Hazard (Edition No. 74) Summer 2012 Victorian Injury Surveillance Unit (VISU) www.monash.edu/miri/visu Monash Injury Research Institute (MIRI)

Adult injury hospitalisations in 16 sports: the codes, other team ball sports, team bat and stick sports and racquet sports

Authors: Erin Cassell, Emily Kerr, Angela Clapperton The Victorian Injury Surveillance Unit is now an independent unit within the newly created Monash Injury In this issue the results of an investigation into the Research Institute (MIRI) at Monash University, operating frequency, incidence, trend and pattern of adult sports- separately from the Monash University Accident Research related injury hospitalisations in four groups of popular Centre (MUARC) which is also part of MIRI. MUARC sports are reported. The 16 sports covered are: football will now concentrate solely on research related to codes (Australian football, combined & league transport injury, whereas VISU is clustered with other and soccer), other team ball sports (, and centres and units working in the home, and leisure ), team bat and stick sports (combined & stream. For more information on MIRI go to: , and ) and racquet sports (badminton, www.monash.edu.au/miri/. THE NEW WEB table , combined squash & racquetball and tennis). ADDRESS FOR VISU IS: www.monash.edu/miri/visu

• Over the 3-year study period 2007/8 - Racquet sports: badminton, table Summary to 2009/10 there were 12,460 adult tennis, squash & racquetball (combined) hospitalisations for sports injury in the 16 and tennis (4%, n=531). For the main analysis, adult (defined as sports: Cases included adults (aged 15 years persons aged 15 years and older) sports injury - Football codes: Australian football, and older) injured in indoor, outdoor, hospitalisations in four groups of sports were rugby union and league (combined) and organised and unorganised participation. selected from the Victorian Admitted Episodes soccer (67%, n=8,357); Dataset (VAED) for the 3-year period July - Other team ball sports: basketball, • The sport with the highest frequency of 1, 2007 to June 30 2010. The VAED covers netball and volleyball (18%, n=2,338); hospitalisations was Australian football admissions to all public and private hospitals - Team bat and stick sports: baseball & (50%, n=6,275), followed by soccer (13%, in . For trend analysis, data for the softball (combined), cricket and hockey n=1,662), basketball (11%, n=1,317), eight-year period 2002/3 to 2009/10 were (10%, n=1,234); and netball (8%, n=937) and cricket (7%, utilised.

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 1 n=832). The 3-year mean hospitalisation hospitalisations, the peak age group was specific body sites among Australian rate (per 100,000 participants) was also 25-34 years followed by 35-44 years. By football hospitalisations were the wrist and highest for Australian football (1,001 contrast, the peak age group for squash/ hand and the head and face (both 23%), hospitalisations per 100,000 Australian racquet ball and tennis hospitalisations followed by the knee and lower leg (20%). football participants), followed by hockey was 35-44 years, closely followed by 45- In cricket and hockey the wrist and hand (329/100,000), soccer (232/100,000), 54 years. Over half of table tennis-related was the most commonly injured body site, basketball (219/100,000), netball hospitalisations were aged 65 years and accounting for 36% of hospitalisation in (156/100,000) and cricket (134/100,000). older. Participation data by age group both sports. In table tennis, the elbow and The hospitalisation rates for rugby union/ were not available for individual sports forearm and hip and thigh were the most league and baseball/softball were not so we could not estimate and compare commonly injured body sites (both 23% of estimated because of the lack of reliable hospitalisation rates. table tennis hospitalisations). participation data from the oversampled Victorian component of the Australian • Hit/struck/crush, mainly due to impact • Fracture was the most common injury ERASS sports participation survey with another person, was the most leading to hospitalisation in the three (ERASS participation data were supplied common cause of injury among Australian football codes (Australian football [53%], by Sport and Recreation Victoria - SRV). football hospitalisations (accounting rugby union & league [49%] and soccer for 45% of cases) and also caused most [49%]), in two of the three team ball sports • Although the trend in the frequency of hospitalisations in baseball/softball (basketball [42%] and volleyball [39%]), hospitalisations increased significantly (45%), cricket (52%) and hockey (63%). the four team bat and stick sports (baseball over the 8-year period 2002/3 to 2009/10 However, hit/struck/crush injury in the & softball [62%], cricket [52%] and for Australian football, soccer, basketball, three latter sports was mainly due to hockey [52%]) and two of the four racquet netball, cricket and hockey, there was no impacts with sports equipment (usually sports (table tennis [58%] and tennis significant increase in the hospitalisation the ball but also the bat or stick) rather [35%]). The most common injury type rate per 100,000 participants in any of these than with another person. Falls (38%), among netball-related hospitalisations sports. This indicates that the observed mostly due to collisions with/pushing was dislocation, sprain and strain (42%), upward trend in frequency was related to by another person, closely followed by whereas injury to the muscle and tendons increased participation. The frequency hit/struck/crush by other person (37%) was the most frequent injury type among of rugby-related hospitalisations also were the major causes of rugby union/ badminton (63%) and squash/racquetball increased significantly between 2002/3 league hospitalisations. Falls was also the (59%) hospitalisations. and 2009/10 but reliable rate data were not major cause of injury among table tennis available. hospitalisations (75%), but the fall injuries • The average length of stay in hospital was in table tennis were mainly on the same relatively high for table tennis (mean stay • Males comprised the majority of level from tripping, slipping or stumbling. of 2.8 days), badminton (1.7) and tennis hospitalisations in 14 of the 16 sports In the remaining sports, the commonest (1.6) participants and relatively low for studied (range 55% for tennis to 98% for cause of injury hospitalisations was Australian football (0.9) baseball/softball Australian football). The exceptions were ‘other unintentional’ injury, mostly (0.9) and hockey (0.7) participants. netball, where females comprised 85% sprains and strains due to overexertion of hospitalisations, and softball because and strenuous movements — soccer • The ICISS injury scale (a threat-to-life when softball and baseball cases were (36% of hospitalisations), basketball scale) was used to measure the severity of disaggregated then females accounted (49%), netball (63%), volleyball (52%), hospitalisations in the sports studied. The for the majority (61%) of softball-related badminton (74%), squash/racquetball sport that scored highest on the ICISS was hospitalisations. When participation (78%) and tennis (51%). table tennis with 25% of hospitalisations by males and females in the sport is classed as ‘serious’ (i.e. assessed as having taken into consideration (based on the • The upper extremity was the most a 6% chance of death on admission). published ERASS participation data for frequently injured body region among There were 40 hospitalisations for table Victoria), it would appear that males are Australian football (43%), rugby union/ tennis injury. Of the quarter (n=10) more at risk of hospitalisation for injury league (38%), baseball/softball (39%), who suffered ‘serious’ injury, nine were than females when playing basketball, cricket (49%) and hockey (46%) hip fracture cases and the other was a volleyball, hockey and badminton. These hospitalisations, whereas the lower serious injury of ‘other and unspecified results should be interpreted cautiously as extremity was most frequently injured intrathoracic organs’ (i.e. heart or lungs). direct comparison should be made using among hospitalisations in the other sports The proportion of serious cases among the number of hospitalisations per 1,000 (soccer [52%], basketball [33%], netball hospitalisations in other sports ranged player hours and this information is not [66%], volleyball [52%], badminton from 1-2% (netball and basketball, available. [84%], table tennis [40%], squash/ respectively) to 6% (rugby union/league), racquetball [68%] and tennis [54%]). with 4% to 5% of serious cases among • The peak age group for hospitalisations injury hospitalisations related to the other related to the football codes (Australian • The knee and lower leg was the most sports. football, rugby union/league and soccer), commonly injured specific body site in other team ball sports (basketball, netball most sports: rugby union/league (23%), • The total direct cost of injury and volleyball) and team bat and stick soccer (44%), basketball (30%), netball hospitalisations in all 16 sports over the sports (baseball/softball, cricket and (59%), volleyball (35%), baseball/softball 3-year study period was $47.8million. hockey) was 15-24 years, followed by (29%), badminton (83%) and squash Australian football hospitalisations 25-34 years. Among badminton-related (67%). The most frequently injured accounted for the highest proportion

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 2 of this cost (48%, $23.1m), followed by soccer (15%, $7.0m), basketball Methods Results (11%, $5.0m) and netball (9%, $4.1m.). However, table tennis hospitalisations Case selection Over the 3-year period 2007/8 to 2009/10, there were 12,460 hospitalisations of adults were the most costly on average ($6,816 Data were extracted from the VAED – Injury per case), followed by tennis ($4,557), aged 15 years and older for sports injury in Surveillance held by VISU for the fiscal years 16 selected popular sports – football codes netball ($4,351) and soccer ($4,187) 2007/8 to 2009/10 for the main analysis and hospitalisations. (Australian football, combined rugby union & 2002/03 to 2009/10 for trends. league and soccer including festal), other team ball sports (basketball, netball and volleyball), The discussion includes summaries of the main Cases were selected if the activity at the time findings of this research, the risk factor research team bat and stick sports (combined baseball of injury was one of the selected popular & softball, cricket and hockey) and racquet that has been undertaken in the sports covered sports AND the person was 15 years or older in this report and of the current evidence on sports (badminton, table tennis, combined AND the first (principal) diagnosis code squash & racquetball and tennis). In sports the efficacy and effectiveness of sports injury was an injury AND the admission was not a prevention measures. Recommendations are injury hospitalisations data, case counts transfer between hospitals or a change of care include persons injured in organised and made on actions that can be taken to reduce type within a hospital. See Box 1 for further sports injury and improve the completeness unorganised games as they are not separately details. Cases included persons injured in coded. and quality of sports injury data. indoor, outdoor, organised and unorganised participation. 1. Football injury Introduction Rates and trends hospitalisations: Hospitalisation rates were estimated for Sports injuries are common but mostly sports with reliable participation data from Australian football, preventable. Injuries requiring hospital the oversampled ERASS participation survey admission (also called hospitalisations) are (held by Sport and Recreation Victoria, SRV). rugby union and at the more serious end of the sports injury The participation data were provided to VISU and are generally more costly to the league combined by SRV. The over-sampled ERASS was health system and the injured individual. This too small to reliably estimate participation report describes adult injury hospitalisations and soccer (N=8,357) in rugby union and league (combined) and in four groupings of sports. The groupings baseball & softball (combined). Trends were are based on the classification of sports and 1.1 Frequency and rate determined using a log-linear regression leisure activities in the WHO International model of the rate data assuming a Poisson Classification of Diseases Version 10 There were 28,366 adult hospital-treated distribution of injury cases. A trend was Australian Modifications (ICD-10-AM) that football injury cases (including all three codes) considered to be statistically significant if the is used to classify hospitalisations recorded in Victoria over the 3-year period 2007/8- p-value of the slope of the regression model on the Victorian Admitted Episodes Dataset 2009/10. The majority (70%, n=20,009) was less than 0.05 (See box 2). (VAED). The groups include most of the were treated in the ED and discharged. Thirty popular organised sports in Victoria: percent (n=8,357) were admitted to hospital Injury severity and cost of and data on these more serious cases were • Football (Australian football, combined hospitalisations analysed further. rugby union & league and soccer) See Box 2 for definitions and methods.

• Other team ball sports (basketball, Table 1A Hospital-treated injury in football codes (Australian football, Rugby netball and volleyball) union/league and Soccer) Victoria, July 1, 2007 to June 30, 2010

• Team bat and stick sports (combined HOSPITAL-TREATED INJURY IN FOOTBALL CODES football1 Rugby Union & Soccer3 All baseball & softball, cricket and hockey) League combined2 N % N % N % N % • Racquet sports (badminton, table tennis,\ Hospitalisations 6,275 30.2 420 33.7 1,662 26.1 8,357 29.5 combined squash & racquetball and tennis). ED presentations - 14,469 69.8 828 66.3 4,712 73.9 20,009 70.5 non-admissions The frequency, pattern and trend of injury All hospital- 20,744 100.0 1,248 100.0 6,374 100.0 28,366 100.0 hospitalisations in the 16 sports included in treated cases these four groups are described in this report. Notes: Data on the number of Emergency Department 1 All ‘football unspecified’ cases were included in Australian football category on the presumption that (ED) presentations (non-admissions) for injury patients playing soccer and rugby would specify they were playing these codes in Victoria. This may not in each sport are included in the first table of be so for all soccer players, so there may be some misclassification. 2 Rugby includes union, league and unspecified rugby cases the section on each sport, but ED presentations 3 Soccer includes outdoor and indoor () cases data are not analysed further. Source: Victorian Admitted Episodes Dataset (VAED - Hospital admissions) Victorian Emergency Minimum Dataset (VEMD - Emergency Department presentations, non-admissions)

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 3 Over the 3-year study period the large majority of football-related hospitalisations occurred Figure 1 Trend in the frequency of football injury hospitalisations: Australian in Australian football (75%, n=6,275), football, rugby union/league and soccer, Victoria 2002/3 to 2009/10 followed by soccer (20%, n=1,662) and rugby union/league (5%, n=420). The 3-year mean hospitalisations rate for Australian football was more than 4 times the rate for soccer (1,000 hospitalisations per 100,000 Australian football participants compared with 232 hospitalisations per 100,000 soccer participants). The hospitalisation rate for rugby union/league could not be estimated as participation data were unreliable due to small numbers.

1.2 Trends Trend in the frequency of football injury hospitalisations, 2002/3 to 2009/10 Figure 1 shows the trend in frequency of injury hospitalisations for the three football codes Source: Victorian Injury Surveillance Unit (VISU) over the 8-year period 2002/3 to 2009/10. Figure 2 Trend in football injury hospitalisations rate per 100,000 participants: Trend analysis showed: Australian football and soccer, Victoria 2002/3 to 2009/10

• Australian football injury hospitalisations increased significantly from 1,693 in 2002/3 to 2,098 in 2009/10, representing an estimated annual change of 3.6% (95% confidence intervals 1.8% to 5.3%) and an overall increase of 33% (15% to 51%);

• Rugby union/league injury hospitalisations increased significantly from 109 in 2002/3 to 138 in 2009/10, representing an estimated annual change of 5.5% (2.3% to 8.4%) and an overall increase of 54% (20% to 91%); and

• Soccer injury hospitalisations increased significantly from 458 in 2002/3 to 558 in 2009/10, representing an estimated annual change of 5.1% (2.4% to 7.5%) and an overall increase of 49% (21% to 79%). Source: Victorian Injury Surveillance Unit (VISU)

Trend in the rate of football injury 1.3 Detailed analysis hospitalisations per 100,000 participants, males and females participating in these 2002/3 to 2009/10 As summarised in Tables 1B & C, analysis of hospital admission data for the three football sports according to the ERASS 2008 and 2009 surveys. The trend in injury hospitalisation rates codes over the 3-year study period (2007/8 to per 100,000 participants for Australian 2009/10) showed: • The age profile of Australian football and football and soccer over the 8-year period rugby union/league hospitalisations was July 2002-June 2010 are shown in Figure • Males comprised 98% and 95% of similar, with two-thirds of hospitalisations 2. Trend analysis showed that there was no Australian football and rugby union/league in both these codes aged between 15 and significant change in injury hospitalisation injury hospitalisations, respectively, and 24 years. By comparison, half of the rates for either Australian football or soccer 89% of soccer hospitalisations. We did not hospitalised soccer players were aged 25 over the study period. This indicates that have oversampled Victorian participation years and older. Victorian participation increased participation explained most of data by gender so could not estimate and data by age group is not available so age- the significant increase in the frequency of directly compare hospitalisation rates for related hospitalisation rates could not injury hospitalisations observed among both males and females in the three football be estimated and compared within and Australian football and soccer participants. codes However, these results appear to be fairly consistent with the proportion of between sports.

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 4 • Hit/struck/crush was the most Table 1B Demographics and major causes of football injury hospitalisations common cause of Australian football (Australian football, rugby and soccer), Victoria July 1, 2007 to hospitalisations (45%), whereas falls on June 30, 2010 the same level, mostly due to collisions with/pushing by another person, was FOOTBALL-RELATED HOSPITALISATIONS (N=8,357) the most common cause of rugby union/ Australian football1 Rugby union/league2 Soccer3 league hospitalisations (38%). Other (n=6,275) (n=420) (n=1,662) unintentional injury (mostly overexertion Gender • Males (98%) • Males (95%) • Males (89%) and strenuous movements) was the most • Females (2%) • Females (5%) • Females (11%) common cause of soccer hospitalisations Age • 15-24 (65%) • 15-24 (66%) • 15-24 (49%) (36%). • 25-34 (26%) • 25-34 (24%) • 25-34 (32%) • 35-44 (7%) • 35-44 (7%) • 35-44 (13%) • The upper extremity was the most • 45-54 (2%) • 45-54 (3%) • 45-54 (5%) frequently injured body region in Australian football and rugby union/ • 55-64 (0.2%) • 55-64 (0%) • 55-64 (1%) league (comprising 43% and 38% of • 65+ (0.1%) • 65+ (0%) • 65+ (0.2%) hospitalisations respectively), whereas the Cause • Hit/struck/crush {45%, • Falls {38%, mostly fall on • Other unintentional {36%, lower extremity was the most frequently mostly hit/struck/kick same level due to collision mostly overexertion & injured body region in soccer (52% of by other person (58%) and with/pushing by another strenuous movements striking against/bumped person (76%) and other fall (43%)} hospitalisations). Around half the upper into by another person on same level (11%)} extremity injuries in Australian football (25%)} • Hit/struck/crush {35%, and rugby union and league were to the • Hit/struck/crush {37%, mostly hit/struck/kick by wrist and hand. Eighty-six per cent of the • Other unintentional {32%, mostly hit/struck/kick other person (58%), striking lower extremity injuries in soccer were to mostly overexertion & by other person (66%) and against or struck by the knee and lower leg. strenuous movements striking against/bumped sporting equipment, mostly (21%)} into by another person ball (20%) and striking • The specific body sites most frequently (29%)} against/bumped into by • Falls {22%, mostly fall on another person (18%)} injured in Australian football same level due to collision • Other unintentional {24%, hospitalisations were the wrist and with/pushing by another mostly overexertion & • Falls {28%, mostly other hand and the head and face (both 23%), person (54%) and other fall strenuous movements fall on same level (36%), followed by the knee and lower leg (20%). on same level (27%)} (28%)} fall on same level due to By contrast, the knee and lower leg was collision with/ pushing by the most frequently injured specific body • Intentional inflicted by • Intentional inflicted by another person (24%) site among rugby union/league (26%) and other {1%, mostly assault\ other {1%, assault by and fall on same level soccer (44%) hospitalisations, followed by bodily force (97%)} bodily force (100%)} from slipping, tripping or stumbling (17%)} by the head and face (19% and 16% respectively) and the wrist and hand (18% • Intentional inflicted by and 11% respectively). (Not shown in other (1%) tables.) Notes: 1 • In all three football codes, fractures All ‘football unspecified’ cases were included in Australian football category 2 Rugby includes union, league and ‘rugby unspecified’ cases accounted for about half of hospitalisations, 3 Soccer includes outdoor and cases but the pattern was different with regard to Source: Victorian Admitted Episodes Dataset the fracture site. In Australian football and rugby union/league, the wrist and hand injuries were the next most common injury was the most common site of fracture, in rugby (8%) and Australian football accounting for 32% of Australian football (7%), whereas for soccer the next most fracture cases (17% of all Australian common injury was muscle/tendon injury football hospitalisations) and 32% of (10%). Intracranial injuries accounted for rugby union/league fracture cases (15% of 5% of soccer injury hospitalisations. all /union hospitalisations). By contrast, the knee and lower leg was • Length of stay in hospital can be used as the most common fracture site in soccer, a proxy measure of injury severity. The accounting for 39% of soccer fracture cases average length of stay was highest for and 19% of all soccer hospitalisations. soccer cases, with an average of 1.3 days Dislocations, sprains and strains was the stay compared with 1.0 for rugby and next most common type of injury in all 0.9 for Australian football. Twenty-five three football codes, accounting for 22% percent of soccer players had a stay of 2-7 of Australian football cases, 20% of rugby days compared with 18% of Australian union/league cases and 25% of soccer football players and 20% of rugby players. cases. The knee and lower leg was the most In all three codes only 1% of cases stayed frequent site of dislocation/sprains/strains 8 or more days. in each of the football codes. Intracranial

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 5 Table 1C Injury type, body region injured and injury severity among football • Injury severity was also measured using injury hospitalisations (Australian football, rugby and soccer), the International Classification of Disease Victoria July 1, 2007 to June 30, 2010 (ICD)-based Injury Severity Score (ICISS) which is a ‘threat to life’ scale. Five FOOTBALL-RELATED HOSPITALISATIONS (N=8,357) percent of Australian football cases, 6% of Australian football1 Rugby union/league2 Soccer3 (n=6,275) (n=420) (n=1,662) rugby cases and 4% of soccer cases were Body site • Upper extremity {43%, • Upper extremity {38%, • Lower extremity {52%, classed as ‘serious’ on this scale i.e. on injured mostly wrist & hand (53%) mostly wrist & hand (47%), mostly knee & lower leg admission they had at least a 6% threat-to- and shoulder & upper arm shoulder & upper arm (86%) and ankle & foot life (i.e. chance of death). (27%)} 32%) and elbow & forearm (10%)} • Head/face/neck {25%, (21%)} • Upper extremity {27%, • The total direct cost of all hospitalisations mostly head/face (91%)} • Lower extremity {29%, mostly wrist & hand (42%) for football injury over the three study years • Lower extremity {23%, mostly knee & lower leg and elbow & forearm was $31.6m: Australian football ($23.1m), mostly knee & lower leg (88%)} (38%)} rugby ($1.5m) and soccer ($7.0m). The (86%)} • Head/face/neck {26%, • Head/face/neck {17%, average cost per hospitalisation was • Trunk (8%) mostly head/face (72%) and mostly head /face (92%)} highest for soccer cases, $4,187 per • Other and unspecified (1%) neck (28%)} • Trunk (4%) admission, compared with $3,677 per • Trunk (7%) • Other & unspecified (1%) admission for Australian football cases Nature • Fracture {53%, mostly • Fracture {49%, mostly • Fracture {48%, mostly knee and $3,652 per admission for rugby cases. of injury wrist & hand (32%), head wrist & hand (32%), knee & g & lower le (39%), elbow & including face (22%) and lower leg (30%) and elbow forearm (19%), wrist knee & lower leg (18%)} & forearm (13%)} & hand (18%) and head • Dislocation, sprain & strain • Dislocation, sprain & strain including face (15%)} 2. Injury {22%, mostly knee & lower {20%, mostly knee & lower • Dislocation, sprain & strain leg (39%), shoulder & leg (40%) and shoulder & {25%, mostly knee & lower hospitalisations upper arm (29%) and wrist upper arm (37%)} leg (63%) and shoulder & & hand (21%)} • Intracranial (8%) upper arm (15%)} in ‘other team ball • Intracranial (7%) • Injury to muscle & tendon • Injury to muscle & tendon • Injury to internal organs (4%) (5%) (10%) sports’: basketball, • Injury to muscle & tendon • Open wound (3%) • Intracranial (5%) (3%) • Injury to internal organs • Open wound (3%) netball and volleyball • Open wound (2%) (2%) • Superficial injury (2%) • Superficial injury (2%) • Superficial injury (1%) • Injury to internal organs (2%) (N=2,338) • Other & unspecified (8%) • Other & unspec. (11%) • Other & unspec. (7%) 2.1 Frequency and rate Length • Less than 2 days (81%) • Less than 2 days (79%) • Less than 2 days (74%) of stay • 2-7 days (18%) • 2-7 days (20%) • 2-7 days (25%) • 8-30 days (1%) • 8-30 days (1%) • 8-30 days (1%) There were 11,135 adult hospital-treated • 31+ days (0.02%) injury cases in ‘other team ball sports’ in Average length of stay: 0.89 Average length of stay: 1.04 Average length of stay: 1.25 Victoria over the 3-year study period 2007/8 days (range: 0 to 42 days) days (range: 0 to 12 days) days (range: 0 to 19 days) to 2009/10, an average of 3,712 per year (Table 2A). Eighty percent of cases (n=8,797) Injury • Serious i.e. high threat-to • Serious i.e. high threat-to • Serious i.e. high threat-to were treated in the ED and discharged home. 4 severity life (5%, n=331) life (6%, n=24) life (4%, n=59) Twenty percent (n=2,338) were admitted - Multiple injuries (28%) - Multiple injuries (29%) - Multiple injuries (27%) to hospital, an average of 779 per year; data - Injury of intra-abdominal - Injury of intra-abdominal - Intracranial injury (22%) organs (16%) organs (21%) - Fracture of skull and facial on these more serious cases were analysed - Intracranial injury (13%) - Intracranial injury (8%) bones (9%) further. - Fracture of skull and facial - Fracture of rib(s), sternum - Injury of other and bones (9%) and thoracic spine (8%) unspecified intrathoracic Basketball participants accounted for over - Injury of urinary and organs (7%) half of the 2,338 hospitalisations (56%, pelvic organs (9%) - Injury of intra-abdominal n=1,317), followed by netball (40%, n=937) organs (7%) and volleyball (4%, n=84) participants. - Fracture of femur (7%) The 3-year mean injury hospitalisation Direct Total cost: $23.1m Total cost: $1.5m Total cost: $7.0m rate was highest for basketball (219 injury cost of Mean cost : $3,677 Mean cost; $3,652 per Mean cost: $4,187 hospitalisations/100,000 participants), admissions (range: $790 to $79,657) (range: $790 to $18,332) (range: $790 to $18,995) followed by netball (156/100,000) and volleyball (55/100,000). Notes: 1 All ‘football unspecified’ cases were included in the Australian football category 2 Rugby includes union, league and rugby unspecified cases 3 Soccer includes outdoor and indoor soccer cases 4 The final injury severity score is the product of the severity score of each of the diagnoses in the patient’s case record. In some cases a single injury diagnosis code is sufficient to attract the classification of ‘serious’, in others it is the combination of injury diagnoses that attracts this classification. Where the classification is based on one diagnosis the specific diagnosis is listed, where the classification is due to multiple diagnoses the term ‘multiple injuries’ is used. Source: Victorian Admitted Episodes Dataset (VAED)

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 6 Table 2A Hospital-treated injury in ‘other team ball sports’ (basketball, netball and volleyball), Victoria July 1, 2007 to June 30, 2010 HOSPITAL-TREATED INJURY IN ‘OTHER TEAM BALL’ SPORTS Basketball Netball Volleyball All

N % N % N % N % Hospital admissions 1,317 21.7 937 20.0 84 21.8 2,338 21.0 ED presentations - 4,759 78.3 3,737 80.0 301 78.2 8,797 79.0 non-admissions All hospital- 6,076 100.0 4,674 100.0 385 100.0 11,135 100.0 treated cases Source: Victorian Admitted Episodes Dataset (VAED - Hospital admissions) Victorian Emergency Minimum Dataset (VEMD - Emergency Department presentations, non-admissions)

Figure 3 Trend in the frequency of ‘other team ball sports’ injury hospitalisations: 2.2 Trends basketball, netball and volleyball, Victoria 2002/3 to 2009/10 Trend in frequency of ‘other team ball sports’ injury hospitalisations, 2002/3 to 2009/10

Figure 3 shows the trend in frequency of injury hospitalisations for basketball, netball and volleyball over the 8-year period 2002/3- 2009/10.

Trend analysis showed that:

• basketball injury hospitalisations increased significantly from 339 in 2002/3 to 431 in 2009/10, representing an estimated annual change of 4.8% (95% confidence intervals 2.2% to 7.2%) and an overall increase of 46% (19% to 75%);

• netball injury hospitalisations increased significantly from 282 in 2002/3 to 312 in 2009/10, representing an estimated annual Source: Victorian Injury Surveillance Unit (VISU) change of 2.7% (1.3% to 4.0%) and an overall increase of 24% (11% to 37%); and Figure 4 Trend in ‘other team ball sports’ injury hospitalisations rate per 100,000 participants: basketball, netball and volleyball, Victoria 2002/3 to 2009/10 • there was no significant change in the frequency of volleyball injury hospitalisations between 2002/3 and 2009/10.

Trend in the rate of ‘other team ball sports’ injury hospitalisations per 100,000 participants, 2002/3 to 2009/10

The trend in injury hospitalisation rates per 100,000 participants for basketball, netball and volleyball over the period 2002/3 to 2009/10 are shown in Figure 4. Trend analysis showed that there was no significant change in injury hospitalisation rates in any of the three sports, indicating that the significant increases in the frequency of hospitalisation found for basketball and netball were due to increased participation. Source: Victorian Injury Surveillance Unit (VISU)

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 7 age-related hospitalisation rates could not Table 2B Demographics and major causes of injury hospitalisations in ‘other be estimated and rates compared within team ball sports’ (basketball, netball and volleyball), Victoria July 1, and between sports. Sport participation 2007 to June 30, 2010 data by age for Australia as a whole (from OTHER TEAM BALL SPORTS-RELATED HOSPITALISATIONS (N=2,338) published ERASS 2008 and 2009 surveys) Basketball Netball Volleyball indicates that only around 6% of volleyball (n=1,317) (n=937) (n=84) participants are aged 45 years and older, Gender • Males (76%) • Males (15%) • Males (74%) which suggests that older participants • Females (24%) • Females (85%) • Females (26%) are over-represented in volleyball-related hospitalisations in Victoria. Age • 15-24 (49%) • 15-24 (36%) • 15-24 (33%) • 25-34 (30%) • 25-34 (32%) • 25-34 (29%) • ‘Other unintentional injury’ (mostly • 35-44 (15%) • 35-44 (25%) • 35-44 (11%) overexertion and strenuous movements) • 45-54 (6%) • 45-54 (6%) • 45-54 (21%) was the most common cause (mechanism) • 55-64 (1%) • 55-64 (1%) • 55-64 (5%) of injury in all three ‘other team ball’ • 65+ (0.3%) • 65+ (1%) • 65+ (1%) sports, accounting for 49% of basketball Causes • Other unintentional {49%, • Other unintentional {63%, • Other unintentional {52%, hospitalisations, 63% of netball mostly overexertion and mostly overexertion & mostly overexertion & hospitalisations and 52% of volleyball strenuous movements strenuous movements strenuous movements hospitalisations. Hit/struck/crush was (44%)} (48%)} (68%)} the next most common injury cause in basketball (26%) and volleyball (25%). • Hit/struck/crush {26%, • Falls {20%, mostly other • Hit/struck/crush {25%, These injuries were mostly due to being mostly hit/struck/kick by fall on same level (42%), mostly striking against other person (47%), striking fall on same level from or struck by sporting hit/struck/crushed or bumped by another against or struck by sporting slipping, tripping or equipment (48%) and hit/ person in basketball, whereas they were equipment, mostly ball stumbling (23%) and fall on struck/kick by other person mostly due to impacts with sporting (29%) and striking against/ same level - collision with (38%)} equipment in volleyball, followed by hit/ bumped into by another pushing by another person struck by another person. Falls was the person (20%)} (10%)} • Falls {20%, mostly other next most common injury in netball (20%). fall on same level (41%) • Falls {25%, mostly other • Hit/struck/crush {17%, and fall on same level • The lower extremity was the most fall on same level (48%), mostly hit/struck/kick from slipping, tripping or frequently injured body region in all fall on same level - collision by other person (41%), stumbling (12%)} three ‘other team ball sports’, comprising with pushing by another striking against or struck by 43% of basketball hospitalisations, 66% person (15%) and fall on sporting equipment, mostly • Other & unspecified (3%) same level from slipping, ball (38%) and striking of netball hospitalisations and 52% of tripping or stumbling against/bumped into by volleyball hospitalisations. The majority of (13%)} another person (16%)} lower extremity injuries in all three sports were to the knee and lower leg: basketball • Intentional inflicted by other (83%), netball (90%) and volleyball (0.3%) (66%). Source: Victorian Admitted Episodes Dataset • In all three sports, the knee and lower leg was also the specific body site most 2.3 Detailed analysis respectively based on participation data frequently injured, accounting for 36% in the web published ERASS 2008 and of all basketball hospitalisations, 59% of As summarised in Tables 2B & C, analysis of 2009 surveys) were compared with the all netball hospitalisations and 35% of all hospital admission data for the three ‘other proportion of injury hospitalisation that volleyball hospitalisations (Not shown in team ball sports’ over the 3-year study period were male in each of these sport (76%, table). 2007/8 to 2009/10 showed: 15% and 74% respectively), it would appear that male basketball and volleyball • Fracture was the most common type of injury among basketball (42%) and • Males comprised 76% of basketball and 74% players are at higher risk of hospitalisation volleyball (39%) hospitalisations. Fracture of volleyball injury hospitalisations. The for injury than their female counterparts. to the wrist and hand was the most frequent gender pattern for netball hospitalisations fracture injury in basketball, accounting for was the opposite with females comprising • The age profile of basketball 35% of basketball fracture cases (15% of 85% of cases, reflecting their much higher hospitalisations was younger than the all basketball hospitalisations) and fracture participation in this sport. Oversampled profile in the other two sports with 49% to the knee and leg was the most frequent ERASS Victorian sports participation of hospitalised basketball players aged fracture injury in volleyball accounting data disaggregated by gender were not between 15 and 24 compared with 36% for 36% of volleyball fracture cases available so we could not reliably estimate of netball players and 33% of volleyball (14% of all volleyball hospitalisations). and directly compare hospitalisation players. Twenty-seven percent of Dislocation, sprain and strain to the knee rates for males and females in the three hospitalised volleyball players were aged and lower leg was the most common injury sports. However, when the proportion of 45 years and older compared with only type in netball, accounting for 71% of males participating in basketball, netball 7% of basketball players and 8% of netball netball dislocation/sprain/strain cases and and volleyball (63%, 17% and 47% players. Victorian sports participation data by age group were not available so 30% of all netball hospitalisations.

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 8 • Volleyball-related hospitalisations had the longest average length of stay in hospital Table 2C Injury type, body region injured and injury severity among injury at 1.25 days, compared with 1.09 days for hospitalisations in ‘other team ball sports’ (basketball, netball and netball cases and 0.99 days for basketball volleyball), Victoria, July 1, 2007 to June 30, 2010 cases. A quarter of netball and volleyball OTHER TEAM BALL SPORTS HOSPITALISATIONS (N=2,338) cases had a stay of 2-7 days compared Basketball Netball Volleyball with 19% of basketball cases. Only 1% of (n=1,317) (n=937) (n=84) basketball and netball cases stayed 8-30 Body site • Lower extremity {43%, • Lower extremity {66%, • Lower extremity {52%, days and no volleyball cases stayed longer injured mostly knee & lower leg mostly knee & lower leg mostly knee & lower leg than 7 days. (83%), ankle & foot (12%) (90%) and ankle & foot (9%)} (66%) and ankle & foot and hip & thigh (4%)} • Upper extremity {26%, (27%)} • Comparing scores on the ICISS, a • Upper extremity {38%, mostly wrist & hand (55%), • Upper extremity {37%, mostly wrist & hand (54%), elbow & forearm (29%) mostly wrist & hand (65%) comparatively higher proportion of elbow & forearm (24%) and and shoulder & upper arm and shoulder & upper arm volleyball hospitalisations (4%) than shoulder & upper arm (22%)} (16%)} (26%)} basketball (2%) or netball hospitalisations • Head/face/neck {16%, • Head/face/neck {5%, mostly • Trunk (6%) (1%) were classed as serious. On the mostly head/face (90%)} head/face (88%)} • Head/face/neck {5%, ICISS scale, patients with serious injury • Trunk (3%) • Trunk (2%) mostly head/face (75%)} have at least a 6% chance of not surviving • Other & unspecified (0.4%) • Other & unspecified (1%) on admission. Nature • Fracture {42%, mostly wrist • Dislocation, sprain & strain • Fracture {39%, mostly knee of injury & hand (35%), knee & {42%, mostly knee & lower & lower leg (36%), wrist • The total direct cost of hospitalisations for lower leg (21%) and elbow leg (71%), wrist & hand & hand (30%) and ankle & injury in the three ‘other team ball sports’ & forearm (19%)} (10%), shoulder & upper foot (12%)} over the three study years was $9.41m: • Dislocation, sprain & strain arm (8%) and ankle & foot • Dislocation, sprain & strain basketball ($5.01m), netball ($4.08m) and {29%, mostly knee & lower (8%)} {37%, mostly knee & lower volleyball ($0.32m). The average cost of a leg (49%), shoulder & • Fracture {26%, mostly wrist leg (36%), wrist & hand hospitalisations over the three study years upper arm (22%) and wrist & hand (38%), elbow & (29%), shoulder & upper was highest for the netball cases ($4,351 & hand (17%)} forearm (26%) and knee & arm (16%) and ankle & foot per admission), compared with $3,807 for • Injury to muscle & tendon lower leg (22%)} (16%)} {14% mostly knee & lower • Injury to muscle & tendon • Injury to muscle & tendon basketball cases and $3,780 for volleyball leg (85%)} {24%, mostly knee & lower {12%, mostly knee & lower cases. • Intracranial (3%) leg (92%)} leg (60%) and hip & thigh • Open wound (2%) • Intracranial (2%) (20%)} • Superficial injury (1%) • Superficial injury (1%) • Open wound (5%) • Eye injury (1%) • Open wound (1%) • Injury to internal organs (1%) • Other & unspecified (7%) • Other & unspecified (5%) • Other & unspecified (6%)

Length • Less than 2 days (80%) • Less than 2 days (75%) • Less than 2 days (74%) of stay • 2-7 days (19%) • 2-7 days (25%) • 2-7 days (26%) • 8-30 days (1%) • 8-30 days (1%) • 8-30 days (0%) Average length of stay: 0.99 Average length of stay: 1.09 Average length of stay: 1.25 days (range: 0 to 16 days) days (range: 0 to 17 days) days (range: 0 to 7 days)

Injury • Serious injury i.e. high • Serious injury i.e. high • Serious injury i.e. high severity 1 threat-to-life (2%, n=27) threat-to-life (1%, n=9) threat-to-life (4%, n=3) - Multiple injuries (22%) - Multiple injuries (22%) - Multiple injuries (66%) - Fracture of skull and - Fracture of femur (22%) - Fracture of femur (33%) facial bones (15%) - Other injuries of spine - Intracranial injury (15%) and trunk, level - Injury of nerves and spinal unspecified (22%) cord at neck level (7%) - Intracranial injury (11%) - Fracture of lumbar spine - Injury of nerves and spinal and pelvis (7%) cord at neck level (11%) - Injury of intra-abdominal - Injury of intra-abdominal organs (7%) organs (11%) - Fracture of femur (7%)

Direct Total: $5.01m Total: $4.08m Total: $0.32m cost of Mean: $3,807 Mean: $4,351 Mean: $3,780 admissions (range: $790 to $18,412) (range: $870 to $18,412) (range: $1,230 to $18,412)

Notes: 1 The final injury severity score is the product of the severity score of each of the diagnoses in the patient’s case record. In some cases a single injury diagnosis code is sufficient to attract the classification of ‘serious’, in others it is the combination of injury diagnoses that attracts this classification. Where the classification is based on one diagnosis the specific diagnosis is listed, where the classification is due to multiple diagnoses the term ‘multiple injuries’ is used. Source: Victorian Admitted Episodes Dataset (VAED)

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 9 3. Injury Table 3A Hospital-treated injury in team bat and stick sports (baseball & softball hospitalisations in combined, cricket and hockey), Victoria July 1, 2007 to June 30, 2010 HOSPITAL-TREATED INJURY IN TEAM BAT AND STICK SPORTS team bat and stick Baseball & Cricket Hockey All Softball sports: baseball & N % N % N % N % softball (combined), Hospital admissions 115 26.0 832 23.6 287 20.6 1,234 23.0 ED presentations - 327 74.0 2,698 76.4 1,109 79.4 4,134 77.0 cricket and hockey non-admissions All hospital- 442 100.0 3,530 100.0 1,396 100.0 5,368 100.0 (N=1,234) treated cases Source: 3.1 Frequency and rate Victorian Admitted Episodes Dataset (VAED - Hospital admissions) Victorian Emergency Minimum Dataset (VEMD - Emergency Department presentations, non-admissions) As shown in Table 3A, there were 5,368 adult hospital-treated injury cases related to team Figure 5 Trend in the frequency of team bat and stick sports injury hospitalisations: bat and stick sports in Victoria over the 3-year baseball & softball combined, cricket and hockey, Victoria 2002/3 to 2009/10 period 2007/8 to 2009/10, an average of 1,789 per. year Of these cases, 23% (n=1,234) were admitted to hospital, an average of 411 per year. Data on the hospitalisations—the more serious cases—were analysed further.

Over the 3-year study period, just over two- thirds of the 1,234 hospitalisations related to team bat and stick sports were for injuries that occurred in cricket (67%, n=832). Twenty- three percent (n=287) were hockey-related and 9% (n=115) were baseball & softball-related. However, the 3-year mean hospitalisations rate for hockey injury was more than twice as high as the rate for cricket injury (329/100,000 hockey participants versus 134/100,000 cricket participants). The hospitalisations rate for baseball & softball injury (combined) could not be calculated because the oversampled ERASS participation estimate for Victoria was Source: Victorian Injury Surveillance Unit (VISU) unreliable.

Figure 6 Trend in bat and stick sports injury hospitalisation rates per 100,000 3.2 Trends participants: cricket and hockey, Victoria 2002/3 to 2009/10 Trend in frequency of team bat and stick sports injury hospitalisations July 1, 2002 to June 30, 2010

Trend analysis for the 8-year period 2002/3 to 2009/10 showed that:

• cricket injury hospitalisations increased significantly from 248 in 2002/3 to 259 in 2009/10, representing an estimated annual change of 3.1% (95% confidence intervals 0.5% to 5.7%) and an overall increase of 28% (4% to 56%);

• hockey injury hospitalisations increased significantly from 71 in 2002/3 to 96 in 2009/10, representing an estimated annual change of 4.1% (1.3% to 6.7%) and an overall increase of 38% (11% to 68%); and Source: Victorian Injury Surveillance Unit (VISU)

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 10 • there was no significant change in the frequency of combined baseball & softball Table 3B Demographics and major causes of injury among injury injury hospitalisations. hospitalisations in team bat or stick sports (baseball & softball combined, cricket and hockey), Victoria July 1, 2007 to June 30, 2010 Trend in the rate of team bat and ball sports injury hospitalisations per 100,000 TEAM BAT OR STICK SPORTS HOSPITALISATIONS (N=1,234) participants, 2002/3 to 2009/10 Baseball & Softball Cricket Hockey (n=115) (n=832) (n=287) The trends in injury hospitalisation rates per Gender • Males (72%) • Males (97%) • Males (75%) 100,000 participants for cricket and hockey • Females (28%) • Females (3%) • Females (25%) over the 8-year period 2002/3 to 2009/10 are Age • 15-24 (44%) • 15-24 (31%) • 15-24 (47%) shown in Figure 6. Reliable participation • 25-34 (24%) • 25-34 (31%) • 25-34 (18%) data were not available for baseball & softball • 35-44 (23%) • 35-44 (24%) • 35-44 (21%) (combined). Trend analysis showed there • 45-54 (9%) • 45-54 (10%) • 45-54 (12%) was no significant change in either hockey • 55-64 (0%) • 55-64 (2%) • 55-64 (3%) and cricket injury hospitalisation rates. This • 65+ (1%) • 65+ (2%) • 65+ (0%) suggests that the significant increase in the frequency of injury hospitalisations observed Causes • Hit/struck/crush {45%, • Hit/struck/crush {52%, • Hit/struck/crush {63%, in both these sports over the period 2002/3 to mostly striking against mostly striking against mostly striking against or struck by sporting or struck by sporting or struck by sporting 2009/10 was due to increased participation. equipment, mostly ball but equipment, mostly ball equipment - ball and stick also bat (79%)} (90%)} (85%) and hit/struck/kick 3.3 Detailed analysis by other person (8%)} • Other unintentional {35%, • Other unintentional {33%, As summarised in Tables 3B & C, analysis of mostly overexertion & mostly overexertion & • Other unintentional {22%, strenuous movements strenuous movements mostly overexertion & hospital admission data for the three team bat (63%)} (50%)} strenuous movements and stick sports over the 3-year study period (40%)} 2007/8 to 2009/10 showed: • Falls {20%, mostly other • Falls {14%, mostly other fall on same level (43%) fall on same level (49%) • Falls {15%, mostly other • Males comprised 72% of the combined and fall on same level and fall on same level fall on same level (33%), baseball and softball injury hospitalisations from slipping, tripping or from slipping, tripping or fall on same level from but when data were disaggregated, stumbling (22%)} stumbling (26%)} slipping, tripping or males comprised 83% of baseball stumbling (21%) and fall on • Other & unspecified (1%) same level - collision with hospitalisations compared with only 39% pushing by another person of softball hospitalisations, reflecting the (17%)} gender difference in participation in these two sports. Almost all (97%) of the cricket • Other & unspecified (1%) injury hospitalisations and three-quarters of the hockey injury hospitalisations were Source: Victorian Admitted Episodes Dataset males. According to sports participation data for Victoria in the published ERASS for 2008 and 2009, an average of 47% • Hit/struck/crush (mostly by sporting whereas in cricket and hockey cases of hockey participants were male, yet equipment – bat/stick/ball) was the most the wrist/hand was most commonly males comprised 75% of hockey-related common cause of injury in all three of the injured (36% of both cricket and hockey hospitalisations between 2007/8 and team bat and stick sports, accounting for hospitalisations). (Not shown in tables.) 2009/10. These results indicate that males 45% of baseball & softball admissions, are more at risk of injury in hockey than 52% of cricket admissions and 63% of • Fracture was the most common injury females. Participation data by gender hockey admissions. ‘Other unintentional type for all three sports groups, accounting were not reliable for the other sports. injury’ (mostly overexertion and strenuous for between 52% and 62% of injury movement) was the next most common hospitalisations in bat and stick sports. • The age pattern of hospitalisations cause of injury in all three sports, However, the pattern was different with in the team bat and stick sports was accounting for 35% of baseball/softball regard to fracture site. Among baseball fairly similar with 68% of baseball & cases, 33% of cricket cases and 22% of & softball hospitalisations, fracture to softball hospitalisations, 62% of cricket hockey cases. the knee/lower leg was the most common hospitalisations and 65% of hockey fracture, accounting for 30% of baseball & hospitalisations aged between 15 and 34 • In all three team bat and stick sports the softball fracture cases (18% of all baseball years. Sports participation data by age upper extremity was the body region most & softball hospitalisations). By contrast, group for Victoria were not available so frequently injured: baseball & softball fracture to the wrist/hand was the most age-related hospitalisation rates could not (39% of cases), cricket (49%) and hockey common fracture in cricket and hockey, be estimated and compared within and (46%). However, the specific body site accounting for 51% of cricket fracture between sports. most commonly injured in baseball & cases (27% of all cricket hospitalisations) softball cases was the knee/lower leg (29% and 57% of hockey fracture cases (30% of of baseball & softball hospitalisations), all hockey hospitalisations). Dislocations,

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 11 sprains and strains were the next most Table 3C Injury type, body region injured and injury severity among injury common type of injury in all three sports, hospitalisations in bat and stick sports (baseball & softball accounting for 16% of baseball & softball combined, cricket and hockey), Victoria July 1, 2007 to June 30, 2010 cases, 21% of cricket cases and 13% of TEAM BAT OR STICK SPORTS HOSPITALISATIONS (N=1,234) hockey cases. Baseball & softball Cricket Hockey (n=115) (n=832) (n=287) • Cricket players had the longest average Body site • Upper extremity {39%, • Upper extremity {49%, mostly • Upper extremity {46%, length of stay in hospital at 0.91 days, injured mostly wrist & hand (51%) wrist & hand (74%) and mostly wrist & hand (77%) compared with 0.86 days for baseball & and shoulder & upper arm shoulder & upper arm (18%)} and shoulder & upper arm softball players and 0.7 days for hockey (31%)} • Lower extremity {23%, (14%)} players. The length of stay was similar • Lower extremity {33%, mostly knee & lower leg • Head/face/neck {30%, among hospitalisations in all three sports, mostly knee & lower leg (81%), ankle & foot (10%) mostly head/face (94%)} (87%)} and hip & thigh (8%)} • Lower extremity {18%, with 83% of baseball & softball and cricket • Head/face/neck {24%, • Head/face/neck {21%, mostly knee & lower leg cases and 85% of hockey cases staying less mostly head/face (89%)} mostly head/face (99%)} (90%)} than 2 days, and 16% of baseball & softball • Trunk (4%) • Trunk (5%) • Trunk (4%) cases and 15% of cricket and hockey cases • Other & unspecified (1%) • Other & unspecified (2%) staying between 2 and 7 days.

Nature • Fracture {62%, mostly knee • Fracture {52%, mostly wrist • Fracture {52%, mostly wrist • Four percent of baseball & softball and of injury & lower leg (30%), wrist & hand (51%) and head & hand (57%) and head cricket cases and 5% of hockey cases were & hand (28%) and head including face (26%)} including face (21%)} including face (24%)} • Dislocation, sprain & strain • Dislocation, sprain & strain classified as ‘serious’ on the injury severity • Dislocation, sprain & strain {21%, mostly wrist & hand {13%, mostly knee & lower scale i.e. on admission they had at least a {16%, mostly knee & lower (42%), shoulder & upper leg (61%) and shoulder & 6% chance of not surviving. leg (39%) and shoulder & arm (27%) and knee & upper arm (24%)} upper arm (33%)} lower leg (26%)} • Open wound {13%, mostly • The total direct hospital costs of injury • Injury to muscle & tendon • Injury to muscle & tendon head including face (68%) hospitalisations in the three bat and (6%) {11%, mostly knee & lower and wrist & hand (24%)} stick sports was $4.4m over the three • Superficial injury (4%) leg (78%) and shoulder & • Injury to muscle & tendon (4%) study years: cricket ($2.98m), hockey • Open wound (3%) upper arm (13%)} • Intracranial (4%) • Intracranial injury (2%) • Open wound (4%) • Superficial injury (2%) ($1.0m) and baseball & softball ($0.45m). • Injury to internal organs • Superficial injury (3%) • Traumatic amputation (1%) However, the average (mean) cost of a (1%) • Intracranial (2%) • Injury to internal organs (1%) hospitalisation was higher for baseball & • Other & unspecified injury • Eye injury (1%) • Injury to nerves & spinal softball cases ($3,944 per hospitalisation) (7%) • Injury to internal organs (1%) cord (1%) than cricket ($3,582) and hockey ($3,475) • Other & unspecified injury (5%) • Other & unspecified injury (8%) cases.

Length • Less than 2 days (83%) • Less than 2 days (83%) • Less than 2 days (85%) of stay • 2-7 days (16%) • 2-7 days (15%) • 2-7 days (15%) • 8-30 days (2%) • 8-30 days (2%) • 8-30 days (0%) • 31+ days (0.1%) Average length of stay: 0.86 Average length of stay: 0.91 Average length of stay: 0.69 days (range: 0 to 8 days) days (range: 0 to 35 days) days (range: 0 to 6 days)

Injury • Serious injury i.e. high • Serious injury i.e. high • Serious injury i.e. high severity 1 threat-to-life (4%, n=4) threat-to-life (4%, n=31) threat-to-life (5%, n=15) - Multiple injuries (50%) - Multiple injuries (19%) - Multiple injuries (40%) - Fracture of skull and - Fracture of femur (16%) - Fracture of skull and facial bones (50%) - Fracture of skull and facial facial bones (20%) bones (13%) - Intracranial injury (20%) - Intracranial injury (13%) - Injury of other and - Fracture of lumbar spine unspecified intrathoracic and pelvis (13%) organs (7%) - Injury of intra-abdominal - Fracture of femur (7%) organs (10%) - Other injuries of spine and trunk, level unspecified (7%)

Direct Total: $0.45m Total: $2.98m Total: $1.0m cost of Average: $3,944 Average: $3,582 Average: $3,475 admissions (range: $790 to $13,797) (range: $790 to $77,453) (range: $790 to $15,207)

Notes: 1 The final injury severity score is the product of the severity score of each of the diagnoses in the patient’s case record. In some cases a single injury diagnosis code is sufficient to attract the classification of ‘serious’, in others it is the combination of injury diagnoses that attracts this classification. Where the classification is based on one diagnosis the specific diagnosis is listed, where the classification is due to multiple diagnoses the term ‘multiple injuries’ is used. Source: Victorian Admitted Episodes Dataset (VAED)

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 12 4. Injury Table 4A Hospital-treated injury in racquet sports (badminton, table tennis, hospitalisations squash & racquetball combined), Victoria July 2007 to June 2010 HOSPITAL-TREATED INJURY IN RACQUET SPORTS in racquet sports: Badminton Table Tennis Squash/ Tennis All Racquetball badminton, table N % N % N % N % N % tennis, squash Hospital admissions 57 32.9 40 43.0 98 28.2 336 33.5 531 32.9 ED presentations - 116 67.1 53 57.0 250 71.8 666 66.5 1,085 67.1 & racquetball non-admissions All hospital- 173 100.0 93 100.0 348 100.0 1,002 100.0 1,616 100.0 (combined) and treated cases Source: tennis (N=531) Victorian Admitted Episodes Dataset (VAED - Hospital admissions) Victorian Emergency Minimum Dataset (VEMD - Emergency Department presentations, non-admissions) 4.1 Frequency and rate Figure 7 Trend in the frequency of racquet sports injury hospitalisations: badminton, table There were 1,616 adult hospital-treated injury tennis, squash& racquetball (combined) and tennis, Victoria 2002/3 to 2009/10 cases related to racquet sports in Victoria over the 3-year period 2007/8 to 2009/10, an average of 539 per year (Table 4A). Of these, 33% (n=531 or 177 per year on average) were hospital admissions (hospitalisations). Among all 16 sports studied, injured table tennis participants had the highest ratio of hospitalisations to ED presentations (43:57).

Hospitalisations were examined in more detail. Over the 3-year study period, tennis accounted for a large majority of the 531 racquet sport hospitalisations (63%, n=336), followed squash & racquetball (combined) (19%, n=98), badminton (11%, n=57) and table tennis (8%, n=40). However, the hospitalisation rate was highest for squash/ racquetball, followed by badminton. The mean injury hospitalisation rate in each of the racquet sports over the 3-year study period Source: Victorian Injury Surveillance Unit (VISU) was: squash & racquetball (64 hospitalisations per 100,000 participants), badminton (48/100,000), table tennis (41/100,000) and Figure 8 Trend in racquet sports injury hospitalisation rates per 100,000 participants: tennis (38/100,000). badminton, table tennis, squash/racquetball and tennis, Victoria 2002/3 to 2009/10

4.2 Trends

Trend in frequency of racquet sports injury hospitalisations, 2002/3 to 2009/10

Trend analysis over the 8-year period showed:

• the frequency of squash/racquetball injury hospitalisations decreased significantly from 36 in 2002/3 to 26 in 2009/10, representing an estimated annual decrease of 6.2% (95% confidence intervals -11.7% to -1.2%) and an overall reduction of 40% (-63% to -9%);

• there were no significant changes in the frequency of injury hospitalisations in the three other racquet sports. Source: Victorian Injury Surveillance Unit (VISU)

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 13 Table 4B Demographics and major causes of injury hospitalisations in racquet sports (badminton, table tennis, squash & racquetball combined and tennis), Victoria July 1, 2007 to June 30, 2010 RACQUET SPORTS HOSPITALISATIONS (N=531) Badminton (n=57) Table Tennis (n=40) Squash & Racquetball (n=98) Tennis (n=336)

Gender • Males (72%) • Males (65%) • Males (82%) • Males (55%) • Females (28%) • Females (35%) • Females (18%) • Females (45%)

Age • 15-24 (14%) • 15-24 (20%) • 15-24 (6%) • 15-24 (6%) • 25-34 (26%) • 25-34 (8%) • 25-34 (21%) • 25-34 (9%) • 35-44 (25%) • 35-44 (10%) • 35-44 (33%) • 35-44 (25%) • 45-54 (18%) • 45-54 (3%) • 45-54 (32%) • 45-54 (24%) • 55-64 (11%) • 55-64 (5%) • 55-64 (6%) • 55-64 (18%) • 65+ (7%) • 65+ (55%) • 65+ (2%) • 65+ (17%) Causes • Other unintentional {74%, • Falls {75%, mostly fall on same • Other unintentional {78%, • Other unintentional {51%, mostly overexertion & strenuous level from slipping, tripping or mostly overexertion & strenuous mostly overexertion & movements (50%)} stumbling (47%) and other fall movements (58%)} strenuous movements (57%)} on same level (33%)} • Falls {23%, mostly other fall on • Hit/struck/crush {13%, mostly • Falls {43%, mostly other fall same level (77%)} • Other unintentional {18%, striking against or struck by on same level (41%) and fall mostly overexertion & strenuous sporting equipment – ball and on same level from slipping, • Hit/struck/crush {4%, striking movements (57%)} racquet (69%) and striking tripping or stumbling (36%)} against or struck by sporting against or struck by other object equipment, (100%)} • Cutting/piercing {5%, contact (23%)} • Hit/struck/crush {5%, mostly with sharp glass (100%)} striking against or struck by • Falls {8%, mostly other fall sporting equipment, mostly ball • Hit/struck/crush (3%) on same level (63%) and fall (78%) and striking against or on same level from slipping, struck by other object (17%)} tripping or stumbling (25%)} • Natural/environmental {1%, • Cutting/piercing (1%) mostly exposure to excessive natural heat (67%)}

Source: Victorian Admitted Episodes Dataset

Trend in the rate of racquet sports injury sports injury, with 65% of cases aged less the elbow and forearm were the most hospitalisations, 2002/3 to 2009/10 than 45 years whereas table tennis players frequently injured body sites (both 23% of had the oldest age profile, with 60% of cases cases). (Not shown in tables.) The trends in racquet sports injury aged 55 and over. Sixty-five percent of hospitalisation rates per 100,000 participants squash/racquetball-related hospitalisations • Injury to the muscle and tendon was the over the 8-year period 2002/3 to 2009/10 were aged between 35 and 54 and the vast most frequent injury type in badminton and are shown in Figure 8. Trend analysis found majority of tennis-related hospitalisations squash/racquetball hospitalisations. Muscle that there was no significant change in injury (84%) were aged 35 or older. Victorian and tendon injuries to the knee and lower hospitalisation rates for any of the racquet sports participation data by age group were leg were most common, accounting for all sports indicating that the observed significant not available so age-related hospitalisation badminton muscle and tendon injury cases decrease in the frequency of squash/ rates could not be estimated and compared (63% of all badminton hospitalisations) racquetball injury hospitalisations was mostly within and between sports. and 95% of squash/racquetball muscle and due to decreased participation. tendon injury cases (55% of all squash/ • ‘Other unintentional injury’ (mostly racquetball hospitalisations). Fracture 4.3 Detailed analysis overexertion and strenuous movement) was the most frequent injury type in table was the most common cause of injury tennis and tennis hospitalisations. Among As summarised in Tables 4B & C, analysis of in badminton (74%), squash/racquetball table tennis-related hospitalisations, hip hospital admission data for the four racquet (78%) and tennis (51%), whereas falls was and thigh fracture was the most common sports over the 3-year study period 2007/8 to the most common cause of injury in table fracture accounting for 39% of table 2009/10 showed that: tennis (75%). tennis-related fracture cases (23% of all table tennis hospitalisations). Fracture • Males comprised the majority of injury • The lower extremity was the most to the elbow and forearm was the most hospitalisations in all racquet sports, but commonly injured body region among common fracture among tennis-related they formed a higher proportion of squash hospitalisations related to all four racquet hospitalisations, accounting for 60% of & racquetball (82%) and badminton (72%) sports. The knee and lower leg was the tennis fracture cases and 21% of all tennis cases than table tennis (65%) and tennis specific body site most frequently injured hospitalisations. (55%) cases. in badminton (83% of cases), squash/ racquetball (67% of cases) and tennis • Table tennis cases had the longest average • Badminton players had the youngest age (46% of cases). Among table tennis length of stay in hospital at 2.7 days, profile among hospitalisations for racquet hospitalisations, the hip and thigh and compared with 1.7 days for badminton

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 14 Table 4C Injury type, body region injured and injury severity among injury hospitalisations in racquet sports, Victoria, July 1, 2007 to June 30, 2010 RACQUET SPORTS HOSPITALISATIONS (N=531) Badminton (n=57) Table Tennis (n=40) Squash & Racquetball (n=98) Tennis (n=336)

Body site • Lower extremity {84%, mostly • Lower extremity {40%, mostly • Lower extremity {68%, mostly • Lower extremity {54%, mostly injured knee & lower leg (98%)} hip & thigh (56%) and knee & knee & lower leg (99%)} knee & lower leg (86%) and hip • Upper extremity (5%) lower leg (38%)} • Upper extremity {15%, mostly & thigh (10%)} • Head/face/neck (5%) • Upper extremity {38%, mostly shoulder & upper arm (60%) and • Upper extremity {35%, mostly • Trunk (5%) elbow & forearm (60%) and elbow & forearm (27%)} elbow & forearm (60%) and shoulder & upper arm (27%)} • Head/face/neck {13%, mostly shoulder & upper arm (31%)} • Head/face/neck {13%, to the head/face (85%)} • Head/face/neck {7%, to the head/face (100%)} • Trunk (3%) head/face (100%)} • Trunk (10%) • Trunk (4%) • Other & unspecified (1%)

Nature • Injury to muscle & tendon {63%, • Fracture {58%, mostly hip • Injury to muscle & tendon • Fracture {35%, mostly elbow & of injury to the knee & lower leg (100%)} & thigh (39%) and elbow & {59%, mostly knee & lower leg forearm (60%) and shoulder & • Dislocation, sprain & strain forearm (30%)} (95%)} upper arm (12%)} {19%, mostly knee & lower leg • Injury to muscle & tendon • Dislocation, sprain & strain • Injury to muscle & tendon (82%)} {13%, mostly elbow & forearm {14%, mostly knee & lower leg {34%, mostly knee & lower leg • Fracture (11%) (40%)} (64%) and shoulder & upper arm (86%)} • Eye injury (5%) • Open wound (8%) (36%)} • Dislocation, sprain & strain • Other & unspecified (2%) • Dislocation, sprain & strain • Fracture {11%, mostly elbow & {19%, mostly knee & lower leg (8%) forearm (36%)} (61%) and shoulder & upper • Eye injury (3%) • Open wound (4%) arm (25%)} • Other & unspecified (14%) • Eye injury (4%) • Intracranial (3%) • Intracranial (2%) • Open wound (2%) • Other & unspecified (5%) • Eye injury (1%) • Other & unspecified (7%)

Length • Less than 2 days (61%) • Less than 2 days (53%) • Less than 2 days (77%) • Less than 2 days (65%) of stay • 2-7 days (35%) • 2-7 days (35%) • 2-7 days (24%) • 2-7 days (32%) • 8-30 days (4%) • 8-30 days (13%) • 8-30 days (0%) • 8-30 days (4%)

Average length of stay: 1.65 days Average length of stay: 2.73 days Average length of stay: 0.98 days Average length of stay: 1.62 days (range: 0 to 11 days) (range: 0 to 11 days) (range: 0 to 5 days) (range: 0 to 17 days)

Injury • Serious injury (5%, n=3) • Serious injury (25%, n=10) • Serious injury (1%, n=1) • Serious injury (5%, n=18) severity 1 - Superficial injury of abdomen, - Fracture of femur (90%) - Fracture of skull and facial - Fracture of femur (44%) lower back and pelvis (33%) - Injury of other and bones (100%) - Intracranial injury (28%) - Fracture of lumbar spine and unspecified intrathoracic - Fracture of lumbar spine and pelvis (33%) organs (10%) pelvis (11%) - Fracture of femur (33%) - Injury of intra-abdominal organs (11%) - Fracture of shoulder and upper arm (6%)

Direct Total: $0.25m Total: $0.27m Total: $0.38m Total: $1.53m cost of Average: $4,326 (range: $1,058 to Average: $6,816 Average: $3,893 (range: $790 to Average: $4,557 admissions $18,269) (range: $904 to $36,713) $14,603) (range: $870 to $24,934)

Notes: 1 The final injury severity score is the product of the severity score of each of the diagnoses in the patient’s case record. In some cases a single injury diagnosis code is sufficient to attract the classification of ‘serious’, in others it is the combination of injury diagnoses that attracts this classification.Where the classification is based on one diagnosis the specific diagnosis is listed, where the classification is due to multiple diagnoses the term ‘multiple injuries’ is used. Source: Victorian Admitted Episodes Dataset (VAED)

cases, 1.6 days for tennis cases and 1.0 days 25% of cases being classified as ‘serious’ tennis ($0.27m), squash & racquetball for squash & racquetball cases. Thirteen i.e. on admission they had a 6% chance ($0.38m) and tennis ($1.53m). Of the percent of table tennis hospitalisations had of not surviving. These were mostly hip four racquet sports, the average cost of a stay of 8-30 days compared with 4% of fracture cases in players aged 55 years and hospitalisations over the three study years tennis and badminton hospitalisations. No older. Five percent each of badminton and was highest for the table tennis cases squash/racquetball cases were in hospital tennis cases and 1% of squash/racquetball ($6,816 per admission) compared with for more than 7 days. cases were classified as ‘serious’. $4,557 for tennis, $4,326 for badminton and $3,893 for squash & racquetball cases. • Table tennis cases had the highest • The total direct hospital cost of the four proportion of serious injuries of the racquet racquet sports was $2.43m over the three sports (and in all the 16 sports studied) with study years: badminton ($0.25m), table

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 15 participation survey conducted biennially formed a high proportion of ‘serious’ injuries Discussion across Australia. There are more refined and in tennis (8/18 ‘serious’ injuries). The age reliable measures of comparative risk—for profile of hospitalisations in these two sports Hospitalisations represent ‘the tip of the example, the number of injuries per 1,000 was very different to that of the other sports iceberg’ of sports injury as many people player hours based on surveillance rather than studied with 55% and 17% of table tennis injured when playing sport are treated in the estimated number of participants based on and tennis hospitalisations, respectively aged hospital emergency departments or by self-report—so the risk ranking reported here 65 and older compared with ≤1% to 2% in general practitioners, physiotherapists and should be interpreted cautiously. the other sports except for badminton (7%). other allied health professionals and sports The enjoyment and general health benefits of trainers. Mitchell et al. (2010) recently The high frequency, incidence and direct playing table tennis and tennis into older age analysed organised sports participation and cost ($43m. per year) of hospitalisations probably outweigh the risk of hip fracture but injury data from the 2005 NSW Population ine the fiv highest ranked sports for injury older players with diagnosed osteoporosis or Health Survey. One-third of the 2414 hospitalisations—Australian football, soccer, other risk factors for hip fracture should weigh respondents aged 16 years or older who basketball, netball and cricket—signify the the health benefits of these sports against their reported they had participated in organised need for urgent preventive action. A focus on personal hip fracture risk. sport in the previous 12 months had been reducing injury in these sports was previously injured during participation. Half of them recommended in edition 51 of Hazard Overall, males and younger adults (aged 15- required formal treatment from a health or published in 2002 that covered a broader 34) were over-represented in sports injury medical practitioner. Only 9% were treated range of sports (Cassell & Clapperton, 2002). hospitalisations in most of the sports studied in a hospital setting including around 3% who Consideration should be given to including but the gender and age profile of hospitalised were admitted to hospital for their injury and hockey (due to its comparatively high cases in the 16 specific sports generally reflects a further 6% who received treatment in an hospitalisation rate per 100,000 participants) the gender and age profile of participants emergency department. By contrast, 27% and rugby union and league (emerging full (Standing Committee on Exercise and Sport, were treated by physiotherapists and 16% by contact sports in Victoria) in the priority list. 2009 & 2010). However, there was some general practitioners. preliminary evidence that males are more at The risk of serious injury in table tennis risk than females when playing basketball, Over the 3-year period July 1, 2007 to June 30, (measured using the ICISS) is also noteworthy volleyball, hockey and badminton, although 2010 there were 23,187 adult hospitalisations – one quarter of the comparatively small we could not estimate injury rates by gender to for injury in sport and active recreation (an number (n=40) of hospitalisations for table make a direct comparison as we did not have average of 7,729 hospitalisations per year). tennis injury were classed as ‘serious’ because participation data in the 16 sports by gender. Adults are defined as persons aged 15 years they were neck of femur (hip) fractures, which and older because ages are generally classified generally result in long hospital stays and may Attention should be paid to the information on into 5-year groups (0-4, 10-14, 15-19 etc.). have serious health and social consequences the pattern and causes of injury in each sport The four groupings of sports covered in for older people. Neck of femur fracture also from injury surveillance and research when this report (16 sports in all) accounted for 54% of these hospitalisations (n=12,460 or an average of 4,153 hospitalisations per year). The case counts in this report include persons injured in indoor, outdoor, organised and unorganised sports participation as these variables are not specifically coded on the VAED. The VAED includes hospitalisations from all Victorian public and private hospitals but the number of hospitalisations for sports injury is underestimated because the activity being undertaken at the time of injury is not specified for approximately 60% of injury hospitalisations recorded on the VAED.

Five sports accounted for around 90% of the sports injury hospitalisations in the 16 sports covered in this report. They were, in rank order, Australian football, soccer, basketball, netball and cricket. These same sports, plus hockey, were also the highest ranked of the 16 sports in terms of serious injury risk, based on their hospitalisation rate per 100,000 participants. The hospitalisation rate for rugby union & league combined and baseball & softball combined could not be estimated due to the lack of reliable participation data from the oversampled Victorian component of the Australian ERASS, the comprehensive sports

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 16 devising prevention strategies and measures soccer and basketball (both considered non- and a narrative text. They acknowledge, which should focus on reducing serious contact sports). Hit/struck/crush also caused however, that these enhancements would be injuries and high frequency-lower severity the majority of hospitalisations in baseball/ difficult to achieve as ICD-10 is a widely injuries such as sprains and strains. Prevention softball, cricket and hockey but the mechanism used international health classification and control of injuries to the knee and lower of injury was contact with sports equipment system and not a purpose-built internationally leg should be a particular focus of sports injury (usually the ball but also the bat/stick) rather harmonised injury surveillance classification prevention strategies and measures, as this than with another person. Falls caused three- systeme lik the WHO-approved International specific body site accounted for the highest quarters of table tennis hospitalisations and Classification of the External Causes of Injury proportion of hospitalisations in 12 of the 16 were mainly same-level falls from tripping, (ICECI). sports studied: rugby league/union, soccer, slipping or stumbling. In the remaining sports, basketball, netball, volleyball, baseball/ the commonest cause of injury hospitalisations There was some talk that the ICECI would be softball, badminton, squash/racquetball and was ‘other unintentional’ injury caused by incorporated into the next Australian version tennis. The exceptions were Australian ‘overexertion and strenuous movements’ of ICD (ICD-11-AM) replacing the ICD-10 football, cricket and hockey in which injuries — soccer (36% of soccer hospitalisations), External Causes of Injury chapter, but this has to the wrist and hand (and for Australian basketball (49%), netball (63%), volleyball been scotched. In Victoria there is probably football also to the head and face) outranked (52%), badminton (74%), squash/racquetball more scope for improving the usefulness of injuries to the knee and lower leg, and also (78%) and tennis (51%). the Victorian Emergency Minimum Dataset table tennis where injuries to the hip and thigh (VEMD)—collected in the emergency and elbow and forearm were more common There are some data limitations. The major departments of 38 major Victorian hospitals— than injuries to the knee and lower leg. Head issue is data completeness. Over the study for sports injury research and prevention and face injuries were comparatively common period, 60% of injury hospitalisations recorded purposes, than the VAED. The most ambitious among Australian football, rugby, soccer and on the VAED did not have an activity code and proposal would be replacing the VEMD hockey hospitalisations and deserve attention. a further 1.7% were coded to ‘activity: sport minimum dataset for injury surveillance with The major injury types in most sports were and recreation’ but the sport or recreational the ICECI in all 38 VEMD hospitals or in a fractures and dislocations, sprains and strains, activity being engaged in at the time of injury representative sample. This would provide except for badminton and squash/racquet was not specified. Hence, the number of injury more informative data on most major injury ball in which injury to muscle and tendons hospitalisations in the 16 sports covered in this issues including sports injury as coding of accounted for around 60% of hospitalisations. report may be grossly underestimated. injury events is more detailed and groups of coded ICECI items form modules on topics The prevention of concussion in sport is Another issue of concern is the small amount such as transport, sport and violence. currently receiving substantial attention from of information on the mechanism and sports injury researchers, practitioners and circumstances of sports injury on the VAED VEMD data on ED presentations for sports sports administrators. The proportion of for prevention purposes. Mitchell & Hayen injury (non-admitted cases) are available on intracranial injury (mostly concussion) was (2006) examined some of the inadequacies of application to VISU. There is currently wide comparatively high among hospitalisations the ICD-10-AM classification system (used variation in coded and narrative data quality related to participation in the football codes to classify and code cases on the VAED) for across the hospitals contributing to the VEMD (5% for soccer through 7% for Australian sports injury prevention, focussing on the but the ED data provide useful additional football to 8% for rugby union/league), ‘hit/struck’ cause of injury codes (W20-W23 information on the less serious sports injury moderate for hockey (4%) and basketball struck by/against or caught between objects cases (mostly less severe/complex sprains and (3%) and comparatively low in the other sports including W21 striking against or struck by strains and fractures) presenting to Victorian studied (2% or less). The evidence related to sports equipment and W50-52 hit/struck etc. hospitals. the effectiveness of strategies and measures to by or against another person). The sub codes prevent head injury and concussion in sport under W21, introduced in 2004, now provide is discussed below. The Zurich Consensus some information on the object causing hit/ Preventing sports Statement on Concussion in Sport (and the struck injury in sports (cases can be coded to attached Sports Concussion Assessment Tool- striking against or struck by bat or racquet, injury SCAT2) should be use by certified trainers, ball, or object or structure on or near sports In a paper published in 1992, van Mechelen coaches and other people involved in the care area) but there is no coded information on any and colleagues proposed a four step prevention of concussed community sports participants other agents related to sports injury. Another model for the reduction of sports injuries: (1) to underpin management and return-to-play example of an uninformative code is the ‘other identify and describe the extent of the problem; decisions. As advised in the Statement, sports unintentional’ injury cause code ‘overexertion (2) identify the risk factors and mechanisms associations should keep abreast of, and and strenuous movements’; it is almost for sports injury by epidemiological research; communicate to all clubs, development in the meaningless in the context of sports injury (3) introduce preventive measures to reduce science of concussion as it is a rapidly evolving as it is too broad to be informative. There is future risk and/or severity of injuries (after field of research (McCrory et al., 2009). also no provision on the ICD-10-AM for the they have been evaluated for effectiveness in inclusion of a case narrative to describe the controlled trials); and (4) evaluate prevention As would be expected of full contact sports, circumstances of the injury. measures (van Mechelen et al., 1992). hit/struck/crush, mainly due to impact with another person, and falls on the same level due Mitchell & Hayen (2006) recommend that an More recently, Finch (2006) advocated that to collision with/pushing by another person assessment should be made of the feasibility this sports injury prevention model should be caused the majority of injury hospitalisations of collecting additional information on sports expanded to include two additional stages: (5) in Australian football and rugby union/league. injuries regarding the agent of injury, the research to describe the intervention context These two causes were also prominent for phase of activity at the time of the injury to inform implementation (i.e. the personal

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 17 and cultural motivators and barriers to the uptake of scientifically proven interventions in sports) and (6) implementation of scientifically interventions in the real world context and evaluation of effectiveness. Finch contends that adherence to the four-stage model provides evidence of the efficacy of a specific intervention (i.e. its performance under ‘ideal’ conditions in trials) but this does not necessarily translate to effectiveness when the strategy/measure is implemented in the real world of community level sports due to barriers operating at the sports association, club, coach, trainer, team and individual player levels.

It is beyond the scope of this study to fully canvass what is known about the risk and protective factors for injury and the efficacy/ effectiveness of prevention measures at the community level of adult participation in each of the 16 sports included in this study. One of authors of this report (EC) was part of a team that completed a series of risk factor and countermeasure reviews for five of the sports included in this report for Sport and Recreation Victoria (published on the VISU webpage). She drew on these reviews and conducted a search in December 2011 of the Cochrane Database of Systematic Reviews, PubMed (MEDLINE), SPORTSDiscus and a Google search of the grey literature to update these reviews, to search for risk factor studies conducted in the other sports of interest and for recently published systematic reviews of the evidence on the effectiveness of prevention measures, to provide a snapshot of the current state of knowledge.

Risk factor studies

The literature search found prospective cohort or case control studies of risk factors for injury among non-elite, mostly adult, participants in only seven of the 16 sports: Australian football (4 studies), rugby union/league (4), soccer (9), basketball (3 studies, two of which involved team and volleyball players), netball (1) and volleyball (1 study, that included basketball and handball players).

There were also two recently published reviews on the effect of playing surface on injury rates.

Football codes

Australian football

There were four prospective cohort studies investigating risk factors for injury in adult non-elite Australian football reported in the

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 18 literature (McManus et al, 2004; Braham et al., collected prospectively from the same cohort identified eight independent risk factors 2004; Gabbe et al., 2004; Gabbe et al 2005). of 320 adult players from the two highest (predictors) of injury: (1) previous injury The Western Australian Sports Injury Cohort grades of five clubs belonging to the Victorian experience (being injured at the beginning of Study monitored injuries and exposure in four Amateur Football Association (Gabbe et al., the season); (2) hours of strenuous physical sports over two seasons (McManus et al., 2002). In the study of risk factors for lower activity a week (more than 39 hours a week); 2004). The cohort included 535 community- extremity injury (Gabbe et al., 2004), older (3) playing position (playing midfield back level Australian football players aged 16-50 age (age 25 years and older), past injury (two compared with front row); (4) cigarette years playing with randomly selected Perth or more injuries in the previous season) and smoking status (being an ex- or current clubs. The incidence of injury in this study non-participation in weight training during smoker); (5) body mass index (BMI<23); was 24 injuries per 1000 player hours. Three– the season were identified from regression (6) years of rugby participation (playing quarters of the total sample sustained at least analyses as independent predictors of lower from zero to 3 years); (7) stress, aerobic and one injury during the two-year study period. extremity injury, but none of these results anaerobic performance; and (8) number of Nearly two-thirds of reported injuries were was statistically significant. The study of push ups (performing between 20-33 push ups to the lower limb (knee, ankle, hamstring and risk factors for hamstring injury (Gabbe et compared with fewer than 19). quadriceps muscles). al., 2005) found that increasing age (age 23 years and older) and decreasing quadraceps In a study conducted in Australia by Hollis et Five risk and protective factors for football flexibility (as measured by the modified al. (2009), the investigation was confined to the injury were identified from stepwise logistic Thomas test) were significant predictors of risk and protective factors for mild traumatic regression modeling. All variables collected time to sustaining a hamstring injury. brain injury (mTBI). A cohort of 3,207 male in the baseline survey of participants were in nonprofessional rugby union players (aged the model. The three significant independent Rugby (union and league) 15-49 years) from Sydney was recruited from predictors of injury were: (1) not wearing school, suburban and grade level competitions sports-specific football boots which increased When single club studies were excluded, and followed for between one and three winter the risk of injury by 40%; (2) having a our literature search found four prospective playing seasons. Demographic information, diagnosed existing back pathology which cohort studies investigating risk factors for history of recent concussion, and information increased the risk of injury by 29%; and (3) injury in rugby played by older adolescents on putative risk and protective factors were having been diagnosed with excessively and adults at the non-elite level [Quarrie et collected pre-season and injuries that met pronated feet which increased the risk of al. 2001 (all injury); Hollis et al., 2009 (mild the study definition of mTBI were recorded injury by 29%. In addition, two protective traumatic head injury); Garraway et al., 1999 by trained injury recorders or nominated factors were identified: (1) not being injured (tackle injury); and Lee & Garraway, 2000 suburban or grade club doctor, physical in the previous 12 months which reduced the (environmental factors)]. therapist or coach. risk of injury by 27%; and (2) cooling down after training which reduced the risk of injury In the most comprehensive of the studies, a The overall incidence of mTBI in the cohort by 5%. cohort of 258 male rugby union players (mean was 7.97 per 1000 player game hours, with 313 age 20.6 years) from five clubs players (9.8%) sustaining 1 or more mTBIs In the second study, a prospective cohort and four secondary schools were followed during the study. Multiple regression analysis study of players competing in a through a full competitive season (Quarrie found that always wearing protective headgear metropolitan league, Braham et al. (2004) et al., 2001). A pre-season assessment during games was a significant protective surveyed injuries and exposure in 301 adult collected individual player data on basic factor and having sustained a previous mTBI a players recruited from nine clubs over football characteristics, health and lifestyle patterns, significant risk factor (predictor) for in-season season 2001. The overall injury incidence playing experience, injury experience, training mTBI. The risk of sustaining an mTBI was rate was 12.1 /1,000 player hours. Over one- patterns, anthropometric characteristics and almost halved in players who always wore third of players were injured during the season performance in a battery of fitness tests. Two protective headgear in games compared with studied. The leg was the most commonly different outcome variables were used to those who did not and the likelihood of mTBI injured body region (lower leg, ankle, thigh/ examine the influence of the preseason factors was more than one-and-a half times higher hamstring and knee). Logistic regression on injury experience during the season: injury among players who reported having sustained analyses (that included age, playing history, incidence rate (per 1000 player game hours) either 1 or more mTBIs within the previous playing position and previous head/neck/ and playing time lost i.e. the proportion of the 12 months. The study authors recommend the dental injury in the model) identified two season a player missed (a proxy measure for use of protective headgear in rugby players independent predictors of injury: (1) older age injury severity). and the proper management of prior or current (players aged 25 years and older were more mTBI to reduce the incidence of mTBI. than twice as likely to sustain an injury than One common (previous injury experience) and younger players); and (2) playing in midfield several different risk factors were associated The study limitations (for example, lack of (midfield players were more than three times with these outcomes. The multiple regression training of data collectors and reliance of self- more likely to sustain an injury than players model that utilised injury incident rate as the report by players of headgear use) indicate playing other positions). Playing history, outcome measure, identified that only two of that it would be wise to wait for confirmation measured in years played, and head/neck/ the variables studied were independent risk of these findings by at least one other well- dental injury in the previous season were not factors for in-season injury: (1) grade (playing designed study among adult amateur players found to predict injury in season 2001. in higher grades) and (2) previous injury before bringing in rules requiring protective experience (being injured at the beginning headgear use in adult rugby. It should be Theo tw other published risk factors studies of the season). The second regression model borne in mind that an earlier RCT of protective focussed on injury at specific body sites (Gabbe that utilised injury severity (i.e. the proportion headgear use conducted in a large cohort of et al., 2004; Gabbe et al., 2005) utilising data of season missed) as the outcome measure 3,686 young New Zealand rugby players

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 19 (ages 12-21) followed over two seasons did regression revealed that there were significant were collected so the injury incident rate was not support the use of standard headgear, month-of-season (P = 0.003), wind strength (P not estimated. or a modified version, for the prevention of = 0.008) and temperature (P = 0.011) effects concussion (McIntosh et al., 2009). on the risk of injury. The results show that Univariate analysis that compared the The case control study by Garraway et al. the month of the season and the weather may frequency of risk factors in uninjured and (1999) was focused on investigating the risk influence the occurrence of rugby injuries, but injured players revealed 17 risk factors which factors for tackle injuries in rugby union, that the state of the pitch does not appear to the authors describe as of “high predictive measuring the influence of selected aspects of have any influence. value” for new injury. The significant risk lifestyle, personality and other player related factors, that covered a wide spectrum, were: factors. In this study the tackling and tackled Soccer body fat proportion, number of previous players (“the cases”) involved in a tackle injuries (>6), pain in the joints, muscular injury were each matched with “control” The risk factors for injury in adult non-elite warm-up (fairly poor or poor), endurance, age players who held the same respective playing or mixed level soccer have been investigated at beginning football, not beginning football positions in the opposing teams. in nine recent studies. Six prospective cohort in a club, more a fighter than a technician studies involved male players (Dvorak et al., when getting past an opponent, self rated A total of 964 rugby matches involving 71 2000; Engebretsen et al., 2010a, 2010b, and dribbling/long passing (average to poor), senior clubs drawn from all districts of the 2010c; Kofotolis et al., 2007; Engebretsen reaction time, life-event stress, and change Scottish Rugby Union (SRU) were observed et al., 2010d) and two prospective cohort to another club. A study weakness was that over one season (1996/97). Each recruited studies and one nested case control study these putative risk factors were not included player completed a questionnaire that gathered were focussed on female players (Ostenberg in multivariate regression analyses to explore demographic data and information on & Roos, 2000; Soderman et al., 2001; Myer et their interrelationship and determine whether lifestyle habits, match preparation, training, al., 2008). All studies included adolescent as any were independent predictors of injury. previous rugby experience, previous coaching well as adult players. in tackling, use of protective equipment Four studies by Engebretsen and colleagues (headgear and mouthguard) and history of A Czechoslovakian prospective cohort study (Engebretsen et al., 2010a, 2010b, 2010c, and recent concussion. A validated battery of investigated the overall and specific incidences 2010d) utilised data from a Norwegian cohort psychological tests assessed players’ trait of injury and risk factors for injury in different of male amateur soccer players originally anger and responses to anger and hostility. age and skill levels in 264 adult and adolescent recruited for a randomised controlled trial of a male players who were followed over a 12 training program designed to prevent injuries A total of 71 tackle injury episodes with month period after a baseline examination to (that did not eventually find any differences correct matching of cases and controls were ascertain possible predictors of injury (Dvorak in injury rates between the intervention and studied. Multivariate analysis showed that et al, 2000). There were 558 recorded injuries control groups). A total of 508 players none of the investigated player related factors in the 216 players who completed the weekly representing 31 amateur teams were tested that were shown to be associated with mTBI follow-up over the year. No exposure data during the 2004 pre-season for potential risk in univariate analysis (alcohol consumption before the match, feeling “below par” through minor illness, the extent of match preparation, previous coaching or practising tackling) were independent risk factors (predictors) for tackle injuries. Also, injured and non-injured players in the tackle did not differ in their disposition toward, or expression of, anger or hostility.

A Scottish cohort study conducted by Lee and Garraway (2001) investigated the influence of weather and pitch conditions on the frequency and nature of rugby injuries. Observers at 26 senior rugby clubs in the Borders District of the Scottish Rugby Union reported all injuries to 1169 (96%) registered players at Saturday home and away matches during the 1993-94 season (August 1993 to April 1994). Weather and pitch conditions at 112 grounds were recorded on 605 occasions; 1268 Borders teams played at these grounds with 344 injury episodes being sustained. Matches were played in dry weather for three-quarters of occasions, consistent for autumn, winter and spring.

There was a moderate association between weather and the state of the pitch with heavier pitches occurring in wetter weather. Logistic

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 20 factors for injury (including ankle, acute The finding from the Norwegian study that 5.5 leg injuries (traumatic and overuse) per knee, hamstring and groin injury) through previous ankle injury was a strong predictor of 1000 hours of soccer. Multivariate analysis a questionnaire on previous injury, function subsequent ankle injury confirmed the results showed hyperextension of the knee joint, a scores (including ankle, knee, hamstring of an earlier cohort study of risk factors for low postural sway, reduced hamstring–to- and groin outcome scores) and clinical ankle sprain injury involving 336 male soccer quadriceps ratio and higher exposure to soccer examinations. Their injuries were recorded players participating in the North Division significantly increased the odds for new over the 2004 season. A total of 505 injuries League of the Greek Amateur Association traumatic leg injury. were reported, sustained by 283 (65%) of the who were followed for two years (Kofotolis 508 players included in the study. The total et al., 2007). Over the study period, 132 ankle A nested case control study by Myer et al. incidence of injuries during the season was 4.7 sprains were documented in the 312 players (2008) investigated the risk factors for anterior per 1000 playing hours for match injuries and included the study; the total incidence of ankle cruciate ligament (ACL) injury in a mixed 2.7 for training injuries. sprain injuries was 2.16 ankle sprains per cohort of high school and collegiate female 1,000 hours of exposure. Multivariate logistic soccer and basketball players. Of the 1,558 The ankle injury study found that 8.5% of regression analyses (in which previous injury, athletes who were prospectively screened, 19 the 508 players experienced at least one age, weight, height and years of training went on to tear their ACLs (15 during soccer acute ankle injury over the season (there were entered into the model) showed that a and 4 during basketball). Four teammates of were 58 ankle injuries in 43 players); the history of ankle sprain was the only significant the nearest height and mass (“controls”) were overall incidence of acute ankle injuries predictor of new ankle injury (P<.05). selected from the uninjured screened athletes was 0.5 injuries per 1,000 playing hours for comparison with each of the 19 injured (Engebretsen et al., 2010a). Multivariate Three further risk factor studies, two subjects (“cases”). regression analyses revealed that the number conducted in Sweden (Ostenberg & Roos, of previous acute ankle injuries was the only 2000; Soderman et al., 2001) and one in Thee fiv measured joint/knee laxity measures significant predictor of new acute ankle injury, the USA (Myer et al., 2008), involved only that were shown to be potential risk factors in with the analysis showing a 23% increase in female players. In the first, a cohort of 123 univariate analyses and prior knee injury were the odds of new acute ankle injury in players elite and non-elite senior players (aged 14-39) entered into the multiple regression model. with previous ankle injury (Engebretsen et al., from eight teams playing at different levels Two of the knee laxity measures were shown to 2010a). The knee injury study reported that of competition in the south of Sweden were predictCL A injury: for every 1.3-mm increase 10% of the 508 players experienced at least followed up over the 1996 soccer season in side-to-side difference in anterior-posterior one acute knee injury in the season (there was and all their injuries registered (Ostenberg & knee displacement, the odds of ACL-injured a total of 61 acute knee injuries in 53 players); Ross, 2003). Prior to the season, players were status increased 4-fold; and a positive measure the overall incidence of knee injuries was 0.6 screened for injuries and took part in a testing of knee hypertension increased the odds of injuries per 1,000 playing hours. None of procedure that included measures of isokinetic ACL-injury status 5-fold. The authors suggest the candidate intrinsic risk factors for acute muscle strength, functional performance, that their screening methods and findings may knee injury (pain, function in daily living, aerobic capacity and physical characteristics assist the identification of high risk female any findings at clinical examination, flexion including body mass index. athletes for ACL injury opening the possibility contraction in range of motion testing and of targeted neuromuscular training. varus stress tests in full extension) were found Sixty five injuries were registered in 47 of the to be associated with acute knee injury risk 123 players giving a total injury rate of 14.3 Other team ball sports (Engebretsen et al., 2010b). per 1000 game hours and 3.7 per 1000 practice hours. Multivariate logistic regression Basketball The hamstring injury study reported 76 acute analyses identified three significant risk and overuse hamstring injuries over the 2004 factors for any injury in the new season: (1) Surprisingly, our literature search found no season affecting 12% of the 508 players in the increased general joint laxity that increased prospective cohort study investigating risk study; the total incidence of hamstring injuries the odds of new injury more than fivefold; (2) factors for injury in non-elite adult basketball was 0.7 injuries per 1,000 playing hours. Only high performance in the functional test square players, although there were several involving previous acute hamstring injury independently hop, that conferred a four-fold increase in the elite college and national players. Basketball predicted hamstring injuries, the analysis likelihood of injury; and (3) age over 25 years players represented 12% of the 1,512 showing that the odds of new hamstring injury (the odds of older players being injured were participants in the Western Australia Sports were doubled for previously injured players 3.7 times those of younger players). General Injury Cohort Study and the small sample size (Engebretsen et al., 2010c). During the season joint laxity was a significant predictor of knee and comparatively low injury incidence rate of soccer studied there were 61 groin injuries injury conferring a fivefold increase in the over the 2-year study period (15 injuries per affecting 51 (10%) of the 508 players in the odds of a player having a knee injury in the 1,000 exposure hours) did not allow risk factor study; the total incidence of groin injuries was new season. analysis specifically for basketball injury 0.6 injuries per 1,000 playing hours. Previous (mostly ankle and finger/thumb injuries) acute groin injury and weak abductor muscles, A later prospective cohort study conducted in (Finch et al., 2002). We found three case determined clinically, predicted groin injury, Sweden over the 1998 season by Soderman control studies, the first investigated the risk with the analysis showing that the odds of new et al. (2001) investigated leg injuries in of ankle injury in mainly recreational-level groin injury were more than doubled among 146 female soccer players adolescent and male basketball players (McKay et al., 2001) players with previous groin injury and more adult players enrolled from 13 teams in ando the tw others investigated the risk factors than quadrupled in players with weak adductor the second and third soccer divisions. Fifty for traumatic knee injury in groups of female muscles (Engebretsen et al., 2010d). players sustained 61 traumatic injuries and athletes that included basketball players (Myer 17 players sustained 19 overuse injuries. et al., 2008; Vauhnik et al., 2008). The overall incidence rate of leg injury was

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 21 In the case control study investigating the registered in all Slovenian divisions of the that injury prevention strategies should focus risk factors for ankle injury conducted in Slovenian National Sports Association who on the development of effective training Australia by McKay et al. (2001), 10,393 were prospectively followed over the 2004 programs that include netball-specific skills, male participations in mainly recreational season. Participants completed a preseason activities and movements. indoor basketball were observed courtside and test and post season questionnaire (the latter 40 ankle injuries documented. The 40 ankle- provided a summary of all traumatic knee Volleyball (indoor) injured players were the designated “cases”. injuries that occurred in the season including A group of non-injured players formed the those not reported the principal researcher at The literature search found no comprehensive control group (n=360). Ankle injuries occurred the time of injury). prospective study of the incidence and at a rate of 3.85 per 1000 participations. At risk factors for injury among non-elite the end of each game all players were asked Twenty (3.7%) of these sportswomen adult volleyball players. There was one about their injury status and injured players experienced traumatic knee injuries in the comparatively large prospective cohort study completed a questionnaire. A control group season. Several sets of data analysis were with injury as the outcome measure but was recruited by administering a questionnaire performed in order to build a model for this involved players competing in the top to entire teams of players who were not injured predicting female traumatic knee injury. two divisions of the Norwegian Volleyball on a particular day but were from the same Multivariate logistic regression analyses Federation and the findings may not apply to competition as the injured players. indicated that higher amounts of training, community level players (Bahr & Bahr, 1997). greater body height and greater knee anterior Other risk factor studies in the literature Comparing cases to controls, multivariate knee laxity were potential risk factors for also involved elite players and were mainly analysis identified three independenttraumatic knee injury. For each additional focussed on investigating shoulder and knee predictors of ankle injury: (1) having a history millimetre of knee anterior laxity present there injury. The one study that met our inclusion of ankle injury increased the odds of ankle was a 30% greater likelihood of sustaining criteria was a prospective study that aimed to injury status 5-fold; (2) wearing shoes with air traumatic knee injury and for each additional investigate knee injury risk factors in female cells in the heel increased the odds of ankle centimetre of player height there was a 10% athletes including volleyball players (Vauhnik injury status more than 4-fold; and (3) not increased likelihood of sustaining a traumatic et al., 2008). stretching before the game more than doubled knee. injury As the predictive model only the odds of ankle injury. There was also a explained 1% of the variability in traumatic This study has already been described in the trend toward ankle tape decreasing the risk of knee injuries among sportswomen, the authors basketball section above. The main purpose ankle injury in players with a history of ankle suggested that much larger studies are needed of the study, conducted in Slovenia, was to injury (p = 0.06). The authors recommended to test the full range of variables associated investigate if knee anterior laxity, measured that the three identified risk factors, and with traumatic knee injury. with an arthrometer, is a risk factor for female landing technique, should all be considered traumatic knee injury but the influence of other when preventive strategies for ankle injuries Netball potential risk factors was also investigated in basketball are being formulated. (Vauhnik et al., 2008). The study found that The cohort study of injury in four sports higher amounts of training, greater body height The case control study of the risk factors for conducted in Western Australia in the late and greater knee anterior laxity increased the ACL injury in female athletes by Myer et al. 1990s is the only published study that has odds of knee injury in female basketball, (2008) is already described in the soccer section investigated risk factors for injury in non- team handball and volleyball players but, as above as it involved both soccer and basketball elite senior netball (McManus et al., 2006). previously stated, these factors only explained players. Of the 1,558 high school and college Three-hundred and sixty-eight non-elite 1% of the variability in knee injury among the players who were prospectively screened for netballers completed a baseline questionnaire sportswomen studied. general and knee laxity measures, 19 went on at the commencement of the 1997 preseason. to tear their ACLs (“cases”) who were each Participants were telephoned each month Team bat and stick sport matched to 4 teammates (“controls”). Only during the 1997 and 1998 playing seasons to four of the 19 cases were basketball players. provide details of their exposure at training Baseball & softball, cricket and hockey The independent predictors of ACL injury and games and any injury experiences in the identified in the multivariate model were: knee previous 4 weeks. The literature search found no prospective hyperextension and side-to-side differences in cohort or case control studies investigating anterior-posterior tibiofemoral translation. About half of the netball cohort sustained one the risk factors for injury in adult non-elite or more injuries over the study period. The baseball, softball, cricket or hockey players. The main aim of the other basketball case incidence of injury was 14 injuries per 1000 The risk factor studies in the baseball/ control study was to investigate if knee player hours. Injuries were most commonly to softball literature were mainly focussed on anterior laxity, measured with an arthrometer, the ankle, knee and finger/thumb (Finch et al., investigating the risk and protective factors for is a risk factor for female traumatic knee injury 2002). The identified risk factors for injury upper limb (mainly shoulder) injury in youth (Vauhnik et al., 2008). Other easily measured were: not warming up before a game (IRR and elite adult pitchers and, in cricket, on variables such as anthropometry, lower leg 1.11, 95% CI 1.00-1.23) and not being open injury patterns in elite and county Australian, characteristics, sport exposure and menstrual to new ideas (IRR 1.04, 95% CI 1.00-1.07). English and South African players and back cycle characteristics were also evaluated as Training for 4 or more hours per week (IRR injuries in elite and young, talented fast and possible risk factors. 0.66, 95% CI 0.45-0.98) and not sustaining an spin bowlers. Hockey was included in the injury in the previous 12 months (IRR 0.58, Western Australian Sports Injury Cohort Subjects were 540 sportswomen aged 11 95% CI 0.43-0.79) were found to protect Study that targeted four community sports but to 41 years from 43 (54%) of the 79 teams against injury. The authors recommended risk factor analysis was not done for hockey in basketball, team handball and volleyball due to the relatively low number of injuries

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 22 and (3) Sport specific surface comparisons.

In Part 1 of the review, four studies were described that evaluated the effectiveness of environmental variables on natural grass surfaces. One study observed lower injury rates in high school (gridiron) games on grass during wet/slippery conditions compared to good field conditions. The three others (all involving elite players playing in the Australian Football League- AFL) found, respectively, significantly higher rates of anterior cruciate ligament (ACL) injuries at venues with drier grass, no correlation between ACL injury rates and ground hardness and a higher risk of ACL injury on Bermuda grass than on rye grass.

Six studies (five in American football and one in football) that compared injury rates on natural grass to “first-generation” (made from very short nylon fibres) reported higher overall rates of injury on these first generation surfaces, one other study reported the emergence of ‘turf toe’ (damage to the ligaments under the big toe) correlated with the more widespread use of artificial turf surfaces and another reported higher incidences of knee and elbow bursitis on artificial turf than natural grass.

Included studies, all conducted in American football, that compared injury rates on natural grass to “second generation” artificial turf (characterised by longer fibre length, sand fill and thicker cushioning layers) showed significantly higher rates of injury on artificial turf including ACL tears and knee sprains, ankle sprains, lower extremity injuries more generally, head, neck and lower extremity injuries, conditioning-related injuries, and Grade 2 concussion. One study showed the recorded in the two-season follow-up period The effect of playing surface game-related injuries but not practice injuries (Finch et al., 2002). The injury incidence on injury risk were more frequent on grass than second rate was 15 injuries per 1,000 hours of generation artificial turf. participation; finger/thumb, knee, ankle and There have been two recently published thigh injuries were almost equally common. reviews on the effect of playing surface Five studies, conducted either in soccer or on injury rate (Dragoo & Braun, 2010; American football, compared injury rates on Racquet sports Williams et al., 2011). The review by Dragoo natural grass to “third generation” artificial and Braun (2010) did not consider study turf surfaces—characterised by longer fibre Badminton, table tennis, squash & quality but included all studies published length, rubber synthetic infill and increased racquetball and tennis in English that presented or referenced an shock absorbency. They generally showed no epidemiological study or provided injury data significant difference in overall injury rates, Our literature search also found no risk factor and directly assessed or referenced the effect but there were some observed differences studies in non-elite adult badminton, table of playing surface on injury rate. In all, 43 in patterns of injury. One study found tennis, squash and racquetball. There were studies met these generous inclusion criteria. significantly higher incidences of head/neck some reports in the literature on the frequency The review was divided into three parts: (1) injuries, lacerations/skin lesions and ankle and pattern of injury in these sports based Studies comparing surfaces for field games – foot and joint injuries among male soccer on data from retrospective studies of injuries natural grass to one of the three generations players and lower incidence of ankle sprains, presenting to hospital Emergency Departments of ‘all weather’ artificial field turf; (2) Studies joint injuries including ligament and cartilage or specialist eye and dental hospitals and from comparing surfaces for court games–natural legions among female soccer players on surveys of high-grade players. grass,, clay acrylic/polyurethane and /or wood; artificial turf compared to natural grass. A prospective study conducted in American

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 23 football found a significantly higher incidence significantly higher on artificial than wooden other surfaces for men. Limited evidence of muscle-tendon overload injuries and floors. The second, that compared injury rates suggested that clay surfaces were safer than significantly greater amounts of abrasions, in female players, found a much the other surfaces on which tennis is played. non-contact injuries and running and sprinting higher overall injury incidence on artificial The authors found no studies comparing injuries during temperatures greater than than wooden floors. basketball injury rates on wooden floors 21°C. on artificial turf. A third study, involving compared to artificial floors. soccer players, reported significantly higher In Part 3 of the review the authors examined risk of ankle sprains and a significantly lower the literature that compared relevant surface The review authors highlighted the weakness risk of lower extremity injuries on artificial types for the four major sports where most of available evidence and recommended turf compared to grass, but the sample sizes of research has been done – American football, further well-designed studies that take account these subsets of injuries were small. soccer, tennis and basketball. Most of the of environmental factors that may contribute to reviewed studies had been discussed in the injury (such as field conditions, shoe/surface In Part 2 of the review, studies comparing three earlier sections. The authors concluded interface and climate) so that more definitive the injury effect of different artificial court that the American football studies failed to conclusions can be drawn regarding the effect surfaces—clay, acrylic/polyurethane and provide consistent evidence about the effect of artificial playing surfaces on injury rates. wood—to natural grass were reviewed. The of artificial playing surfaces on football injury literature was sparse. Two tennis studies (one rates although the majority of studies had A later review by Williams et al. (2011) involving senior and the other professional found higher injury rates on artificial surfaces was confined to studies that compared the tennis players) found, respectively, lower rates compared to natural grass. For soccer, the incidence, nature and mechanisms of injuries of knee problems among players who spent authors concluded that the majority of studies sustained in three football codes (American their careers on clay courts and a significantly revealed no significant difference in overall football, soccer and rugby union/league) on higher frequency of injuries requiring injury rates on artificial surfaces compared newer generation artificial turfs (third and treatment during match play on grass and to natural grass but that gender-specific fourth generation) and natural turfs. They hard courts than clay courts. Two studies difference in sub-groups of injuries were found 11 studies that provided 20 cohorts of compared the risk of injury on wooden floors observed in several studies. In tennis, there players and the authors compared the risk to rubber coated floors (that are mainly used was weak evidence at the elite level of play of injury between the surfaces using injury for indoor sports such as basketball, handball that there were fewer incomplete matches on incidence rate ratios. Their analysis found and floorball). The first, that compared the grass and a higher rate of incomplete matches that 16 of the 20 cohorts showed only minor ACL injury rate in team handball, found that on Australian hardcourts than other surfaces differences in the overall injury incidence rate the incidence of ACL injuries in women was for women and on American hardcourts than ratios between surfaces; taken together they provide strong evidence of comparable rates of injury on the newer generation artificial turfs compared to natural turf. However, they found that there was increased risk of ankle injury associated with playing on artificial turf with eight cohorts showing incidence rate ratios ranging from 0.7 (one study found a decreased risk) to 5.2 (a five-fold increase in risk). They concluded that ankle injury prevention should be a priority for athletes playing on artificial turf. Evidence concerning risk of muscle and knee injuries on the two surfaces was inconsistent and was dependent on the football code, gender of athlete and whether injuries occurred during matches or training.

Sport and Recreation Victoria has just published a general and sport specific guide for local government, sporting associations, clubs and schools that provides guidance, as well as highlighting ‘best practice’, for the planning, design, selection, installation, management, maintenance and replacement of artificial grass surfaces for sport (Sport and Recreation, Victoria 2011). This comprehensive and practical guide includes sections on injury prevention and can be downloaded from the web (see entry in reference list).

A recently published study by Twomey et al. (2012) described the risk and details of injuries associated with ground hardness in

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 24 community Australian football. One hundred Two of the authors involved in the symposium presented that showed that the wearing of ground hardness assessments were undertaken had earlier conducted a review of the literature full faceguards has resulted in a significant using a Clegg hammer at nine locations across since 1980 on the prevention of head and reduction in eye injury incidence and a non- 20 grounds over the 2007 and 2008 seasons. neck injury in sport focusing on catastrophic significant reduction in the concussion rate in Injury data were prospectively collected by and brain injury and identifying the range of junior and in oculofacial injuries in trained primary data collectors. A total of injury prevention methods in use (McIntosh youth baseball. 402 injuries were recorded on the assessed &, McCrory 2005). They found few studies grounds, of which 352 were able to be related evaluating injury prevention interventions The review authors reported that the new rugby to ground hardness. Despite the low number and speculated that the reasons for the lack of scrum laws were widely viewed as useful of injuries, the risk of sustaining an injury on formal evaluations may be that interventions in the prevention of cervical injury in rugby low/normal and unacceptably hard grounds were introduced precipitously in response to but that law changes that were designed to was significantly greater than on the preferred an acknowledged problem with no time for eliminate an impact in engagement and permit (optimum) range of hardness. However, a trial, or that blanket rule changes or new the orientation of each front row had not been the severity of injuries on unacceptably protective equipment were introduced across formally evaluated. [Since this review there hard grounds was lower than for the other a sport or a league so there was no study has been further encouraging evidence from categories of hardness. comparison group available. French, New Zealand and Australian studies to support the effectiveness of the new rugby Evidence-based preventive The review found some evidence to indicate scrum rules (Bohu et al., 2009; Quarrie et al., measures that softer balls reduce the risk of injury in 2007; Carmody et al., 2005)]. baseball, lightweight helmets are effective The literature search found recent reviews on in preventing head and neck injury in pedal The authors concluded that current designs the effectiveness of measures to prevent head, cycling, and emerging evidence that helmets of helmets do not appear to reduce the risk face and neck injury, oral-dental injury and are effective in reducing head injury risk in of concussion and found that other potential lower limb injuries (thigh, knee and ankle). skiing and snowboarding. The review also interventions such as rule and environmental found that the combination of laws to control changes to slow down games thereby reducing Preventing head, face and the use of head in blocking and the use of impact forces, training interventions to neck injury standards-approved helmets have decreased improve strength, fitness and skills, pre- fatalities and serious head injury in American participation screening and tackle training Readers are referred to the recently published football, and that the introduction of rigorous have not been implemented and evaluated for article Sports helmets now and into the future standards for helmets and increased wearing effectiveness. They recommend that “research in the British Journal of Sports Medicine that rates have reduced fatal and serious head methods involving epidemiological, medical, reports on a symposium on sports helmets held injury in ice hockey. There was also evidence and human factors are required in combination at the 2011 International Olympic Committee World Conference on Prevention of Injury and Illness in Sport (McIntosh et al., 2011). The main conclusions of the Symposium were:

• Helmets play an important part in head injury prevention and control. Research has shown that are very efficacious and effective in reducing moderate to severe head injury risks in a range of sports (American football, ice hockey, bicycling and alpine sports). • Helmet standards are important and there needs to be further development. • Helmets for competitive (elite) athletes and separately for recreational athletes need to be designed and developed. • Helmets for children, adults and older athletes, because of the different tolerances to impacts and impact profiles, need to be designed and developed.

The Symposium identified deficiencies in the evidence base for head injury risks and helmet efficacy and effectiveness in some sports, recognised the difficulties in designing helmets that were suitable to prevent severe head injuries and concussion, and identified the need to evaluate helmet performance in oblique impacts and incorporate this into standards.

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 25 with biomechanical and technological to stretch and lengthen hamstring muscles, rigorous and sufficiently powered RCTs to approaches to further reduce head and neck propioceptive balance training, chiropractic, determine the effects of commonly used injury risks in sport”. spinal manipulative therapy, massage and interventions and to justify the continued mobilisation, functional training and sports inclusion of stretching protocols within Preventing oral-dental/facial specific drills. prevention programs. They also recommended injury research on the effectiveness of injury Seven randomised controlled trials (RCTs) awareness education especially in relation The authors of a recent comprehensive review involving a total of 1919 participants met the to the role of poor running technique and of the effectiveness of mouthguards (Knapik et inclusion criteria of the review. The trials hamstring injury and to develop and evaluate al., 2007) concluded that the modern materials involved predominantly young adult players multimodal approaches for the prevention of used to construct mouthguards all appear to participating in regular sporting activities: injuries that address more than one risk factor. perform equally, as they all can be manipulated soccer (four studies), Australian football (one to provide the range of necessary protective study), high school basketball (one study) Knee characteristics. Evidence from impact studies and recreational running (one study). These indicated that, compared with no mouthguard, studies evaluated four different interventions: The comprehensive systematic review by mouthguards composed of many types of hamstring muscle group strengthening; Thacker et al. (2003) that assessed the materials reduce the number of fractured manual therapy; proprioceptive training; and a evidence on the effectiveness of the range teeth and head acceleration. The authors warm-up/cool down/stretching program. The of interventions to prevent knee injuries recommended that consideration be given to authors commented on the poor methodology is out-of-date as it covered the published the expected nature/force of the impact in the of some of the included trials. literature to September 2001. At that time sport being played (hard or soft objects) and the reviewers found 13 intervention studies the characteristics of the mouth (e.g. brittle Four trials, including 287 participants, that met their inclusion criteria. They were incisors, more rugged occusal surfaces of examined interventions directly targeted assessed as weak-to-moderate in terms of molars, soft gingiva) when choosing the most at preventing hamstring injuries (the other study design and methodology. The review appropriate material for fabrication. three targeted lower limb injuries including authors concluded that there was no consistent hamstring injuries). Three of the hamstring- evidence to support knee bracing from the The reviewers commented that studies on specific trials (all involving soccer players), five studies conducted in American football the effectiveness of mouthguards in relation that tested hamstring strengthening players. Inadequate reporting of methodology to injury prevention were of highly variable protocols, had contradictory findings and the prevented the reviewers from drawing quality. However, despite problems caused review authors could not conclude whether any conclusions on the injury prevention by the use of different injury definition and strengthening exercise programs were effectiveness of alternative cleat designs (two methodological weaknesses, their meta- beneficial or not. The other trial (involving a studies) and of adjustments in the ski boot/ analyses of grouped data from similar relatively small number of Australian football binding system (one study). They concluded studies indicated that the risk of an orofacial participants) found that manual therapy from reviewing the six prospective studies sports injury was 1.6-1.9 times higher when (involving manipulation, massage and specific that addressed the impact of conditioning and a mouthguard was not worn than when one stretches to joints and muscles of the spine training programs that proprioceptive and was worn. They found insufficient evidence and leg) may prevent injuries of leg muscles, neuromuscular training for protection against to support the proposition that mouthguards including the hamstrings. knee injury appeared promising but that more protect against concussion. Their overall rigourous evaluations were required before conclusion was that mouthguards should Three trials (conducted in soccer, basketball widespread implementation. continue to be promoted and used in sport and running) tested interventions for preventing activities where there is significant risk of leg injuries for which data for hamstring injury The focus of most interventions in the 2000s orofacial injury. were available. Results were inconclusive. in the area of knee injury prevention has The trials in soccer and basketball found no been on preventing catastrophic anterior Preventing lower limb injury evidence for an effect for balance training on a cruciate ligament (ACL) injury in female wobble board (proprioceptive protocol). The athletes, concentrating on the modification Thigh (hamstring muscles) other trial found no evidence of effectiveness of the neuromuscular risk factors by for a warm up/cool down/stretching protocol preventive exercise. Epidemiological studies Hamstring muscle injuries usually occur as a for distance runners. in basketball, soccer and handball show result of either direct trauma (a blow of some that female athletes have a four- to six-fold kind) or indirect trauma (a single powerful The review authors concluded that no specific increased risk of ACL injury compared to their muscle contraction) and are common in intervention can be recommended at this stage male counterparts playing at similar levels in football codes and basketball. They are for decreasing the risk of hamstring injuries in the same sports (Hewett et al., 2006). often serious and require long periods of sport. They assessed that the most promising rehabilitation and have a high re-injury risk. intervention was the manual therapy Systematic reviews by Hewett et al. (2006) and The Cochrane review, Interventions for intervention conducted in Australian football Yoo et al (2010) yielded six and seven articles, preventing hamstring injuries, by Goldman (Gabbe et al, 2005) but recommended that the respectively, that met their study inclusion & Jones (2010) examined the evidence for results would need to be confirmed in larger criteria viz. either a RCT or prospective cohort the effectiveness of various interventions trials before widespread implementation. study that investigated a neuromuscular targeting the prevention of injuries to the Player adherence to this onerous stretching preventive training intervention in female hamstring muscles. The review found that program was an issue that would require athletes. In the Hewitt review, four of the six many preventive interventions are currently attention in the design of future trials. More interventions significantly reduced knee injury being implemented such as exercise therapy generally, the review authors recommended incidence and three of them significantly

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 26 reducedCL A injury incidence. Data from semi-rigid orthosis (three trials), air-cast brace Most of the new studies included in this all six studies were pooled and the meta- (one trial) or high top shoes (one trial); ankle updated review (15 in all) investigated the analysis demonstrated a significant effect of disk training; taping; muscle stretching; boot effect of multifaceted injury prevention neuromuscular training programs on ACL inserts; a health education programme and programs (combining rule changes, balance injury incidence in females (test for overall controlled rehabilitation. The main finding training, sports specific technical training/ effect, Z=4.31, P<.0001). was a significant reduction in the number of balance and sports specific training/tailored ankle sprains in players allocated external exercise program including balance and The Yoo et al. review included the six studies ankle support i.e. semi rigid orthosis or air strength training), balance training, strength in the review by Hewett and colleagues plus cast brace (RR 0.53, 95% CI 0.40 to 0.69). training or a combination of these. Of these a later study conducted in a mixed group The reduction in risk (of approximately 50%) studies, ten reported significantly lower risks of female soccer, volleyball and basketball was greater among persons with a previous of ankle ligament injury in the intervention players, which supported the efficacy of the history of ankle sprain, but still observed for group compared with the control group (RRs prevention program. The pooled analysis of those without prior sprain who participated in of ankle sprain rates range 0.15 to 0.4 in favour data from all seven studies showed a 40% sports with high risk of ankle ligament sprains of the intervention group). Five studies found reduction in the odds of ACL injury in the such as basketball and soccer. The available no significant preventive effect. Several of the trained group compared to the untrained group evidence was inconclusive in relation to the successful trials reported a greater preventive (Odds Ratio 0.40, 95% CI 0.27, 0.60). effectiveness of the other interventions. effect on recurrent sprains in previously injured athletes, rather than first-time sprains. The weight of evidence from these reviews A recently published review by Verhagen There was some evidence to indicate that indicate that multiple neuromuscular & Bay (2010) examined the literature on proprioceptive/neuromuscular (balance board) pre- and in-season training components— interventions to prevent ankle ligament training was the most effective component of incorporating plyometric (high intensity injuries using the outdated Cochrane critical successful preventive programs as the two jumping movements) training and review as their starting point and model. The trials that included strength training alone strengthening exercises—may provide some authors used the same search strategy and showed no difference in ankle sprain incidence levelCL of A risk reduction. The small size of stringent inclusion criteria which included the density between the strength training group the studies meant that most were statistically application of a quality assessment test. They and the control group. underpowered, the fact that only two of the included seven of the eight studies that were studies were RCTs, and other methodological in the earlier Cochrane review (excluding one The authors concluded that, based on current weaknesses of the individual studies indicate trial of an external boot brace that was not evidence, a combination of an external support that more and larger well-designed trials in conducted in the sports setting) and found 18 measure (tape or brace) with proprioceptive/ multiple sports with longer follow up periods additional studies published between 1999 and neuromuscular training will achieve the best are needed before neuromuscular training April 2010. Two of these were excluded as preventive outcomes with minimal burden for interventions can be broadly implemented at they did not meet the methodological quality the athlete as each of these interventions is the community sports level in high risk sports cut-off score. In total, the review analysed 23 linked to an approximately 50% reduction in for female ACL injury. The optimal intensity studies: 16 RCTs, four CTs, two prospective ankle sprain recurrence risk. The evidence of the program and the incentives required to cohort studies and one retrospective study. indicates that the strongest benefit will be increase acceptability (and player compliance) observed in previously ankle-injured athletes. also requires further research. Both groups of Four studies investigated the effect of shoe reviewers concluded that impractical protocols type/design on ankle sprain risk but the Large-scale interventions that are time-consuming and expensive evidence on effectiveness was inconclusive. cannot be realistically implemented and that Another four studies investigated the injury at the community level of motivation is enhanced when preventive prevention effects of taping and found a two- sports participation training programs have a performance to-fourfold reduction in ankle sprain rates for enhancing effect. taped ankles compared to untaped ankles, with The challenge of translating evidence of injury a stronger effect observed in athletes with prevention effectiveness gathered under ideal Ankle previous ankle injury. The evidence from two conditions in the research setting to large-scale of these studies suggest that bracing, which is interventions, taking account of the resources Arguably, the most common injury in sport the less expensive option, may be as or more available at the community level of sport, and physical activity is the ankle sprain, most effective than taping but the review authors is currently been addressed in community frequently acute lateral ligament sprain. The assessed that the superiority of one method level soccer and rugby. In both sports, injury Cochrane review Interventions for preventing over the other had not been established. Six prevention experts have partnered with ankle ligament injuries (Handoll et al., 2001) studies investigated the effectiveness of community sports association and clubs was withdrawn from the Cochrane Database bracing. Of these, four RCTs evaluated the and national sports insurance companies in 2011, because it was substantially out- effectiveness of a semi-rigid brace against a to develop and evaluate comparatively of-date. A new title of the same name has control group, three of which reported that simple, time efficient, low cost multifaceted been registered on the Cochrane Database braces were effective for the prevention of interventions to reduce common injuries in by the authors, the protocol for which will be ankle sprains with RRs of ankle sprain rates these sports. There is also a similar initiative published in 2011. ranging from 0.15 to 0.5 in favour of the in junior community netball that has not yet braced groups. Evidence from two of these been evaluated and a large-scale intervention The (withdrawn) review found 14 randomised studies indicates that bracing may only be is currently being implemented and evaluated trials of interventions in physically active effective for the prevention of recurrent ankle in community Australian football. individuals that included the application of sprains. an external ankle support in the form of a

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 27 Soccer: ‘The 11’ teams (14.2; 95% CI, 12.8-15.6) and a 17.2% Since the Swiss trial, “The 11” program has lower incidence compared with 2004 (15.2; been further improved by F-MARC into a 15 A country-wide campaign to prevent soccer 95% CI 14.3-16.1). The incidence of training exercise warm-up program known as “11+”. injuries “The 11” (subsequently expanded injuries in the last 4 weeks decreased from For full details and resources see website: to “11+”), a FIFA-Medical Assessment and 2004 (2.44/1,000 hours of player exposure to www.f-marc.com/11plus. The efficacy of the Research Centre (F-MARC) warm up exercise training; 95% CI, 1.72-2.24) to 2008 by 18.9% “11+” program has been studied in a cluster program developed by a team of experts, was in teams that performed “The 11” (1.98; 95% randomised trial in Norway involving 1,892 trialed in amateur soccer teams involving CI 1.72-2.24) but it increased by 8.6% in teams young female football players in 125 clubs adolescent and adult players aged 14-65 in not doing “The 11” (2.65; 95% CI 11.5, 13.7), (Soligard et al., 2008). Though the primary Switzerland between 2004 and 2008 (Junge et so the reduction in injury incidence in 2008 outcome of reduction in lower extremity al., 2011). was 25.3% in the teams performing “The 11” injury did not reach statistical significance, than teams not doing the warm-up program. the risk of severe injuries, overuse injuries and “The 11” was composed of 10 evidence-based injuries overall was reduced. The program has or best practice exercises designed by experts The results are promising but the evaluation now expanded to other European countries, to reduce the most frequent and common relied on injury data supplied by coaches and Japan and the Americas. types of soccer injury such as ankle sprains, was subject to reporting and recall biases. hamstring and groin strains and ligament The national Swiss accident insurance Although persons seeking Australian injuries to the knee, plus the promotion of fair company data provides additional evidence accreditation as Youth Football Certificate 1 play. The target group was male and female of a beneficial effect – the company registered coaches are advised to familiarise themselves soccer players aged 14 years and older of an increase of 1.7% in the cost of soccer withA the FIF “11+” warm up program and it all levels of play below the first league. No injuries between 2003 and 2007 despite a 10% is a listed resource on the Football Federation equipment was required, except for a ball, and increase in the number of players registered in Australia website, there does not appear to be the program was designed to be implemented the Swiss national football association. any real push to promote “11+” by Football in 15 minutes at every training session. All Federation Australia or Victoria. coaches in the Swiss national soccer league received information (DVD, booklet and Rugby: RugbySmart poster) and practical instruction on ‘The 11” from autumn 2004 onwards during their A similar initiative has been implemented and compulsory basic or compulsory biennial evaluated in community rugby. RugbySmart, refresher courses. At the completion of their a nationwide coach- and referee-delivered training, they were instructed to incorporate rugby union injury prevention program ‘The 11” in all club training sessions. developed in New Zealand, promotes physical conditioning, improved tackling and The program aimed to reduce the incidence scrimmaging technique, injury management, of soccer injury by 10%. The evaluation, mouthguard wear and warm up and cool performed by an independent research group, down. The program was launched in 2001 and consisted of a pre- and post- structured implemented at all levels of the community computer assisted telephone interview (CATI) game from under 6 grade to senior adults survey of a representative sample of coaches (Gianotti et al., 2009). conducted prior to the start of the intervention (in May 2004) and four years later (May Injury insurance claims data were collected 2008). In the pre-intervention interview, 1027 over the implementation period (seasons coaches were asked about the frequency and 2001-05) by the Accident Compensation characteristics of match and training injuries Corporation (ACC), the body that administers in their teams in the four weeks prior to the the New Zealand national injury insurance interview. In the post-intervention CATI scheme, but unfortunately pre-intervention survey, 1015 coaches were interviewed (310 of (baseline) injury claims data were not whom had participated in the pre-intervention collected in 2000. The evaluation showed interview) and asked about their knowledge that the RugbySmart program was associated and implementation of ‘The 11”, as well as with a significant decrease in moderate-to- about the training and match injuries in their serious injury claims per 100,000 rugby team in the previous 4 weeks. players between 2001 and 2005 in most areas the program targeted—neck/spine, knee The evaluation results indicated that in and leg injury but not shoulder where there 2008, 80% of the coaches knew ‘The 11” was a non-significant decrease in claims or and 57% stated they currently performed the ankle injury—and had a negligible impact on program (instructing their team on average 3.7 non-targeted injury sites. These promising exercises for a mean duration of 13 minutes results were supported by data from pre- and mostly once or twice a week). In 2008, teams post intervention player behaviour surveys performing ‘The 11” had an 11.5% lower which found improvements in the areas of incidence of match injuries in the last 4 weeks safer tackle, ruck and scrum behaviours and (12.6 injuries/1000 hours of player exposure the performance of cool down. The South to matches; 95% CI, 11.5-13.7) than other

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 28 African Rugby Union Association version of perform correct landing technique, whereas If successful, FootyFirst and the lessons the program—BOKSmart—is currently the the challenges included insufficient delivery learned from this trial including the facilitators subject of implementation research with the time at training, the drills were too difficult and barriers to full implementation will be focus of the evaluation on whether the acquired and there were too many sets and repetitions. adapted so that it can be rolled out nationally injury prevention knowledge and skills are The second paper reported the results of a through the AFL and then introduced in other being translated by coaches and referees to survey of 278 junior female players prior to sports. Injury data will be collected by a club- their practices on the field (Verhagen & Finch, their participation in the program that found nominated trainer at each participating club 2011). that they had positive intentions and attitudes using SportsInjury Tracker, an existing on-line towards learning safe landing techniques community sports injury surveillance system SmartRugby, the Australian version of (White et al., 2012). developed by Sports Medicine Australia RugbySmart, is a compulsory occupational (http://www.sportsinjurytracker.com.au/), health and safety training program for every Australian football: NoGAPS and the feasibility of widespread adoption rugby union coach and referee at every level (FootyFirst) of this system to reliably record injuries in from U6 grade upwards in Australia (http:// community Australian football will also be www.vicrugby.com.au/). The initial face- The Australian NoGAPS (National Guidance evaluated. to-face training course covers balance and for Australian Football Partnerships and stability, tackle and tackle laws, support the Safety) is a major initiative to develop and maul and scrum, mayday procedures, lineout disseminate an exercise-training program to Conclusion and restarts, foul play and injury prevention. prevent leg injuries in community sport with an The qualification must be updated every 2 or initial focus on community Australian football. This analysis of the frequency and incidence 4 years depending on the coaching certificate This National Health and Medical Research of serious injury in 16 popular sports indicates level. Implementation research is not being Council (NHMRC) funded partnership project that Australian football, soccer, basketball, undertaken by the Victorian Rugby Union has a total budget of $1.6m comprised of cash netball and cricket (and perhaps hockey and but a survey of coaches and referees could be and in-kind contributions from a wide range rugby union & league) should be prioritised developed to be administered when coaches of partners and is led by Professor Caroline for community sports injury prevention in and referees undertake their compulsory re- Finch at Monash Injury Research Institute Victoria, although all community sports qualification course on-line. No similar (MIRI) (personal communication Caroline associations and clubs have a duty of care to program has been developed in Rugby League. Finch, Chief Investigator). prevent injury among players. Our literature search found only a modest number of proven Netball: Down to Earth (D2E) In the initial phase of NoGAPS, the latest risk factors for injury in community sports available research was used to identify the with most risk factor research having been A program to reduce lower limb injuries in most common injuries in Australian Football done in community soccer, probably because community netball is being promoted by Netball and to develop an evidence-based and of the international nature of the sport. The Victoria for widespread implementation at the industry-endorsed exercise-training program results of risk factor studies show that, in junior level of the sport, but there are currently (“FootyFirst”) to prevent them. If implemented general, a strong focus should be placed on no plans to evaluate its effectiveness due correctly, FootyFirst has the potential to reduce the treatment and full rehabilitation of injury to lack of resources. The “Down-to-earth” the risk of lower limb joint (ankle and knee) before return to play because previous injury (D2E) program is an evidence-informed and muscle (hamstring, groin and hip) injuries was found to be a potent risk factor for new lower limb injury prevention training program that are the major cause of missed games and injury in a range of studies whether they for teaching safe and effective landing. The long-term disability (e.g. osteoarthritis) among were focused on specific sports (Australian D2E program was developed from research community Australian football players. Small- football, rugby, soccer, basketball and netball) evidence on the prevention of knee and ankle scale evaluations of similar exercise training or specific injury types (for example, ankle, injury and laboratory based studies conducted programs implemented in other sports have hamstring and groin injury). at the University of that supported demonstrated that they can reduce the risk of the efficacy of landing training in reducing lower limb injuries by 39%, the risk of acute There are some interventions of proven ACL injury risk factors (Sanders, 2006). The knee injuries by 54%, and the risk of ankle efficacy under trial conditions and promising core program consists of 6-weeks of coach- sprain injuries by 50%. effectiveness under ‘real-world’ conditions led training in correct landing techniques at the community level of adult sports incorporated into regular training sessions One hundred and eighty teams (4000 participation – helmets or combination of supplemented by 6 weeks of home-based players) drawn from five large community helmets/laws for head injury, faceguards, training, done in conjunction with the coach Australian football leagues in Victoria (the mouthguards for orofacial injury, and directed training, and the inclusion of six Ballarat, Bellarine, , Southern and a combination of an external support balance-related exercises in regular warm-up Maryborough-Castlemaine District Football measure (tape or brace) with proprioceptive/ sessions (Saunders et al., 2006). Leagues) have agreed to participate in the neuromuscular training for ankle ligament sprains and neuromuscular training for knee project. The major focus of the trial is (ACL) injury in females. The multifaceted There have been two published papers on translating sports research knowledge into F-MARC 11+ soccer injury prevention the D2E program. The first reported positive practice at the community level of sport program and the RugbySmart (SmartRugby) feedback from a survey of the 31 junior and the research team is working hard to rugby union injury prevention provide coaches involved in implementing the program encourage and support community football blueprints for the development of low cost, in Victorian junior netball in the 2003/4 clubs, coaches and players to adopt and time efficient, evidence-informed sports injury season (Saunders et al., 2010). The identified implement FootyFirst. prevention interventions that are feasible to benefits included improved player ability to implement at the community level of sports

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 29 participation. The NoGAPS research program Protective measures (bracing, taping, hospitalisations recorded on the Victorian will provide evidence on the effectiveness of a padding and rehabilitative exercise as Admitted Episodes Dataset (VAED). similarly framed intervention to prevent lower prescribed) should be implemented. limb injuries in Australian football. Return to play during a match or after a • The Victorian Department of Health break in participation due to injury should should be approached to consider the not be influenced by the pain threshold of feasibility of introducing the WHO- Recommendations players or the importance of matches to the approved International Classification of club. External Causes of Injury (ICECI) in • Based on the findings of this study the emergency departments of all or a of sports injury hospitalisations in 16 • Prevention of injuries to the knee and representative sample of the 38 Victorian popular sports, priority should be given lower leg should be a particular focus hospitals currently contributing data to the to reducing injury in Australian football, of community sports injury prevention Victorian Emergency Minimum Dataset to soccer, basketball, netball and cricket strategies and measures, as this specific provide more informative data on sports (and perhaps hockey and rugby union & body site accounted for the highest and other injuries. league), although all community sports proportion of hospitalisations in 12 of the associations and clubs have a duty of care 16 sports studied (rugby league/union, to prevent injury among players. soccer, basketball, netball, volleyball, baseball/softball, badminton, squash/ • All clubs should call for volunteers or racquetball and tennis) and a substantial appoint team trainer/s to collect and record proportion of hospitalisations in Australian player injury data on a weekly basis using football, cricket and hockey. SportsInjury Tracker, the free online injury surveillance system developed by • The prevention of wrist and hand injuries Sports Medicine Australia (http://www. should be a priority for prevention in sportsinjurytracker.com.au/). These data community Australian football, cricket should be monitored weekly, analysed and hockey, as should the prevention of mid-season and at the end of the season, head and face injuries in the three football to learn more about the frequency and codes and hockey. pattern of injury among club players. Injury surveillance reports should be used • Ankle injury prevention should be a priority to underpin evidence-informed strategies for athletes playing on artificial turf and in and measures implemented by coaches, sports played on other surfaces where ankle trainers and players to prevent first injury ligament sprains are common. Research and reduce recurrent injuries. evidence indicates that a combination of an external support measure (tape or • The feasibility of conducting an on- brace) with proprioceptive/neuromuscular line survey of coaches and referees training will achieve the best preventive (administered at the time of their outcomes with minimal burden for the compulsory reaccreditation) on their on- player. field implementation of ‘SmartRugby’, the rugby union injury prevention program, • The use of mouthguards should be should be considered by Victoria Rugby promoted and enforced in sport activities Union. Rugby League should consider where there is significant risk of orofacial the applicability of components of injury, as should eye protection in squash. RugbySmart to their sport.

• The F-MARC “11+”soccer injury • The Zurich Consensus Statement on prevention warm up program should be Concussion in Sport and the attached considered for statewide implementation Sports Concussion Assessment Tool- at the community level of play by Football SCAT2 should be understood and used by Federation Victoria. certified trainers, coaches and other people involved in the care of concussed sports • The netball “Down-to-earth” (D2E) participants to underpin management and program, the evidence-informed lower return-to-play decisions and, as advised limb injury prevention training program in the Statement, sports associations for teaching safe and effective landing to should keep abreast of developments in junior netballers, should be evaluated for the science of concussion as it is a rapidly effectiveness. evolving field of research.

• Previous injury is a significant risk factor • VISU and the Victorian Department of for new injuries and clubs should ensure Health should work intensively with that injured players are fully recovered hospitals to improve the coding of before they are allowed to return to play. ‘activity’ in the case records of injury

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 30 Engebretsen AH, Myklebust G, Holme L et Hollis SJ, Stevenson MR, McIntosh AS et References al. Intrinsic risk factors for groin injuries al. Incidence, risk, and protective factors among male soccer players: a prospective of mild traumatic brain injury in a cohort Bahr R, Bahr IA. Incidence of acute volleyball cohort study. The American Journal of of Australian nonprofessional male rugby injuries: a prospective cohort study of Sports Medicine 2010;38(10): 2051-7. players. The American Journal of Sports injury mechanisms and risk factors. Medicine 2009; 37(12):2328-33. Scandinavian Journal of Medicine and Finch C, da Costa A, Stevenson M et al. Science in Sports 1997;7(3):166-71. Sports injury experiences from the Junge A, Lamprecht M, Stamm H et al. Western Australia sports injury cohort Countrywide campaign to prevent soccer Braham R, Finch C, McIntosh A, McCrory study. Australia and New Zealand Journal injuries in Swiss amateur players. Am J P. Community level Australian football: a of Public Health 2002;26(5):462-467 Sports Med. 2011 Jan;39(1):57-63. profile of injuries. Journal of Science and Medicine in Sport 2004;7(1): 96-105 Finch C. A new framework for research Knapik JJ, Marshall SW, Lee RB et al. leading to sports injury prevention. Mouthguards in sport activities: history, Bohu Y, Julia M, Bagate C, Peyrin JC et al. Journal of Science and Medicine in Sport physical properties and injury prevention Declining incidence of catastrophic 2006;9:3-9. effectiveness. Sports Medicine 2007 cervical spine injuries in French rugby: 37(2): 117-44. 1996-2006. The American Journal of Gabbe B, Finch C, Wajsweiner H, Bennell K. Sports Medicine 2009;37(2):319-23. Australian football: injury profile at the Kofotolis ND, Kellis E, Vlachopoulos SP. community level. Journal of Science and Ankle sprain injuries and risk factors in Carmody DJ, Taylor TK, Parker DA, Medicine in Sport 2002:5(20):149-160. amateur soccer players during a 2-year Coolican MR, Cumming RG. Spinal period. The American Journal of Sports cord injuries in Australian footballers Gabbe BJ, Finch CF, Wajswelner H, Bennell Medicine 2007 Mar;35(3):458-66. 1997-2002. Medical Journal of Australia KL. Predictors of lower extremity injuries at the community level of Australian 2005:6;182(11):561-4. Lee AJ, Garraway WM. The influence of football. Clinical Journal of Sports environmental factors on Medicine 2004;14(2)56-63. Cassell E, Clapperton A. Preventing injury injuries. Journal of Sports Science in sport and active recreation. Victorian 2000;18(2):91-5. Injury Surveillance & Applied Research Gabbe B J, Finch CF, Bennell KL, Wajswelner H. Risk factors for hamstring injuries in Unit, Hazard Edition 51, Winter 2002. McCrory, P Meeuwisse W, Johnston K et community level Australian football. al. Consensus Statement on Concussion Dragoo JL, Braun HJ. The effect of playing British Journal of Sports Medicine 2005;39:106-110. in Sport. 3rd International Conference surface on injury rate: a review of the on Concussion in Sport held in Zurich, current literature. Sports Medicine November 2008. Clinical Journal of 2010;40(11):981-90. Garraway WM, Lee AJ, Macleod DA et al. Factors influencing tackle injuries in Sport Medicine. 2009;19:185–200. Dvorak J, Junge A, Chomiak J et al. Risk factor rugby union football. British Journal of Sports Medicine 1999;33(1):37-41. McIntosh AS, McCrory P, Finch CF et al. analysis for injuries in football players. Does padded headgear prevent head Possibilities for a prevention program. Gianotti SM, Quarrie KL, PA. injury in rugby union football? Medicine The American Journal of Sports Medicine and Science in Sports and Exercise 2000;28(5 Suppl):S69-74. Evaluation of RugbySmart: a rugby union community injury prevention programme. 2009;41(2):306-313. Journal of Science and Medicine in Sport. Engebretsen AH, Myklebust G, Holme I, 2009;12(3):371-5. McIntosh AS, McCrory P. Preventing head Engebretsen L, Bahr R. Intrinsic risk and neck injury. British Journal of Sports factors for acute ankle injuries among Goldman EF, Jones DE. Interventions for Medicine 2005 39(6): 314-8. male soccer players: a prospective cohort preventing hamstring injuries. Cochrane study. Scandinavian Journal of Medicine Database of Systematic Reviews 2010, McIntosh AS, Anderson TE, Bahr R. et al. and Science in Sports. 2010;20(3):403-10. Issue 1. Art. No.: CD006782. DOI: Sports helmets now and in the future. 10.1002/14651858.CD006782.pub2. British Journal of Sports Medicine Engebretsen AH, Myklebust G, Holme I et 2011;45:1258-1265. al. Intrinsic risk factors for acute knee Handoll HHG, Rowe BH, Quinn KM, de Bie injuries among male football players: a R. Interventions for preventing ankle McKay GD, Goldie PA, Payne WR, Oakes prospective cohort study. Scandinavian ligament injuries. Cochrane Database of BW. Ankle injuries in basketball: injury Journal of Medicine and Science in Sports. Systematic Reviews 2001, Issue 2. Art. rate and risk factors. British Journal 2010;21(5):645-52. No.: CD000018. DOI: 10.1002/14651858. Sports Medicine 2001;35(2):103-8. CD000018. Engebretsen AH, Myklebust G, Holme L et al. McManus A, Stevenson M, Finch C, et al. Intrinsic risk factors for hamstring injuries Hewett TE, Ford KR, Myer GD. Anterior Injuries and risk factors for injury in among male soccer players: a prospective cruciate ligament injuries in female non-elite Australian Football. Journal cohort study. The American Journal of athletes: Part 2, a meta-analysis of of Science and Medicine in Sport Sports Medicine 2010;38(6):1147-53 neuromuscular interventions aimed at 2004;7(3):384-391. injury prevention. The American Journal of Sports Medicine. 2006;34(3):490-8.

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 31 McManus A, Stevenson MR, Finch CF. Saunders NA. Characteristics of the female van Mechelen W, Hlobil H, Kemper HC. Incidence and risk factors for injury in landing pattern. Victoria: University of Incidence, severity, aetiology and non-elite netball. Journal of Science and Ballarat, 2006. Thesis. prevention of sports injuries. A review of Medicine in Sport 2006; 9(1-2):119-24. concepts. Sports Medicine 1992;14(2):82- Söderman K, Alfredson H, Pietilä T, Werner 99. Mitchell R, Hayen A. Sport- or leisure- S. Risk factors for leg injuries in female related injury hospital admissions: do soccer players: a prospective investigation Vauhnik R, Morrissey MC, Rutherford we need to get more out of being struck? during one out-door season. Knee Surgery OM et al. Knee anterior laxity: a Journal of Science and Medicine in Sport. Sports Traumatology and Arthroscopy. risk factor for traumatic knee injury 2006;9(6):498-505. 2001;9(5):313-21 among sportswomen? Knee Surgery, Sports Traumatology, Arthroscopy Mitchell R, Finch C, Boufous S. Counting Sport and Recreation Victoria, Department 2008;16(9):823-33. organised sport injury cases: evidence of of Planning and Community incomplete capture from routine hospital Development. Artificial grass for sport. Verhagen EA, Bay K. Optimising ankle sprain collections. Journal of Science and Guide. DPDC Melbourne, February prevention: a critical review and practical Medicine in Sport. 2010;13(3):304-8. 2011. http://www.dpcd.vic.gov. appraisal of the literature. British Journal au/__data/assets/pdf_file/0007/58912/ of Sports Medicine 2010; 44(15):1082-8. Myer GD, Ford KR, Paterno MV et al. The ArtificialGrassForSportGuide.pdf effects of generalized joint laxity on risk of Verhagen E, Finch CF. Setting our minds to anterior cruciate ligament injury in young Standing Committee on Exercise and Sport implementation. British Journal of Sports female athletes. The American Journal of 2008. Participation in Exercise, Sport and Medicine 2011;45(13):1015-6. Sports Medicine 2008;36(6):1073-80. Recreation Survey Annual Report 2008. Australian Sports Commission, Canberra White PE, Ullah S, Donaldson A et al. Ostenberg A, Roos H. Injury risk factors in 2009. http://www.ausport.gov.au/__data/ Encouraging junior community netball female European football. A prospective assets/pdf_file/0004/304384/ERASS_ players to learn correct safe landing study of 123 players during one season. Report_2008.PDF technique. Journal of Science and Scandinavian Journal of Medicine and Medicine in Sport 2012;15(1):19-24 Science in Sports. 2000;10(5):279-85. Standing Committee on Exercise and Sport 2009. Participation in Exercise, Sport and Quarrie KL, Alsop JC, Waller AE et al. New Recreation Survey Annual Report 2008. Williams S, Hume PA, Kara S. A review Zealand rugby injury and performance Australian Sports Commission, Canberra of football injuries on third and fourth project VI. A prospective cohort study 2010. http://www.ausport.gov.au/__data/ generation artificial turfs compared of risk factors for injury in rugby union assets/pdf_file/0005/377069/ERASS_ with natural turf. Sports Medicine football. British Journal of Sports Report_2009.PDF 2011;41(11):903-23. Medicine 2001;35(3):157-66. Thacker SB, Stroup DF, Branche CM et al. Yoo JH, Lim BO, Ha M et al. A meta- Quarrie KL, Gianotti SM, Hopkins WG, Journal of Sports Medicine and Physical analysis of the effect of neuromuscular Hume PA. Effect of nationwide injury Fitness 2003; 43(2): 165-79. training on the prevention of the anterior prevention programme on serious spinal cruciate ligament injury in female athletes. injuries in New Zealand rugby union: Twomey DM, Finch C, Lloyd D et al. Ground Knee Surgery, Sports Traumatology, ecological study. British Medical Journal hardness and injury in community level Arthroscopy 2010;18:824-830. 2007;334(7604):1150. Australian football. Journal of Science and Medicine in Sport (2012), doi:10.1016/j. Saunders N, Otago L, Romiti et al. Coaches’ jams.2011.12.003. . perspectives on implementing an evidence- informed injury prevention programme in junior community netball. British Journal of Sports Medicine 2010;44(15):1128-32.

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 32 Box 1: Data sources and case Team bat and stick sports U51.0 “Baseball” OR U51.3 “Softball” = Baseball & selection Softball (combined) U51.1 “Cricket” = Cricket (1) Hospital admissions (hospitalisations) U51.22 “” OR U51.23 “” OR U51.29 “Hockey unspecified” = Hockey Source: Victorian Admitted Episodes Dataset (VAED) Racquet sports U59.0 “Badminton” = Badminton Hospital admissions (hospitalisations) for injury and U59.3 “Table tennis and ping pong” = Table Tennis poisoning that contain an external cause code are extracted U59.1 “Racquetball” OR U59.2 “Squash” = Racquetball from the VAED by the Victorian Department of Health & Squash combined (DH) and supplied in de-identified unit record format to U59.4 “Tennis, clay/grass/hard court= Tennis VISU every six months. The dataset includes admissions to all Victorian public and private hospitals. From July 1998 cases recorded on the VAED are coded to the ICD- (2) Emergency Department Presentations - 10-AM, the WHO International Statistical Classification non-admissions of Diseases and Related Health Problems, Tenth Revision, Australian Modification. The external causes chapter of the ICD-10-AM describe the causes of injury, poisoning and Source: Victorian Emergency Minimum Dataset (VEMD) adverse events (complications of medical and surgical care). Adverse events and sequelae (late effects) of external causes Emergency Department presentations for injury and of morbidity and mortality are usually not included in VISU poisoning are extracted from the VEMD by the Victorian reports. Department of Health (DH) and supplied quarterly in unit record format to VISU. From January 2004, VEMD Case selection: data are collected by all 38 Victorian public hospitals that 1. Admissions recorded on the VAED that occurred provide a 24-hour ED service. The VEMD contains both between 1 July 2002 and 30 June 2010 (for trend admitted and non-admitted cases. Presentations that are analyses only) and July 1, 2007 and June 30, 2010 (for treated and discharged from within the ED within 4 hours main analyses) from the time patient management commences are classified 2. The person injured was aged 15 years or older as non-admissions and cases that are treated for 4 hours or 3. The first (principal) diagnosis code asw an injury (ICD- more in the ED or a short stay ward attached to the ED or 10-AM range S00-T98) and the mode of admission depart from the ED to an inpatient bed or are transferred has any value except those indicating that transfer from to another hospital campus are classified as hospital another hospital has occurred or that the record is a admissions. Admissions for injury and poisoning recorded ‘statistical separation’ – a change of care type within a on the VEMD are not usually included in VISU injury hospital. The aim of these exclusions is to reduce over- surveillance reports if admissions are also being selected counting of cases and to provide an estimation of the from the VAED because cases would then be over-counted. incidence of injury hospitalisations in the sports being studied Case selection: 4. Cases with any of the following ICD-10-AM activity • Hospital ED presentations (non-admissions) recorded codes: on the VEMD that occur between July 1, 2007 and June Football 30, 2010 U50.00 “Australian Rules Football” OR U50.09 • Initial presentation only “Unspecified football” = Australian football • Age of presenting person has a value of 15 years or older U50.01 “Rugby Union” OR U50.02 “Rugby League” • Narrative data (short case descriptions) were used to OR U50.03 “Rugby unspecified” = Rugby Union & select cases through keyword searches as there are League combined no codes to identify specific sports-related injury cases U50.04 “Soccer” = Soccer on the VEMD. Keywords used to select cases included: Other team ball sports football, AFL, Australian Rules, footy, rugby, soccer, U50.1 “Basketball” = Basketball futsal, basketball, netball, volleyball, baseball, softball, U50.30 “” OR U50.39 “Netball other and cricket, hockey, badminton, table tennis, ping pong, unspecified” = Netball tennis and spelling variations of these. U50.5 “Volleyball, includes beach, indoor & outdoor” = • All case narratives were then read to check for relevance Volleyball and irrelevant cases were excluded.

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 33 Box 2: Special analyses: definitions Costing and methods The National Hospital Costs Data Collection (NHCDC) is based on the principles of Casemix costing analysis which Estimating rates & trends is a scientific approach to the classification of patient care whereby each hospital admission is assigned an Australian Trends were determined using a log-linear regression Refined Diagnosis Related Group (AR-DRG)2. AR-DRGs model of the rate data assuming a Poisson distribution provide a clinically meaningful way of relating the types of of injuries. The statistics relating to the trend curves, patients treated in a hospital to the resources required by the slope and intercept, estimated annual percentage change, hospital. The NHCDC contains component costs per DRG estimated overall change, 95% confidence intervals around and enables DRG Cost Weights and average costs for DRGs these estimated changes and the p-value, were calculated (National and state/territory specific) for acute in-patients using the regression model in SAS® 9.1.3. A trend was to be produced. The types of component costs included considered to be statistically significant if the p-value of the are ward medical, ward nursing, non-clinical salaries, slope of the regression model was less than 0.05. pathology, imaging, allied health, pharmacy, critical care, operating rooms, ED, ward supplies and other overheads, Injury severity: definition of ‘serious’ injury specialist procedure suites, on-costs, prostheses, hotel and depreciation. For this analysis the average Victorian To examine the severity of sports injury hospitalisations cost per AR-DRG (for the relevant year of admission) was each hospital record was given an International applied to each admission to estimate the hospital costs Classification of Disease (ICD)-based Injury Severity associated with traffic-related pedestrian injury in Victoria. Score (ICISS) (Davie et al., 2008)1. The ICISS involves estimating probability of death using the ICD injury 1 Davie G, Cryer C, Langley J. Improving the predictive diagnosis codes recorded in a person’s hospital record. ability of the ICD-based Injury Severity Score. Injury Determining which injuries are ‘serious’ involves Prevention 2008; 14; 250-255. calculating a survival risk ratio (SRR) for each individual 2 National Hospital Cost Data Collection. Hospital injury. An SRR is the proportion of cases with a certain Reference Manual Round 11 (2006-07), September injury diagnosis in which the patient does not die, or in 2007. Australian Government Department of Health and other words, a given SRR represents the likelihood that a Ageing patient will survive a particular injury. Each patient’s final ICISS is the product of the SRRs associated with all the diagnoses listed on the patient hospital record. An injury is considered serious if the ICISS is less than or equal to 0.941, this is equivalent to a survival probability of 94.1% or worse – meaning the injured person has a probability of death (when admitted) of at least 5.9%.

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 34 - INDEX - Subject...... Edition Ladders...... 63 Asphyxia...... 60 Lawn mowers...... 22 Assaults...... 55,73 Marine animals...... 56 Babywalkers, update...... 16,20,25,34 Martial arts...... 11 Baseball...... 30 Mobility scooters...... 62 Boating-related recreational injury...... 56 Motor vehicle related, non-traffic, vehicle jack injuires...... 20,63 Bunkbeds...... 11 Motorcycles...... 64,65 Bicycles - Bicycle related...... 6,31,34,44,65 Needlestick injuries...... 11,17,25 - Cyclist head injury study...... 2,7,8,10 Nursery furniture...... 37,44 Burns - Scalds, Burns prevention...... 3,12,25 Older people...... 19 - Unintentional burns and scalds in vulnerable populations...... 57 Off-street parking areas...... 20 Child care settings...... 16 Pedestrians...... 71,72 Client survey results...... 28 Playground equipment...... 3,10,14,16,25,29,44,61,65 Cutting and piercing (unintentional)...... 52 Poisons - Domestic chemical and plant poisoning...... 28 Data base use, interpretation & example of form...... 2 - Drug safety and poisons control...... 4 Deaths from injury (Victoria)...... 11,38 - Dishwasher detergent, update...... 10,6 Dishwasher machine detergents - Update...... 18 - Early Childhood, Child Resistant Closures...... 27,2,47 DIY maintenance injuries...... 41 - Adult overview...... 39 Dog bites, dog related injuries...... 3,12,25,26,34,69 Power saws, Chainsaws...... 22,28 Domestic architectural glass...... 7,22,25 Roller Blades, Skateboards...... 2,5,25,31,44 Domestic Violence...... 21,30 School...... 10 Drowning/near drowning, including updates...... 2,5,7,30,34,55 Shopping trolleys...... 22,25,42 Elastic luggage straps...... 43 Smoking-related...... 21,25,29,44 Escalator...... 24 Socio-economic status and injury...... 49, 70 Exercise bicycles, update...... 5,9 Sports - child sports, adult sports, surf sports, Falls - Child, Older Persons, Home...... 44,45,48,59 snow sports...... 8,9,44,15,51,56,66,74 Farm, Tractors...... 30,33,24,47,68 Suicide - motor vehicle exhaust gas...... 11,20,25,41 Finger jam (hand entrapment)...... 10,14,16,25,59 Trail bikes...... 31 Fireworks...... 47 Trampolines...... 13,42,61 Geographic regions of injury...... 46 Trends in road traffic fatality and injury in Victoria...... 36,65 Home...... 14,32,59,65 Vapouriser units...... 43 Horse related...... 7,23 Venomous bites and stings...... 35 Infants - injuries in the first year of life...... 8 VISS: How it works, progress, A decade of Victorian injury Injury surveillance developments, inc. ICD10 coding...... 30,43 surveillance...... 1,26,40 Intentional...... 13 VISAR: Celebration of VISAR’s achievements, VISAR name change to VISU...... 50,61 - First 3 months, Injury surveillance & prevention Work-related...... 17,18,58 ...... 9, March 1992, Feb 1994

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 35 Acknowledgements Contact VISU at: Guest Editors MIRI – Monash Injury Research Institute Professor Caroline Finch, Director, Australian Centre for Research into Injury in Building 70 Sport and its Prevention (ACRISP), Monash Injury Research Institute Monash University Victoria, 3800 VISU Staff Phone: Director: Ms Erin Cassell Enquiries (03) 9905 1805 Research Fellows: Ms Karen Ashby Co-ordinator (03) 9905 1805 Ms Angela Clapperton Director (03) 9905 1857 Research Officers: Ms Emily Kerr Fax (03) 9905 1809 Mr Isaac Dunn Email: [email protected]

Participating hospitals From October 1995 From July 1996 Austin & Repatriation Medical Centre Alfred Hospital All issues of Hazard and other Ballarat Base Hospital Monash Medical Centre information and publications of the The Hospital Campus From September 1996 Monash Injury Research Institute can Box Hill Hospital Angliss Hospital Base Hospital be found on our internet home page: The Geelong Hospital From January 1997 www.monash.edu/miri/visu Base Hospital Royal Melbourne Hospital Maroondah Hospital From January 1999 Base Hospital Werribee Mercy Hospital The Northern Hospital Royal Children’s Hospital From December 2000 St Vincents Public Hospital Rosebud Hospital Base Hospital From January 2004 & District Base Hospital Hospital Western Hospital - Footscray Central Health Service (Sale) Western Hospital - Sunshine Hamilton Base Hospital Williamstown Hospital Royal Women’s Hospital Base Hospital Sandringham & District Hospital From November 1995 Swan Hill Hospital VISU is a unit within Monash Injury Dandenong Hospital West Gippsland Hospital () Research Institute (MIRI) and is funded Regional Health Group by the Department of Health From December 1995 Royal Victorian Eye & Ear Hospital From January 2005 Hazard was produced by the Frankston Hospital Mercy Hospital for Women From April 2005 From January 1996 Injury Research Institute Latrobe Regional Hospital Casey Hospital Victorian Injury Surveillance Unit (VISU) How to access VISU data: Illustrations by Debbie Mourtzios ISSN-1320-0593 VISU collects and analyses information on injury problems to underpin the development of prevention strategies and their implementation. VISU analyses are Printed by Lithocraft, publicly available for teaching, research and prevention purposes. Requests for Truganina information should be directed to the VISU Co-ordinator or the Director by contacting them at the VISU office. Layout by Inspired by Graphics, Caroline Springs

VICTORIAN INJURY SURVEILLANCE UNIT HAZARD 74 page 36