [email protected] (R. Mitchell).
In Australia, information regarding the magnitude of sport- or leisure-related injuries can be obtained
1440-2440/$ — see front matter © 2006 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jsams.2006.05.005
Getting more out of being struck from a number of sources, including hospital admissions, sports medicine clinics, general practitioners, insurance records, surveys, and local sportsspecific data collections.1 Many of these data sources use different taxonomies to collect information on sport- or leisure-related injuries. Internationally, the International Classification of Disease, 9th (ICD-9)2 or 10th Revision (ICD10),3 is often used to classify hospital admissions. However, detailed information regarding the type of sport or leisure activity is not available from these classification systems. In New South Wales (NSW) and throughout Australia, injury-related hospital admissions are classified using the International Classification of Disease, 10th Revision, Australian Modified (ICD-10-AM).4—6 Within chapter XIX of the ICD-10-AM on Injury, poisoning and certain other consequences of external causes certain injury types are identified by three character codes, and a fourth and fifth character identified the place of occurrence and the activity of the individual at the time of the incident, respectively, in the first edition4 and in supplementary variables in later editions.5—7 Since 1998, sport- and leisure-related activities have been able to be recorded, but not until more recent editions have particular sports (e.g. soccer, cricket, Australian rules) been able to be reported. Unfortunately, information obtained from hospitalisation data is not always ideal for injury prevention purposes since it often lacks specificity or is absent.8 Evaluation of data quality from hospital collections has been published elsewhere,9 but use of hospitalised data to identify injury prevention strategies has not been examined in NSW following the implementation of ICD-10-AM. Using one of the most common injury mechanisms for sportor leisure-related hospitalisation in NSW, struck by/struck against injuries, (A. Hayen unpublished analysis) an illustration of the type of information available at a state level upon which to base the development of prevention strategies is provided.
NSW inpatient statistics collection (ISC) The NSW Inpatient Statistics Collection is a census of all services for admitted patients to public and private hospitals, private day procedures, and public psychiatric hospitals in NSW.10 Hospitalisation data include information on episodes of care in hospital, which end with the discharge, transfer, or death of the patient, or when the service category for the admitted patient changes. Also included within the ISC are hospitalisations of NSW residents that occurred in another state or territory. How-
499 ever, these data were not available for 2003—2004, but were estimated to include 233 or 3% of all cases of hospitalisation for struck by/struck against injuries based on the average of the past 4 years. Hospitalisations relating to transfers or statistical discharges were excluded using the separation mode variable in order to attempt to partly eliminate ‘multiple counts’, which occur when an individual has more than one episode of care for a given injury. These exclusions refer to transfers between hospitals or changes in the on-going clinical care requirements (e.g. from acute to rehabilitation) for a patient during the one period of stay in a single facility, respectively.10
Identification of sport- or leisure-related struck by/struck against injuries Sport- or leisure-related injuries of NSW residents were identified in the NSW ISC for the 1999—2000 to 2003—2004 financial years using the following criteria: • the hospitalisation was for a patient who was a resident of NSW; • a principal diagnosis was in the ICD-10-AM range S00-T98; and • an activity code in the fifth character subdivision in ICD-10-AM version 1 was equal to 0 or 1 or the activity code in ICD-10-AM versions 2 and 3 indicated the activity of the injured person was sport- or leisure-related (i.e., U50—U72); or • a location code in the fourth character subdivision in ICD-10-AM version 1 was equal to 3 or the location code in ICD-10-AM versions 2 and 3 was equal to ‘sports and athletic areas’ (i.e., Y92.3). Struck by/struck against injuries refer to the contact made between one person and another person(s) or object(s) as the result of an unintentional event.3 For example, struck by injuries can refer to the type of unintentional contact that may occur between players participating in a sporting activity. Struck by/struck against injury hospitalisations were identified using the ICD-10-AM external cause codes in the range W20—W23 for struck by injuries and W50—W52 for struck against injuries.
Changes to ICD-10-AM During the period under study there were three changes to the coding frames used by ICD-10-AM to classify both the activity and location of the injurious incident, largely in the form of creating additional codes. As a result, different cod-
500
R. Mitchell, A. Hayen
ing frames were used to classify the activity and location variables during the three time periods of 1999—2000,4 2000—2001 to 2001—2002,5 and 2002—2003 to 2003—2004.6 For consistency, place of occurrence was classified using ICD-10-AM version one.4 Activity at time of incident was only analysed for the years 2002—2003 to 2003—2004 using ICD-10-AM version 3.6 As appropriate, injury types were grouped by bodily region.
Data analysis Analysis was performed using SAS.11 Age- and sexspecific population estimates as at 31 December of each of the years under study were obtained from the NSW Health Department. These estimates are based on the Australian Bureau of Statistics (ABS) population estimates as at 30 June.12 Directly age standardised rates were calculated using the estimated Australian residential population as at 30 June 2001 as the standard population. Ninety-five per cent confidence intervals (95% CI) were calculated assuming a Poisson distribution.13 To examine the association between age group and body location of injury, a chi-squared test of independence was used.14 Differences between age-standardised rates for males and females were assessed for significance using the method described in Armitage et al.14
Results and discussion While information needs from hospitalised injury data can vary depending on who requires the information and for what purpose, all injury professionals seek information of high quality and of enough detail to allow the identification of injury prevention priorities and/or to assist in evaluation of the effectiveness of injury prevention programs.
Basic demographics The type of information that can be obtained from the NSW ISC to assist with the identification of prevention measures for sport- or leisure-related injuries, includes basic demographic information (such as age and sex), which can be used in conjunction with demographic data to calculate injury incidence rates to illustrate the magnitude of the issue. For example, overall during 1999—2000 to 2003—2004, there were 35,416 hospitalisations for struck by/struck against injuries, at a rate of 107.8 per 100,000 population (95% CI 106.6—108.9). Of these, 8181 (23.1%) were recorded as sport- or leisure-related, at a rate of 25.1 per 100,000 population (95% CI 24.5—25.6). Males had a significantly higher hospitalisation rate of sport- or leisure-related struck by/struck against injuries than females (2 = 4221.9, d.f. = 1, p < 0.0001) with over three-quarters of the struck by/struck against hospitalised injuries of males (85.9%) (Table 1). Males aged 15—19 years had the highest age adjusted hospitalisation rates for struck by/struck against injuries during 1999—2000 to 2003—2004. For females, girls aged 10—14 years had the highest hospitalisation rates during this timeframe (Fig. 1).
Injury mechanism and type of injury Fundamental information regarding the type of injury mechanism and information regarding the type of injury experienced can be obtained from the NSW ISC. For sport- and leisure-related struck by/struck against injuries these included: being hit, struck, kicked, twisted, bitten or scratched by another person (33.9%), striking against or struck by sports equipment (29.4%), and striking against or bumped into by another person (24.9%) were the most common types of struck by/struck against injuries (Table 2). Head injuries (38.3%), wrist and
Table 1 Sport- or leisure-related struck by/struck against injury hospitalisations by year, NSW, number, rate and 95% CI, 1999—2000 to 2003—2004 Year
Persons
Males
N
Rate
1999—2000 2000—2001 2001—2002 2002—2003 2003—2004
1469 1287 1396 2063 1966
22.8 19.9 21.4 31.2 29.8
Total
8181
25.1
a b
a
95% CI
b
Females a
N
Rate
21.7—24.0 18.8—21.0 20.3—22.5 29.8—32.5 28.5—31.2
1290 1125 1202 1741 1671
39.7 34.3 36.4 51.9 50.1
37.5—41.9 32.3—36.3 34.4—38.6 49.4—54.4 47.7—52.5
24.5—25.6
7029
42.5
41.5—43.5
Age-standardised rate per 100,000 population. 95% confidence interval.
95% CI
b
Ratea
95% CIb
179 162 194 322 295
5.6 5.1 5.9 9.8 9.0
4.8—6.5 4.3—5.9 5.1—6.8 8.8—11.0 8.0—10.1
1152
7.1
6.7—7.5
N
Getting more out of being struck
501
Figure 1 Age adjusted rates for sport- or leisure-related struck by/struck against injury hospitalisations by sex, NSW, 1999—2000 to 2003—2004.
hand injuries (19.6%), and injuries to the knee and lower leg (15.8%) were the most common principal diagnoses following admission for a struck by/struck against injury (Fig. 2). Injury type varied significantly by age group (2 = 458.5, d.f. = 90, p < 0.001), with head injuries common in individuals aged less than 24 years. However, for sport- and leisure-related struck by/struck against injuries, the lack of specific information regarding, for example, the type of object that struck the person, or the circumstances surrounding or contributing to each injury mechanism, limits injury prevention activities aimed at reduc-
ing the incidence of injuries due to struck by/struck against injuries. There are several ways in which improvements could be made, such as by expanding the information that is currently recorded in the existing inpatient data collection on injuries. One option to achieve this would be an assessment of the feasibility of collecting relevant additional variables, such as those specified in the International Classification of External Causes of Injuries (ICECI), particularly regarding the type of object that struck the person.15 The ICECI is a multi-axial, modular, hierarchical system for classifying external causes
Table 2 Mechanism of sport- or leisure-related struck by/struck against injury hospitalisations, NSW, number and per cent, 1999—2000 to 2003—2004 Mechanism Struck by thrown, projected or falling object (W20)
Number
Per cent
182
2.2
Striking against or struck by sports equipment (W21) Striking against or struck by bat and racquet (W21.0) Striking against or struck by ball (W21.1) Striking against or struck by object or structure on or near sports area (W21.2) Striking against or struck by other sports equipment (W21.8) Striking against or struck by unspecified sports equipment (W21.9)
2408 117 758 28 243 53
29.4 4.9 31.5 1.2 10.1 2.2
Striking against or struck by other objects (W22) Caught, crushed, jammed or pinched in or between objects (W23) Hit, struck, kicked, twisted, bitten or scratched by another person (W50) Striking against or bumped into by another person (W51) Crushed, pushed or stepped on by crowd or human stampede (W52)
487 197 2775 2036 96
6.0 2.4 33.9 24.9 1.2
Total
8181
100.0
502
R. Mitchell, A. Hayen
Figure 2 Principal body location of injury of sport- or leisure-related struck by/struck injury hospitalisations by sex, NSW, per cent, 1999—2000 to 2003—2004.
of injuries. It was developed by an international collaboration of injury experts to complement the International Classification of Diseases (ICD) coding of injuries and enable further detail on the circumstances of the incident to be recorded, including information on the risk and (lack of) protective factors that led or contributed to the injury event occurring.
Activity and location of the injury event Some information regarding the activity and location of the injury event is available in the NSW ISC. For sport- and leisure-related struck by/struck against injuries, common activities reported to be undertaken at the time of the incident were team ball sports (58.3%) and team bat or stick sports (11.7%) (Table 3). Sports and athletic areas (82.2%), schools, other institution and public administrative areas (2.6%), and the home (1.3%) were the most common specified locations where the injury occurred. Unspecified locations accounted for only 9% of the places where the struck by/struck against injury occurred (Table 4). In recent editions of ICD-10-AM, there has been particular attention to the addition of activity codes to indicate different types of sport- or leisure-related activities that were performed at the time of the injurious incident. For example, at
present there are around 260 types of sport- and leisure-related activities (such as, cricket, rugby union, basketball) that can be classified. Unfortunately, there is no specific detail regarding the type of activity being performed at the time of the incident, which is often necessary to identify effective preventive actions. For example, if the sport- or leisure-related activity was identified as cricket, no information is available regarding whether the player was batting, bowling, or fielding when the struck by/struck against injury occurred, nor what phase of activity was being undertaken, such as training or competition. Nevertheless, having information about the sport played at time of injury is a major advance for sports injury researchers in Australia, but this has not yet been adopted universally internationally in ICD-10.
Sources of additional information Another option to gain further detail regarding the injurious incident may be to combine or link different data sources together to obtain a more comprehensive picture of injury events and outcomes.16,17 Data linkage involves combining information from two or more different data sources either through manual methods, deterministic procedures, or probabilistic linking techniques.18 Combining information from different data sources regarding the
Getting more out of being struck
503
Table 3 Activity of incident for sport- or leisure-related struck by/struck against injury hospitalisations, NSW, number and per cent, 2002—2003 to 2003—2004 Activity
Number
Team ball sports Football (including Australian Rules, Rugby Union, Rugby League, Soccer) Basketball
2350 2146 124
58.3 53.3 3.1
471 57 269 64 34
11.7 1.4 6.7 1.6 0.8
6 35
0.1 0.9
Individual water sports Fishing Surfing and boogie boarding Swimming
238 15 127 61
5.9 0.4 3.2 1.5
Ice and snow sports Individual athletic sports Aesthetic activities (e.g. dancing) Racquet sports (e.g. tennis) Target and precision sports Combat and martial arts Power sports Equestrian activities Adventure sports (e.g. hiking) Wheeled motor sports Non-wheeled motor sports (e.g. cycling, roller balding, scooters) Aero sports (e.g. hang gliding) Other school-related sports Other sport, exercise and leisure activities Other activities Not known/missing
43 29 8 33 58 146 19 12 4 11 42 4 25 403 86 6
1.1 0.7 0.2 0.8 1.4 3.6 0.5 0.3 0.1 0.3 1.0 0.1 0.6 10.0 2.1 0.1
4029
100.0
Team bat or stick sports Baseball Cricket Field hockey Softball Team water sports (e.g. water polo) Boat sports
Total
Per cent
Table 4 Location of incident for sport- or leisure-related struck by/struck against injury hospitalisations, NSW, number and per cent, 1999—2000 to 2003—2004 Location
Number
Home Residential institution School, other institution and public administrative area Sports and athletics area Street and highway Trade and service area Industrial and construction area Farm Other specified places Unspecified place Not known/missing
108 10 216 6723 14 38 2 3 330 736 1
Per cent 1.3 0.1 2.6 82.2 0.2 0.5 0.0 0.0 4.0 9.0 0.0
Total
8181
100.0
504 same injury event is often able to provide an increased understanding of an event and/or outcome as the combined data sources can often provide a temporal ordering to the sequence of events from the causal factors leading to a injury-related incident, to a description of the incident itself and of its outcome. The addition of a narrative text field to complement the coded data may be another alternative that has been investigated elsewhere with some success.19,20 The use of data text mining of narrative text fields is not a new concept to the NSW Health system as this technique has been used to examine records of presentations to a number of emergency departments and has potential use for injury prevention efforts.21 Additionally, the NSW ISC does not capture information relating to date of injury. If such a variable were introduced, it could be used with probabilistic data linkage methods (or a unique patient identifier), to determine whether a given episode of care for a patient was the first admission to hospital for a particular injury. This would enable better estimates of the incidence of sports- and leisurerelated activity.
Limitations This paper only reviewed one type of injury, sportor leisure-related injuries, and one type of injury mechanism, struck by/struck against injuries. It is possible that analyses of other injury mechanisms will reveal additional issues and outline further recommendations to improve hospitalisation data for injury prevention purposes. This paper also only reviewed a number of variables from the NSW ISC and does not represent the total information that is available for analysis. One limitation of the analysis conducted is the use of the whole population estimates to calculate injury incidence rates as these are likely to result in an underestimate of the true incidence rate. In fact, only 28.4% of NSW residents reported taking part in some form of organised sporting activity in the 12 months prior to 1999—2000.22
Conclusion This examination of sport- or leisure-related struck by/struck against injury hospitalisation data sought to provide an indication of several areas for improvement both in the collection of hospitalisation data in NSW and in the ICD-10-AM external causes classification system. In NSW, during
R. Mitchell, A. Hayen 1999—2000 to 2003—2004 injury was the sixth leading cause of hospitalisation with 606,954 hospitalisations, representing 6.3% of all hospitalisations of NSW residents (A Hayen and R Mitchell unpublished analysis). Injury is thus an important area for the identification of appropriate prevention strategies. Although, these recommendations for improvement have been illustrated using NSW hospitalisation data, they have national and international implications as ICD is one of the most widely used classification system for hospital morbidity records.23
Practical applications • To develop targeted injury prevention strategies for hospitalised sport- or leisure-related struck by/struck against injuries additional information is required. • Both the activity and location codes in the existing ICD-10-AM classification system require further enhancement to allow for additional detail to be collected. • Use of the International Classification of External Causes of Injuries would augment hospital injury-related morbidity data currently using ICD-10-AM. • Injury-related hospitalisation data would benefit from the addition of narrative text fields and from data linkage with other data collections that include information on the circumstances of injury.
Acknowledgements The authors are supported by the NSW Injury Risk Management Research Centre, with core funding provided by the NSW Health Department, the NSW Roads and Traffic Authority and the Motor Accidents Authority. Rebecca Mitchell is also supported by a Ph.D. scholarship from Injury Prevention and Control Australia. The authors wish to thank the Centre for Epidemiology and Research at the NSW Health Department for providing the data from the Health Outcomes and Information Statistical Toolkit (HOIST) analysed in this study.
References 1. Australian Government Department of Health and Ageing. Sports safety in Australia. Canberra: Commonwealth of Australia; 2003.
Getting more out of being struck 2. World Health Organization. ICD-9 International Classification of Diseases, 9th revision. Geneva: World Health Organization; 1977. 3. World Health Organization. ICD-10 International Classification of Diseases, 10th revision. Geneva: World Health Organization; 1992. 4. National Centre for Classification in Health. ICD-10-AM. 1st ed. Sydney: National Centre for Classification in Health; 1998. 5. National Centre for Classification in Health. ICD-10-AM. 2nd ed. Sydney: National Centre for Classification in Health; 2000. 6. National Centre for Classification in Health. ICD-10-AM. 3rd ed. Sydney: National Centre for Classification in Health; 2002. 7. National Centre for Classification in Health. ICD-10-AM. 4th ed. Sydney: Commonwealth Department of Health; 2004. 8. Langley J, Chalmers D. Coding the circumstances of injury: ICD-10 a step forwards or backwards? Inj Prev 1999;5:247—53. 9. MacIntyre C, Ackland M, Chandraraj E. Accuracy of injury coding in Victorian hospital morbidity data. Aust N Z J Public Health 1997;21(7):779—83. 10. Population Health Division. The Health of the People of NSW. Report of the Chief Health Officer. Sydney: NSW Health Department; 2004. 11. SAS Institute. SAS: statistical software, version 8.02. Cary, North Carolina: SAS Institute; 2000. 12. Centre for Epidemiology and Research. Health Outcomes and Information Statistical Toolkit (HOIST). NSW Department of Health; 2006. 13. Dobson A, Kuulasmaa K, Eberle E, Schere J. Confidence intervals for weighted sums of Poisson parameters. Stat Med 1991;10:457—762.
505 14. Armitage P, Berry G, Matthews J. Statistical methods in medical research. 4th ed. Cornwell: Blackwell Science; 2002. 15. WHO Collaborating Center on Injury Surveillance. International Classification of External Causes of Injury World Health Organization; 2005; Available from: URL:http:// www.iceci.org/. [cited 29 June 2005]. 16. Graitcer PL. Injury surveillance. In: Halperin W, Baker E, editors. Public health surveillance. New York: John Wiley & Sons Inc.; 1992. p. 142—56. 17. Horan JM, Mallonee S. Injury surveillance. Epidemiol Rev 2003;25:24—42. 18. Clark DE. Practical introduction to record linkage for injury research. Inj Prev 2004;10(3):186—91. 19. Lincoln AE, Sorock GS, Courtney TK, Wellman HM, Smith GS, Amoroso PJ. Using narrative text and coded data to develop hazard scenarios for occupational injury interventions. Inj Prev 2004;10(4):249—54. 20. Williamson A, Feyer A-M, Stout N, Driscoll T, Usher H. Use of narrative analysis for comparisons of the causes of fatal accidents in three countries: New Zealand, Australia, and the United States. Inj Prev 2001;7(Suppl. I):i15— 20. 21. Muscatello D, Churches T, Kaldor J, Zheng W, Chiu C, Correll P, et al. An automated, broad-based, near real-time public health surveillance system using presentations to hospital Emergency Departments in New South Wales, Australia. BMC Public Health 2005;4(141). 22. Australian Bureau of Statistics. Participation in Sport and Physical Activities. (Cat. No. 4177.0) Canberra: ABS; 2000. 23. Harrison J. Injury classification: balancing continuity and utility. Inj Control Saf Promot 2000;7(1):51—63.