EPIDEMIOLOGY OF CATASTROPHIC RUGBY INJURIES IN NEW SOUTH WALES

Tai Rotem

This dissertation is submitted in fulfilment of the requirements of the degree of Masters by research thesis.

Faculty of Medicine, University of New South Wales, 2007 Publications and official reports generated from this research

Published articles in refereed journals

Rotem TR, Lawson JS, Wilson SF, Engel S, Rutkowski SB, Aisbett CW. 1998. Severe cervical spinal cord injuries related to and league football in New South Wales, 1984-1996. Medical Journal of Australia 168(8): 379-381.

Rotem TR, Davidson R. 2001. Epidemiology of Schoolboy football injuries. International Sports Med Journal 2(2). [Online] Available at: www.esportmed.com/ismj

Lawson J, Rotem T, Wilson S. 1995. Catastrophic injuries to the eyes and testicles in footballers. Medical Journal of Australia 163(5): 242-244.

Published letters in refereed journals

Wilson SF, Atkin PA, Rotem T, Lawson JS. 1996. Spinal cord injuries have fallen in rugby union in New South Wales. British Medical Journal 313(7071):1550.

Published conference papers

Geffen S, Wilson S, Rotem T, Lawson J. 1999. Severe injury and fatality in football. Abstracts from the Football Australasia Conference. Journal of Science and Medicine in Sport 2 (1): 178.

Official reports

Rotem T, Wilson S, Lawson J. 2001. injury mechanism study. Report to the New South Wales Sporting Injuries Committee Research and Injury Prevention Program.

Lawson J, Wilson S, Rotem T. 1995. Head and neck injuries in sport: Report to the New South Wales Sporting Injuries Committee Research and Injury Prevention Program. Epidemiology of catastrophic rugby football injuries in New South Wales

ABSTRACT

Aims and objectives: To investigate the epidemiology, risk factors, and preventive strategies for serious head and spinal injuries related to rugby league and union football in New South Wales (NSW).

Methods: The three main components of this study included.

1. A retrospective analysis of clinical and compensation case file records during the 16-year period, 1984-1999. 2. A review of game rules, weights and heights of players, over the past 100 years. 3. A review of film and video footage of rugby football games spanning most of the 20th century. A method was developed to allow valid comparisons in style of play over a 70-year period.

Results: The estimated annual incidence rate of SCI for rugby league players was 1.9 (95%CI 1.3 - 2.8) per 100,000 estimated players per year, and 3.3 (95%CI 2.2 - 4.8) per 100,000 estimated rugby union players per year over the study period 1984 to 1999. There was no significant change in the incidence of rugby league related serious spinal cord injuries (1984-1999), fatalities (1984-1999) or serious head injuries (1984-1999). There was a small but significant decline in rugby union related serious spinal cord injuries (1984-1999, P<0.05). However, the relative risk of spinal cord injury was 1.34 times greater for rugby union compared to rugby league over the entire study period (95%CI 1.01 - 1.67, P<0.05).

For rugby football, the evidence suggested that the force of impact between participants was a key causal risk factor for serious injury. Elite rugby league and union players from 1999-2000 teams were significantly heavier (P<0.05) and taller (P<0.05) than players pre 1950. Players in modern elite games post 1989 were more likely to be tackled by multiple opponents (P = 0.000), tackled head on (P<0.05), at chest level (P<0.05) and at greater speeds than their earlier counterparts pre 1958. However, modern players appear to be no more aggressive or perpetrate greater foul play than their predecessors.

Conclusions: There was a continued annual occurrence of catastrophically serious injuries leading to permanent brain damage and quadriplegia associated with rugby league and union in NSW. The critical risk factors found to be associated with serious injury in rugby football suggest relatively novel approaches to the development of preventive strategies. ACKNOWLEDGMENTS

I would like to acknowledge with much appreciation the support and guidance of Professor Anthony Zwi and Associate Professor Mary-Lou McLaws, my academic supervisors. Their guidance has been very helpful in completing the present dissertation report.

I would also like to convey deep appreciation to Professor James Lawson, my original academic supervisor. I was privileged to work with Professor Lawson and Dr Stephen Wilson in the conduct of the original study in 1998 from which this dissertation evolved. The late Dr David Garlick (School of Physiology and Pharmacology, UNSW) provided valuable advice and support as my co supervisor.

I am indebted to the NSW Sporting Injuries Committee for the funding support they have provided reflecting their strong commitment to injury prevention research. Special thanks to Mr John Anderson and all the staff at the NSW Sporting Injuries Committee who provided ongoing support and access to the benefits paid case data.

Dr George Argyrous (School of Social Science and Policy, UNSW) offered most helpful advice on statistical analysis and research methods.

Dr Sue Rutkowski and Dr Stella Engel provided access to spinal unit medical records. Mr Simon Drake and Ms Naomi Lamb at Screensound Film and Sound Archive Australia assisted in obtaining archival football video footage. Similarly, Mr Alex Byron at Sports Recording Services assisted in providing footage of more recent games.

I acknowledge with appreciation the work completed by Mr Jordan Telfer. Using the designed recording tool, Mr Telfer was employed to review the archival film and video footage gathered for this study independently, with no knowledge of the study hypothesis.

Finally, I am very grateful to my parents for the interest and support they have provided over the years.

1 TABLE OF CONTENTS

LIST OF APPENDICES 7 LIST OF FIGURES 7 LIST OF TABLES 8

TABLE OF ABBREVIATIONS 9

CHAPTER ONE – INTRODUCTION 10

Background 10 Importance of sport and recreational activities 10 The risk of injury associated with sport and recreational activities 11 The burden of sporting injuries 12 The role of research and government policy in sporting injury prevention 13 Objective 15 Research objectives 15 Organization of thesis structure 17

CHAPTER TWO – LITERATURE REVIEW 19

SECTION ONE Public health approaches to preventing serious injuries in Australia 20 Establishment of injury prevention programs 21 Resources and coordination for injury prevention research 24 Road related injury and fatality prevention 26 Childhood pool drowning prevention 28 Sport and recreational injury and fatality prevention 30

SECTION TWO Developing and evaluating injury prevention strategies in sport and recreational activities 37 Study design in epidemiological studies of sport injuries 38 Constraints in data collection 40

SECTION THREE Background to serious injuries occurring in sport 42 Spinal Cord Injury 42 Clinical definition and management of spinal cord injury 42 Risk Factors for spinal cord injury 45 Epidemiology of spinal cord injuries in Australia 46 Head and Brain Injury 50 Clinical definition and management of head and brain injury 51 Risk factors for head injury 54 Epidemiology of head and brain Injuries in Australia 55 2 Fatalities 57 Risk factors for fatal injuries 57 Epidemiology of fatalities in Australia 58 Non-traumatic fatalities in sport and recreational activities 60

SECTION FOUR Serious rugby football injuries 62 Explanation of rugby football codes 62 Rugby league rules 62 Rugby union rules 64 SCI in rugby football 68 Brain injuries in rugby football 70 Prevention strategies for serious rugby football injuries - safety programs, rule change and enforcement 71 Rugby union 71 Rugby league 74 Effectiveness of prevention strategies in reducing serious injury 75 Legal consequences of rugby football injuries 77

SECTION FIVE Risk factors for serious injury in rugby football 79 Overview 79 Extrinsic risk factors 82 Types and style of play that increase the risk of serious injury 82 Spinal cord injury 82 Head injury 83 Mechanism of tackling injury 83 Force of impact 85 Direction of impact 88 Multiple tacklers 90 Level of impact 91 High tackles risk of injury to the ball carrier 91 Low tackles increased risk of injury to the tackler 92 Mechanism of ruck and maul injuries 93 Mechanism of scrum injuries 94 Acute management for suspected serious injuries 97 Spinal cord injury 97 Brain injury 98 Illegal/foul play 99 Number of players 103 Stage of season 104 Other external risk factors 105 Protective equipment for serious injury in football 106 Spinal cord injury 107 Head and brain injury 108

3

Intrinsic risk factors 110 Player characteristics that influence players’ serious injury risk 110 Anthropometric and physiological characteristics 110 Pre-existing conditions 114 Physical preparation 116 Skills, experience and technique 118 Level of competition, competitiveness and aggression 121 Age 127 Player position 128 Risk Exposure duration 130

SECTION SIX The historical and sociological development of rugby football and its effect on participant aggression and competitiveness 131

SECTION SEVEN Style of football played in modern times (post 1989) compared to that played in an earlier era (pre 1958) 139 Number of tacklers 140 Impact force 141 Dangerous scrums and tackling technique 141 Illegal/foul play 144 Number of players 145 Players strength, fitness and size 146

CHAPTER THREE – METHODOLOGY 152

SECTION ONE Injury definitions of data included in study population 152 Spinal cord injury 152 Head injury 153 Fatality 153 Study design 154 Study group 156

SECTION TWO Sources of data 158 NSW spinal units 158 NSW Sporting Injuries Insurance Scheme 160 Rules of play 162 Identification of the literature 162 Archival film and video footage 163

SECTION THREE Statistical analysis 169 Incidence rate analysis 169 Trend analysis and code comparisons 170 Risk factor analysis 171

4 Archival video and player size analysis 171 Data limitations 173 Research design 173 Incidence rate calculations 173 Statistical analysis 173 Coverage 173 Risk factor detail 175 Film and video analysis 175

CHAPTER FOUR – FINDINGS 177

SECTION ONE Overview of serious injury related to sport and recreational activities in New South Wales 1980-1999 177 Spinal cord, head and fatal injuries compensated by the New South Wales Sporting Injuries Committee Insurance Scheme 1980-1999 177 Sport and recreational admissions to NSW Spinal units 1984-1999 181

SECTION TWO Serious rugby football injuries 183 NSW Spinal units and NSW Sporting Injuries Committee Insurance Scheme rugby football SCI cases 1984-1999 183 NSWSIC rugby league brain injuries and fatalities 1984-1999 188 Risk factors (Extrinsic) 191 Types and style of play that increase the risk of serious injury 191 Spinal cord injury 191 Brain injury 193 Fatalities 193 Multiple tacklers 194 Level of impact 195 Rucks and mauls 197 Scrums 197 Illegal play 199 Injury management for suspected serious injuries 200 Protective equipment for serious injury 201 Risk factors (Intrinsic) 201 Weight/height and body type 201 Physical preparation 202 Skills, experience and technique 202 Player position 202 Age 203 Country vs Metropolitan competitions 203 Level of competition 205 Time in game and season 205 Position on field 206 Training 206

5

SECTION THREE Development of Rugby football 207 Style of football played in modern times (post-1989) compared to that played in earlier era (pre-1958) 207 Number of tacklers 207 Level of impact 208 Direction of impact 209 Impact speed 210 Ball carrying style 210 Dangerous scrums and tackling technique 211 Comparisons of rugby football players’ age, height and weight, Pre-1950 with post-1998 212

CHAPTER FIVE – DISCUSSION 219

Relative contribution of various organised sports and recreational activities to serious injury and fatalities in NSW 219 NSW Sport injuries Committee 1980-1999 219 NSW Spinal unit admissions 1984-1994 221 Rugby football SCI, brain injury and fatality 1984 to 1999 222 Risk factor analysis 224 Interrelationship of risk factors 225 Extrinsic risk factors 228 Intrinsic risk factors 236 Historical, sociological, playing style, player characteristics developments in rugby football 244 Implications of study findings for injury prevention infrastructure in Australia and methodological approaches to rugby football injury prevention 246

CHAPTER SIX – CONCLUSION AND RECCOMENDATIONS 251

Conclusion 251 Recommendations 255 Decreasing forces in impacts 257 Player selection and training 258 Illegal play 260 Risk exposure 260 Injury management 260 Protective equipment 261 Injury surveillance 262

REFERENCES 263

6

LIST OF APPENDICES

Appendix A – ASIA impairment scale i Appendix B – Major definitions used for the pilot Australian rugby union injury survey in 2000-2002 iii Appendix C – Medical record/Case file data collection form v Appendix D – NSW Sporting Injuries Committee membership and premium rates viii Appendix E – Video incident review form and coding manual xii Appendix F – Glossary xvi Appendix G – Estimates of participant populations for miscellaneous sports xxiv Appendix H – Statistical analysis xxvi

LIST OF FIGURES

Figure 1: Multiple tacklers in rugby league 90 Figure 2: Head high tackles - league 91 Figure 3: Head high tackles - league 91 Figure 4: Head high tackles - league 92 Figure 5: Tackler injured through head hitting hip of opponent – league 92 Figure 6: Spear tackles - league 102 Figure 7: Spear tackles - league 102 Figure 8: Spear tackles - league 102 Figure 9: Young players practicing scrummaging very low against scrum machine in the 1930s 143 Figure 10: Early conceptions of ideal tackling technique Illegal/foul play 144 Figure 11: Early conceptions of ideal tackling technique Illegal/foul play 144 Figure 12: Study group chart 157 Figure 13: Flow chart of hypothesised interrelationship between risk factors for serious injury and fatality in rugby football 227

7 LIST OF TABLES

Table 1: Traumatic SCI by cause 1995-2001 49 Table 2: Potential risk factors for serious injury in rugby football 79 Table 3: Haddon matrix for risk factors in rugby football serious injuries and fatalities 81 Table 4: Levels of research design and suitability to research questions posed 155 Table 5: Availability of data by source, injury and code 162 Table 6: Intra-rater reliability of recording tool 166 Table 7: Inter-rater reliability of recording tool 166 Table 8: NSW Sporting Injuries Committee compensated brain, neck injuries and fatalities 1980-1999 180 Table 9:: Acute sport and recreational admissions to the main spinal injury units in NSW 1984 to 1994 182 Table 10: Spinal unit admissions and cervical spinal cord injury in rugby union and league footballers in NSW 1984–1999 183 Table 11: Summary of spinal unit admissions and cervical spinal cord injury in rugby union and league footballers in NSW 1984–1999 184 Table 12: Summary of trends in spinal unit admissions and incidence of cervical spinal cord injuries in rugby league 185 Table 13: Tests of proportional difference between the incidence rate in the baseline period and each subsequent period for rugby union SCI 187 Table 14: Site of cervical spinal cord injuries in rugby football players in NSW from 1984-1999 187 Table 15: Summary of brain injuries and fatalities for rugby union and league footballers in NSW 1984–1999 188 Table 16: Rugby league brain injuries and fatalities from the NSW Sporting Injuries Committee benefits paid cases 1984-1999 189 Table 17: Cause of SCI, spinal admission, brain injury and fatalities by rugby football code in NSW 1984-1999 192 Table 18: Level of first impact by ball carrying style 196 Table 19: Descriptive age analysis of seriously injured rugby football players in NSW 1984-1999 203 Table 20: Team region of all injured players by code 204 Table 21: Distribution of all rugby football serious injuries and fatalities over months of the year 205 Table 22: Number of tacklers by period 208 Table 23: Level of initial by period 209 Table 24: Direction of initial tacklers impact by period 209 Table 25: Ball carrying style by period 211 Table 26: Separated codes comparing age, height and weight for early and modern players 212 Table 27: Comparisons of age, height and weight by codes and period 213 Table 28: Comparisons of age, height and weight by player position and code 213 Table 29: Comparisons of age, height and weight by player position by code and period 214

8 TABLE OF ABBREVIATIONS

ABS Australian Bureau of Statistics AIHW Australian Institute of Health and Welfare AIPN Australian Injury Prevention Network ARL ARU/ARFU Australian Rugby Union/Australian ASC Australian Sports Commission ASCIR Australian Spinal Cord Injury Register ASIA American Spinal Injury Association ASIPT Australian Sports Injury Prevention Taskforce CDHFS Commonwealth Department of Health and Family Services CDHSH Commonwealth Department of Housing Services and Health DHFS Department of Health and Family Services ECG Electrocardiogram IRFB International Rugby Football Board Kg, m/sec, km/h Kilograms, metres per second, kilometres per hour NHMRC National Health and Medical Research Council NISU National Injury Surveillance Unit NRL National Rugby League (Australia) NSW New South Wales NSWRFL/NSWRL NSW Rugby Football League/NSW Rugby League NSWRFU/NSWRU NSW Rugby Football Union/NSW Rugby Union NSWSIC New South Wales Sporting Injuries Committee NSWSU New South Wales Spinal Units (acute admission centres) NZ New Zealand PHA Public Health Association of Australia PPE Pre-participation physical evaluation SCI Spinal cord injury TBI Traumatic brain injury

9 CHAPTER ONE INTRODUCTION

BACKGROUND

The importance of sport and recreational activities

Sport and recreational activities fulfil many important social and economic roles and are an integral part of healthy living. Public and private institutions recognise the important role of these activities and provide different forms of support (Australian Bureau of Statistics [ABS] 2001a; Australian Sports Commission [ASC] 2006; Orchard & Finch 2002; Sport and Recreation Ministers' Council 1997; Truswell 1997). These institutions recognise the positive social contribution of sport and recreation to social cohesion and community development. Involvement in sport is regarded as important for skill development, stress reduction, enjoyment, fitness, prevention of chronic diseases and the promotion of both physical and mental health.

In Australia, there is a high level of participation among school children and adults in a variety of sporting and physical activities. Australia’s history of involvement in sport and recreational activities has long seen it distinguished as a sporting nation, particularly in codes of contact football such as rugby league, rugby union, and Australian Rules (ABS 1999a; ABS 2002a). Interest and participation has only heightened in recent years by Australia's success in international sporting events such as the Olympic games, Commonwealth games, Soccer and Rugby World cups (ASC 2006).

10 The risk of injury associated with sport and recreational activities

Despite the significant health, social and economic benefits to be gained by participating in sport and physical activities (World Health Organisation [WHO] 2004), these activities are not without risk of injury (Finch & Owen 2001; Fricker 1999; Gotsch et al. 2002; Rotem et al. 1998; Sherry 1998; Watt & Finch 1996). Injuries associated with contact sports and recreational activities range from minor, acute and chronic injuries to the ankle, knee and shoulder as well as lacerations and muscle strain (ABS 1995; Seward et al. 1993; Sherry 1998), to catastrophic injuries to the spine, brain and vital organs. Most sporting injuries are musculoskeletal. Brain injuries, spinal cord injuries and other internal injuries are much less common, but carry greater risk of death or other serious and lifelong consequences (Harrison 1999; Lawson et al. 1995; Rotem et al. 1998).

Minor and chronic injuries are so common in most sports that their management is considered integral to the sport (Almekinders 1999; Browning & Donley 2000; Fadale & Hulstyn 1997; Gregory & Van Valkenburgh 1990; Locke 1999; Lynch & Renstrom 1999). An athlete’s ability to recover from these injuries often determines their degree of success over time. In response, developments in sports medicine increasingly address recuperation of athletes from minor and chronic injury along with associated pain management (Sherry 1998).

Catastrophic sporting injuries result in permanent disabling conditions (those that severely limit functional ability), shortened life expectancy, or death (Sherry 1998). Catastrophic injuries to the spinal cord or to the brain are most often associated with recreational diving, horse riding and contact sports such as rugby league and union (Blanksby et al. 1997; Cripps 2004, O'Connor 2000b, O'Connor 2002b; Rotem et al. 1998; Watt & Finch 1996).

11 Catastrophic injuries and their associated costs cannot be accepted in the same manner as bruising, strains and lacerations. These injuries need to be prevented rather than just managed because of their grave consequences. It is expected that the priority for prevention of injuries will be related to their degree of severity and not merely to their level of incidence (Van Mechelen 1997b). The future popularity of high-risk sports such as rugby football depends on the ability of sporting organisations to improve safety for players. This expectation is particularly strong at the community level where support for elite forms of the game is generated (Sports Medicine Australia 2002).

The burden of sporting injuries

Sporting injuries are one of the most frequent injuries in developed countries (Parkkari et al. 2001). Between 1989 and 1993, sport and recreational activities accounted for 20% of all child and 18% of all adult injury presentations to emergency departments in Australia (Finch et al. 1998). As treatment may be sought from the private system, or may not be sought at all, it is likely that the true incidence of sport and recreational injuries is even higher, representing a considerable demand on health resources.

While the total number or incidence rate of sports injuries occurring annually in Australia is unknown, the significant associated costs to the health care system and losses through time off work due to injury were estimated in 1990 to be around one billion dollars (Egger 1991). A more recent estimate, extrapolated from cost estimates made in a 1998 Victorian study, estimated that sports injuries now cost at least $1.65 billion a year to the Australian community (Orchard & Finch 2002).

12 Similar patterns are observed in the United States, where an estimated 4.3 million (95% CI = 3.7-4.8 million) sports and recreation-related injuries were treated at hospital emergency departments from July 2000 to June 2001. This comprised 16% of all unintentional injury-related emergency department visits (Gotsch et al. 2002).

Severe injuries carry a particularly large financial and social burden. The average cost of the long-term care of a person with spinal cord injury (SCI) in Australia has been estimated to range from $600,000 for a paraplegic, to $4,000,000 for a ventilator-dependent tetraplegic (Walsh & DeRavin 1995). Beyond the costs incurred for care, severe injuries represent a financial cost to the community through loss of productive members. Those who suffer a SCI due to trauma are commonly young people at the peak of their productivity and earning capacity, 60% to 70% of whom will not return to work (Athanasou & Murphy 1996).

The role of research and government policy in sporting injury prevention

The importance of research into the nature of sporting injuries and their prevention has been reported in various literary sources and documented in government policy (ASC 1997; Finch et al 1995; Finch & Owen 2001; O’Connor 2000a; Public Health Association 1999; Orchard & Finch 2002). The responsible promotion of sporting and recreational activities necessitates consideration of injury-prevention principles. Activities that inherently carry the greatest risk of severe injury need to be identified. The associated risks need to be addressed through preventive strategies including education of the public concerning the risk of permanent injury from certain sport-related activities. Increasing awareness of the need to reduce the risk of sporting injuries over recent decades has seen the introduction of various injury prevention strategies (Blanksby et al. 1996; Blitvich et al. 2000; Rotem et al. 1998; Silver 2002; Taylor & Coolican 1987; Watt & Finch 1996).

13

The investigation and a better understanding of the factors responsible for sporting injury is an essential antecedent to the development of evidence based preventive measures (Bottini et al. 2000; Gibbs 1994; Milburn 1993). Proper epidemiological evidence to assess the magnitude of the problem, and evaluation of the effectiveness of existing and planned injury prevention strategies is essential for changing current trends (Noakes & Jakoet 1995). While an evidence-based public health approach to sporting injury prevention is relatively new, similar approaches have demonstrated marked success in the prevention of injury and death on Australian roads in recent decades (Australian Transport Safety Bureau 2001; ABS 2002b; McDermott 1985; Lawson 2001; Orchard & Finch 2002).

14 OBJECTIVES

This study aims to investigate serious injuries in rugby football in NSW, a major contributor to serious injuries in Australia. This study focuses on fatalities, and brain and spinal injuries associated with rugby football in NSW during the period 1984 to 1999. Due to the relatively rare nature of these events, three sources of retrospective surveillance data are utilised, NSW Spinal Injury Units medical records, NSW Sporting Injuries Committee benefits paid case files, and archival footage of rugby football games. These data sources are used to address the following objectives:

Research Objectives

This study seeks to: 1. Determine the frequency of fatalities and acute, permanently disabling spinal injury and brain injury in sport and recreation over the study period in NSW. 2. Determine the rate of fatalities and acute, permanently disabling spinal and brain injuries for the total number of registered NSW rugby football players over the study period. 3. Determine whether the rate of these injuries and fatalities have changed over this period. 4. Identify the extrinsic and intrinsic risk factors common to seriously injured players over the study period. 5. Detail the development of football codes over the last century in order to identify changes in the rules, styles of play and player size that may be associated with changes in the frequency of injury. 6. Determine the number of injuries reported in NSW since the introduction of safety measures and rules. 7. Establish new techniques for assessing the critical factors in style of play that contribute to injury risk. 8. Assess whether current injury prevention strategies could be improved.

15 To investigate serious injuries in rugby football, this thesis takes a population perspective to support the development of preventive strategies using both quantitative and qualitative examination. Attention is given to injury rates and risk factors as well as the effectiveness of existing and potential sporting injury prevention strategies. While the biomechanics of injury and secondary prevention strategies are considered, such as post injury management, the thesis does not address the clinical implications of these injuries.

Several methodological approaches are incorporated from a range of disciplines including epidemiology, biomechanics and sociology. These varied but interrelated conceptual approaches provide complementary findings that facilitate the development of strategies for prevention.

16 ORGANISATION OF THESIS

The thesis is divided into six chapters, each focusing on a different stage in the research process.

Chapter 1, Introduction, outlines the significance of the study, research aims and the organisation of the thesis.

Chapter 2, Literature Review, consists of a review of the literature concerning the effectiveness of a public health approach to injury prevention. This is followed by an examination of the existing sporting injury prevention and surveillance program infrastructure in Australia. An overview of epidemiological and clinical evidence associated with serious injuries provides perspective for assessing the contribution of sport and recreational activities. Available epidemiological evidence concerning rugby football injury is reviewed including a consideration of associated risk factors. Archival evidence is reviewed to establish how rules, styles of play and player safety have changed since the inception of both rugby codes.

Chapter 3, Methodology, presents the methods employed in each section of this study. Methodological approaches from disciplines including epidemiology, biomechanics, and sociology are incorporated.

Chapter 4, Results, begins with a description of serious injuries from all sport and recreational activities, then focuses on rugby football injuries from a 16-year surveillance database. Rugby football rules are examined in order to identify developments that correspond to changes in the incidence of injury over a period of 16 years. Observational findings from an examination of video taped games are used to assess changes in elite playing style over the 20th century.

17 Chapter 5, Discussion, considers the study’s findings within the Australian rugby football context, for serious injury and fatality prevention, and takes into account implications for stakeholders.

Chapter 6, Conclusion and recommendations, concludes the thesis with a final assessment of the implications of the findings and provides recommendations for future research.

CHAPTER ONE SUMMARY

The significant health, social and economic benefits to be gained by participating in sport and physical activities are well evidenced. However, these activities are not without the risk serious injury which pose a significant burden to Australian society.

The responsible promotion of sporting and recreational activities needs to address the associated risks through evidenced based preventive strategies. This approach has demonstrated marked success in the prevention of injury and death on Australian roads in recent decades.

This study aims to investigate serious injuries in rugby football in NSW, a major contributor to serious injuries in Australia. Three sources of retrospective surveillance data are utilised, NSW Spinal Injury Units medical records, NSW Sporting Injuries Committee benefits paid case files, and archival footage of rugby football games.

18 CHAPTER TWO LITERATURE REVIEW

Overview

This chapter provides a critical review of the literature pertaining to research and development in sporting injury prevention, with particular reference to rugby football. The review sets the scene for this study and highlights the critical issues it faces concerning definitions, methodological approaches, risk factors and preventive strategies. The material is organised into eight major sections as follows:

Section one The effective public health approaches to injury prevention and the development of an Australian infrastructure for sporting injury prevention.

Section two The methodological considerations for effective sporting injury research and prevention.

Section three The clinical and epidemiological background to serious sporting injuries.

Section four An overview of rugby football (league and union) rules, epidemiology of serious injuries, prevention strategies used and their effectiveness.

Section five The extrinsic and intrinsic risk factors of serious rugby football injury and the evaluation of the effectiveness of protective equipment for serious injury prevention.

Section six The historical and sociological developments in the rugby codes that have influenced established risk factors attributed to injury.

Section seven Changes in player characteristics that have influenced established risk factors attributed to injury.

19 SECTION ONE

PUBLIC HEALTH APPROACHES TO PREVENTING SERIOUS INJURIES IN AUSTRALIA

This section reviews the literature relating to public health approaches to serious injury prevention in Australia. It presents the key principles underlying these approaches, and the positive safety outcomes demonstrated in areas such as road related injury and fatality prevention, childhood pool drowning prevention, and sports injury prevention. Finally, this section provides an overview of the development of an Australian infrastructure for sporting injury prevention.

The steady increase in Spinal Cord Injury (SCI) and head injury, from all causes in Australia during the 1970s and early 1980s, led to concerted efforts by researchers to identify risk factors and preventive strategies (Bedbrook 1992; Finch 1995; Lawson & Bauman 2001; O’Connor 2002a, 2002b; O'Connor & Cripps 1996; Rotem et al. 1998; Taylor & Coolican 1987; Woodward et al. 1984; Yeo 1993 1998a). A growing focus on injury as a public health issue in Australia developed in response to significant numbers of people being hurt or killed as a result of injury. The need for action in this area was accentuated by the success of disease prevention and control programs in improving Australian's life expectancy (Australian Institute of Health and Welfare & Commonwealth Department of Health and Family Services [AIHW&DHFS] 1997). During the 1990s, injury in Australia was responsible for more than 7,000 deaths annually. The number of related hospitalisations reached 400,000, and related medical costs were estimated at $2.6billion (National Health and Medical Research Council [NHMRC] 1999).

20 The success of injury prevention strategies in Australia and overseas led to the rejection of the concept of ‘accidental’ injury, as this implied an inability to predict or prevent injuries. Rejecting this notion of injuries being accidental, the British Medical Journal banned the use of the word ‘accident’ in its publications, and discouraged its use elsewhere (Davis & Pless 2001). This principle has since been accepted in relation to rugby football SCI, as definable risk factors have now been established (Silver 2002).

From a public health perspective, primary prevention of serious injuries such as SCI is always preferred. Despite significant advances in medicine, complete recovery from these types of injuries is not expected in the near future (Bedbrook 1992; O'Connor 2001b; Yarkony et al. 1997). The sheer volume of traumatic injuries, and the high incidence among young healthy people, has led to the recognition of injury prevention as an important area of study and public health concern.

A public health approach to injury prevention and control embraces evidence based on primary and secondary prevention of injuries, using a careful sequence of monitoring, evaluation and intervention.

Establishment of injury prevention programs

Recognition of injury as an important public health issue during the 1980s led to the establishment of several injury surveillance and prevention initiatives. This recognition is particularly evident in the road safety sector, and in moves towards the establishment of ongoing infrastructure for injury prevention within health agencies (AIHW&DHFS 1997; Australian Injury Prevention Network [AIPN] 2001; NHMRC 1999; PHA 1999).

21 Better Health Commission In 1986, the Better Health Commission published a report that is now considered Australia’s first endeavour to develop an all encompassing, national orientation to injury prevention (Better Health Commission 1986). This work provided the basis for the development of a revised set of goals and targets in relation to injury prevention, prepared for the Commonwealth Department of Health, Housing and Community Services by Nutbeam et al. (1993). One year later, the Australian Government officially recognised injury control as an area of major public health importance (Commonwealth of Australia 1994). However, despite these improvements, a report documenting the implementation of national goals and strategies for injury prevention published in 1997, found little systematically planned development in the Australian public health sector to enable identification and prevention of injury (AIHW&DHFS 1997).

National Injury Surveillance Unit The Australian Spinal Cord Injury Surveillance System (ASCIR) was established in 1995 by the National Injury Surveillance Unit (NISU). The NISU is responsible for public health surveillance of injury at a national level and compiles a range of relevant information from the six Australian spinal injury units. A register of incident cases was established by NISU, and as a result, uniform core surveillance data is now incorporated into the routine registration of cases by spinal units. A central collection of core data is managed by NISU. The unit also prepares regular statistical summaries from a national perspective and provides an ad hoc information service. The NISU aims to improve the surveillance of injury by planning and developing surveillance methods and by supporting users of the Injury Surveillance Information System. It also encourages and conducts research into the cause and prevention of injury and provides an information exchange service to practitioners (O'Connor 2000a, 2002, 2002a; O'Connor & Cripps 1996 1999).

22 The Australian Injury Prevention Network The Australian Injury Prevention Network (AIPN) was formed in 1996, as the peak national body advocating for injury prevention in Australia. Its aim was to establish a framework for collaboration between injury researchers, policymakers and practitioners, and to provide coordination to injury prevention professionals. It has broad-based representation among the community and professionals and national coverage. The scope of the AIPN interest areas includes: injury prevention; treatment and rehabilitation; research and surveillance; education and training; and information exchange. The AIPN has four main objectives, which include: organisational development; information exchange; workforce development; and partnership development (AIPN 2001).

Australian Health Ministers The Australian Health Ministers reaffirmed the importance of injury prevention in the late 1990s. The Health Ministers identified injury prevention as a National Health Priority Area using performance indicators such as the annual incidence rate of SCI from traumatic causes (AIHW&DHFS 1998; O’Connor 2000a). However, though the importance of SCI has been recognised and selected as one of the indicators for injury prevention and control, no target has yet been set for its reduction. The ASCIR enables the patterns and trends in SCI to be monitored and can assist in defining an appropriate target (O'Connor & Cripps 1999).

The National Health Priority Areas Report The National Health Priority Areas Report – Injury Prevention and Control (1997) states: ‘The health sector’s role in injury is to: plan and implement prevention strategies to reduce the incidence of injury; provide treatment and rehabilitation for those affected by injury; identify priority issues for collecting and analysing relevant data and promote inter-sectoral cooperation on injury issues.’ (Moller & Elkington 1997; p15).

23 Strategic Injury Prevention Partnership Further important developments in this area include the establishment of the Strategic Injury Prevention Partnership representing all health departments in 2000, to implement the Federal Government’s Health Policy Plan – National injury prevention policies priorities for 2001–2003 (Orchard & Finch 2002).

Resources and coordination for injury prevention research

Despite these policy initiatives and infrastructure developments over the 1980s and 1990s, injury prevention in Australia remains under-resourced in comparison to other health priority areas (Orchard & Finch 2002). The Public Health Association of Australia (PHA) has identified the need to actively lobby governments to accord a higher priority to injury control providing resources proportional to the magnitude of the problem. The PHA argues that there is still a need for better injury surveillance and inter-sectoral links aimed at monitoring and preventing injury (PHA 1999). Comments from the 2nd National Injury Prevention Conference held in Melbourne in 1998 suggested that resources available for injury prevention initiatives in Australia were linked to levels of public interest and political commitment. The difficultly for injury prevention is its multi-disciplinary and inter- sectoral nature. This impedes the ability to develop an effective political constituency to lobby for sufficient resources (Orchard & Finch 2002).

It is argued that in contrast to disease related research, where profit-oriented pharmaceutical companies often provide funding for independent randomised controlled trials, limited support is available to injury intervention studies. The Strategic Research Development Committee of the National Health and Medical Research Council (NHMRC), in its assessment of new research directions for injury prevention (NHMRC 1999), recommend that new partnerships between researchers and health administrators were needed to address this area. The Committee argued that injury research is required to shift from traditional health research paradigms to incorporate contributions from a wider range of disciplines and different health system structures.

24 The injury research sector could still maintain its independence in research, assisting in the setting of standards, certifying interventions, and obtaining resources for pilot interventions from the implementing agencies of the health sector (NHMRC 1999).

The Strategic Research Development Committee also concluded (NHMRC 1999) that not withstanding the great value of developments in injury prevention policies and infrastructure, they have not achieved the goal of an integrated national response. Although the many local, state and issue-specific groups responsible for injury prevention in Australia have been successful partly because of the focused nature and their relatively homogenous memberships; there are limitations to an unplanned fragmented approach that is increasingly becoming evident. In areas such as road and occupational injury, relevant, administrative structures have created more systematic approaches to research and intervention. In areas where these structures have been lacking there has been a shortage of guidance and limited progress (NHMRC 1999).

Most importantly, the current approach does not provide a peak body with clear responsibilities to develop an infrastructure to build a quality assured injury prevention workforce, facilitate new research and implement coordinated prevention programs across sectors, interest groups and geographical areas. The political strength of injury issues has lacked power in the absence of such a peak body representing community interests (NHMRC 1999).

Despite the acknowledged importance of injury as an issue, the evidence needed for confident implementation of interventions is still lacking for many causes and types of injury. There is commonly only limited incidence and nature of injury data available for many sports and recreation activities, particularly at the non elite community level (Finch & Mitchell 2002; Finch et al. 1998; NHMRC 1999). In-depth research and analytical studies to identify the aetiology of injury and intervention strategies are lacking outside the road injury sector. Examples of concerted public

25 health inquiry in other injury areas including road related injuries and fatalities, and in the prevention of childhood drowning in pools, demonstrate the potential contribution of such effort to the development of successful injury prevention strategies.

Road related injury and fatality prevention

From 1925 to 1970 there was a steady increase in road fatalities in Australia, with the exception of periods during the Great Depression and the Second World War. To address the issue, a wide range of population-wide preventive interventions has been introduced over the last thirty years. These have proven to be highly successful in the reduction of road related injuries, the most common cause of serious injury in Australia. Since 1970 the number of fatalities per year has declined significantly, particularly during the 1980s and 1990s (Australian Bureau of Statistics [ABS] 2002b). The rate of motor vehicle mortalities in Australia dropped from 30 per 100,000 in the 1970s to 11 per 100,000 in 1991. There was also a decrease in the severity of injury in traffic collisions (Lawson & Bauman 2001). In 1999 road fatalities had decreased to less than half the 1970 rate. However, after steadily falling from the mid 1980s to 1997, the road toll from 1997 to 2001 has been effectively constant with a slight decline in the last four years (Australian Transport Safety Bureau [ATSB] 2005).

This reduction in road traffic fatalities can be attributed to improvements to roads and vehicles, enactment of road safety legislation, intensive public education and enhanced police enforcement technology (ABS 2002b). Preventive strategies include the introduction of compulsory wearing of seatbelts for motorists and wearing of helmets for motorcyclists, tough drink-driving rules, random breath testing, tougher speeding laws, speed cameras, safer vehicles and improved road design (ABS 2002b; Chorba 1991; Lawson 1991).

26 These safety measures were developed and introduced through research and the engagement of community support for previously unpopular measures. Epidemiological investigations have revealed other valuable data such as ‘black spots’ to be targeted for improvement and particular drivers or situations at higher risk (e.g. occupants of forward control vehicles are at greater risk of injury following crashes than occupants of conventional passenger cars) (Lawson 1991).

This valuable epidemiological information provides important lessons and directions for current and future preventive actions. The prevention of road injuries in Australia has seen an effective inter-sectoral cooperative approach involving relevant government portfolios, police, as well as health, research and education bodies (Lawson 1991). Important to the effective implementation of these strategies is the need for behavioural and attitudinal changes amongst the public (Trinca 1987). This change and the necessary popular support were only reached through extensive efforts at generating awareness through school programs and mass media campaigns (Lawson & Bauman 2001).

However, despite the success of these measures, Australia’s traffic crash fatalities still rate quite highly among comparable developed countries (Fingerhut et al. 1998; La Vecchia et al. 1994). Although Australia's rate of 9.4 road traffic-related fatalities per 100,000 persons in 1998 is comparable to that of Canada (9.7) and Japan (8.5), it is considerably lower than the rates for the USA (15.3) and France (15.1). Australia's rate is, however, markedly higher than for the UK and Sweden, both of whom have recorded six road traffic-related fatalities per 100,000 persons in 1998 (ATSB 2001). Wide variations in reported injury fatality rates between industrialised countries suggest that further reductions from preventive measures can be achieved (Smith 2001). As noted by Wigglesworth (2001), reduction in road related injuries and fatalities in Australia while considerable, do not parallel the estimate of 95% reduction in the death rate from infectious and parasitic diseases resulting from medical research in Australia in the 20th Century.

27 Childhood pool drowning prevention

The importance of prevention of swimming pool drowning, particularly amongst babies and young children falling into pools is well recognised (Carey et al.1994; Cass et al. 1996; Pitt & Cass 2001; Williamson & Schmertmann 2002; Wintenmute & Wright 1991). The importance of pool fencing to restrict unsupervised pool access to children has been reiterated in the literature reviewed (Fisher & Balanda 1997; Ley 1991; Pitt & Balanda 1998; Thompson & Rivara 2000). A Brisbane study of presentations to a Children's hospital emergency department concluded that the risk of drowning or near-drowning involving access to an unfenced pool is 3.76 times higher than the risk associated with a fenced pool (Pitt & Balanda 1991).

In NSW the Swimming Pools Act established in 1992 has endeavoured to make pools safer for children. Warning notices must bear the words ‘Young children should be supervised when using the swimming pool’, together with detail of resuscitation techniques (Swimming Pools Act 1992). In addition, public health education campaigns regarding the dangers of diving into shallow water (eg. ‘look before you leap’ style education programs disseminated through the media, cinema/public transport advertising and school based programs) were effectively implemented during the 1980s (Yeo 1993).

Regulations and Australian standards for pool fencing and gates which bear in mind child safety have since been introduced (Australian Standards 1992). However, the introduction and subsequent repeal of retrospective fencing legislation in NSW due to public opposition suggests that public health practitioners cannot be complacent when safety legislation is introduced (Carey et al. 1994; Mitchell 2002).

28 Similarly, a recent study in Western Australia (Stevenson et al. 2003) highlighted inadequacies in legislation and enforcement procedures concerning barriers surrounding private swimming pools. The review of the coroner's data over a twelve-year period found that almost two thirds of childhood drowning in Western Australia occurred in private pools with inadequate fencing thus emphasising the continued need to monitor preventive strategies.

Inadequacies in legislation and enforcement were also found in a Victorian study that reviewed coroners’ records of childhood drowning and near drowning from 1992 to 1997 (Blum & Shield, 2000). The majority of children studied drowned in unfenced pools and spas. Where access was gained to fenced pools, the majority did so through defective or inadequate gates, or through gates that were propped open. There were no cases reported where a child gained unaided access to a pool fitted with a fully functional gate and fence that met the Australian standard.

A Northern Territory study recently identified the need to improve local prevention measures for childhood drowning. The study revealed that the incidence rate of childhood drowning in the Northern Territory was higher than in the rest of Australia and that unlike other states it has not been significantly decreased (Edmond et al. 2001).

Various states’ experiences with the prevention of childhood drowning through fencing legislation, reveal both success and potential for further improvements in public health approaches to prevention. Some improvements have not been sustained because growing complacency affects compliance with safety measures or because there are weaknesses in the prevention strategies. This stresses the importance of the continual cycle of evaluation and intervention implied in the public health approach.

29 Sport and recreational injury and fatality prevention

In the last decade, sporting injury and its prevention have received increasing attention as a public health issue, possibly due to a growing awareness of the financial and community burden of sports-related injuries (Condie, et al. 1993; Finch & McGrath 1997; Finch & Owen 2001; Lawson & Bauman 2001). The direct and indirect cost of sporting injuries in Australia was estimated in the early nineties at $1billion per year (Egger 1991). In a more recent study from Victoria, sports injuries have been found to account for 21% of all injury costs (Watson & Ozzane- Smith 1998). Extrapolated from this study is a current estimate of direct costs of $1.65 billion to the Australian community annually due to sport and recreational injuries (Orchard & Finch 2002). It has been reported that the NSWSIC alone has paid $5.8million to injured rugby league players and $1.66million to rugby union players since 1979 (Sydney Morning Herald [SMH] 28 December 1999).

Sporting organisations, academic departments and sports/recreation government portfolios have increasingly become involved with injury prevention and surveillance, promoting best practices while monitoring and developing needs and types of preventive strategies (Finch 1995; Finch et al. 1995; Finch & McGrath 1997; Ozanne-Smith et al. 1994; PHA 1999). This has resulted in less player hours lost due to injury and improved the public profiles of many sports (Lawson & Bauman 2001).

The infrastructure in Australia for sporting injury prevention demonstrates a considerable network of departments, bodies, organisations and associations, whose roles have been developed and integrated throughout the 1990s to implement sport safety initiatives (Australian Sports Commission [ASC] 1997; Finch & McGrath 1997). The following is a brief outline of some of the key components of this infrastructure.

30 The Sporting Injuries Insurance Scheme

In 1978, the Sporting Injuries Insurance Scheme was established in NSW to provide compensation for people seriously injured while participating in a sporting activity. An important catalyst for the Scheme was the increase in serious head and neck injuries associated with rugby union and league football. The Scheme is a not-for-profit arrangement that derives its funding from premiums paid by member organisations. It represents a wide variety of sports ranging from popular Australian sports such as cricket and football to more specialised sports such as judo and hang gliding. In 1991 the NSW Sporting Injuries Committee, which administers a sporting injuries insurance scheme that compensates permanently injured players, established an injury research and prevention program providing ongoing funding for projects devoted to the reduction of serious sporting injury (NSW Sporting Injuries Committee 1992).

Sports Medicine Australia

Sports Medicine Australia is a non-profit professional and community education organisation, made up of a variety of professional groups all interested in the many sides of sports medicine including: medicine, physiotherapy, physical education, nutrition, massage therapy, podiatry, psychology sports training, and sport and exercise science. The organisation provides best practice information and policies through programs such as Smartplay and SportSafe Australia to promote sport safety and injury surveillance (Sports Medicine Australia [SMA] 2002).

Smartplay

Smartplay is a sport safety and injury prevention program, operating in several Australian states, which aims to reduce the incidence and severity of sport and recreation injuries. The program uses the slogan 'Warm Up, Drink Up, Gear Up' which represents simple yet important injury prevention practices for all sports participants, coaches and administrators. Smartplay provides news, resources,

31 events and feature articles relevant to sports safety. Each of the Smartplay programs operating around the country has its own slight differences in targets and activities. The Smartplay Central website has been developed to allow each of the state based Smartplay programs to come together, yet still maintain their local characteristics (Smartplay 2002).

SportSafe Australia

SportSafe Australia (no longer active), a joint initiative of the Commonwealth Department of Health, Aged Care, and the Australian Sports Commission through its Active Australia Participation Division, was initiated following recommendations from the Australian Sports Injury Prevention Taskforce to establish a program to promote the identification and funding of quality sports injury prevention research (Finch & McGrath 1997). SportSafe was aimed to be the major advocate for sports safety at a national level but its efforts have not been sustained. This highlights the current need for government to ensure this vital role is filled working towards: providing and encouraging an evidence-based approach to sports safety; developing national approaches to sports safety; and undertaking and facilitating strategic research.

The Australian Sports Commission

The Australian Sports Commission (ASC) organises and funds sports injury research such as the establishment of the National Sports Research Centre, and SportSafe on behalf of the federal government. It is the foundation of an integrated national sporting system that encourages sport and physical activity for all Australians. Its Active Australia initiative is aimed at increasing the long-term involvement of Australians in sport and physical activity (Sport and Recreation Ministers' Council 1997).

32 Rugby football associations

Overseas the success of rugby union surveillance systems has already been noted (Garraway et al. 1991; Garraway et al. 1999; Garraway et al. 2000; Lee et al. 2001; Quarrie et al. 1996; Quarrie et al. 2001; Quarrie et al. 2002). Although a unified approach was lacking (Orchard & Finch 2002), various organised sporting associations attempted to establish their own systems for monitoring injuries and implementing safety programs in Australia. However there is now little evidence to suggest that these attempts have been sustained. In 1996, a serious injury register was set up by the Australian Rugby Union Ltd (M Robilliard. 1996, pers. comm., 22 November). In 2000, a broader injury surveillance system co-funded by the NSW Sporting Injuries Committee was piloted and conducted through the UNSW Sports Medicine Unit (refer to Chapter 1, Section four for more detail) (Orchard et al. 2002). The Australian Rugby League Association also attempted to implement their own injury surveillance systems (H Hazard 2000, pers. comm., 18 July).

Both the Australian national rugby codes set-up safety committees e.g. NRL Head and Neck Committee and the ARFU Safety Committee. These were aimed at overseeing the continual examination and revision of game rules in the interests of player safety making systematic use of epidemiological data provided by injury surveillance systems. Injury prevention strategies for rugby union and league football in Australia, aimed at SCI and brain injury have included numerous rule changes, safety awareness campaigns and improved player selection and training (Rotem et al. 1998; Taylor & Coolican 1987; Yeo & Walsh 1987 1993) (refer to Chapter 1, Section four for more detail).

However attempts by both codes appear to have failed to come to effective fruition as there is as yet no published reports in peer reviewed journals or evidence that such committees are active or making regular recommendations for consideration by the NRL/ARL.

33 University based units

Several universities in Australia have units that have extensive involvement in sporting injuries research and surveillance. These academic bodies have made numerous independent investigations and collaborations with relevant stakeholders in the interests of sporting injury prevention. These include;

Monash University Accident Research Centre – notable for research into the quality improvement of sporting injury surveillance (Finch 1995 1997; Finch et al. 1995; Watt & Finch 1996).

Monash University Sports Injury Prevention Research Unit – notable for research into safety attitudes of athletes (Eime & Finch 2002; Eime et al. 2002; Finch et al. 2002; Finch & Mitchell 2002).

Flinders University Research Centre for Injury Studies – notable for administering NISU and ASCIR (O'Connor & Cripps 1996, 1997, 1999).

Deakin University School of Human Movement – notable for research into golf and skiing injury prevention and assessing the safety practices of sporting clubs and centres (Finch & Hennessy 2000; Finch & Kelsall 1998; Sherman & Finch 2000).

Deakin University Sports Injury Prevention Research Unit – notable for research into golf, cricket, gymnastic and rugby union injury prevention as well as quality improvement of sporting injury surveillance (Daly et al. 2001; Finch et al. 2001; Finch, Elliott & McGrath 1999; Finch & Kenihan 2001; Finch & Owen 2001; Finch, Valuri & Ozanne-Smith 1999; Fradkin et al. 2001).

34 University of NSW Sports Medicine Unit – notable for sports medicine research and involvement in an ARU injury surveillance pilot program (Garlick 2001; Orchard et al.1997; Orchard et al. 1998; Orchard et al. 2002).

University of NSW School of Safety Science – notable for research into the effectiveness of helmets in cycling and rugby football (McIntosh et al. 1998; McIntosh & McCrory 2000, 2001).

The New South Wales Injury Risk Management Research Centre - notable for research into the nature of injury risk in community level sports and leisure (Boufos et al. 2006).

The Australian Sports Injury Prevention Taskforce

In the late nineties, a national sports safety framework was developed by a federal government partnership. The Australian Sports Injury Prevention Taskforce (ASIPT) (1995-1997) sought to bring together the various sectors involved in sports safety so that a unified approach to sports injury prevention could be developed. A national approach was developed by ASIPT and recommended that sports injury prevention activities were greatly facilitated by sport safety plans that identify, evaluate and manage the risks of injury. ASIPT also developed and implemented a national data collection system for sporting injuries (Finch & McGrath 1997).

35 The Australian Sports Injury Data Working Party

The Australian Sports Injury Data Working Party established guidelines for sports injury surveillance and released a working data dictionary utilising existing classification systems used in various settings (Australian Sports Injury Data Working Party [ASIDWP] 1997). However, there has been no national body funded to implement these guidelines leaving a lack of direction for an Australia-wide approach to sports injury surveillance (Orchard & Finch 2002). Furthermore, the federal government’s National injury prevention policies priorities for 2001–2003 plan does not list the prevention of sports injuries as a priority (Strategic Injury Prevention Partnership [SIPP] 2001).

The lack of federal government policy and funding priority for sporting injuries may reflect the lack of evidence regarding the burden and risk factors associated with sporting injuries compared with other causes of injury such as road trauma and falls (Orchard & Finch 2002). As discussed earlier in section one, this might be attributed to a lack of centralised administrative infrastructure associated with sport and recreational activities, such as in road related and occupational settings that have provided more fluent development of responsible peak bodies for injury prevention and control.

Orchard and Finch (2002) argue that the New Zealand approach to managing sports injuries, where a central body, the Accident Compensation Corporation, is responsible for monitoring sporting, road related and occupational injuries is superior to the fragmented Australian approach. The authors recommend that Australian government bodies concerned with health and sport need to establish a body with national responsibility for sports safety and injury surveillance.

36 SECTION TWO

DEVELOPING AND EVALUATING INJURY PREVENTION STRATEGIES IN SPORT AND RECREATIONAL ACTIVITIES

Having examined the usefulness of a public health approach to injury prevention in Australia, section two focuses on sporting injuries and explores relevant methodological and definitional issues inherent in the public health approach to injury research and prevention. The section then considers the infrastructure developed in Australia to facilitate sporting injury research and prevention activities.

Although the implementation of injury and disease prevention measures has been associated with improved health in Australia, evaluating the effectiveness of health promotion is often difficult (Garrard 1992). The epidemiological approach to sports injury research is still relatively new and not yet widely used in Australia. However, its application locally has already been demonstrated to be extremely useful in documenting the magnitude of injury problems in some sports, for identifying high- risk groups for injury and in helping to identify variables that may be modified to prevent injury (Finch et al. 1995; Quarrie et al. 2002; Rotem et al. 1998; Watt & Finch 1996). Appropriate injury prevention strategies develop from the establishment of injury trends and associated risk factors. Monitoring epidemiological data is needed also to evaluate strategies that have been implemented.

Van Mechelen (1997a) points out that sports injury prevention based on the public health approach is a continuous cycle. The problem of injury is identified, the aetiology and mechanism of injury is established leading to the introduction of preventive measures, which are then evaluated for effectiveness. This is then followed by the introduction of new or improved preventive measures and the complete cycle continues.

37 As reiterated extensively in literary sources (Carmody et al. 2005; Finch 1995, 1997; Finch et al. 1995; Haylen, 2004; Larkins 1995; Noakes & Jakoet 1995; van Mechelen 1997a; Wigglesworth 1987), the prevention of sporting injuries requires effective injury surveillance that provides the necessary information for developing and evaluating injury prevention strategies. Accurate surveillance data is also essential in order to plan health and associated services (Blumer & Quine 1996). Determination of the information needs of potential users and the feasibility of meeting these requirements should precede the establishment of new surveillance systems (Blumer & Quine 1996).

Study design in epidemiological studies of sports injuries

Lack of clarity and disagreement over injury definitions and study methodology must be resolved to ensure effective surveillance and prevention. For example, disagreement concerning definitions of concussion stands in the way of establishing agreement on the risk factors. This in turn makes prevention strategies difficult to implement (McCrory & Berkovic 1998). The comparability of studies is often adversely affected by lack of consistency and agreement on methodology and data definitions (de Loes 1997).

The success and wide scale applicability of any sports injury surveillance system is dependent on valid and reliable definitions of sports injury, injury severity and sports participation (Finch 1997). Van Mechelen et al. (1992) argue the importance of including uniform definitions of injury severity to allow comparability across studies. Severity is usually described, depending on the aims or research, through the nature of the injury, the duration and nature of treatment, sporting time lost, working time lost, permanent damage and monetary cost. Van Mechelen argues more severe injuries should be given higher priority for prevention, regardless of the incidence with which they occur (Van Mechelen 1997b).

38 Definition of injury in surveillance systems may dramatically affect results as seen in contrasting findings of studies focussing on serious injuries compared to minor injuries (Wiggleswoth 1987). The standard definition for minor injuries, usually missing a subsequent game, may underestimate certain concussion injuries that appear to quickly resolve.

In discussing the epidemiology of sports injuries, de Loes (1997) argues that by definition, epidemiological studies must include the population at risk, and not merely describe the number of injuries observed. The author outlines four levels of study design employed in epidemiological studies of sports injuries: 1) Clinical case series 2) Community based surveys 3) Studies including participation data but no exposure data 4) Studies with exposure data

The clinical case series has major limitations due to the low frequency of severe injury. This design cannot identify athletes at risk or risk factors for injury and thus it is not truly an epidemiological form of study. The community-based surveys relate injury data to a defined population and also include a non-active component. However, the many confounders in populations being compared limit potential for generalisation. This design may only reflect the patterns in that defined population and do not identify the risks across sports. The third level of study defines the population at risk as the active participant population. With this design there is greater potential for comparability and generalisation as it controls many aforementioned confounders but it is still limited in its ability to assess actual risk without exposure data. The highest level of study design utilises exposure data and estimates the risk per time unit, giving the most accurate reflection of actual risk, usually per 1000 or 10,000 hours of player exposure (de Loes 1997; Edgar 1995; Garraway et al. 1991).

39 In reality, studies based on exposure data are rare because of the time consuming, complex, and arduous nature of the data collection process (de Loes 1997). Depending on the purpose of research, alternative designs may suffice study objectives.

Constraints in data collection

Collecting appropriately detailed information in sporting injury surveillance systems is crucial for the effective development of injury prevention strategies. Personal experience, drawn from assisting with the implementation of a pilot injury surveillance system for the Australian Rugby Union, revealed the practical difficulties in collecting detailed injury data for large numbers of players. The assistance of many people associated with various football clubs, who were not formerly trained data collectors, was required. In addition, the absence of resources to provide payment for work or anything more than rudimentary training for recorders made it difficult to obtain consistent levels of data quality. Some data collectors were more vigilant than others in their follow up of players who missed games and in the collection of injury mechanism details. The difficulties for all data collectors in obtaining accurate and detailed injury mechanism data highlighted the need for sufficient resources to ensure that they are adequately trained and motivated.

In the mid 1990s detailed data on sports injury was not widely available in Australia requiring more detailed surveillance (Larkins 1995). Rugby associations can only give estimations of their player populations from the 1980's which also threatens the accuracy of calculating incidence trends. The National Health and Medical Research Council report on head and neck injuries in football noted the absence of reliable and comprehensive Australian data on sports injuries despite injury to the head and neck being widespread. It recommended a national registry of football injuries in Australia (NHMRC 1994).

40 An Australian feasibility study of applying improved data collection methodologies for sports injuries concluded that a lead agency was needed to guide sports injury surveillance activities at national and state levels. The study coincided with the establishment of the Australian Sports Injury Prevention Taskforce (Finch et al. 1995). Barriers to collecting sports injury data included the lack of clear guidance about how to undertake injury surveillance and what information should be collected. Many stakeholders identified a standardised data collection methodology, including an easy to use data collection form, as a key requirement.

In response, the Australian Sports Injury Prevention Taskforce established the Australian Sports Injury Data Working Party, which developed a data dictionary to support Australian sports injury data collection through treatment centres and sporting clubs (ASIDWP 1997). However, as discussed in the preceding section ‘Sport and recreational injury and fatality prevention’, there is still no national body funded to implement this framework resulting in inadequate guidance for an Australia-wide approach to sports injury surveillance (Orchard & Finch, 2002).

A classification system designed specifically for sporting injury called the Orchard Sports Injury Classification System has demonstrated success in the field and has recently been shown to be more effective than generic systems such as the International classification of diseases (Rae et al. 2005).

The thesis has attempted to establish the importance of injury prevention and the requirements of well supported research and coordinated action. A brief account is now provided of relevant clinical definitions and injury management issues for SCI and brain injury.

41 SECTION THREE

BACKGROUND TO SERIOUS INJURIES OCCURRING IN SPORT

Section three examines the aetiology and epidemiology of SCI, brain injuries and fatalities. The contribution of sports and recreational activities to the incidence of serious injury from all causes is considered along with the costs of these injuries.

Spinal Cord Injury

Clinical definition and management of spinal cord injury

The spinal column consists of a series of interconnected bones, the vertebrae, which enclose the spinal cord, an integral part of the central nervous system. The spinal cord, through its attached nerve roots, provides the means by which to breathe, move and sense. Between each vertebra are discs of cartilage, which act as shock absorbers and allow the spinal column a degree of flexibility. The spine is divided into various parts: the cervical spine (neck) which is made up of 7 vertebrae; the thoracic spine (chest) includes 12 vertebrae; the lumbar spine (back), 5 vertebrae; the fused vertebrae of the sacrum; and a small vertebra called the coccyx (Yeo 1998b).

Traumatic SCIs are often termed as hyperflexion or flexion injuries. These refer to the mechanism of injury caused by forces applied to the spine when the head is forced backwards or forward, respectively. The spinal vertebrae are vulnerable to dangerous compression, which may result in fracture when pressure is applied in the flexed or hyper flexed position because natural flexibility is reduced. Fracture may also occur from vertical impacts on the top of the head (Bauze & Ardran 1979; Scher 1981). However, the vertebrae are not always fractured – other injuries can include dislocation and subluxation which are terms used to describe injuries involving twisting of the spinal cord vertebrae. The interlocking vertebrae may be

42 completely twisted out of position (bilateral or unilateral dislocation) or may return back to position (subluxation). All of these injuries will cause varying degrees of damage to the spinal cord (Bauze & Ardran 1979).

Any injury to the spinal cord has severe ramifications for the ability to function normally. Damage to the cord may cause quadriplegia (alternatively known as tetraplegia), paraplegia, or chronic painful conditions, dependent on the location of the injury. Generally, lesions high in the cervical spine are fatal. Damage to the spinal cord further down to the level of the first thoracic vertebrae usually indicates quadriplegia in varying degrees. Lesions down to the lower thoracic vertebrae may give rise to paraplegia. A quadriplegic is affected in all four limbs. A paraplegic is only affected in the legs and trunk, while arms still have some function.

Even if the casualty is not affected to these degrees of severity, spinal injury causes chronic back pain and restricted spinal flexibility (Yeo 1998b). Permanent cases of SCI from traumatic causes are distinguished as an important national health priority area because of the devastating neurological impairments involved in these cases (AIHW & CDHFS 1997).

An injury can be complete or incomplete. In a complete injury, a patient has no movement or sensation below the level of the injury. An incomplete injury is one in which there may be some sensation and/or movement, in varying degrees that are often classified, spared below the level of injury (Waters et al. 1991). In addition to motor and sensory deficits, spinal cord patients also experience loss of control over bowel, bladder and sexual function. SCI is now commonly classified under the American Spinal Injury Association (ASIA) impairment scale which ranges from A to D (see Appendix A for detail).

43 Spinal shock is an injury where the spinal column is subject to a forceful blow causing concussion but with no obvious lesion occurring. The response of the nervous system can mimic a severed spine, with identical symptoms. Some time later, the casualty gradually resumes the use of the limbs (Yeo 1998b).

Patients are prone to both acute and chronic medical complications, mostly because of immobility (Yarkony et al. 1997). Respiratory failure and pneumonia are frequent, mainly in quadriplegics (Hartkopp et al. 1997). As spinal cord patients recover neurologically, they go from a flaccid state to a spastic state. Extreme spasticity can be painful and may exacerbate contractures and decubitus ulcers (Yeo 1998b). Urinary infection and severe constipation occur due to the patients’ loss of bowel and bladder control, and immobility (Bergman & Yarkony 1997).

Neurologic recovery after a SCI occurs over a period of 18 months. The greatest amount of recovery occurs within the first 3-6 months. However, the care of the SCI patient is usually a life long endeavour.

The mean life expectancy of persons with complete tetraplegia has been estimated to be reduced by around 25% (10% for those with complete paraplegia) (Yeo et al. 1998). Mortality among SCI patients from causes such as septicaemia, pneumonia, influenza, diseases of the urinary system and suicide are significantly higher than in the general Australian population (Soden et al. 2000).

Since the early 1990s, there have been advances in the acute management of SCI. However, in spite of on-going research, there is currently no cure for SCI (Bedbrook 1992; O'Connor 2001b; O'Connor 2005; Yarkony et al. 1997).

44 Risk factors for spinal cord injury

The most common cause of traumatic SCI in Australia is road related injuries often contributing around 50% of traumatic SCI annually (O’Connor 2002a). A considerable amount of spinal injuries also occur due to falls (often from heights or among the elderly) or from violence (often involving firearms). Sport and recreational activities, such as aquatic activities and rugby football, are usually the next biggest cause of SCI (Lawson & Bauman 2001; O’Connor 2000b, 2001a, 2002c; Selecki et al. 1986; Yeo 1993).

The existing pattern in recent decades has shifted away from the traditional predominance of occupational injuries (Lawson 1991). These have declined steadily this century largely due to major transformations in the nature and safety of work. Those injured tend to be younger than in the past while the predominance among males continues (Lawson 1991). A study of occupational SCI in Australia using ASCIR data found they accounted for about 12% of all traumatic cases of SCI from 1986 to 1997 (O'Connor 2001b). The incidence rate was four occupational SCI per million of population per annum, over the period from 1986 to 1997, based on estimates of the Australian labour force. This is substantially less than the estimates for the general Australian population which are often close to 14 SCI per million of population per annum (Cripps & O’Connor 1998; O’Connor 2001b).

However some occupational groups have considerably higher SCI incidence rates than other professions. The occupational incidence rate was highest amongst farmers (17.0 per million) of which 14% were caused by falling from a horse. The age group incidence rate was highest amongst those aged 25-34 years (4.9 per million) with the vast majority of all cases being male (95%). Close to 50% of the cases studied received their injury due to a fall and close to 20% were motor vehicle occupants involved in crashes. Despite the SCIs studied being associated with work activities, 36% of cases were found not to receive any compensation for their SCI representing a significant burden to the health and welfare system

45 (O'Connor 2001b). Similarly, Carmody et al. (2005) argue there is inadequate compensation available for Australian rugby footballers sustaining quadriplegia (generally $180,000-$300,000 total payment, compared to up to $4 million for similar injuries that occur on the road).

The dangers inherent in diving for spinal cord injuries are well documented in Australia and around the world. A considerable proportion of diving related SCI have been reported in such countries as Japan (Katoh et al. 1996; Noguchi 1994), Israel (Ohry & Rozin 1982), Germany (Raymond 1988; Steinbruck & Paeslack 1980), Portugal (Gaspar & Silva 1980), Romania (Soopramanien 1994), South Africa (Scher 1978 1995) and the USA (Bailes et al. 1990; Kluger et al. 1994).

Of all recreational activities, diving usually accounts for the highest frequency of SCI. In Australia, these diving related injuries are closely followed by rugby union and rugby league which have been responsible for a large proportion of sporting spinal injuries, almost exclusively among males (Rotem et al. 1998; Taylor & Coolican 1987). Horse riding activities also cause a considerable amount of SCI respective to the population of riders and actual time spent at risk (Cripps 2000; Edixhoven et al. 1981; Pounder 1984; Watt & Finch 1996).

Epidemiology of Spinal Cord Injuries in Australia

Evidence suggests an increasing rate of spinal injury due to motor vehicle and sporting injuries throughout the 1970s in Australia, raising concerns about these alarming injuries (Lawson 1991; Taylor & Coolican 1987; Ozanne-Smith et al. 1994; Yeo 1993). The following studies indicate the incidence of traumatic SCI peaked and levelled out over the 1980s and appeared to reduce in the early 1990s, perhaps due to concerted efforts aimed at prevention. However, this diminution appears not to have been sustained throughout the 1990s.

46 One study estimated that SCI due to motor vehicle, contact sport and water related injuries dropped by 20% between 1986-1992 in NSW (Yeo 1993). This study of traumatic SCI admissions to the Royal North Shore and Prince Henry Hospitals (central spinal units in NSW) reported 111 SCI in 1986 with an incidence of 19.2 injuries per million of population in NSW. This compared with a drop to 97 SCI in 1992 with an incidence of 15.6 per million of population. This approached an annual incidence of 100 patients per year in NSW.

In 1988, it was estimated for all of Australia there was a prevalence of 3480 paraplegics and 2520 quadriplegics, with an expected incidence of 400 spinal cord injuries per annum (Walsh 1988).

The Australian Spinal Cord Injury Surveillance System reported a total incidence of 188 SCI (157 being traumatic SCI) for the year 1995-1996 (Table 1) reflecting a decrease in the number of expected annual SCI from 1988. Around 46% of traumatic SCI involved road users. Sporting activities accounted for around 8% of the traumatic SCI with over half of these identified as occurring due to football. Around half of the SCI attributed to sport occurred during rugby scrums or tackles or during football matches. Other sporting activities accounting for the remaining cases included horse-related activities, judo, go-carting, hockey, soccer, and racing. Diving-related SCIs accounted for 5% of all SCI, occurring mainly in the surf and in swimming pools. Some diving-related SCI cases occurred in rivers and off piers (O’Connor & Cripps 1996).

Of the injuries in 1996-97 (Table 1), 38% of persisting SCI cases were motor vehicle occupants and unprotected road users. An alarming 11% were attributed to diving or other water-related activities, almost twice the number recorded in 1995- 96. For 15-24 year olds, activities of this type were only second to transport related injuries. Horse riding related SCI accounted for an increasing proportion, of around 3% of all persisting SCI cases (Cripps & O'Connor 1998).

47 The age adjusted incidence rate of persisting cases of SCI in 1997/98 was estimated to be 15.2 per million of population, an increase from 1996/97 (13.2 per million of population). Eight per cent (n=22) were attributed to water-related activities including diving. There was a substantial increase (76%) from 1996/97 to 1997/98 in transport related SCI in the age group 15-44 years (O'Connor & Cripps 1999). The age adjusted incidence rate of SCI in 1998/99 decreased slightly from 1997/98 and was estimated to be 14.5 per million of population (Table 1). Transport related injury accounted for 43% of the cases of SCI, 9% in an aquatic environment and around 5% engaged in a sporting activity.

No evident trend in the rate of SCI can be detected over the 1990s despite slight year-to-year variations. Improved retrieval, early management and rehabilitation of the victims of road crashes, may have decreased fatalities but increased the number of severely injured survivors (currently being assessed by the Research Centre for Injury Studies) (Cripps 2000).

The age adjusted incidence rate of SCI in 1999/2000 was estimated to be 14 per million of population, a small decrease from 1998/99 (14.5 per million of population) (Table 1). Transport related injury accounted for 50% of the cases of SCI, 10% engaged in a sporting activity and around 5% in an aquatic environment (O'Connor, 2001a). The age adjusted incidence rate of SCI in 2000/01 was estimated to be 13.6 per million of population, a small decrease from 1998/99 (14 per million of population) (Table 1). Transport related injury accounted for 56% of the cases of SCI, 8% engaged in a sporting activity and around 3% in an aquatic environment (O'Connor 2002c).

Because of advances in the management of spinal injury, there is now a growing prevalence of people living with spinal injuries accompanied by ageing in the population of prevalent cases (O’Connor & Cripps 1996).

48 Table 1: Traumatic SCI by cause 1995 to 2001 Year Traumatic SCI % Road % Sporting % Aquatic per million of users activities activities population 1995-1996 14.7 per million 44% 8% 5% 1996-1997 13.2 per million 38% Not specified 11% 1997-1998 15.2 per million 50% Not specified 8% 1998-1999 14.5 per million 43% 5% 9% 1999-2000 14 per million 50% 10% 5% 2000-2001 13.6 per million 56% 8% 3% Source: ASCIR annual reports - SCI in Australia

It was estimated in 1993 that the lifetime costs for paraplegia approached $1million, and $2 million for quadriplegia, costing the NSW community in the order of $150 million a year (Yeo 1993). Earlier cost estimates for all of Australia, based on an incidence of 25 new cases of SCI per million of the population per annum, place the annual cost of SCI exceeding $250m by the year 2006. This figure would come down to $188 million if an alternative projection model based on an incidence of just 12.5 new cases of SCI per million of population was used, which is a closer approximation to current incidence trends (Walsh 1988). These figures obviously do not take into account the immense non-economic human costs of quality of life and reduced life expectancy, which continue throughout the life of all persisting cases of SCI. Considering that the substantial estimated cost of the long term care of SCI, increase in SCI nationally represents the potential for a substantial increase in associated costs (O'Connor & Cripps 1999).

49 Head and Brain Injury

Clinical definition and management of head and brain injury

The Australian Sports Medicine Federation Draft Policy of 1992 defines head injury as encompassing skin lacerations, soft tissue or bone injury and brain injury (Mc Crory & Dicker 1992). Brain injury specifically refers to primary brain injury where the spectrum ranges from concussion (with no structural damage) through to catastrophic head injury (with disruption of neural tissue). Catastrophic brain injury includes conditions such as sub-dural haematoma, intracranial haemorrhage and diffuse axonal injury. These conditions are potentially life threatening and usually result in permanent neurological sequelae (Mc Crory & Dicker 1992).

Primary clinical features of brain injury include loss or alteration in consciousness, orientation and responsiveness, followed by a period of post traumatic amnesia. Other features depend on the nature, severity and complications of the injury. Injuries may be closed or open. Closed injuries usually result from blunt impacts or translation of dynamic forces to the head. Closed head injury is graded into mild, moderate and severe depending on duration of loss of consciousness, period of post traumatic amnesia, and initial Glasgow Coma Scale (based on motor, verbal, eye responses). About 80% of brain injuries in Australia are defined as mild (Geffen et al. 1998).

Concussion, while less severe, is quite common in contact sports, causing a wide variety of neurological symptoms. Confusion and amnesia are characteristic of concussion, but may take some time to evolve. Loss of consciousness does not necessarily need to have occurred. Concussion is a direct response to an injury that is a temporary condition not resulting in structural brain damage and consequently does not lead to permanent neurological sequelae (Mc Crory & Dicker 1992). The definition and severity of minor cerebral injuries is widely misunderstood which has contributed to a lack of reliable epidemiological information about their actual frequency in sporting activities (Clarke 1998).

50

Although concussion usually completely resolves, several authors have argued that there may be delayed consequences with long term sequelae. Although this is usually associated with repeat concussion before an earlier injury has resolved, it highlights the importance of not underestimating the sequelae of mild concussion (Cantu & Voy 1995; Geffen et al. 1998; Maddocks & Saling 1995; Saunders & Harbaugh 1984).

Repeated mild brain injuries occurring over an extended period can result in cumulative neurologic and cognitive deficits, but repeated mild brain injuries occurring within a short period can be catastrophic or fatal. The latter phenomenon has been termed the ‘second impact syndrome’ (Cantu & Voy 1995; Gronwall & Wrightson 1975; Saunders & Harbaugh 1984).

Successive episodes of concussion may have cumulative effects with a second, even lesser impact possibly having an unduly large effect (Cantu & Voy 1995). Repeated concussion, such as seen in professional boxers, may lead to the permanent structural damage of chronic traumatic encephalopathy (Geffen et al. 1998). An athlete with a history of concussion is up to four times more likely to receive a further injury than one with a clear history. Psychometric tests indicate greater initial impairment and extended recovery of mental functions in individuals with previous episodes of concussion (Gronwall & Wrightson 1975).

One study looking at the standardised assessment of concussion, found that concussed players as a group, scored significantly below the non concussed controls and below their own baseline (pre-injury) performance in standardised tests of orientation, immediate memory, concentration and delayed recall, despite all having been considered by the trainers to have suffered mild, grade 1 concussion (McCrea et al. 1997). In studies of sports-related traumatic brain injuries in the United States, there was an increased risk for subsequent traumatic brain injury among persons who have had at least one previous traumatic brain

51 injury (Annegers et al. 1980; Salcido & Costich 1992). Cumulative post concussive symptoms, such as poor co-ordination or balance, impaired concentration, judgement, and fatigue may predispose players to further injuries (Cantu & Voy 1995).

However, it has been argued that the evidence regarding the potential for cumulative damage from concussion relates only to individuals who sustain a second concussive injury before recovering from the initial injury. There is no evidence that several concussions over a football career will result in permanent damage contrary to the dissimilar experience of boxing related head injury (Mc Crory & Dicker 1992). It is argued that the term ‘second impact syndrome’ is therefore misleading and the phenomenon should be referred to as diffuse cerebral swelling (McCrory 2001a, 2002; McCrory & Berkovic 1998).

There are little evidence-based recommendations available to guide to the exclusion of players sustaining multiple episodes of concussion. Although there is evidence for the accumulative damage effect of multiple concussive injuries in boxing and motor vehicle crashes, this has not been effectively demonstrated in rugby football in which the mechanism and forces involved in injury vary considerably. Lack of evidence has lead to arbitrary three concussion episode exclusion policies. While erring on the side of safety, these mandatory exclusions are not evidence-based approaches and may unnecessarily detriment a player’s sporting career/participation (McCrory 2001a). In the absence of appropriate scientific evidence, the only alternative is individualised clinical and neuropsychological assessment of recovery. This kind of individualised assessment is becoming more accessible through web-based computerised neuropsychological test batteries.

A study of all published cases of definite, probable, and possible ‘second impact syndrome’ found over reporting of recalled episodes of concussion in team mates when compared with self reports and videotape analysis. Based on these case

52 reports, the study concluded that the claim that ‘second impact syndrome’ is a risk factor for diffuse cerebral swelling was not established (McCrory & Berkovic 1998).

As catastrophic head injury can initially mimic concussion, accurate diagnosis through detailed neurological assessment is necessary (Mc Crory & Dicker 1992). All head injuries require systematic assessment by the sports physician to recognise and manage the acute consequences in an athlete including the prevention of secondary brain damage, safe transport to an appropriate facility when necessary, close monitoring of the athlete for at least 24 hours and participation only when the athlete has fully recovered. Players should make a progressive return to activity so symptoms can be carefully monitored (Geffen et al. 1998).

Skull fractures can lead to brain compression, intracranial bleeding or infections, although these may occur without fracture. Brain injuries may cause loss of central nervous system control such as breathing. All patients with head injury are at an increased risk of late onset epilepsy. This usually follows traumatic injury with prolonged periods of unconsciousness and occurs in about 23% of hospital admissions of sport related head injury. Post traumatic epilepsy may develop within days or months. Concussive convulsions seen soon after injury are rare, occurring with an approximate incidence of 1 case per 70 concussions. They are transient and don’t necessarily lead to the development of epilepsy (Geffen et al. 1998). These episodes are often confused with post traumatic epilepsy (McCrory & Berkovic 1998).

Mortality in severe brain injury ranges from 30% to 70% (Brandstater et al. 1991). Survivors may not have a shorter lifespan but will have to deal with enduring cognitive, physical, behavioural, or emotional deficits, which will necessitate rehabilitation services. Rehabilitation from more serious traumatic brain injury includes cognitive and motor deficit therapy, behavioural management, as well as developing communication, social and other support (Cobble et al. 1991).

53

These injuries can often affect the personality of the injured, creating profound disturbances to social adjustment and quality of life (Brandstater et al. 1991). For example, a retrospective study of patients from a Melbourne rehabilitation hospital, observed that many patients with permanent head injuries were more likely than those in the community in general to have problems maintaining close attachment figures and social networks (Kinsella et al. 1989).

Risk factors for head injury

Most acute head injuries in Australia are due to road related injuries occurring to young males (Lawson 1991). As with SCI, this is followed by falls and recreational activities (Lawson 1991; O'Connor 2002c). Similar injury patterns have been observed in the USA (Brandstater et al. 1991). International studies suggest that sport-related head injury accounts for approximately one tenth of all head injuries presenting to hospitals and neurosurgical units (McCrory & Dicker 1992). In the USA an estimated 300,000 traumatic brain injuries of mild to moderate severity, most of which can be classified as concussion, are sports-related (Sosin et al. 1996). Trauma to the head region is a common occurrence in Australian sports with two-thirds of mild brain injury the result of sporting activities (Geffen et al. 1998). Although the risk for potential catastrophic head injury exists, most head and brain injuries are mild and do not result in permanent damage (McCrory & Dicker 1992).

Different contact sports vary greatly in the incidence of brain injuries sustained. The highest risk of head injury in recreational activities is in contact sports that create high velocity collisions, such as in football and combat sports like boxing where the head is a legitimate target. Horse riders are also at particular danger of head injury due to their height from ground, speeds travelled and the horse’s unpredictability. The NHMRC report on head and neck injury in football indicated that the rugby codes and Australian football posed a greater danger for concussion than soccer (NHMRC 1994). Comparison of football codes at the junior level in NSW shows that while soccer has the largest player population it has the lowest frequency of

54 head injuries (Northern Sydney Area Health Service [NSAHS] 1997). One study of touch football injuries found the injury rate was much less than in other football codes and that less than 3% of injuries affected the head or neck (Neumann et al. 1998).

In the United States, 81% of patients with traumatic brain injury (TBI) and 96% of patients with SCI reported pre injury alcohol use. The rate of pre injury heavy drinking for both groups was alarmingly high. Fifty-seven percent of persons with SCI and 42% of persons with TBI were heavy drinkers. Implications for risk identification, treatment, and future research are discussed (Kolakowsky-Hayner et al. 1999).

Epidemiology of Head and Brain Injuries in Australia

Head injuries are the single most common cause of death and major disability among Australians up to the age of 44 years. (Lawson 1991; Woodward 1984). Non-fatal permanent head injuries cause severe long-term financial and personal suffering to the injured, their family and carers.

Data for all hospital separations in Australia for injuries in the year 1999-2000 noted around 68,000 cases of injuries to the head and neck (17% of all injury related separations). Around 20,000 of these occurred in NSW alone. About a third were attributable to falls and a quarter to road related injuries. This was followed by assault, and exposure to mechanical forces. For all injury hospital separations in Australia, it is noted that sports activities account for around 4% and leisure activities 3% (AIHW 2001).

It is difficult to access detailed epidemiological data on head injuries in Australia, which is not the case for SCI. This is because head injuries are not officially notifiable conditions, have no national register and are difficult to define. Head injury rates in sports and recreational activities are not often reported in traditional sources of injury data. For example, accessible statistics compiled from the 1995

55 National Health Survey and Hospital Separations reports do not clearly delineate head injuries occurring in sport and recreational activities (ABS 1995; AIHW 2001). The limitations of ICD E-coding schemes for identifying sport injuries has been noted by others (Finch et al. 1995), and means that specific data on sports activities or mechanisms of injury cannot be obtained.

Although road fatalities have been halved in Australia since the mid 1980s (O’Connor 2002a), it has been argued that many deaths remain potentially preventable. The Consultative Committee on Road Traffic Fatalities in Victoria recently reported up to 34% of head related motor vehicle injuries between 1992- 1997 were potentially preventable. It identified improved delivery and quality of trauma care as an important means for reducing fatalities (Rosenfeld et al. 2000).

The identification in the USA that mild traumatic brain injury or concussion in sport constituted a chief health consideration led to improved recognition of its clinical syndromes as well as development of consensual injury management and return to play criteria (Bailes & Cantu 2001). An Australian study argued that Australia trailed the United States in developing a planned system of care for head injuries requiring a flexible coordinated total care plan from time of injury to long term community follow up (as has been established for spinal injury) (Burke 1987).

56 Fatalities

Risk factors for fatal injuries

Road related injuries account for the majority of acute injury fatalities in the general community. Associated with these are a great number of fatalities involving motorcyclists and bicyclists, mainly from head injuries. The pursuit of bicycling can be considered a sport as many riders on the road choose to do so for reasons other than purely transportation such as fitness and socialising. However, as there are many studies in the literature addressing bicycling injury (mostly looking at issues surrounding helmet usage) and interrelated issues of road related injury, they have not been considered in this study (Attewell et al. 2001; Dorsch et al. 1987; Finvers et al. 1996; Jacobson et al. 1998; McDermott 1992; McDermott et al. 1993; Povey et al. 1999; Thomas et al. 1994; Thompson et al. 1996).

Injuries appear to account for only a small proportion of deaths, and therefore sports and recreational activities contribution to fatality rates would be expectantly low (ABS 1999b).

Many severe SCI and head injuries result in death or significantly reduced life expectancy for those afflicted. However, cardiac failure, exacerbated by congenital weakness, over-exertion or acute injury is the most common cause of sporting fatalities. Although deaths are most common in high contact, intensive sport and recreational pursuits, they can occur in a vast array of activities.

From a study by Taylor and Coolican (1987) of SCI in Australian footballers from 1960-1985, it could be extrapolated that a case fatality rate of around 7.5% existed for all codes combined. Australian rules footballers, while not contributing the most injuries, displayed the highest case fatality rate (11.8%), followed by rugby union (7.9%) then rugby league (5.1%). A study of football related deaths using coroners’ records in Victoria 1968-1999, found 25 fatalities mostly related to Australian rules football and only three from rugby union (AFL is a vastly more popular code in

57 Victoria). Nine (36%) of the deaths were attributable to brain injury (McCrory et al. 2000).

In the USA, football head and cervical spine fatalities have been estimated to have been associated with around 85% of all football fatalities from 1945 to 1994. There has been a dramatic reduction in these types of fatalities during the last two decades, largely attributed to rule changes that prohibit initial contact with the head and face when blocking and tackling, better coaching in the techniques of blocking and tackling, helmet standards introduced at the college level and improved medical care (Mueller 1998).

Epidemiology of fatalities in Australia

The age standardised death rate (standardised to 1991 population) in Australia has declined from 860 per 100,000 population per year in 1979 to 760 per 100,000 population per year in 1989 to 590 per 100,000 population per year in 1999. In NSW, similar rates have been observed with the age standardised death rate declining from 870 per 100,000 population per year in 1979, to 780 per 100,000 population per year in 1989, to 590 per 100,000 population per year in 1999 (ABS 2002b).

Over the past century, the average life expectancy of a new-born in Australia has increased by over 20 years (ABS 1999b). The reduction in mortality in the early part of this century has been attributed to improvements in living conditions, such as better water supply, sewage systems, food quality and health education. The continuing reduction in mortality in the latter half of the century has been attributed to improving social conditions, and to advances in medical technology such as mass immunisation and antibiotics. The past two decades in particular have seen further increases in life expectancy. These increases are due in part to lower infant mortality, fewer deaths among young adults from motor vehicle accidents and fewer deaths among older men from heart disease. The reduction in the number of deaths from heart disease has been related to behavioural changes, such as

58 dietary improvements and reduced smoking (ABS 1999b). Life expectancy at birth of Australians now compares favourably with other developed countries and was only exceeded by Japan and France (ABS 1999b).

Figures for 1999 provided by the Australian Bureau of Statistics indicate malignant neoplasms and ischaemic heart diseases (IHD) were the leading causes of death, accounting for 27% and 22% respectively of total deaths registered (ABS 1999b). During the last decade, IHD and cancer remained the two leading causes of death. In recent years cancer has overtaken IHD as the leading cause of death for both men and women. This has been the result of the long-term downward trend in the standardised death rate for IHD, declining by 59% for males and 53% for females from 1981 to 2001. Over the same period the standardised death rate for malignant neoplasms declined by just 13% for males and 6% for females.

In western countries, injuries remain the leading cause of death in young adults (Jennett 1996). Injuries and poisoning are a significant source of preventable illness, disability and mortality in Australia, and place a heavy burden on health services (ABS 1999b). In 1999, 7% of all deaths were due to external causes with an incidence of 44.4 per 100,000 population per year.

Suicide and transport accidents presently account for more than half of all injury deaths in Australia. Suicide numbers and rates have risen in recent decades, especially for young and middle-aged men, bringing this topic substantial public and government attention since the mid-1990s. In contrast, the annual number of road deaths has dropped to about half the number in 1970, despite large increases in population and the amount of travel on roads (ABS 1999b).

59 Non-traumatic fatalities in sport and recreational activities

Fatal arrhythmia appears to be the most common mechanism of death in cases of sudden death in athletes (Futterman & Myerburg 1998), often precipitated by cardiac conditions and over exertion (van Camp et al. 1995). This is paradoxical in that athleticism has long been associated with cardiovascular health (Paffenbarger et al. 1984). Participants with congenital conditions or poor health may be at a higher risk of heart failure.

Australian Aborigines have high rates of cardiovascular mortality (Hoy et al. 1996; Thomson 1991). The high risk this group has for sport-related sudden cardiac death due to ischaemic heart disease has been identified, particularly in Australian Rules football (which tends to be the predominant football code played in areas with relatively large Aboriginal populations) (Young et al. 1999).

Appropriate detection of athletes at high risk of sudden cardiac death and their exclusion from vigorous physical activity may prevent sudden death (Franklin et al. 1997; Young et al. 1999). Pre participation screening of children (Bratton 1997) and adults in sport has been recommended to discover any disqualifying conditions such as risk of cardiovascular complications (Finch & McGrath 1997). Underlying coronary artery disease, the greatest risk for sudden cardiac death, is often only detected after sudden deaths (Meldahl et al. 1988).

This would suggest an aggressive evaluation of those with significant risk factors. However, it has been estimated that only 30% to 40% of patients who die suddenly can be identified as likely candidates before the event. Risk factors in these asymptomatic subjects include a familial history of coronary artery disease, high blood cholesterol levels, hypertension, smoking and an abnormal ECG at rest or during exercise. However, the predictive value of these abnormalities is too low to warrant more detailed clinical investigations in the absence of symptoms (Meinertz et al. 1991).

60 In 1994, the 26th Bethesda Conference addressed recommendations regarding the eligibility for competition in athletes with cardiovascular disorders. It concluded decisions should be based mainly on symptoms experienced as well as the intensity and type of sport played. Athletes with conditions associated with sudden cardiac death should not be permitted to participate in sports or only at a very low workload (Thompson et al. 1994).

In the USA, a range of relevant American health organisations have developed recommendations for cardiovascular screening of student athletes as part of a comprehensive sports pre participation physical evaluation (PPE). These recommendations can help physicians make informed decisions about the eligibility of an athlete to participate in a sport (Lyznicki et al. 2000). At the same time, the low yield of important findings during these examinations has prompted reassessment of when, where, and by whom they might be most efficiently conducted (Rowland 1986).

There are both quality and compliance issues for PPE due to the rarity of sudden cardiac death. The prognostic value of a positive test result is likely to be exceedingly low and a number of healthy individuals would be excluded from participation. In addition, many ‘negative’ responses to questions and normal physical examinations serve to dull the judgment of even the most methodical clinicians. Considerable resources are required to perform screening examinations, yet they are not covered by insurance reimbursement (Bader et al. 2004, Hulkower et al. 2005, O'Connor et al. 2005).

The previous literature review sections have provided the context for the remaining review chapters which will focus specifically on epidemiological evidence concerning serious rugby football injuries, the historical, sociological and playing style development of the rugby football codes since their inception, and the implications of this for injury prevention.

61 SECTION FOUR

SERIOUS RUGBY FOOTBALL INJURIES

Section four provides an overview of the rugby football codes' rules and the burden associated with SCI and brain injuries for both codes of rugby in Australia. An overview of existing injury prevention strategies in rugby football and their effectiveness are considered as are the legal consequences of serious injuries.

Explanation of rugby football codes (adapted from rule books) (Australian Rugby Football Union [ARFU] 1995; Australian Rugby League [ARL] 1995.)

Rugby League rules

Players – Each team has 13 players, with another 4 to 6 pre named players available for 12 interchanges throughout the game at the elite level. Backs – Full Back, Right and Left Wing, Right and Left Centre, Fly Half, Half Back Forwards – 2 x Prop, Hooker, 2 x Second Row, Lock.

Playing area – The playing area comprises of the ‘Field of Play’ and an ‘In-Goal’ area at either end of the ‘Field of Play’. The ‘Try’ line is the line separating the ‘Field of Play’ and the ‘In-Goal’. On each ‘Try’ line, halfway along, is the ‘Goalmouth’. The ‘Goalmouth’ is made up of two tall uprights joined by a crossbar part way up.

Scoring points – A team scores points by scoring tries, conversions, and field goals. Grounding the ball over the opponents’ Try line scores four points. The Try scoring team can score an additional two points by converting a – a kick made from a placed ground position through the field posts and above the crossbar at any distance, but at the equivalent position to where the Try was scored. A field goal is where the ball is dropped to ground while in play and kicked from the ground through the field posts and above the crossbar scoring one point. Players can run with the ball, pass the ball backwards (but not forward) or kick the ball.

62 Players can also try and stop opposing players with the ball (must be carrying ball) by tackling them or pushing them over the sideline.

Tackling – All thirteen opposing players on the field can stop their ball-carrying target getting over their goal line by tackling and stopping movement of the ball. A tackle is completed when the ball-carrying arm of the player has touched the ground and a tackler is touching them. The defenders must get off the person with the ball as they bring it back into play. Once a player is grounded or called held by the referee they then will stand up and begin to ‘play the ball’ by placing the ball in front of the foot and then rolling it backwards with one foot to the next receiver on the team. The defending team must go back ten metres from the restart of play, apart from two defenders who guard the ‘play the ball’ area with one standing behind the other. They can advance once the foot of the tackled player rolls the ball back. A team is allowed to play the ball six times in an attempt to score before they surrender possession back to the other team. Most teams prefer to kick to gain better field position rather than surrender possession.

Passing the ball – There is no limit to how many times the ball is passed from one player to another, but at no time can the ball be passed forward or fumbled forward with the ball touching the ground.

Kicking - The attacking team may kick the ball in general play and attempt to regain the ball, only if the players contesting the ball are behind the kicker when it is first kicked.

Losing possession – When a player fails to catch a ball cleanly and knocks it forward, or is carried over the sideline while , the end result is a scrum with the opposing team controlling the feed of the scrum into the middle. A team also loses possession when the ball is kicked and rolls over the sideline (or into touch).

63 Scrums – Scrums are formed when six forwards from each side interlock with each other and the opposite team. The referee will then instruct the ‘Scrum Half’ to release (or feed) the ball into the middle of the scrum where the two packs of forwards push against each other to control the direction and the outcome of the ball. In rugby league, unlike rugby union, the non-feeding team often concedes possession of the ball in a scrum without it being fiercely contested.

Offside – If a player is in front of a team-mate who kicks forward, he is offside. He is back onside if the receiver who gets the ball runs ten metres, or the kicker overtakes him.

Foul play – High tackles (contact with the head) are not allowed, and incur an immediate penalty of a kick for goal or six more plays. Stealing the ball from a player when there is more than one person in the tackle also incurs a penalty, as does impeding or blocking a man when he hasn't got the ball. Spear tackles in which the tackled player is lifted off the ground and dropped or driven on to the ground are not allowed.

Rugby union rule explanation

Positions – Each team has 15 players. For international matches a Union cannot nominate more than six replacements/substitutes except for under 21 teams where the maximum is seven.

Forwards – 2x Props, Hooker, 2x second row (locks), 2x Flankers, Eightman Backs – Scrum Half, Fly Half, 2x Wingers, Inside Centre, Fullback, Outside Centre.

Playing area – Same as for rugby league. See rugby league playing area above.

Scoring points – A team scores points by scoring tries, conversions, and field goals. Grounding the ball over the opponents' Try line scores five points. The Try scoring team can score an additional two points by converting a place kick where a kick is

64 made from a placed ground position through the field posts and above the crossbar at any distance but at the equivalent position to where the Try was scored. A field goal where the ball is dropped to ground while in play, and kicked from the ground through the field posts and above the crossbar scores three points. Players can run with the ball, pass the ball backwards or kick the ball. Players can also try and stop opposing players with the ball by tackling them or pushing them over the sideline.

Tackling – When a player with the ball is tackled he must immediately pass the ball or leave the ball on the ground and move away. The player making the tackle must also move away from the ball. After a player is tackled, any player, other than the two involved in the tackle, may play at or pick up the ball if they are the only players at the ball. The player must be on his feet when he plays at or picks up the ball. If two or more players are at the ball and the ball is on the ground, a ruck is formed. In a ruck a player can only play at the ball with his feet.

Maul and Ruck – These are loose scrum-like plays where players join and push together without creating a formal scrum. A ruck, which can only take place in the field-of-play, is formed when the ball is on the ground and one or more players from each team are on their feet and in physical contact, closing around the ball between them. If the ball in a ruck is on or over the goal line the ruck is ended. Rucking is the act of a player who is in a ruck using his feet to retrieve or retain the ball. A maul, which can only take place in the field-of-play, is formed by one or more players from each team on their feet and in physical contact closing round a player who is in possession of the ball. A maul ends when the ball is on the ground or the ball or a player carrying it emerges from the maul or when a scrummage is ordered. If the ball in a maul is on or over the goal line the maul is ended.

Lineout – When the ball or player goes over the Sideline a Line-out is formed. A Line-out has two or more players from each side lined up one metre apart, five metres in and at right angles from the point where the ball or player passes over the sideline. The ball is thrown in from the sideline down the middle of the Line-out.

65

Passing the ball – There is no limit as to how many times the ball is passed from one player to another but at no time can the ball be passed forward or fumbled forward with the ball touching the ground.

Kicking – The attacking team may kick in general play and attempt to regain the ball, only if the players contesting the ball are behind the kicker when it is first kicked.

Losing possession – The ball is said to have been 'Knocked-On' when a player propels the ball forward with his hands towards the opposition’s Try line. If, when the ball is ‘Knocked-On’, the opposite team gains immediate possession of the ball, the referee orders a scrum. A scrum is also called when one player unintentionally passes the ball forward to a player closer to the opposing Try line.

Scrum – A scrum is when eight players from each team interlock with each other and the opposite team, and the ball is rolled through the middle of the two front rows of players. Players can only play for the ball with their feet while in a scrum and no player can leave a scrum until the ball has left. Possession of the ball is fiercely contested and opposing packs attempt to drive each other back.

‘Scrum Popping’ – This occurs at the time of scrum engagement in the front row forward player in the ‘tight head’ position. The opposing front row forward in the ‘loose head’ position, in combination with the opposing hooker, lever the ‘tight head’ front row forward upwards and out of the scrum. The resultant forced flexion of the cervical spine followed by rapid extension of the cervical spine occurs as the head explodes from the scrum as a cork would ‘pop’ from a bottle thus the term ‘popping’. It is believed scrum popping can occur accidentally.

‘Scrum Wheeling’ – This refers to a turning action (hence wheeling) which changes the angle of a scrum. This is sometimes used as a technique to manoeuvre key players further away from important areas of play.

66 ‘Crotch binding’ – Holding around the upper leg and groin of fellow team mates in a scrum.

Offside – Generally the same as for rugby league.

Foul play – High tackles (contact with the head) are not allowed, and incur an immediate penalty. Spear tackles in which the tackled player is lifted off the ground and dropped or driven on to the ground are not allowed. Collapsing a scrum formation or rushing into a ruck and maul are not allowed.

67 SCI in rugby football

Rugby football is the most common cause of cervical spinal fractures in Australian sport. However, thoracic spinal fractures are very rare. The anatomical and physiological bracing of the thoracic spine explains the scarcity of fractures to this area of the spine (Geffenet al. 1997).

In the late 1970s and early 1980s there was an apparent increase in the incidence of rugby football related cervical SCI in Australia (Taylor & Coolican 1987; Yeo 1998a) and in other countries (Armour et al. 1997; Burry & Gowland 1981; Carvell et al. 1983; Hoskins 1978; Scher 1998; Silver 1979, 1984; Williams & McKibbin 1987). This prompted sporting organisations to make concerted attempts towards improving player safety through rule modifications, stricter enforcement of rules, better injury management, player selection and training (Burry & Gowland 1981; Silver 1992; Taylor & Coolican 1988; Wilson et al. 1996; Yeo 1998a).

Subsequent studies of rugby union and league related cervical spinal cord injuries in New South Wales (NSW), over the period 1984 to 1996, (Rotem et al. 1998; Wilson et al. 1996) revealed that while there was a small but significant decline in rugby union SCI cases, rugby league had not experienced a significant decline. Admissions to hospital spinal injury units in NSW associated with both rugby union and league injuries (all of which may not have resulted in permanent deficits) were not found to have significantly reduced from 1984 to 1996 (Rotem et al. 1998; Wilson et al. 1996). Disturbingly, eight cases of SCI involving rugby football were reported in Australia for 2000-2001 by ASCIR (O'Connor 2002c). A recently published study of football injuries based on the ASCIR monitoring of NSW spinal units, found only a small, but non-significant decline in the incidence rate of SCI in rugby union and rugby league from 1986 to 2003 (Berry et al. 2006).

68 When the relative number of players is accounted for in comparison to injury rates, most studies worldwide have found a greater relative risk of SCI for rugby union players (Armour et al. 1997; Rotem et al. 1998; Scher 1998; Silver 1984). In Australia, unlike many other rugby football playing countries, rugby league is the predominant code. However, both codes experience similar numbers of SCI, although there is less than a third as many rugby union players as league players.

Several studies have described the occurrence of ‘near miss’ SCI incidents in rugby football where an injury caused by hyperflexion or vertex impact damage to the spine in a tackle, ruck, maul or scrum resolves so that no permanent neurological deficit is experienced (Kew et al. 1991; Noakes & Jakoet 1995; Scher 1998). Other studies have found that these ‘near miss’ injuries can lead to degenerative changes such as spinal canal narrowing (stenosis) which potentially endangers further SCI in heavy contact situations (Odor et al. 1990; Scher 1991a; Torg et al. 1993).

69 Brain injuries in rugby football

The NHMRC report on head and neck injury in football drew attention to the danger of head and brain injuries occurring in both rugby codes (NHMRC 1994). Minor injuries to the head are common in both rugby codes at the elite and community level (Gabbett 2000; Gibbs 1994; MacDougal & Osbourne 1992; Seward et al. 1993). At the elite level the most common injuries are head and facial lacerations followed by concussion (8%) (Seward et al. 1993). Over 25% of the total injuries (40.6 per 1000) sustained during the three-year period were to the head and neck (Gabbett 2000). While traumatic brain injuries in the rugby football codes are rare (Gibbs 1994; McCrory 1998), a Victorian study of deaths due to brain injury among footballers from 1968 to 1999 found nine (36%) of the deaths were attributable to brain injury (McCrory et al. 2000).

Although rugby union players appear to be at a greater risk of SCI than rugby league players (due to a greater emphasis on scrummaging play), a greater relative risk for other serious injuries such as permanent brain injury in rugby union compared to league players has not been demonstrated in Australia. However, at the elite level, rugby union has accounted for more minor head injuries than rugby league (Seward et al. 1993).

Despite recognition of the dangers of brain injury associated with rugby football and considerable debate on definitions of concussions and appropriate management protocols, there is little published epidemiological evidence on brain injuries in rugby footballers in Australia.

70 Prevention strategies for serious rugby football injuries - safety programs, rule change and enforcement

In response to an increasing incidence of SCI injury since the 1970s, strong medical representation in several countries influenced administrative bodies such as the International Rugby Football Board to accept some of the suggested changes to the rules of the game. Rule changes aimed at preventing injury in football have traditionally focussed on modifications for junior play. However, the success of these has seen their extension to senior levels of the games.

Several education campaigns have disseminated information to coaches, players and school children such as Necksafe who promote awareness of the basic anatomy of neck, how damage can occur, appropriate neck exercises and management protocol. Guidelines have also been established for injury management and first aid (McCrory 1997; NHMRC 1994; Wilberger 1998). This has complemented the efforts of other prevention strategies. Some of the attempts by rugby union and rugby league organisations to modify and enforce rules in the interest of safety are detailed below.

Rugby union

ARFU Safety Committee, created in 1984 to investigate the laws of the game, developed law modifications primarily designed for junior level rugby union. These under 19-year-old law modifications were progressively introduced after 1985. These included such measures as increased emphasis on the management of scrum engagement (i.e. crouch-touch-pause-engage technique), disallowing crotch binding and ‘popping’ an opponent in a scrum, limiting pushing to 1.5 metres and by limiting wheeling to 45o. In addition, halfbacks were no longer allowed to follow the ball in the scrum (M Robilliard 1996, pers. comm., 22 November).

71 Players competing to the under 19 rules must also keep the upper trunk and shoulders above the level of the waist in the scrummaging position. This tends to avoid players collapsing to the ground. First or second row players can only be replaced in scrummage if there is a suitable replacement. There are also temporary suspensions to diffuse heated situations. Coaches are encouraged not to select players with inappropriate body types as forwards. De-powering of scrum engagement forms the crux of the under 19 law variations.

Key points for ruck/maul training and management include: head and shoulders must not be lower than hips; keep on your feet; head up and eyes open; no charging, jumping; collapsing ruck/maul; or dragging players out. Boots are not to be placed on opposing player (stomping, raking rucking, stepping not tolerated). Key points for tackling include: head behind the backside; drive with legs; ensure a firm shoulder contact and arm grip. Guidelines are provided for referees in the management and decision-making process such as dealing with indeterminate ruck/mauls.

A junior rugby pathway exists with levels of rule variation from under 7 years and upwards. Main features of variations include team size, playing area ball size playing time, scoring, tagging/tackling, mini scrums and lineouts.

The International Rugby Football Board adopted new rules in 1986 for senior level players in which collapse of the scrum, prolonged mauls or rucks, and spear tackles were banned in the interests of safety. Spear tackles in rugby football refer to tackles in which a player is lifted and driven or dropped headfirst onto the ground as opposed to the usage of the term in American football where it refers to tackling with the vertex of the head as the first point of impact.

72 There has been tougher enforcement of existing rules, particularly regarding foul play such as head high tackles (Lawson et al. 1995) and increased emphasis on player selection and physical preparation. Laws of the game require suitably trained and experienced players in the front row of scrums although requirements are not well defined (Quarrie et al. 2002).

The International Rugby Board acknowledged the success of the ARFU under 19 Law variations when it internationally introduced many of the critical changes at junior levels for senior players in 1990. There has been, as in junior rugby union, a greater emphasis on the management of scrum engagement incorporating the crouch-pause-engage technique and verbal instructions from the referee. Other variations to the rules include the widening of the gap between players in the lineout (1992) and the banning of tackling players when they are off the ground (1993).

The ARFU has instituted a number of programs to ensure the correct techniques and all participants follow rules. These include coaching materials such as video and brochures widely distributed to all rugby playing schools, junior clubs, referee associations and at coaching courses. These materials demonstrate the correct laws and techniques associated with scrums, rucks, mauls and tackles. Instruction in coaching accreditation courses and coaching manuals is partly devoted to safety techniques.

In 1996, a serious injury register was set up by the Australian Rugby Union Ltd (M Robilliard 1996, pers. comm., 22 November). In 2000, a broader injury surveillance system co-funded by the NSW Sporting Injuries Committee and conducted through the UNSW Sports Medicine Unit covering all injuries resulting in missed games was piloted among 17 teams (Orchard et al. 2002). In 2001, 32 teams were surveyed which represented a threefold increase in the number of player hours covered in 2000. There were four elite teams (Wallabies and the three Super 12 teams), eight city teams, nine country teams and eleven school teams. The pilot

73 surveys second year found an average of one injury every second game over the season for a team that kept the player out of the succeeding match or matches. The most common injuries were ankle sprains, knee medial ligament sprains, and hamstring strains (Orchard et al. 2002) (see Appendix B for definitions used in survey).

Rugby league

There have been continual rule changes in an attempt to reduce injury since the formation of the rugby league. In addition to these changes, mostly concerned with the replacement and interchange rules, there has been a greater emphasis on rule enforcement and on improving coaching and training techniques. Since 1982 there has been compulsory accreditation of all coaches and 1985 saw the beginning of stricter enforcement of existing rules particularly concerning foul play. The NSWRL judiciary has been increasing the severity of penalties administered for infringements, such as head high tackles, throughout the 1990s (Orchard & Seward 1994).

After experimentation with the four-tackle rule in 1967, the five-tackle rule was settled on in 1973. This was designed to stop teams overly dominating possession of the football, and served to increase the general speed of the game. The 10 metre rule introduced in 1993, requiring players to fall back a distance of 10 metres after a tackle has arguably resulted in more open play. However, this may be resulting in more tackling injuries as players build more speed before impacts or alternatively are caught running back in defence and are not fully prepared for impacts (SMH 16 November 1993, p. 48).

74 The National Rugby League purportedly implemented an Injury Surveillance System at all elite clubs. All cervical spine injuries and concussions must be reported via the system to the Chief Medical Officer. Clubs are also requested to submit videos of the incident (H Hazard 2000, pers. comm., 18 July). However this system has yet to provide published reports in peer reviewed journals or evidence that such information is being effectively utilised to make regular recommendations for consideration by the NRL.

At the junior level (under 12 years of age) there have been modifications to the game that include smaller footballs, fields and teams, as well as less emphasis on scrums and on competition. Over the study period, new rules were introduced intermittently, and existing rules were more strictly enforced. Safety programs were introduced banning players with long thin necks from taking dangerous positions, such as the front row of scrums. In 1996 there was a pilot program experimenting with depowering the scrum for junior rugby league (M Meredith 1996, pers. comm., 10 December).

The ‘Safeplay’ code for NSW junior rugby league players seeks to remove foul play and poor on-field behaviour by giving the benefit of the doubt to the non-offender and protecting the small and/or inexperienced player (e.g. a player on the ground is not to be dived on, a simple hand will suffice and ball carriers can surrender in tackle). It has proven successful in improving the safety and sportsmanship of junior players (Brentnall 1995).

Effectiveness of prevention strategies in reducing serious injury in rugby football

In rugby union, the current sequential scrum-engagement technique used at junior levels with increased control by referees has proven effective in reducing serious injuries in this phase of play. This is evidenced from an apparent decline in the number of schoolboys sustaining SCI in scrums during rugby union across several countries (Davidson 1987; Lee & Garraway 1996; Rotem et al. 1998; Roux, et al.

75 1987; Silver & Stewart 1994; Williams & McKibbin 1987; Taylor & Coolican 1987). The term ‘Schoolboy’ players will be defined as those players at primary and secondary school age usually playing in school organised competitions (i.e. not playing in open age competitions).

Changes to under 19 rules introduced in 1985 and altered laws for adults were mostly designed to prevent scrum collapse. It has been argued (Taylor & Coolican 1988), however, that these are inadequate, because scrum collapse is less significant than the danger of impact forces generated by the engagement of two packs. A more effective strategy would be to ‘depower’ initial impact forces by front rows packing separately and then being followed in succession by the second and back rows.

As discussed previously in section two, suitable surveillance strategies are needed to ensure the development and implementation of effective injury prevention initiatives and to allow for their regular evaluation (Blumer & Quine 1996; Finch 1995, 1997; Finch et al. 1995; Noakes & Jakoet 1995; Larkins 1995; Van Mechelen 1997a; Wigglesworth 1987). However, the contribution of changes in game laws for preventing serious injury in rugby football is difficult to assess for several reasons (Quarrie et al. 2002). These include a lack of standardised definitions of injuries; the absence of central collation of detailed data concerning serious injuries occurring in rugby league and union; lack of consistent risk factor information; the lack of participant and exposure data; a lack of measurement of the changing patterns of activity resulting from law changes; and the continuing evolution of the sport. As a result there is no appropriate incidence data to conclusively assess changes in injury rates over time in respect to law changes.

Several other factors confound the assessment of the success of law changes. The rare and independent nature of serious injury events means that variations over time are expected. It is also difficult to account for the progressive implementation of these law changes, which have occurred over many years. The effectiveness

76 with which the original intent of the law changes is implemented is also unclear. Many of these laws take some time to translate into changes in player and referee behaviour on the field and also may become less vigilantly adhered to over time. It is also difficult to account for the effect of other prevention strategies, regulations, and policies apart from law changes that may have impacted the incidence of injury. Similarly, other factors such as changing style of play and player characteristics that impact injury risk are difficult to account for.

However, as Quarrie, Cantu and Chalmers (2002) argue it is certain that: "Any law change is likely to result in changes to the typical structure and patterns of activity that are representative of the sport. Part of the reason for this is that after a law is modified, players and coaches generally attempt to use it to gain a competitive advantage over their opposition." (p.643)

Legal consequences of rugby football injuries

The application of medico/legal principles to foul play in rugby union, where the rules of the game are interpreted in general legal terms, has seen increasing numbers of civil law suits brought against players accused of deliberate foul play (Grayson 1996). The institution of the NSW Sporting Injuries Insurance Scheme was largely a response to a civil court action in 1977 (Taylor & Coolican 1987).

There are several cases noted in the press in Australia where on-field aggressive behaviour has resulted in prosecutions for assault. For example, one league player from a suburban club was jailed for a football tackle in which he apparently intentionally elbowed the ball carrier in the head. He was charged with maliciously inflicting grievous bodily harm receiving nine months periodic detention (Gilmore 1998). This is possibly a reflection of a growing trend in societal intolerance to foul play violence in sports (Dunning & Sheard 1979) as part of a gradual ‘civilising’ process and through exposure to increasing media coverage.

77 The NSW Court of Appeal has decided that sports administrators can be sued for making rules that cause injury to players. Two Australian quadriplegics injured in scrums in 1986-87 upheld that administrators from the international rugby board, the determiners of game rules, were negligent for “failing to enforce or modify rules so as scrummaging could take place safely,” (Phelan 1998; SMH, 20 April 1998, p. 5).

The Watson vs Haines case of 1987, stemming from a schoolboy rugby league SCI occurring in 1984, was successful in suing the NSW Department of Education for allowing a player with a long thin neck in a scrum (Australian Torts Reports 1987). In another case, a player made an out of court settlement with his club and team coach due to injury in a ‘flying wedge’ (which is now an outlawed manoeuvre).

Silver (2002) argues that because the school has duty of care, a student can take civil action against the school if injury was caused by negligence. This may occur through inadequate coaching/training, mismatch of players, lack of physical preparation, unsuitable playing conditions, unsuitable injury management, inadequate refereeing and over-psyching up.

These examples illustrate the immense financial, political and popular pressure sports administrators are under to modify, evaluate and enforce rule changes in the interests of player safety.

78 SECTION FIVE

RISK FACTORS FOR SERIOUS INJURY IN RUGBY FOOTBALL

This section considers the available evidence regarding risk factors associated with serious rugby football injuries and fatalities and offers a unified framework for understanding these risk factors. The potential risk factors for serious injuries in rugby football are commonly divided into intrinsic (personal characteristics) and extrinsic (environmental characteristics) factors as presented in Table 2.

OVERVIEW

Table 2: Potential risk factors for serious injury in rugby football Extrinsic risk factors Intrinsic risk factors Code played Risk exposure (time spent at risk) Laws of the game Grade Law enforcement Age/maturity Referees control of game Experience Coaches instructions and training Position Phase of play Skill Illegal play Anthropometric characteristics Speed of play Physiological characteristics Force of impact in player contact Psychological characteristics Direction and level of impact Information processing ability Number of tacklers Visual acuity Number of players Impairment – fatigue/alcohol/drugs Mismatch of players Endogenous medical conditions Location on field Congenital abnormalities Time in game Previous injury Stage of season Exertion Importance of game/ intensity of match Gender Environmental conditions Ethnic origin Pitch condition Acute injury management Protective equipment Adapted from figure in Quarrie, Cantu and Chalmers, (2002).

79 The Haddon Matrix has been utilised as a research tool by injury epidemiologists in a variety of injury prevention settings (Chorba 1991; Conroy & Fowler 2000; Runyan 1998; Short 1999). The matrix for injury events consists of two axis: three rows representing time stages (before the event, during the event, and after the event) and four columns representing the characteristics of the person, the environment, energy agent and the vector resulting in the abnormal energy exchange. It provides a framework for conceptualising risk factors that allows a consideration of temporal issues for prevention and likens injury to epidemiological concepts of disease. This framework provides a systematic point of application for research and prevention strategies (Lett et al. 2002). Injury prevention strategies suggested by this model are considered pre event (reducing the probability of injury), during event (reducing the severity of the injury) and post event (reducing the consequences of the injury).

Lett, Kobusingye, and Sethi (2002) argue that injury specialists have failed to successfully persuade policy makers and the community that injuries are preventable partly due to the lack of a unified understanding of injury control. The authors suggest that the two most important models utilised in injury control, Haddon's Matrix and the Public Health Approach, should be combined to provide a unified framework for understanding injury prevention. The complementary contribution of these two models offers a more coherent, and comprehensive explanation of injury prevention. The Public Health Approach lacks a methodical point of application, which is provided by the Haddon Matrix. Conversely, the Haddon Matrix lacks a systematic action plan. The integration of these models offers the Public Health Approach systematic strategy for the theoretical framework of Haddon's Matrix. There are no published accounts of the Haddon Matrix being utilised to assess risk factors for serious injuries and fatalities in rugby football. To provide a unified framework for understanding serious injury and fatality prevention issues in rugby football, a theoretical framework of risk factors using the Haddon's Matrix has been developed (Table 3).

80 Table 3: Haddon Matrix for risk factors in rugby football serious injuries and fatalities

Characteristics of the person Energy agent Vector Environment Pre injury event Gender Endogenous processes Mismatch of players Laws of the game Ethnic origin Exertion through Intensity of match Importance of game Age/maturity physiological demands of Grade Competition level Experience game and/or heat stress Protective devices Stage of season Skill Trauma to body through Pitch condition Time in game Position impact with another Environmental conditions Coaches instructions and Time at risk person(s), ground, ball or Force of impact in player training Anthropometric characteristics goalposts occurring in; contact Phase of play Physiological characteristics Tackles Direction and level of impact Number of players Psychological characteristics Scrums Number of tacklers Law enforcement Congenital abnormalities Rucks Illegal play Referees control of game Previous injury Mauls Pitch condition Information processing ability Collisions Environmental conditions Fatigue/alcohol/drugs Falling/slipping/tripping Visual acuity During injury Fatigue/alcohol/drugs Attendance by trained Protective devices Injury management facilities event Experience personnel Force of impact in player e.g. first aid kits, stretchers, Skill Injury management contact immobilisation devices, Anthropometric characteristics procedures Direction and level of impact ambulances, proximity to Congenital abnormalities Number of tacklers hospitals Previous injury Law enforcement Referees control of game Pitch condition Environmental conditions Post injury event Age Injury treatment and Acute injury management Anthropometric characteristics rehabilitation procedures Long term injury management Physiological characteristics Congenital abnormalities ___ Severity of injury

81 EXTRINSIC RISK FACTORS

Types and styles of play that increase the risk of serious injury

As might be expected in contact sport, the highest risk plays are those involving forceful impacts. Common causal patterns of serious injury have been observed across rugby football playing countries around the world. The following section reviews available evidence from other studies.

Spinal cord injury

Rugby football is the most common cause of cervical spinal fractures in sport, whereas thoracic spinal fractures are very rare due to the anatomical and physiological bracing of the thoracic spine (Geffen et al. 1997). Rugby union players appear to be most at risk of SCI in a scrum, maul or ruck whereas rugby league players are more likely to receive a SCI (or other serious injury) in a tackle (Berry et al. 2006). This reflects the different emphasis and resulting forces exerted in different aspects of each of the two rugby codes. A major study of Australian footballers found 69% of SCI in rugby league occurred in tackles mainly to the ball carrier compared to only 22% in rugby union tackles (Taylor & Coolican 1987). A more recent study covering NSW 1986 to 2003 found the most common causes of injury were tackles for rugby league (78%) (Berry et al. 2006). Injury to ball carriers in tackle related SCI is more common than to the tackler (Silver 1984; Williams & McKibbin 1987) although not always significantly so (Armour et al. 1997; Wilson et al. 1999). A study of American football SCI found most cervical injuries occurred to defensive players while tackling (Cantu & Mueller 2000).

Although tackles pose the greatest risk in league and American football it has been found that scrums alone account for 45% of rugby union related SCI in NZ (Armour, Clatworthy, Bean, Wells & Clarke 1997), 38% in Britain (Silver 1984) and 44% in Wales (Williams & McKibbin 1987). Australian studies have found as many as 35% to 64% of rugby union SCI in NSW being caused by scrums compared to around

82 only 20% in rugby league (Berry et al. 2006; Rotem et al. 1998; Wilson et al. 1996). Nonetheless, the risk of direct vertex impact injuries to the cervical spine during tackling in rugby union has been identified (Scher 1981). Tackles, rucks and mauls may account for increasing proportions of serious injuries in rugby union across the world as playing styles change and as moves are made to improve scrum safety (Edgar 1995; Jakoet & Noakes 1998; Scher 1991c; Silver & Stewart 1994).

Head injury

Although serious head injuries can occur in scrums through head clashes of front row players, this is relatively rare and injury is more likely to occur in a tackling situation from the player’s head hitting a part of another player’s body or from impact with the ground (Milburn 1993; Mueller 1998; Wilson 1998). In elite rugby league, while over 50% of all injuries are to the lower limbs, head injuries such as concussion, laceration and fractures are common. However, serious head injury is rare (Gibbs 1994).

Mechanism of tackling injury

The classic tackle is performed when the tackler hits the ball carrier at thigh height with his shoulder, at the same time grasping the ball carrier’s legs. Grasping the upper body and arms is allowed but stiff-arm tackling, i.e. striking the ball carrier above the shoulders with an outstretched or bent arm, is illegal. Head high tackles are also penalised in both codes with systematic review of suspected cases (ARFU 1995; ARL 1995; Dunning & Sheard 1979). The NHMRC (1994) and others (Silver 1984; Taylor & Coolican 1987) have clearly identified the danger of high tackles, spear and stiff tackles for serious head and neck injuries. Mistimed tackles have been identified as a major cause of less serious head injuries, indicating a need for greater emphasis on tackling skills early in the season (MacDougal & Osbourne 1992).

83 In rugby league, after a tackle is completed, play is less continuous than in rugby union as play is effectively restarted allowing players time to get back into position. Since 1993, league players have been required to fall back at least 10 metres from the tackled player as the ball is played after a completed tackle. In rugby union, tackled players taken to ground are not able to retain possession of the ball and must release it so it can be put back into play. Often a ruck or maul will form around the tackled player as opposing forwards try to gain possession of the ball. Ideally the play flows from phase to phase as a player is tackled, a ruck or maul forms, the ball is cleared out and put back into play.

The increased risk of injury observed in tackles may intuitively be attributed to the higher speeds and thus force of impact involved in tackling situations (Garraway et al. 1991). For both codes, serious injury to the tackler is usually a result of a mistimed tackle where either the head is swept up by the thigh or struck by the body/limbs of the targeted player or by players falling on them after initial impact. Tackling injury can be due to inadequate technique in which the tacklers head is incorrectly placed or a ‘crash’ tackle is attempted resulting in considerable forces to the head during impact.

When carrying the ball, injury often occurs because of an illegal tackle (around head/neck) or from the head striking the ground. Injury can also occur when the ball carrier runs with their head down into impacts with tacklers. Tackling injuries in rugby league occur with a similar mechanism to rugby union apart from the fact that completed tackles, in which the ball carrier is held, are not followed by rucks or mauls as play is restarted.

In American football the axial loading mechanism of SCI was identified in 27% of tackling injuries, and it has therefore been recommended that players refrain from tackling with the head down (using the head as a battering ram) and alternatively use the shoulder for blocking and tackling (Cantu & Mueller 2000).

84 Comparisons between all football codes in a survey of Australian high school students (NSAHS 1997) revealed that soccer, with the highest participation rate, reported the lowest proportion of injuries to the head. Similarly at higher levels of competition, the rugby codes have the highest rates of concussion (McCrory & Seward 1992). The rugby tackle acting as a common cause of head injury probably explains this comparatively lower risk of head injury for soccer. Similar lower concussion rates have been observed for soccer compared to the rugby football codes at the elite level. As head injury is less frequent in soccer, this suggests other codes may have to change rules to make their games safer (NHMRC 1994).

Force of impact

The mass and velocity of colliding objects determine their momentum. Changes in momentum are due not only to the magnitude of external forces but also the lengths of time that each of these forces act. The product of this force and time is impulse which is effectively the change in motion of two impacting objects. In a collision with a stationary opponent, the momentum of the tackler will not be dissipated. Rather most of the momentum will be transferred to the opponent while the tackler retains some. In this way the momentum between both players will be conserved and the total momentum before the impact will equal the total momentum after the impact (Hall 1995).

A much more realistic collision however is one in which the tackler meets his opponent head on and both players have a value of momentum. In this collision there are two possible situations. The first being that both players have the same product of momentum i.e. mass x velocity. An example of this is the tackler having a mass of 80kg and a running speed of 5m/sec. The opponent however has a mass of 100kg and a running speed of 4m/sec. In this situation both momentums are the same (momentum = 400kg m/sec) therefore although differently constructed, the impact and stopping effect on both players will be the same.

85 After the collision both players will rebound off each other with the same momentum. However, the lighter player because of a smaller mass will be seen to have a greater or faster rebound effect. When the two bodies undergo a direct collision, the difference in their velocities immediately after impact is proportional to the difference in their velocities immediately before impact (Hall 1995). As their respective momentums have been conserved in the collision, both players will experience the same impact forces. The question arises as to the protective factor for serious injury that increased player size might have.

In another example, one player has a greater momentum than his opponent i.e. the opponent has a mass of 80kg and a running speed of 4m/sec (momentum = 360kg m/sec). The tackler however weighs 100kg and has a running speed of 5m/sec (momentum = 500kg m/sec). In this situation, the tackler has a greater momentum and so will have the advantage in the collision. The result would be the tackler stopping and forcing back the opponent.

The greater the force exerted in the tackle the more likely injuries are to occur (Hall 1995). Playing styles in which the velocity on impact of players is maximised thus carry an increased risk of injury. The desire to maximise players’ impact velocity, to more effectively tackle or resist being tackled, is traditionally emphasised in both codes training techniques (Faccioni 1994; Wakelam 1937).

One study described the nature and circumstances of injury occurring in rugby union using data from the New Zealand Rugby Injury and Performance Project, and supplementary information on the nature of tackles involving injury from analysis of videotape of tackle injury events. It found that both players were most often in motion in the tackle at the time of injury with approximately 70% of injuries occurring when the injured player was running or diving/falling to the ground (Wilson et al. 1999).

86 In a prospective case-control study, the tackling and tackled players involved in a tackle injury were each matched with ‘control’ players who held the same respective playing positions in the opposing teams. Either the tackling or tackled player was sprinting or running in all of these injury episodes. One third of injuries occurred in differential speed tackles, where one player was travelling much faster than the other at impact. The player with the lower momentum was injured in 80% of these cases. The study noted that comparative information on the circumstances of the vast majority of tackles in which no injury occurs is required before any changes are considered to reduce injuries in the tackle (Garraway et al. 1999).

Variations in rules between codes mean that league players are more likely to have the opportunity to build momentum for tackle situations as the restart of play allows them to regain their respective attacking/defensive positions. Forwards in rugby union are often tied up in scrums, ruck or mauls while the backs continue play (however modern styles of play increasingly see forwards taking on similar roles to backs and spending less time indisposed in rucks and mauls). The increased momentum usually involved in rugby league tackles, compared to rugby union, is reflected in the predominance of tackling injuries in rugby league players. However, as discussed later, rucks and mauls pose their own dangers of serious injury risk for union players.

Thus a paradoxical situation is set up in which optimising personal momentum in impacts (‘playing hard’) provides some personal protection while increasing the overall injury risk for all players involved in the impact. Injury rates for touch football in which forceful tackles, scrums, rucks and mauls are removed have been found to be much lower than for other codes of football. Much fewer injuries are a result of contact between players and very few are to the head and neck (Neumann et al. 1998).

87 Consistent with these principles, a study by Norton, Craig and Olds (1999) on the development of AFL football concluded; "The increased player size, coupled with increased games speed, inevitably lead to increased momentum and collision forces. Basic allometric considerations point to a likely increase in injury rates given the evolution of the game. The energy dissipated in contact between two or more players is proportional to the combined kinetic energy of the players. Some of this energy is lost as sound and heat, but some is absorbed by deformed structures: flesh, ligaments, tendons, and bones. In the process, those structures may be damaged." (p. 403)

Silver (2001) draws an analogy with rugby football injuries and vehicle crash injuries, where forces involved and speed of deceleration are major factors in determining the severity of injury. The recognition of the fundamental importance of the dynamics of impact in vehicle crash injuries has led to successful advancements in motor vehicle safety technology. These have seen the application of physics and engineering principles to contain the forces experienced by occupants in collisions (Olney & Marsden 1986; Welcher & Szabo 2001). However, the same opportunities for restraining persons and creating impact attenuation do not exist for rugby football so other strategies are required.

Direction of impact

Calculating the energy/force exerted in an impact also requires the determination of the relative velocity of two impacting objects. The respective velocities of the objects, as well as the relative direction in which they are moving determine this. The examples considering force of impact given above are of head on impacts, but this is not always the case.

If two objects are moving forward on the same line in space, one behind the other, and are moving at different speeds (the first slower than the second), they will eventually impact. If the first object is travelling at 5km/h and the second at 10km/h, the relative velocity of impact is in fact 5km/h. If the objects were travelling at the same speed the relative velocity would be zero and in effect there would be no impact (i.e. no energy/force produced).

88 Players who impact head on, therefore produce considerably more energy/force than the equivalent players (with the same mass and velocity) who impact through moving in the same direction, because their relative velocity is drastically reduced. When two players collide in the same direction and plane of motion, the change in velocity is usually positive and thus the tackle is effective usually by making the opponent unstable and fall over.

However, when a head on collision occurs, both players’ velocities are drastically reduced in a short amount of time creating a large impulsive force. The speed of change in velocity normally determines the risk of injury. Variations in league rules from union in which play is restarted after tackles appear to increase the likelihood of players colliding head on in tackles because players are more likely to regain field positions for offensive or defensive runs.

In New Zealand, analysis of videotape of tackle injury events demonstrates the majority of tackle injuries were associated with stopping tackles that were from the front (63%), rather than from the side or behind. Thus consideration should be given to coaching strategies or to rule changes which reduce the likelihood or prohibit front-on tackles (Wilson et al. 1999). A Scottish study found forceful tackles resulting in injury occurred most frequently head on or within the tackled player's side vision. However, the study also identified the danger of high-speed tackles coming in behind the tackled player’s line of vision (Garraway et al. 1999).

89 Multiple tacklers

The increased risk of injury to the ball carrier in multiple tackler situations has been noted in the literature (Carmody et al. 2005; Scher 1983c; Silver 1984, 1994) drawing particular concern to the increasing prevalence of this technique in modern play. The potential for increased serious injury risk from multiple tacklers could be intuitively explained by the multiplied forces involved and opposing directions of impact creating sheering forces (by virtue of the fact that multiple tacklers cannot impact at the same point on the ball carrier). In addition, the ball carrier has a restricted ability to prepare their body for multiple impacts with players and with the ground. The tackled player is often unable to protect with hands, put arms out to break falls or brace properly. Body movement and muscle splinting cannot dissipate forces and the head may strike the ground. Figure 1 depicts a multiple tackle in rugby league.

Figure 7 depicts a multiple tackle in rugby league in which two opposing players tackle the ball carrier around the shoulders and waist. This restricts body movement, the ability to pass the ball and ultimately the ability to use the arms to attenuate impending impact with the ground i.e. the ball carrier loses control and cannot anticipate or protect himself from impacts.

Figure 1 - Multiple tacklers in rugby league

90 Level of impact

High tackles risk of injury to the ball carrier

A well-documented and recognised cause of injury to the ball carrier is the above the shoulders (Edgar 1995; Scher 1978, 1983a, 1991b). As discussed, this form of tackling is not allowed in both football codes with heavy penalties and video judicial panels investigate suspected incidents. The risk of injury to a player who is struck in the neck or head with anything at speed (let alone the force of an entire opponent) is obvious. The inherent danger of high tackles and the importance of their prevention is evidenced by both codes commitment to the policing of high tackle infringements (Edgar 1995; NHMRC 1994).

Figures 2-4 portray head high tackles in rugby league in which the tackler contacts the ball carrier above the shoulders from behind and attempts to pull him back to stop him. All these pictures show tackles that were penalised by referees for being dangerously high. Head high tackles may also occur head on, usually when a tackler leads with a raised elbow, forearm or shoulder which then contacts with the ball carriers head. Sometimes this may occur unintentionally if the ball carrier ducks down at the last moment before impact.

Figure 2: Head high tackle -league Figure 3: Head high tackle – league

91 Figure 4: Head high tackle – league Figure 5: Tackler injured through head hitting hip of opponent - league

Low tackles increased risk of injury to the tackler

Tackling low on the ball carrier has been traditionally taught to players aiming to impact at waist level then slide down in order to hold together the ball carrier’s legs causing them to fall (allowing a much smaller man to tackle a larger one). However, this technique requires careful skills training.

Players should keep their heads up so that they have sufficient awareness to prepare for impact. Tacklers who tackle too low on a ball carrier often increase their own risk for injury. Often the knee, leg or body of the ball carrier is driven into the tackler’s head. If the tackler’s head is down, taking the brunt of the impact, there could be an even greater chance that the force will be transferred to the spine in such a way as to damage it (direct vertex impact). Similarly, ball carriers preparing for impact when being tackled by ducking their heads, are put at increased risk of both head and SCI injury. Tackling players need to keep their heads up and eyes focusing on an impending impact to optimise their readiness.

Figure 5 depicts a low tackle in rugby league when the tackler was subsequently injured (concussion) after his head received a vertex impact with the ball carriers hip. Note the tackler has his head facing down and is not looking ahead.

92

Mechanism of ruck and maul injuries

Rucks and mauls are a phase of play existing in rugby union only. When a player is tackled and the ball touches the ground, it must be released. The forwards of both sides then struggle for possession. This is called a ‘ruck’ and players must come in from their own side of the ball. Only their feet may be used to ‘hook’ the ball back to the scrum half. If the tackled player stays on his feet and continues with the ball in hand, the forwards again struggle for possession but the hands may now be used, then being called a ‘maul’. If the ball goes to ground it then becomes a ruck (ARFU 1995).

Maul and ruck situations have been identified as particularly dangerous for rugby union players (Scher 1983b; Silver 1992; Yeo 1993). Spinal injuries occur through forced flexion to the ball carrier’s neck or to a player at the bottom of the ruck or maul. Serious injury may also occur due to players charging into the ruck or maul (Scher 1983b).

It is often difficult to determine whether certain rugby union injuries occur in a tackle or in the ensuing ruck or maul. This is because players may be injured on impact, when there is a pile up, or on hitting the ground (Silver 1984). Rule changes allowing players to distribute the ball from the ground have seen the increasing popularity of taking the ball to ground thus resulting in more rucks occurring (Silver & Stewart 1994). The banned tactic of charging into a ruck or maul at speed is of particular danger to both those already bound together and to the player rushing in, especially if his head is down ready to pack in. However, rules banning charging are difficult to enforce. Rucks and mauls have been argued to be an ill defined area of law that causes violence among players. This was a primary reason for its abolishment early on in the development of rugby league rule variations (Dunning & Sheard 1979).

93 Mechanism of scrum injuries

Scrum related injuries occur at the engagement of the two packs or when the front rows collapse. They may also occur when a player in the front row, usually the hooker, is ‘popped’ out of the scrum by having their body forced upwards by the force of the competing packs, while their head is pushed down by the interlocking formation with the opposing pack.

The power generated in an initial scrum engagement is potentially greater than is safe for players’ spine. Controlled experimental measurement of forces at scrum engagement show that the forces experienced by the front row players (particularly the hooker), can surpass the structural limits of the cervical spine (Milburn 1993; Taylor & Coolican 1988). The anatomy of the cervical spine is such that dislocation can be produced experimentally in cadaveric spines by applying a pure vertical load of an order far less than that produced in power scrummage (Bauze & Ardran 1979; Wetzler et al. 1996).

Scrum collapse may occur because of instabilities caused by a mismatch of skill, experience and strength. However, it may be intentionally collapsed despite being an illegal tactic. In Australia, collision rather than collapse is the primary mechanism of most scrum injuries with at least 60% of such injuries occurring as an axial load is applied when the two front rows engage (Milburn 1993; Taylor & Coolican 1987, 1988). A recent Australian study covering 1997 to 2002 found six out of seven rugby union scrum related SCI cases occurred at the engagement of opposing teams (Carmody et al. 2005). Frequently SCI occurs before the ball is even placed into the scrum. Often injury at engagement occurs when players are unready or inexperienced and the head of the player is not properly aligned. This can result in head clashes with opposing players (Scher 1981), the head being driven into the torso of an opposing player (Scher 1982) or the player being ‘popped’ from the scrum. Similar support for the predominance of collision phase scrum injuries (as opposed to scrum collapse) can be found in other countries (Burry & Calcinai 1988; Wetzler et al. 1996; Wetzler et al. 1998).

94 The speed of engagement, weight and number of players involved determine the forces generated in scrummage. An impulsive force would be larger in stopping a scrum with a greater mass (weight), as would be expected with an international team compared to a local club team. Experimental reproduction of scrum forces using an instrumented machine (Milburn 1990) showed a large forward impulsive force on engagement, followed by a drop and stabilisation of force exerted. Peak force alone has not been found to be a good indicator of the risk of SCI. The most consistent measure was the impulse of the impact force (i.e. the force in relation to its stopping time) (Milburn 1993).

Milburn (1993) argues from a biomechanical point of view, that the tendency for a scrum to collapse is a reflection of the instability of vertical and lateral forces. Older and more experienced players are better able to maintain the stability of the scrum following engagement (as opposing teams sustain a push) but this appears to be independent of player size. Wheeling also causes danger of excessive horizontal sheer forces. The entire front row experiences a downward force at engagement, as a proportion of their body weight, as they move forward of their base of support and balance to engage opponents. Misalignment of players, mismatches of strength and experience, fatigued players, charging packs, opposing props pulling down will increase these forces. A tightly bound scrum may also increase the risk of SCI in rugby union (Milburn 1993; Scher 1977).

When a front row player is injured on impact it is quite likely the scrum will collapse due to the instability this will cause (Armour et al. 1997). The study conducted by Taylor and Coolican (1987) used interviews with football players receiving SCI in scrums, who had surprisingly good recollections of the events leading to injury, in order to distinguish the phase of scrum at which injury was occurring. They found engagement to be the point at which most players were injured. However, scrum collapse still directly accounts for many scrum related rugby football SCI, indicating the importance of ensuring scrum stability after engagement, in addition to controlling initial impact forces.

95 Scrum engagement injuries suggest the need for ‘depowering’ the initial engagement, justifying the ‘crouch-pause-engage’ sequence introduced in 1992 for senior players and earlier in the mid 1980s for junior players (‘Touch-crouch-pause- engage’). In addition to sequenced engagement rules, other changes made in 1985 to rugby union rules for under 19 year olds aimed at improving scrum safety including limiting pushing to 1.5 metres and wheeling to 45 for the sake of stability. Intentional ‘popping’ of an opponent in a scrum was banned. So too was ‘crotch- binding’ which increased pushing force, particularly downwards, risking collapse. Players were required to keep the upper trunk and shoulders above the level of the waist in the scrummaging position to avoid collapse. In addition, halfbacks were no longer allowed to follow the ball in the scrum. Studies (Milburn 1993; NHMRC 1994) have argued that a time limit should be imposed on the duration of scrums to decrease the chance of collapse, which exposes players to potential SCI. Referees are encouraged to minimise the duration of scrums particularly at junior levels.

However, the effectiveness of these law changes often depends on the referee’s discretion. For example, it is often difficult to determine what constitutes charging at scrum engagement or when to safely terminate a scrum. A referee may not be able to tell if a player is being tightly bound in a scrum.

96 Acute management for suspected serious injuries

Appropriate management of suspected serious injury before hospitalisation can markedly improve recovery outcomes and prevent secondary injuries from occurring. Sports Medicine Australia recommends the early diagnosis and management of injury regardless of level of competition. This would require provision of qualified first aid personnel at all sporting events (Larkins 1995). This proposal is also supported in recommendations from the NHMRC football injury report (NHMRC 1994), arguing sporting organisations and facilities should have pre-arranged, continually reviewed injury management plans. The following section discusses important considerations for the acute management of SCI and brain injury.

Spinal cord injury

The immediate management of patients with suspected spinal injury can often critically influence their prognosis. Appropriate first aid for suspected traumatic SCI patients requires restricting movement that may cause further injury to the spine (secondary prevention of injury) (Yarkony et al. 1997). Among the main objectives of pre-hospital care are spinal stabilisation to avoid further injury and rapid transfer to an SCI management system (Wilberger 1998). While diagnosis of SCI in the unconscious patient may be difficult, all unconscious patients should be suspected of having both a head injury and SCI until both have been excluded. Particular attention needs to be paid to potential spinal injury in concussed patients (McCrory & Dicker 1992). Consensus on the management of suspected SCI contrasts with the divergence of opinions over the appropriate assessment and management of suspected brain injury evident in the literature reviewed.

97 Evacuation procedures in NSW have developed in recent years with greater amounts of suspected injuries receiving appropriate treatment (S Engel PH/POW Spinal unit director May 1998, pers. comm., S Rutkowski RNSH Spinal unit director May 1998 pers. comm.). However, many community level games do not have attending medical personnel trained in acute SCI management with this job left to coaches who have little first aid knowledge (Glaun et al. 1984; NHMRC 1994). Alarmingly, one British study even found that 64% of sports medicine doctors were not proficient at basic life support, assessment and management of seriously injured patients with potential spinal injury (Lavis et al. 2001) highlighting the need for appropriate first aid training for serious injuries.

Brain injury

Suspected concussion in a conscious player requires history taking about loss of consciousness and memory of events before and after the impact, motor and sensory changes, and pain. In an unconscious player, the airway, breathing, circulation, and cervical spine should be checked (Fick 1995).

Pressure to return to play after concussion, while common, is extremely dangerous and not in a players’ best interest. The NHMRC report advised concussed players should be removed from the field and should return to match play only after symptoms have disappeared, having been tested under the exertion of training. It recommended that more research be done on concussion and the use of headgear as a protective measure. With the importance of post injury management for concussion, concern was drawn that non-elite players will rarely have appropriately trained medical personnel to treat and assess injuries, recommending at least one first aid trained person at every match (NHMRC 1994).

98 Illegal/foul play

Several studies across the world have found that foul or illegal play contributes to serious injuries (Kew et al. 1991; Lawson et al. 1995; Milburn 1987; Scher 1978, 1991b; Silver 1984; Taylor & Coolican 1987). An expert panel established by the NHMRC estimated that illegal play accounted for half the serious head and neck injuries in Australian rugby codes. The NHMRC report on head and neck injuries in football indicated dangerous tackling techniques such as high tackles and ‘spear’ tackling (picking an opponent up in a tackle and dropping them on their head) as well as foul play in general as problematic. Key recommendations included the expansion of safety training programs to reduce illegal play and tougher penalties for illegal play. It also recommended the outlawing of gang tackles (NHMRC 1994). Sports Medicine Australia also argues that rule infringements need to be carefully monitored and penalised (Larkins 1995).

In the fast-paced full contact environment of a football game a fine distinction lies between playing admirably hard or being excessively aggressive. Although a player seeks to physically dominate their opponents, it is hoped they do not wish to purposefully harm them. The effectiveness of rules designed for player safety depends on referee enforcement and appropriate penalty deterrents for transgressions.

At the elite level, judicial panels review video footage of incidents to decide penalties for players usually in the form of match suspensions or fines with formal judicial processes being developed in recent years. The NRL has very recently experimented with having two referees at 1st grade matches. The media has increasingly covered elite level foul play incidents and there is large-scale acceptance of foul play sanctions among the public. However, it is much more difficult to maintain this level of vigilance at the community level where most footballers play.

99 The following is a brief discussion of particularly dangerous illegal plays:

High tackles – have long been identified as a risk for serious injury (Scher 1978, 1991b) with intentional contact above shoulders banned in early rulebooks in both codes (NSW Rugby Football League [NSWRFL] 1913; NSW Rugby Union [NSWRU] 1926). However, it is only in recent years that this rule has been thoroughly enforced and penalised.

Intentional scrum ‘popping’ or collapse – is an outlawed tactic in rugby union that is still sometimes perpetrated. It is the action of purposefully ‘popping’ a scrum to collapse the opposition’s players. This action has proven to have tragic consequences with the hooker most often at greatest risk of SCI. Players might also seek to pull down opposing players by their jerseys to collapse the scrum (Milburn 1987).

Wedge formations and other banned attack strategies. Throughout the history of both rugby codes a number of playing strategies have been developed that were eventually banned in the interests of player safety (Ryan 1983). A wedge formation or a ‘Flying V’ was one of several attack strategies developed to break over a Try line when an attacking team was unable to break the defensive line. It involved bunching together in a V formation (the point of the V pointing forward) and barge through the opposition, often with the attacking players putting their heads down as if scrummaging (Macklin 1962). Several serious injuries occurred that forced the respective rugby codes rule making bodies to officially ban this and other dangerous strategies (ARFU 1995; ARL 1995).

100 Spear tackles. Spear tackling (where a player is lifted off the ground and dropped, or driven, head first into the ground) has caused a number of serious injuries (Close et al. 1993) and has been officially banned with increased attention from referees to enforce this (ARFU 1995; ARL 1995). It has been targeted along with head high tackling as a very serious offence. Figures 12-14, showing spear tackles, illustrate how multiple tacklers are often involved.

Rushing a maul. The practice of players rushing into a maul (running in to pack in with excessive force) or ruck has been a recognised cause of injury to both those who rush in and those already in the maul or ruck (Milburn 1993; Scher 1983b). Rucks and mauls are inherently unstable and often players have fallen in amongst them waiting for the play to clear. Rushing mauls and rucks is discouraged in training and can be penalised by referees (ARFU 1995).

Power scrums. In Rugby Union this kind of scrummage came into prominence throughout the 1970s and marked a greater focus by trainers on the force exerted by players in a scrum (Quarrie et al. 2002). Players’ ability to dominate scrums greatly influences ball possession in rugby union so it is a valued skill. This increased focus on scrummaging has markedly amplified the forces being exerted on both sides of the scrummaging players. Rule changes to sequentially engage the scrum under tighter control by the referee were introduced in an attempt to reduce the excessive forces and instability created in power scrummage (Milburn 1993; Taylor & Coolican 1988;). Players, particularly at the junior levels are prevented from packing too low.

101 Figure 6: Lifting of ball carrier Figure 7: Impact with ground after prior to spear tackle. spear tackle.

Figures 6-8, showing spear tackles in rugby league, illustrate how players are lifted upside down and dropped or even driven to the ground often by multiple tacklers. These pictures show spear tackles that were penalised by referees.

Figure 8: Spear tackle with ball carrier’s head driven into ground

102 Number of players

Studies have considered player contributions to force in scrummaging. Although estimates have varied, the front and second row each account for around 40% of the force exerted. Flankers contribute around 20-38% to forward force, and the lock position contributes little additional force (Milburn 1993). Based on these results, flankers rather than locks would be the best players to remove from a scrum formation to reduce forces experienced. Locks provide stability to the scrum formation while adding little extra forward force. Flankers contribute considerable force and may increase lateral forces that cause instability.

Whether the number of players on a field effects injury rates has not been directly studied. It may be partly inferred from comparing injury rates in games adopting different numbers of players. Rugby league has two less players than in rugby union and experiences a relatively lower risk of some serious injuries. However, many other variations in game rules confound this comparison.

More suitable evidence is gained within rugby union where there is an international competition running with variations to the 15-man-a-side rules (also with significantly shorter games) i.e. super sevens. Under these rule variations, many more tries are scored as increased room between players makes break through runs more likely. Players appear more likely to be tackled while being chased rather than in head on impacts. While reducing the intensity and number of impacts, the reduction in player numbers emphasises other more skilled aspects of the game such as running, passing and kicking with more exciting, open and continuous plays. However, there are no published studies comparing rates of injury between persons playing under different variations of game rules.

103 Stage of season

High proportions of all injuries in rugby football around the world occur early in the playing season (Alsop et al. 2000; Garraway et al. 1991; Garaway & Macleod 1995). Similarly, SCI has been found to occur most frequently early in the season (Armour et al. 1997; Kew et al. 1991; Williams & McKibbin 1978,1987).

It is likely though that the relationship between the stage in the season and the frequency of injury is a reflection of other underlying causes that affect injury risk which have previously been discussed in considerations of player skill, experience, physical preparation and aggression. Little research exists to validate these causes, however, several suggestions why injuries are more likely to occur very early in the season have been made by Quarrie, Cantu and Chalmers (2002) and include: lack of skill and experience in new players taking up the sport, unfamiliarity with new positions, lack of familiarity within teams, lack of continued practice prior to the start of season, lack of physical and impact conditioning, mismatch of player skill and size in trial games and the aggressive nature of trial games where players vie for team positions.

However, one study has shown an increased risk of injury for players attending pre- season training for a longer period of time, although this may have been confounded by characteristics of the people attending more training e.g. more intense participation in game (Lee et al. 2001). Studies in Australia and Argentina have also found large proportions of injuries occurring in the latter stages of the season (Gabbett 2000; Taylor & Coolican 1987) suggesting an important contributory role of fatigue or accumulative microtrauma to injury risk. These late season injuries might also be related to the increased fervour and aggression among players as competitions are coming to a conclusion and finals games are played (Taylor & Coolican 1987). Differences in findings across these studies from a range of countries are explained by varying study populations and severity of injury under study.

104 Other external risk factors

Other external factors that may be associated with injury risk are the stage of the game and position on field (i.e. proximity to Try lines). Several studies have reported a predominance of injuries early in games possibly because of a greater degree of overzealousness exercised by players when in the early stages of a game or because players are not adequately warmed up (Garraway et al. 1991; Seward 1993). However, other studies have found that most injuries were sustained in the second half of matches (Bottini, et al. 2000; Gabbett 2000; Wekesa, et al. 1996) suggesting that fatigue may contribute to injuries. Differences in findings between these studies are possibly explained by varying study populations and injury definitions used.

A study of all injuries during a Rugby World Cup pre-qualifying tournament found slightly more injuries occurred in the defensive half of the field of play (53.2%) than in the offensive half (46.8%) (Wekesa et al. 1996). A study of minor injuries in rugby league found that slightly more injuries occurred in the injured players own half of the playing field (Gibbs 1994). This might be a reflection of the increased force and aggression used when players are defending their own half.

105 Protective equipment for serious injury in rugby football

Injury prevention strategies inevitably seek the use of protective equipment as a means of reducing injury. The effectiveness of protective equipment for preventing serious injuries has been demonstrated for bicyclists (Attewell et al. 2001; McDermott 1992), reducing the risk of cycling related brain injuries in Australia by up to 88% (PHA 1999).

Sports Medicine Australia recommends protective equipment should always be used when appropriate (Larkins 1995). The Public Health Association of Australia recommends a systematic investigation by governments and peak sporting bodies about the use of protective headgear in sports, particularly where their use is not standard (Public Health Association [PHA], 1999).

In rugby football protective equipment can be effective for protection against minor injuries, but there is much less evidence concerning the effectiveness of protective equipment for serious injuries to the head and cervical spinal cord (Gerrard 1998; MacDougal & Osbourne 1992; McIntosh & McCrory 2000; Wilson 1998) requiring further research (NHMRC 1994). There is little evidence to suggest that protective padding modifies the risk of serious injury in rugby in any respect. Proper epidemiological studies assessing the risk of injury before and after the introduction of protective equipment are needed to conclude their usefulness or harm before recommendations about usage can be made (Hrysomallis & Morrison 1997; Quarrie & Chalmers 2001).

American football has a long history of attention to injury prevention and has seen the standard adoption of extensive protective equipment for players. However, this led to a cycle in which ever more equipment was needed to ensure player safety and many aspects of the game had to significantly change including player selection and general playing tactics (Hrysomallis & Morrison 1997).

106 A Scottish study (Garraway et al. 2000) assessing the impacts of professionalism on rugby injuries concluded that the universal adoption of protective equipment was largely responsible for an increase in injuries. They assert that elite players expect that protective equipment minimises the consequences of impacts and therefore use increasing force against their opponents. However, the conclusions from this study have been called into question because of lack of supporting evidence (Quarrie & Chalmers 2001).

One AFL study argued that protective equipment potentially slows players down through added weight, less heat dissipation and movement restriction. This means slower games with subsequently less impact force, so there would likely be a greater number of collisions but possibly less damage as a result (Norton et al. 1999).

Spinal cord injury

SCI is rare among football players. It is important to consider that introducing new protective equipment, intended for intervention of one rare problem (i.e. cervical quadriplegia), may lead to other more common injuries (Bishop 1996).

For SCI, the use of helmets can reduce the effects of impact by spreading the load and increasing the duration of impact. Studies of helmet use in motorcycle riders confirm that full face helmets have a significant effect in transmitting impact away from the fragile cervical spine (Yeo 1979, 1980). However helmets pose their own risk for injuring other players. Milburn (1993) argues it is almost impossible to protect against the effects of acceleration and deceleration loading on the spine in football. Neck fixation devices aimed at immobilising the cervical spine have not been found to provide adequate fixation or shock absorption. Only with devices that fix the collar firmly to the chest (such as in a cervicothoracic orthosis) could cervical motion be sufficiently restricted, but this is clearly contrary to the mobility needs of a rugby football player although they have been employed for elite rugby league players with previous injuries. (Milburn 1993).

107 Head and brain injury

The scientific literature supports the use of mouthguards (Barth et al. 2000; McCrory & Dickers 1992) with custom fabricated mouth guards being found to be the most effective (Chalmers 1998). Level of acceptance and use is relatively high in rugby codes but is not compulsory in Australia. In New Zealand rugby union mouthguards are compulsory and referees can stop a game if the player takes to the field without one. (Quarrie et al. 2001). While mainly utilised to protect teeth, correctly fitting mouth guards have been argued to provide some degree of protection against concussion in rugby football by reinforcing the stability of the jaw (McCrory & Dicker 1992). However, a recent review of available evidence concluded that there was actually not sufficient evidence to support that mouth guards can protect players from any type of injury, particularly concussion (McCrory 2001c). It is clear that any protection provided by mouth guards against serious brain injury is of a very limited nature.

Soft helmets have been shown to be effective in reducing more minor head injuries such as abrasions, lacerations and possibly some concussion but research is still ongoing. Protection from injury caused by forceful impact is limited (Wilson 1998). An Australian study tested commercially available football head protectors for the ability to reduce the likelihood of concussion. Laboratory tests indicated that the football headgear tested would not reduce the likelihood of concussion. The study argued that internationally recognised standards for soft headgear were needed (McIntosh & McCrory 2000).

All types of helmets increase mass and thus head momentum, and make the head a bigger target for impact. They can possibly (but not necessarily) obstruct vision and give a false sense of security encouraging reckless behaviour (Geffen et al. 1998). Following the introduction of hard helmets in American football, which were introduced to prevent head injuries, the number of neck injuries increased dramatically (Geffen et al. 1998).

108 To develop effective protective headgear, it is necessary to define the nature of the impacts that result in concussion. The important parameters for definition include impact site, velocity and energy. It would be impossible to measure these parameters in the filed of play with standard measuring instruments. However, one Australian study estimated impact parameters from videotape evidence of head injury events. This information can be used to establish performance criteria for assessing the impact energy attenuation properties of headgear (McIntosh & McCrory 2000).

Where headgear is worn, referees need to ensure that it is not used to manipulate an opponent. Players, coaches and administrators must recognise the danger and discourage the use of the head as an implement or targeting the headgear of opposing players. More research is needed into the effectiveness of headgear in light of the altered behaviour its use may produce before full endorsement. Attitudes to usage must be thoroughly assessed, particularly regarding reasons for current low levels of acceptance (Wilson 1998) and the effect that wearing helmets and other protective equipment have on players’ sense of resiliency to injury and how this in turn affects the intensity of the game (Quarrie & Chalmers 2001). A study of Australian schoolboy rugby players found safety was the main reason for wearing headgear and that players felt they played more confidently in headgear (Finch et al. 2001).

109 INTRINSIC RISK FACTORS

Player characteristics that influence the risk of serious injury

Several interrelated factors specific to each player may influence vulnerability to serious injury. These include body size (weight and height), body type, physical preparation, age, experience, and skill. Other factors that might be considered include position played and the level of competition played. The importance of appropriate fitness, preparedness and suitable body type for playing both rugby league and rugby union in order to reduce the risk of injury has been identified extensively in the literature (Kew et al. 1991; Milburn 1993; Silver 1992; Williams & McKibbin 1978, 1987).

Anthropometric and physiological characteristics

Individual rugby league players have been shown to cover distances of approximately 5000 to 8000m a game, and be involved in 20 to 40 tackles. Forwards generally have higher body mass, subcutaneous fat and fat free mass levels than backs (Brewer & Davis 1995). For rugby union players there is a lack of comprehensive research into anthropometrical and physiological characteristics of players and game requirements. This is possibly explained by traditional concern in rugby union for aspects of skill rather than physical and physiological requirements. However elite players are now being placed under greater physical demands with the introduction of professionalism, with many more competitions and longer seasons (Nicholas 1997; Silver 2001, 2002; Williams 2002).

Larger players are traditionally considered better equipped to play rugby football and withstand serious injury. Certainly, an impact between a smaller and larger opponent will see the advantage of inertia for the larger player. This highlights the importance of matching players appropriately for size, which is discussed below in the section concerning age. However, players are usually of similar size to their opponents. The larger players have more force that is likely to be exerted in

110 impacts (Kew et al. 1991). As force of impact has shown to be a crucial factor in determining the risk of injury (Kew et al. 1991; McIntosh et al. 2000; Wetzler et al. 1996), it might be expected that games with larger players have an increased risk of injury. This is reflected in the findings of several studies that players in higher grades of play, who are larger than those at lower grades, are more likely to be injured despite being more skilled, experienced and physically developed (Armour et al. 1997; Jakoet & Noakes 1998; Kew et al. 1991; Lee & Garraway 1996; Seward et al. 1993).

Studies of rugby union played in the United States found a relatively high incidence of scrum injuries compared to other nations was partly accounted for by American players’ greater size and strength (Carlson et al. 1994; Wetzler et al. 1996; Wetzler et al. 1998).

The increased risk of injury this additional force will bring must be balanced by any protective factor that size provides to a player against injury. However, it is not clear how much protection body size can provide from serious injury to vulnerable body areas such as the head and cervical spine. Increased muscle mass may provide protective benefits through muscle splinting in an impact, but would not be sufficient to resist the loading on the cervical spine potentially occurring in an average scrum or tackle (Bauze & Ardran 1979; Taylor & Coolican 1988; Wetzler et al. 1996). Similarly, increased strength may reduce the acceleration of the head following impacts acting as a protective factor for brain injury (Geffen, cited in Sherry 1998) but is unlikely to guarantee protection against the potential force of impacts experienced in a typical tackle.

111 Weight alone is not the best indicator of body type as a person’s weight has varying proportions of fat and muscle composition. It could be expected that higher muscle composition could provide limited protection due to the benefits of muscle splinting in an impact (Taylor & Coolican 1988). It is generally accepted that in most sports a high level of body fat has an adverse effect on performance because increased mass decreases acceleration with fat representing a non-contributing load, offering no assistance to mechanical movement. Heat load experienced by athletes with greater fat mass will also be higher than leaner athletes (Woolford et al. 1993).

One British study tried to determine whether there was an association between players’ physique and susceptibility to injury. It found that endomorphic (obese) players were more likely to be injured than ectomorphs (linear) even when controlling for age. However, forwards were more likely to be heavier players possibly confounding the effect of body type (Lee et al. 1997). Larger players tend to be injured more than smaller players but this is also associated with the grade played. A New Zealand study found that after controlling for position and grade, the thinner club players as measured on a body mass index missed a greater amount of playing time due to injuries than their bigger counterparts (Quarrie et al. 2001).

Other body characteristics may also influence vulnerability to serious injury. Several authors have drawn attention to the increased risk of spinal injury to players with long thin necks (particularly while scrummaging) (Burry & Calcinai 1988; Milburn 1993; Yeo 1993). Several studies have found an increased risk of injury for players when playing out of usual positions (Lee et al. 1997; Silver 1984, 1992), which is often due to an inappropriate body type for the position in addition to inexperience. Placing players in positions appropriate to body type is a key recommendation of the NHMRC report on Football Head and Neck Injury (NHMRC 1994). This has also been recognised by the legal system for many years as evidenced from the Watson vs Haines case of 1987, stemming from a schoolboy rugby league SCI incident, which was successful in suing the NSW Department of Education for allowing a player with a long thin neck in a scrum (Australian Torts Reports 1987).

112

The recognition of this has led to improved selection of appropriate players, particularly in non-elite football. Excluding players with unsuitable body types (i.e. long thin necks in forward positions) has formed a part of the basis of SCI prevention education campaigns for rugby football in Australia (Yeo 1998a).

However, the authors of Australia’s most comprehensive investigation into football spinal injuries have argued that there is no suitable evidence that a long slender neck inherently increases the risk of neck injury (Taylor & Coolican 1988). Similarly, it is argued that strong neck musculature has not been shown to protect against these injuries as dislocation has been shown to occur from axial loading below that is experienced by front row players in scrummaging (Bauze & Ardran 1979, Wetzler et al. 1996).

Taylor and Coolican (1988) argue that a “strong neck musculature no more protects against dislocation of the spine than strong shoulder muscles would avert shoulder joint dislocation,” (p.224). Children whose bones and muscles are yet fully developed (Fricker 1999) still seem to be at a much lower risk of SCI (Armour et al. 1997; Jakoet & Noakes 1998; Kew et al. 1991; Lee & Garraway 1996; Quarrie et al. 2001; Seward et al. 1993; Silver 2001, 2002; Silver & Stewart 1994) despite having less appropriate body types than older players. A comprehensive study of New Zealand football SCI 1976-1995 found all builds equally at risk of serious SCI (Armour et al. 1997).

Although scientific evidence is lacking, Quarrie, Handcock and Toomey's (1996) comprehensive review of the rugby football SCI literature led them to wisely suggest: "Until more information is collected about the effect of physique on risk, it is safer to err on the side of caution and continue to recommend that thin players (or players with long/thin necks) do not take up positions in the front row." (p. 55)

113 However, regardless of whether SCI is more likely to occur in individuals with slender body types than in individuals who are muscular and/or have short thick necks under the same loads – the protection anthropometric and physiological characteristics afford for SCI are not sufficient to prevent their occurrence amongst those with ideal characteristics, for example among elite players (Armour et al. 1997; Jakoet & Noakes 1998; Kew et al. 1991; Lee & Garraway 1996).

The vulnerability of all body types to serious injury suggests that preventive efforts should additionally be directed at minimising the forces involved in contact play, rather than purely ensuring players are of suitable strength and body type to withstand these impacts, the force of which is likely to increase with player size.

Pre-existing conditions

Existing disease or physical weakness may indicate preclusion or special consideration when playing football. The risk for competitive young Aboriginal sportsmen may be particularly high. A study reported that between 1982 and 1996, there were eight sudden cardiac deaths due to Ischaemic heart disease and related to sporting activity among Aboriginal sportsmen under 40 years of age in the Northern Territory. Education of athletes, heat-stress reduction strategies, and cardiovascular screening were recommended to reduce the incidence of sudden cardiac death in sport (Young et al. 1999) (See section three for discussion of sudden death in athletes).

Although more research is required, several authors in South Africa and New Zealand (Quarrie et al. 2002; Scher 1983a, 1990a, 1990b) have drawn attention to the risk of rugby football players with congenital abnormalities or degenerative changes to the spine sustaining SCI. Accordingly it has been recommended that screening procedures are developed and that players sustaining a neck injury more serious than a muscle strain obtain clearance from a spinal or orthopaedic specialist prior to return to play. Research for American football has developed a number of recommendations regarding return to play guidelines for a number of

114 conditions involving congenital and degenerative abnormality of the spine (Torg & Ramsay-Emrhein 1997).

Failure to recover from even minor injury may place players at increased risk of serious injury through being unable to perform at peak performance. Several studies have shown that previous injury is a predictive factor for further injury in sport (Lee et al. 2001; Quarrie et al. 2001; Quarrie et al. 2002)

Recommendations concerning preventing long-term sequelae of minor head injuries in American football, state that an athlete should not return to play while any symptoms remain, regardless of how minor the injury was. Symptoms that do not resolve rapidly suggest evaluation by a neurologist, as the potential long-term consequences of minor head injury remain to be clearly determined (Wilberger 1993). Team medical officers need to be aware of the prospect of concussion in athletes who do not realise their injury or may want to conceal it (Fick 1995).

Elite rugby league clubs in Australia have their own, often disparate, injury management policies. Sydney metropolitan clubs require injured players to produce a doctor’s certificate indicating fitness to resume before being permitted to play. County clubs often do not adopt this as standard practice. Rugby union clubs tend to have a more consistent approach to injury management as they follow safety guidelines set out by the International Rugby Board. The fact that elite sport practices set the standard for community level players highlights the importance of higher level teams adopting safest practices (SMH 25 February 1995, p. 68).

However as discussed in a preceding section, 'Clinical definition and management of head injury’, of this chapter, mandatory exclusion periods for concussed players have been argued to be excessively arbitrary with each player’s condition needing to be individually assessed (McCrory & Berkovic 1998). Several elite rugby league club doctors do not support mandatory exclusions but have recommended the usefulness of establishing psychometric testing performance for more accurate assessment of cognitive effects post injury.

115 Repeated baseline assessment (with reliable testing tools) before injury to reflect a player's potential, allows measurement of impaired performance post injury providing a quantitative basis for decisions regarding return to play (Hinton-Bayre et al. 1999; McCrory 1997). One recent Australian pilot study suggested that both the number of post concussive symptoms and their duration might be used as a measure of injury severity and a guide for return to play (McCrory et al. 2000). Decisions regarding return to play are executed most effectively if a clear plan for handling these situations is in place before the injuries take place (Porter 1999).

The recent formulation of guidelines for the management of concussion in sports adopted by the American Academy of Neurology specifically calls for the development of a standardised, systematic sideline evaluation for the immediate assessment of concussion in athletes. A standardised sideline examination of this type can be useful in detecting concussion and determining fitness to return to play (McCrea et al. 1997).

Physical preparation

As mentioned earlier, adequate player preparation and fitness is traditionally emphasised as one of the most important factors in reducing injury (Sherry & Wilson 1998; Williams 2002). A study estimating the association of strength on sport injury in Turkey found that back and leg muscle power was greater in non- injured athletes than injured ones. The authors concluded that because the back and leg muscles help maintain equilibrium, the power of these muscles may affect ability to avert sporting injuries (Dane et al. 2002).

Several studies (Alsop et al. 2000; Armour et al. 1997; Garaway & Macleod 1995; Garraway et al. 1991; Kew et al. 1991; Seward et al. 1993; Williams & McKibbin 1978, 1987) have found that injuries are more likely to occur early in the playing season or early in the first half of the game, which may reflect players not being sufficiently warmed up and generally physically prepared (Upton et al. 1996). Strengthening exercises for the neck, warming up exercises before games and

116 choice of players have formed part of the basis of football SCI prevention education campaigns in Australia (Yeo 1998a). In terms of head damage, it is not possible to condition the brain against injury but fitness and strength may reduce acceleration of the head following impacts (Geffen, cited in Sherry 1998).

However, several studies in rugby football and other sports have found little evidence to suggest that fitness, flexibility and strength provided significant protection from serious injuries (Lee et al. 2001; Silver 2001, 2002). Many studies have also found a greater risk of injury at higher levels of competition where players are strongest and fittest (Armour et al. 1997; Jakoet & Noakes 1998; Kew et al. 1991; Lee & Garraway 1996; Quarrie et al. 2001; Seward et al. 1993; Silver 2001; 2002; Silver & Stewart 1994).

Although physical preparation is undoubtedly of importance to the risk of injury, speed of play, size of players (Roux et al. 1987; Silver 2002) and forces of engagement (Taylor & Coolican 1987) are important aetiological factors in the majority of rugby injuries. Of note, is the fact that players’ fitness and strength influence the velocity and force they are able to exert in an impact and in turn injury risk (Kew et al. 1991). It also influences their ability to keep up with play and position themselves to be able to build momentum for head-on impacts.

117 Skills, experience and technique

Many key studies of rugby football SCI around the world have concluded that in rugby union the cause of SCI is largely a function of technique (Burry & Gowland 1981; Milburn 1993; Scher 1977; Silver 1984, 1992). The importance of impact technique lies not only in the prevention of those injured due to poor technique but also from those who might be protected from injury by improved impact technique in what is inherently a dangerous situation (i.e. tackles or scrums). Skills training in tackling technique to minimise the risk of injury has been long recognised by football trainers and often form a fundamental part of training sessions (particularly for younger players). The ability to perform a tackle, and many other aspects of the game of rugby union and league relatively safely, is a highly adept skill that only comes with training, practice and experience.

The inexperience of players has been identified as contributing to injury, for example when a 2nd row continues to push after the collapse of a scrum, or when inexperienced players cause scrum collapse (Kew et al. 1991; Silver 1984; Williams & McKibbin 1978). This often occurs when players are out of their normal playing position tending to cause misalignment and instability in the scrum due to unfamiliarity or inexperience (Milburn 1993). Studies of rugby union played in the United States found inexperienced players, lack of coaching, and practice facilities for scrummaging lessened players ability to coordinate engagement and maintain control of the scrum leading to a high incidence of scrum injuries (Wetzler et al. 1996; Wetzler et al. 1998).

Players tackling or running with the ball with their heads down, making the head the most likely first point of impact and decreasing visual capabilities thus readiness for impact, are put at greater risk of serious SCI or head injury. Tackling too low with poor technique increases the risk of injury to the tackler. A tackle should be aimed at the waist then to slide down to the legs. In addition, the head should always be up, looking ahead at all times, with the head slightly to the side ready to take the impact with the shoulder. Care must be taken when tackling low of upswinging

118 knees and that the head does not lead the impact. As previously discussed, evidence from NSWSU/NSWSIC analysis suggests tackling low with the head down is most often the cause of injury for seriously injured tacklers.

In American football most severe cervical spine injuries share the common mechanism of application of an axial load to the straightened spine. Avoiding techniques that employ head-down ‘spear’ tackling markedly reduce the risk of serious injury (Thomas et al. 1999). Note that for American football, ‘spear’ tackling usually refers to the tackler (or sometimes the ball carrier) launching themselves head first, helmeted, into an impact. This is opposed to the rugby codes concept of ‘spear’ tackling that refers to a ball carrier being dumped head first into the ground by tacklers.

Australian Rugby League associations are well aware that most spinal cord injuries occur in tackles. Courses for coaches emphasise the necessity to teach young players the importance of the ‘eyes open – head up’ technique in tackling. Players are encouraged to keep tackling below the armpits, keep their head behind the backside, drive with the legs, and ensure a firm shoulder and arm contact. Another important skill is the ability to predict how a tackled player will fall in order to correctly position the body in preparation for impact and falling to ground (Wakelam 1937).

In scrums, the inherent dangers posed by force loadings on the front row players’ spine in scrums emphasises the importance of correct alignment of the head, neck, and trunk, along with adequate strength to maintain body position during engagement. The ideal pushing positioning is with head, trunk and legs in alignment. Usually the greater the angle between the trunk and thigh, or the lower the packing position of the front row, the greater the forward force.

119 Studies considering binding techniques have concluded that the practice of a hooker binding over and forward of each of the props’ shoulders, increases the risk of being popped upwards and increases forward force on the hooker. The hip binding (2nd row binding around the hip of the prop and grasping his jersey above his hip) and crotch binding (second row binds onto the props jersey by reaching between his legs) techniques were compared by Milburn (1987). The crotch binding technique added to the downward force at engagement and was considered to increase the risk of scrum collapse, assumedly caused by additional downward pull on the prop’s jersey.

At the non elite level some children and teens have particular problems with their tackling and scrummaging technique possibly warranting special attention by coaches to judge their overall suitability and extra training needs. Vertical forces have been found to be greater in lower level, and younger teams increasing risk of scrum collapse, reinforcing the need for correct body alignment (shoulders above the level of the hips) during a scrum particularly for non-elite players (Milburn 1993). The danger of ‘horse play’, where foolishness results in injury, has also been noted (Silver 1984, 1992).

Although skill does afford some measure of protection against SCI, unskilled players are unlikely to play as hard as those that are more skilled. The greater degree of force exerted by more skilled players may paradoxically act to amplify the likelihood of injury (Silver 1984). As discussed below, in the consideration of level of competition, elite players are at no less risk of SCI than players at lower levels of competition. Silver (2001, 2002) argues that greater skill does not protect against SCI in rugby football players. To highlight the importance of size, weight, speed and force in rugby football injuries, Silver (2001) has drawn an analogy with vehicle crash injuries where forces involved and speed of deceleration are major factors in determining the severity of injury.

120 Level of competition, competitiveness and aggression

Several studies have found a greater number of injuries among more skilled players due to a more aggressive approach to the game and greater forces exerted in impacts (Armour et al. 1997; Jakoet & Noakes 1998; Kew et al. 1991; Lee & Garraway 1996; Lee et al. 2001; Orchard & Seward 1994; Quarrie et al. 2001; Seward et al. 1993; Silver 2001, 2002; Silver & Stewart 1994; Rotem & Davidson 2001). Similarly, studies of AFL player injuries have concluded that more injuries occur at higher grades of play where the game is faster and more intense (Norton et al. 1999; Seward et al. 1993).

Many top players are highly paid full-time athletes. Elite players have increased their physical preparation and size with the growing demands of professionalism in all codes of football (Garraway, et al. 2000; Junge et al. 2000; Nicholas 1997; Silver 2001, 2002; Williams 2002). In elite games, players are becoming fitter, larger, stronger, can run faster, weigh more and impact with greater force and accordingly are at greater risk of SCI and other injuries (Lee et al. 2001; Silver 2001, 2002). Coaches for elite rugby football teams in both codes are increasingly using larger mesomorphic (muscular) players rather than endomorphic (obese) or ectomorphic (linear) players (Huxley 2003). Shorter playing careers for professional players in recent years has been speculated by the media to be attributed to the increasing size and speed of players causing more injuries (BBC Sports News online 2002).

Similarly in AFL, Norton, Craig and Olds (1999) concluded that for elite players; "…the overall increase in game tempo, and the opportunity to recruit specialist players, is likely to result in selection pressure for players with high anaerobic powers and capacities…injury rate is at least, in part, a function of the overall game tempo [and player size/preparedness]." (p.403)

121 The fiercely competitive nature of the game at elite levels, over ‘psyching up’ of players and condoning of aggression have been identified as growing problems in the ever competitive senior grade rugby football teams around the world (Bottini et al. 2000; Burry & Calcinai 1988; Kew et al. 1991; Silver 1984, 2001; Silver & Stewart 1994; Taylor & Coolican 1987; Williams & McKibbin 1978, 1987).

A prospective Scottish cohort study, conducted on schoolboys and senior rugby union clubs for one season, concluded schoolboy rugby is much safer than senior club rugby and the outcome of injuries that do occur are less disruptive (Lee & Garraway 1996). A subsequent study of the same senior Scottish clubs found the proportion of players injured nearly doubled after rugby union was professionalised there (Garraway et al. 2000). Subsequent studies in Scotland also found an increased risk of injury for professional players compared to lower levels of competition (Lee et al. 2001). In a major New Zealand study, the majority of players with SCI were injured in competition games especially senior club games (Armour et al. 1997).

Correspondingly in South Africa, (Kew et al. 1991) 69% of SCI in one study occurred in adult A team or senior 1st team players. This study estimated that adult players were 10 to 12 times more likely to sustain a SCI than schoolboys due to increased aggression, size and physical strength of players making them capable of generating greater forces during impacts. A Welsh study found mid level clubs had more SCI than elite and schoolboys clubs (Lee & Garraway 1996).

A comparison of the incidence of less serious injuries among elite levels of rugby league and rugby union teams in Australia found senior teams had a higher rate of injury than in schoolboy teams (Seward et al. 1993). An analysis of the frequency and nature of injuries to players during the 1995 Rugby World Cup claimed to have reported the highest frequency of injury yet recorded in any group of rugby players. It was concluded that the risk of rugby injury was therefore greatest in the best players in the game, challenging the view that superior fitness, skill and experience can reduce the risk of rugby injury. One player suffered a paralysing SCI resulting

122 in an incidence of SCI in the tournament of 4.6 per 10,000 player hours (Jakoet & Noakes 1998).

Many early reports of spinal injuries in rugby football in the late 1970s and 1980s, while tending to be small case series reports and scientific journal letters expressing concern, drew attention to the apparent high number of schoolboy rugby SCI (Carvell et al. 1983; Hoskins 1978; Silver 1979, 1984; Taylor & Coolican 1987; Williams & McKibbin 1987). As discussed, more recent studies have found that adults are more likely to be injured. This is possibly due to a reduction in schoolboy SCI achieved through prevention strategies and a development of more rigorous methods for assessing the relative risk between schoolboy and adult players.

The major difference in junior and senior level games is the amount of force exerted by players. Apart from player size and experience, several rule modifications are aimed at reducing the forces exerted by players at junior levels of the game. When comparing rates of injury among children and adults it is clear that young players, while seemingly vulnerable to injury through lack of muscle and bone development, are unlikely to be seriously injured, as they do not often generate the necessary forces to cause serious injuries (Davidson 1987; Orchard & Seward 1994; Taylor & Coolican 1988; Rotem & Davidson 2001).

An Australian study (1969-1986) of the casualty room in a rugby union playing school found little serious SCI risk for schoolboys. However, minor injury rates were highest in older boys in the highest grades of play (Davidson 1987). Other studies in Australia and in the UK (Carvell et al. 1983; Hoskins 1978; Silver 1979, 1984; Williams & McKibbin 1987) have found significant schoolboy risk of SCI particularly for rugby union. A study of Australian footballers with SCI from 1960 to 1985 found 11 were playing schoolboy rugby union mostly after 1977, and six were playing schoolboy rugby league (Taylor & Coolican 1987).

123 A New Zealand study (Burry & Gowland 1981) found that the extremely high level of competition play at elite levels is spreading down to the schoolboys’ level bringing with it a greater risk of SCI for young players. Those schoolboys injured tended to be players in higher grades with most injuries occurring in competitive games, confirming the risk associated with greater competition.

Two studies looking at all injuries in defined player populations found that significantly more injuries occurred at players’ home fixtures (Davidson 1987; Lower 1994). This possibly reflects the greater risk of injury when playing harder (more competitively) at one’s home fixture.

However, recent studies in Australia have shown a dramatic drop in the incidence of schoolboy rugby union SCI (Rotem et al. 1998; Rotem & Davidson 2001; Wilson et al. 1996). It is worth noting here that all of these studies found virtually no SCI for schoolboy players under 16 years of age. As discussed, the lack of SCI in players less than 16 years might be explained by the fact that they do not have the strength at this age to produce the magnitude of force that may cause injury.

However, sometimes pose the problem of poor supervision and commitment to safety (AIHW & DHFS 1997). An NHMRC working party argued that amateur football players are at a greater risk of concussion than elite players due to better surveillance of foul and dangerous play at higher levels of competition (NHMRC 1994).

The contrasting findings of these studies can be explained by differences in study populations and in their definitions of injury. Whereas the Newington College survey recorded prospectively all injuries reporting to a school casualty station, the other studies retrospectively gathered data from hospital admission records of more severe injuries for all grades of players (Davidson 1987). The injury rates cannot be accurately compared, because the latter studies were not able to calculate injury incidence per player hours. Estimating the incidence of schoolboy rugby injury inevitably leads to the question as to what is an ‘acceptable’ injury level

124 in such a contact sport at junior levels (Wigglesworth 1987). Clearly, injuries such as permanently disabling SCI and head injuries, although rare, are intolerable in light of their devastating consequences.

The importance of players’ emotive influence on competitiveness and aggression levels and the subsequent effect on injury risk has been demonstrated in many studies. Several studies in the United States, frequently investigating injuries in American football, have indicated the influence of psychological factors on sports injuries (Junge 2000). These studies usually focus on the effect of life events, personal characteristics and other psychological attributes such as control beliefs, self-concept, and type A behaviour (i.e extroverted, risk taker) on the risk of injury. Conclusions from these studies, while occasionally conflicting due to differences in sports studied and methodological approaches, generally confer that events resulting in major changes in the life of those affected, such as marriage, the death of a loved one, career change or job loss can influence the risk of injury in athletes (Junge 2000).

Although several studies showed that injured players were significantly more troubled with life events compared with uninjured players, few however were able to conclusively show that positive life events had any direct effect on sports injuries. Additionally, the relationship between life events and sports injuries may be moderated by the degree of social support and several personal characteristics such as a sensation-seeking nature (Junge 2000). From the numerous psychological attributes that have been investigated in relation to sports injuries, only competitive anxiety has been shown to be associated with injury occurrence (Junge 2000). A personality profile typical of the ‘injury-prone’ athlete does not exist (Junge 2000; Lysens et al. 1989). However, several studies have shown a certain readiness to take risks such as a lack of caution and an adventurous spirit on the part of injured athletes (Taimela et al. 1990; Thompson & Morris 1994).

125 Andersen and Williams (1988) developed a model based on stress theory in which the response of the athlete to a potentially stressful athletic situation is critical in determining whether injury will result. The stress response of the athlete depends directly or indirectly on several psychological factors and can be influenced by psychological interventions (Hamilton et al. 1989; Junge 2000).

Other external factors that may influence player competitiveness are the location of the games (home vs away), the time the player participates in the game or season and their position on the field (i.e. proximity to Try lines). Two studies looking at all injuries in defined player populations found that significantly more injuries occurred at players’ home fixtures (Davidson 1987; Lower 1994). This possibly reflects the greater risk of injury identified when playing harder (more competitively) at one’s home fixture.

A predominance of injuries early in games noted by several studies may be because of a greater degree of overzealousness exercised by players when in the early stages of a game. In contrast to earlier discussed findings that injuries tend to predominate early in the season (due to inadequate preparation) a comprehensive study of Australian football SCI injuries in all codes over several decades found that these injuries tended to peak in the latter part of the season (as discussed, differences in study populations and injury definitions may explain this difference). These late season SCIs might arguably be related to the increased fervour and aggression among players as competitions are coming to a conclusion resulting in escalating injury risk (Taylor & Coolican 1987).

A study of more minor injuries in rugby league found that slightly more injuries occurred in the injured players own half (Gibbs 1994). This intuitively could be a reflection of the increased force and aggression used when players are in critical stages of a game i.e. close to scoring or defending a Try. Age

126 Age will largely determine a player’s experience, skill, body type and size. Complete musculoskeletal development and attainment of peak bone mass is generally only achieved in late adolescence or early adulthood (Fricker 1999). All these factors would suggest young players should be at particular risk of serious injury. However, as discussed earlier, adolescents very rarely experience SCI in contact sports most likely due to the inability to produce the dangerous levels of force in impacts (Davidson 1987; Orchard & Seward 1994; Rotem & Davidson 2001; Taylor & Coolican 1988). Forces experienced are also decreased by rule modification specifically designed for younger players to ‘depower’ impacts such as reduced players in the scrum. This emphasises the effectiveness of decreasing impact forces as a measure against SCI as opposed to emphasising increasing player strength and thus their ability to create greater forces in impacts.

The greater danger for junior players is that of mismatch of skill, experience, size and/or strength particularly in the scrum (Milburn 1993, Silver 1979, 1984; Silver & Stewart 1994). Although player size matching in teams is part of junior league and rugby rules it is not always strictly adhered to. The Daily Telegraph (2000) has reported that it has continued, “…to receive complaints from worried parents about the sometimes outrageous weight diversity in junior football. Players in under-12s and under-14s where the average weight of players is around 40kg and 65- 70kg respectively are being hammered by kids close to double those dimensions. Any number of parents say kids are regularly taken to hospital by ambulance on weekends and one group insists it has documentary evidence that one outer western suburbs casualty ward averages six cases every Saturday and Sunday. It must be getting close to time for the NSW Rugby League to conduct a probing inquiry into the claims. Tutt-tutting and "we'll look into it" assurances from that body are wearing painfully thin at the game's grass roots where other codes are waiting in the wings to take over.” (Daily Telegraph: 2000:6)

127 The importance of matching players appropriately for size, skill and experience has long been recognised particularly for junior players. This is evidenced as far back as 1907 in competition rules for Australian public schools rugby union football. Maximum average and individual weights were prominently specified for the different grades of junior teams. Players needed special permission to play in lower grades having played at least two matches at a higher grade (Australian Public Schools Athletic Association [APSAA] 1907).

Player position

In rugby union, first row forwards (props and hookers) are predominantly at risk in scrums. Several studies around the world have found significantly higher incidence of SCI among forwards than in backs for rugby union and rugby league (Armour et al. 1997; Carmody et al. 2005; Orchard & Seward 1994; Secin et al. 1999; Taylor & Coolican 1987; Wetzler et al. 1996). An Australian study (Taylor & Coolican 1987) of rugby union injuries found 43% of rugby union injuries were to hookers. An Argentinean study (Secin et al. 1999) found as many as 50% and an American study (Wetzler et al. 1996) found 48% of rugby union SCIs were to hookers.

The vulnerability of hookers to SCI in scrums is particularly noted for rugby union, but they can occur in rugby league (Armour et al. 1997; Silver 1984, 1994; Wetzler et al. 1998; Williams & McKibbin 1987).

In New Zealand 1976-1995, 83% of SCIs were to forwards, which was significantly more than backs. Of known positions, 33% were hookers. The front row including the hooker accounted for 59% of SCI (Armour et al. 1997). In rugby union, back line players not involved in scrums tend to be injured more by tackling injuries, mainly while carrying the ball (Burry & Gowland 1981; Kew et al. 1991; Silver 1984, 1994; Taylor & Coolican 1987).

128 A four-year prospective study of all injuries from one professional rugby league club compared the differences in the incidence of injury between forwards and backs. Forwards had the highest rate of injury to all body sites with the exception of ankle injuries and were more likely than backs to be injured in possession of the ball or when tackling. It was concluded that this was most likely the result of their greater physical involvement in the game, both in attack and in defence (Gissane et al. 1997).

However, the risk of SCI for back line players in rugby league may be greater than that in rugby union because of the greater risk of tackling injuries for rugby league. As in rugby union, rugby league backs are more likely to be injured in tackles because they are not involved in scrum activity but are responsible for critical defensive tackling and attacking runs. One Australian study found that 40% of SCI in rugby league were to back line positions compared to only 8% in rugby union (Taylor & Coolican 1987).

Studies looking at less serious injuries for both league (Lower 1994) and schoolboy rugby union (Davidson 1987) reported fullbacks as being the most injured player position. Another study of elite rugby league and union players found most minor injuries were to forwards (mostly comprising injury to upper body) while other types of injury such as ankle and knee damage were more common to backs (Seward et al. 1993).

Several studies have found an association between anthropometric characteristics and player position (Carlson et al. 1994; Lee et al. 1997; Nicholas 1997, Quarrie et al. 1996). Front row players are usually more endomorphic than other forwards and forwards are generally larger than backs. However, coaches for elite rugby football teams in both codes are increasingly using larger mesomorphic players rather than endomorphic or ectomorphic players. Associated with this is progressively less distinction between the traditional body types associated with various playing positions (Huxley 2003).

129 Risk Exposure duration

Reducing each player’s total playing time or time spent in specific phases of play diminishes the probability of injury occurring by reducing exposure to risk and minimising player fatigue. Of particular danger is the practice of players participating in more than one full game in the same day. As seasons become extended into pre-season training and trial matches, with ever growing professional pressures for both codes, players’ injury risk exposure time is increasing (Lee et al. 2001; Silver 2001, 2002; Williams 2002). The English Professional Rugby Players Association chief executive, Damian Hopley, has warned that elite players are now becoming unable to sustain long-term careers because of the increasing number of games demanded of them inevitably leading to more injuries (BBC Sports News Online 2002).

Apart from reducing absolute playing hours, identifiably high-risk plays such as scrums should have their numbers minimised and their duration limited (NHMRC 1994). A spokesman for the ARU has related that in the last couple of years the laws of rucks and mauls have changed and this has resulted in a 40% reduction in the number of scrums. Referees are instructed to ensure that scrums are completed quickly and occur less often (Connell & Zuel 1994).

The contextual background to serious rugby football injury has been described in previous sections. Attention was then turned to the existing epidemiological evidence, established risk factors and injury prevention strategies concerning serious rugby football injuries. Sections six and seven will now turn attention to the literature for an historical and sociological perspective of the rugby codes development.

130 SECTION SIX

Section six assesses the historical development of the game and takes a sociological perspective in order to understand the origins of the game’s inherently dangerous nature. It provides perspective to current injury prevention strategies in light of past efforts to improve player safety and modify game rules to increase its entertainment value for spectators.

The historical and sociological development of rugby football and its effect on participant aggression and competitiveness

Rugby descended from the winter ‘folk-games’ that were a deeply rooted tradition in pre-industrial Britain. These games were undeniably violent and brutal affairs. Rules for ‘foote balle’ were almost non-existent and play took place over large areas and was somewhat akin to real warfare. The means of getting the ball back to the goals were not specified but carrying it, kicking it and hitting it with sticks and clubs were the most popular. Some of the players were on horseback while others carried swords, clubs and staves. Many people were injured for life or killed and it was an ideal way of settling feuds. In many cases, ambushes were set to enable private duels to be settled. Play for the day was only abandoned at sunset (Dunning & Sheard 1979; Woolgar 1999).

These games had a harmful effect on business and other national duties with villagers neglecting their responsibilities to play ‘foote balle’. The result being that it was banned by Royal decree thirty one times in three hundred years with several Kings claiming the aftermath from the game’s violence was too great a toll to be tolerated. However, these bans and criticisms did little to curb the game’s popularity until the coming of the Industrial revolution in Britain which saw it slowly lose popularity as it changed to a ‘street game’ of catch (Woolgar 1999).

The public schools of in the eighteenth and nineteenth centuries were responsible for the revival of the sport (Woolgar 1999). Around 1750 boys took up antecedents of the modern games of rugby in the public school system in England.

131 It was common for boys at public schools of the day to object to being ordered around by socially inferior teachers. Some would rebel on a regular basis, even arming themselves with guns, swords and explosives which could occasionally necessitate the army to restore peace. To help the boys release their aggressive tendencies several schools tried adapting the ‘street game’ to school sports. Three schools (Charterhouse, Westminster and Rugby) continued to develop this game but Charterhouse and Westminster had hard quadrangles. Tackled participants got abrasions and broken limbs, so it was decided to leave the ball on the ground and kick it. Rugby school stayed with their modified version of the street game because they were the only school with big green open playing fields. Thus it was playing on surfaces that largely dictated the evolution of the two games – rugby and soccer (Woolgar 1999).

The maul was the dominant feature of the Rugby school game. The main idea was for a forward to gain possession by wedging the ball between his knees, and around him the forwards of both sides packed, still erect. Scrumming down was illegal and punishable with an uppercut (Woolgar 1999). Early accounts of the football played at Rugby school indicate that the ball was relatively unimportant to the game, with the focus on scrummaging – which at the time were indiscriminate kicking matches (Dunning & Sheard 1979).

Despite opposition from earlier players, the ‘passing game’ caught on. The game became more like a group contest that provided the enjoyment of real fighting with reduced chance of serious injury. Pleasure in playing began to be derived less from brute force and more from force transformed by the use of complex skills, e.g. passing, kicking and running with the ball. The game required the development of intricate balance between force and skill, spontaneity and control, individuality and teamwork (Dunning & Sheard 1979). By 1850, the game played in the public schools began to be subject to more stringent and formal regulation with footballers required to exercise greater self-control (Dunning & Sheard 1979).

132 The game entered another stage as past students attending university began to organise their own competitions forming . Disputes quickly arose regarding some of the rules, the main area of disagreement being whether to allow ‘hacking’ which the original meeting had wanted to exclude (Dunning & Sheard 1979). In those days hacking was considered very manly and you were allowed to kick an opponent between the ankle and the knee provided it was face to face and that your opponent was not being held at the time. Another variation of hacking happened after the game when players were paired off with opponents and lined up facing each other. When the whistle went they hacked away at each other until the referee called a halt. It was considered cowardly to retreat or flinch and in order to make hacking more effective, players wore special boots with metal toe- caps.

The proponents of the carrying code were resolute that this be allowed to continue as well as tripping, much to the abhorrence of the advocates of the kicking game (Griffiths 1982; Woolgar 1999). The adherents to the carrying code broke away from the Football Association in 1871 and formed another body calling it a Union and embracing the rules applying to the game at Rugby School thereby giving the sport its present name. The union became the Rugby Football Union and the rules became the Laws (Griffiths 1982).

The clubs then found that their playing strength was continually being depleted by injuries through hacking and tripping, so much so, that these were banned as well as that of bringing down a player with a kick to the body which until then was an accepted form of tackling. These decisions created a storm of protest from the veterans of the day who were most indignant of the developments and flooded the letter pages of London’s leading newspaper The Times denouncing the effeminate fellows who had succeeded in abolishing this practice (Dunning & Sheard 1979).

133 Up until the introduction of these new laws the maul or scrummage had continued to be the cornerstone of the game taking up to twenty minutes for the ball to emerge from the scrum. There was resistance to letting the backs have the ball. They were to run with, or kick the ball, as at that time passing was not common. The new laws introduced a number of factors that changed the nature of the scrumming. First was the art of wheeling the scrum which reduced the pushing contest and made the skill of breaking away from the scrum and of dribbling the ball at the forward’s feet an important part of the game. Then came the emergence of the practice of heeling the ball back for the backs to use and finally a further change in the law which introduced the scrum at the place where any breach of the laws was not otherwise dealt with (Griffiths 1982; Woolgar 1999).

Within twenty five years after the formation of the Rugby Union, radical developments had taken place in the style of play and there was a need to regularly update the laws. The game spread rapidly with separate controlling bodies being established in Scotland, Ireland, Wales, New Zealand, Queensland, New South Wales, Rhodesia and South Africa (Griffiths 1982).

By the end of the 1880s none of the home unions would play against England as a result of dissatisfaction with the Rugby Football Union, which the English controlled. The outcome of this development was the formation of a new body, the International Rugby Board, in 1890, which was to govern the game right throughout the world (Griffiths 1982).

The inevitable pressure towards professionalism for the game of rugby was fiercely resisted from the rugby union establishment who were committed to an amateur game in which it was argued that players’ considerations would take precedents to spectators (Dunning & Sheard 1979). However, the International Rugby Board professionalised elite level rugby in 1995 (Silver 2002).

134 The code of rugby league was formerly created when the Northern Union split from the rugby union establishment in 1895 to start their own competition motivated by a division over whether players should be financially compensated for missed work time (Macklin 1962). Although the first two seasons of the new rugby league code were played under the same rules, progressively rule variations were introduced. These were largely in the interest of removing scrappy, contentious play and untidy charging and barging. From early on the need to compete for spectator support, with the ever-growing popularity of soccer, increased the need to make the game more open and exciting (Macklin 1962).

Variations included the abolishment of lineouts and punt outs, and the requirement that scrum halves retire behind the scrummage. In 1899, the play the ball rule changed and in 1904 clubs were ordered to take steps to prevent players packing down with more than three players in the front row, designed to stop barging and wheeling. In 1906, the number of players was reduced to 13 by removing two forwards to lessen ‘scrambling and destructive’ play of which the forwards were most involved (Dunning & Sheard 1979)

Despite these changes, a report of the Northern Unions first Annual General Meeting shows that the Northern Union authorities still believed the roughness of rugby league to be one of its problematic aspects (Dunning & Sheard 1979). A sub- committee had been appointed to make the game less violent, more attractive to spectators and to increase its chances of competing successfully with soccer (Dunning & Sheard 1979; Macklin 1962). Inherent violence constituted one of the main sources of spectator appeal but administrators were faced with several needs to reduce violence. Importantly, players had become valuable assets with increasing need to maintain teams at peak performance. There was also a pressure from the developing ‘civilising’ process in society to reduce violence without emasculating the game (Dunning & Sheard 1979).

135 Dunning and Sheard (1979), looking at the sociological development of rugby football, argue that there are distinguishable stages of development in the game of rugby, each characterised by more formal, more complex rules and organisation. Each stage involved the demand for increasingly orderly and restrained behaviour as the game experienced continual societal pressures that might be considered a ‘civilising’ process.

The development of rugby football has undergone a continual ‘civilising’ process. Rugby football developed consistently with the demand for greater orderliness and more civilised behaviour, characteristic of the advanced industrial societies in which it was played. Central elements of the ‘civilising process’ include the refinement of social standards regarding the control and conduct of social relations with growing social pressure for people to exercise self-control. It is inherent in this process that people must be subject to stricter self-control over violence and aggression (Dunning & Sheard 1979)

However, many studies have found increasing levels of injury accompany the development of professionalism (Dunning & Sheard 1979; Garraway et al. 2000; Nicholas 1997; Silver 2001; Williams 2002). The inevitable development of rugby football into the modern sporting codes of rugby union and league has been greatly influenced by the pressures of professionalism and the evolving attitudes to sport of society in general. Societal trends worldwide over the last century have arguably tended towards growing competitiveness, seriousness of involvement and ‘achievement-orientation’ in most modern sports (Dunning & Sheard 1979).

Players in modern games, particularly at higher levels of competition, are required to make increasing commitments to the game regarding training, preparation, and injury management with increasing financial incentives and viable playing career opportunities. Professionalism and growing competitiveness has placed increasing emphasis on strength and fitness training with considerable resources spent on maximising athletes’ performance (Garraway et al. 2000; Junge et al. 2000; Nicholas 1997; Silver 2001; Williams 2002). This has resulted in the top players

136 becoming highly paid full-time professional athletes with the intensity of training increasing, and the impact between players dramatically increased due to increased size, speed and power (Williams 2002).

It has been suggested that the increasing violence observed in some modern sports may be associated with the growing seriousness and competitiveness of modern players (Dunning & Sheard 1979; Silver 1984). As previously discussed in the section five ‘Level of competition, competitiveness and aggression’, increasing aggression and competitiveness associated with professionalism may increase the likelihood of serious injuries occurring. Concerns from administrators about unsportsmanlike behaviour, allegedly fuelled by professionalism, were raised early in Australia’s rugby league history (Cooke 1923).

Although it has been argued that professionalism encourages increased competitiveness, unsportsmanlike behaviour and in turn injury risk (Armour et al. 1997; Silver 2001; Silver & Stewart 1994) due to the shifting of playing motivations towards financial incentives, it has also influenced the games in other ways. Modern players shifting loyalties to teams (reducing player rivalry), increased number of games to play (more than a few important matches per season) and career pressures to avoid injury may act to reduce the intensity of modern play.

The growth of professionalism has seen elite players become more concerned with avoiding injury in order to maintain a long-term career and meet increasing playing season demands (Nicholas 1997; Silver 2001; Williams 2002). The accompanying growth of media coverage has also seen the reduction in acceptability and increase in accountability for unsportsmanlike behaviour.

137 Player loyalty and pride in their team are important emotive influences predicting the intensity of play. As discussed earlier in section five of the literature review, several studies have demonstrated the influence of psychological factors on sports injuries. In both rugby union and league, matches between teams from different origins (such as state of origin or international tests) are played more intensely with fights and aggressive behaviour often more common than in regular competition games. Professionalism has arguably seen player loyalties shift from team affiliation to contractual obligations (Garraway et al. 2000; Junge et al. 2000; Nicholas 1997; Silver 2001; Williams 2002).

The need to modify the games’ rules in response to societal pressures to reduce violence, improve player safety, and increase the games entertainment value has long been evident. It is worth considering then that sustained efforts to ‘improve’ the games should not be viewed as an attempt to break with the games’ tradition but rather as a continuance of their natural progression. Undoubtedly, the future popularity and ultimately survival of both codes at the elite as well as the community level depends on such adaptability as it has in the past.

138 SECTION SEVEN

Section seven explores how styles of play relevant to key risk factors for serious injury assessed earlier in chapter six, have developed over the years in both codes. These factors include the number of tacklers, impact speed, dangerous scrums and tackling technique, illegal/foul play, number of players on the field and average player strength, fitness and size.

Style of football played in modern times (post 1989) compared to that played in an earlier era (pre 1958)

Although there was undoubtedly an increase in serious injuries worldwide resulting from both codes of football through the 1970s and early 1980s, it is not clear what the rate of these injuries were in earlier times. Most studies concerning serious rugby football injuries have not been able to review prior to professionalism in the 1960s. Efforts to compare injury rates with those in the first half of the century are confounded by advances in injury recognition and management together with developments in injury surveillance systems.

However, limited evidence does suggest that the previous incarnations of the games in England were dangerous. Very early records on rugby union players from 1890-91 to 1892-3 reveal that were an astounding 71 deaths and 366 broken legs, arms, collar bones and other injuries. Evidence exists that rugby league was also a dangerous game after having split from rugby union around the turn of the century. At its first general meeting, the Northern Union (the first rugby league association) records several compensation payouts for families of men who received fatal injuries while playing. The perceived risk for players was reflected in the difficulty the Northern Union had finding insurers for its players (Dunning & Sheard 1979).

139 In recent years, while safety awareness has improved, a number of injury prevention strategies have been introduced and foul play has become a greater priority, other developing factors may have acted to adversely affect the risk of injury for rugby footballers.

A retrospective study analysing video footage of elite Australian rules football games and statistics on player size back to 1961, found the average game tempo had increased along with player size. It concluded that this had probably contributed to increasing player injuries (Norton et al. 1999). A comparable pattern in the development of rugby football codes may have had a similar affect on injury rates over the years (i.e. advances in injury prevention offset by faster games and bigger players).

Number of tacklers

Multiple tacklers targeting the ball carrier has become an increasingly emphasised part of modern game tactics and training. The objective is to optimise the ability to ‘kill the ball’ and stop attacking play from continuing. In practice, this means stopping the ball carrier from unloading the ball to a team-mate as the tackle is being implemented. This is often achieved by ‘wrapping a player up’ (i.e. one tackler goes down low to stop forward motion and at least one other targets the mid to upper body to stop last minute passing). An increase in the number of tacklers has been implicated as a possible factor in the apparent relative increase in the occurrence of tackle related serious injury (Edgar 1995, Scher 1991c; Silver & Stewart 1994).

140 Impact force

The modern game of rugby football is played at a faster tempo than in earlier times (Heads 2000; Silver 2001, 2002; Woolgar 1999). Professionalism demands increasing competitiveness, size, speed and power from players (Garraway et al. 2000; Nicholas 1997; Silver 2001, 2002; Williams 2002).

A study analysing video footage of elite Australian rules football since 1961 (Norton et al. 1999) concluded the average game tempo has increased possibly contributing to increased player injuries. Modern games were judged to be faster and more intense but with longer rest periods which facilitate shorter high intensity play periods. These factors were amplified by increases in the number of interchange players. Faster game speed increases the risk of high velocity player impacts. The study concluded that over the last 30-40 years, the kinetic energy of players has increased due to an increase in the mass of players, and an increase in the speed of the game. This may account for increased injury rates observed in AFL despite the introduction of preventive measures, specialist rehabilitative techniques and medical expertise.

Dangerous scrums and tackling technique

The ‘power scrum’ developed in the modern rugby union game increases the impulsive force experienced by players in a scrum and is more likely to cause collapse, wheeling and ‘popping’. It thus poses greater risk of serious injury than in earlier styles of scrum formation. The ‘power scrum’ involves emphasis and intensive training to get down lower, pushing hard in unison and generally focusing on scrummaging as a tactic to dominate ball possession. The development of this technique is argued to have been led by rugby law changes in 1963 that resulted in the desirability of longer scrums (Quarrie et al. 2002).

141 Rugby union experienced increased incidence of injury related to scrum activity after 1970, while in rugby league, where the scrum was ‘depowered’ through its importance being de-emphasised, there was a reduction in the incidence of scrum related injuries. Following determined efforts in rugby union to modify scrummaging rules in the interest of player safety in 1987, there was a significant reduction in serious scrum related injuries occurring.

Available archival evidence regarding the style, training of players, and rules concerning scrummaging in earlier games suggest little concern for player safety (Cooke 1923; Anon. How to play rugby 1910; NSWRFL 1913; NSWRU 1926; Wakelam 1937). Official discussions of rules provided by the Australian Rugby Football League board of Control (NSWRU 1926) in the 1930s illustrates that many recommended practices and terminology used in the games' early existence would not be suitable today. For example, the board advised: “It is permissible for a player at the rear of a scrum to detach himself and pick up the ball to bore his way through the pack, or for a player in the scrum to hold the ball with his knees, or for a player on either side in the scrum to screw the pack and take the ball with them.” (p. 23)

Early players’ guides for rugby union provide descriptions and photographs of advised scrummaging technique showing a dangerous emphasis on getting down as low as possible with legs extended far back (see figure 9). Several playing guides stretching as far back as the 1930s, commonly emphasised the importance of tackling ‘as hard and low as you can’ allowing a smaller player to tackle a larger one and increase the chances of an incomplete tackle knocking the ball carrier off balance. Considerable attention was given to safe effective tackling technique. (Anon. How to play rugby1910; Wakelam 1937). However, there was an awareness of the safety value of tackling technique, such as appropriate head placement in anticipation of the direction of ball carriers’ fall to the ground (see figures 10-11).

142 Figure 9: Young players practicing scrummaging against scrum machine in the 1930s. Note extremely low position taken that would now be considered dangerous. (Wakelam 1937)

143 Figures 10 and 11: Early conceptions of ideal tackling technique Illegal/foul play (Anon. How to play rugby 1910)

Illegal/foul play

As early as 1923, a paper read at a rugby football league conference discussed the problem of unsportsmanlike behaviour and disregard of rules. It was speculated that this was fuelled by the influence of professionalism (Cooke, 1923). Some studies have partly attributed observed increases in serious injury to a growing aggressive and violent attitude among players as football competitiveness increases (Garraway, Lee, Hutton, Russell, & Macleod, 2000; Silver, 1979; Silver, 1994; Williams, 2002). A study of Australian footballers found 26% of SCI were due to illegal play (Taylor and Coolican, 1987). However, both codes have paid growing attention to issues of foul play throughout their history.

144 The games have experienced a gradual process of ‘civilisation’ since their inception as rather brutal and lawless ‘folk’ pastimes. Many accounts exist, throughout the history of both codes, of the commonplace nature of what would now clearly be considered brutal foul play (Dunning & Sheard 1979; Heads 2000). The analysis of archival film and video footage confirmed that many foul play incidents such as head high tackles and brawls were relatively common in games pre 1990. During the 1970s, rugby union was criticised by the media for an increase in aggressive and deliberately dangerous play (Sharp et al. 2001).

Unprecedented attention to foul play was assured when the games became widely televised, particularly for rugby league. Many play infringements were now clearly visible to both officials and spectators and were no longer tolerated by the viewing public (Heads 2000). The application of medicolegal principles to foul play in rugby union, where the rule of law on the field should reflect the wider law, has seen increasing numbers of civil law suits brought against players accused of deliberate foul play (Grayson 1996). It is thus likely that foul play, just as tackling infringements, like head high tackles, are no more likely in modern games than in earlier times despite possibly growing levels of competitiveness in modern players.

Number of players

Over the last century, several variations in the number of players per team have been made from rugby union’s original fifteen players a side. Pressures to reduce player numbers in rugby league were evident early in the history of the Northern Union. By 1902, it was argued that 15 players were excessive resulting in mauling, scrambling, destructive and dull games. Many advocated a 12-a-side competition with several games being played under this variation. A compromise was reached in the 1906-7 season where it was decided in the interest of open play to settle on 13 players a side (Macklin 1962).

145 Rugby union has more recently experimented with different competitions such as the ‘sevens’ tournaments where there is a reduction of players in each team on the field. As discussed earlier in the literature review these examples provide untapped opportunities to evaluate whether reducing player numbers on the field decreases the risk of players incurring serious injury.

Players strength, fitness and size

Demands of professionalism require increasing fitness, size, speed and power (Garraway et al. 2000; Huxley 2003; Nicholas 1997; Silver 2001, 2002; Williams 2002). These factors increase the forces exerted in impacts in modern games. Because size and fitness provide limited protection against serious injury, the increased forces in impacts may actually be increasing the overall risk of serious injury (Armour et al. 1997; Kew et al. 1991; Silver 2001; Silver & Stewart 1994). Silver (2001) has drawn an excellent analogy with vehicle crash injuries where forces involved and speed of deceleration are major factors in determining the severity of injury.

Although selected statistics note the increased participation in sport and recreational activities in recent decades (ABS 2002a) others have noted the increasingly sedentary nature of modern lifestyles (Cordain et al. 1998; Egger et al. 2001; Powell & Balie 1994). There is a lack of suitable evidence available to conclude whether any trends towards healthier lifestyles, better nutrition and increasing achievement orientation among the general public in modern times suggests that increases observed in average player size and fitness at the elite level might also be inferred to community level players.

146 CHAPTER TWO SUMMARY

Recognition of injury as an important public health issue during the 1980s led to the establishment of several injury surveillance and prevention initiatives. This recognition is particularly evident in the road safety sector, and in moves towards the establishment of ongoing infrastructure for injury prevention within health agencies.

In the last decade, sporting injury and its prevention have received increasing attention as a public health issue with considerable growth in Australia's sporting injury prevention infrastructure. However, the difficultly for sporting injury prevention is its multi-disciplinary and inter-sectoral nature with limited support available for injury intervention studies (unlike research in the pharmaceuticals industry). This impedes the ability to develop an effective political constituency to lobby for sufficient resources.

Appropriate injury prevention strategies develop from the establishment of injury trends and associated risk factors. Monitoring epidemiological data is needed also to evaluate strategies that have been implemented. While studies incorporating time based exposure data are optimal, they are rare because of the time consuming, complex, and arduous nature of the data collection process. Depending on the purpose of research, alternative designs may suffice study objectives.

Lett, Kobusingye, and Sethi (2002) argue that injury specialists have failed to successfully persuade policy makers and the community that injuries are preventable partly due to the lack of a unified understanding of injury control. The authors suggest that the two most important models utilised in injury control, Haddon's Matrix and the Public Health Approach, should be combined to provide a unified framework for understanding injury prevention.

147 To provide a unified framework for understanding serious injury and fatality prevention issues in rugby football, a theoretical framework of risk factors using the Haddon's Matrix and the Public Health Approach has been developed in this study.

Sport and recreational activities, particularly rugby football codes, have continued to contribute disturbingly to annual SCI and TBI figures in Australia. While treatment of severe SCI and TBI has progressed over the decades, 'cures' to these debilitating injuries still elude researchers and clinicians. The social and economic costs of these injuries is an enormous burden to Australian society.

In response to an increasing incidence of SCI and TBI since the 1970s, strong medical representation in several countries influenced administrative bodies such as the International Rugby Football Board to accept changes to the rules of the game. Rule changes and education campaigns aimed at preventing injury in football have traditionally focussed on modifications for junior play. However, the success of these has seen their extension to senior levels of the games.

As might be expected in contact sport, the highest risk plays are those involving forceful impacts. Common causal patterns of serious injury have been observed across rugby football playing countries around the world. Rugby union players appear to be most at risk of SCI in a scrum, maul or ruck whereas rugby league players are more likely to receive a SCI (or other serious injury) in a tackle. This reflects the different emphasis and resulting forces exerted in different aspects of each of the two rugby codes. Although serious head injuries can occur in scrums through head clashes of front row players, this is relatively rare and injury is more likely to occur in a tackling situation from the player’s head hitting a part of another player’s body or from impact with the ground.

148 The critical factors in the mechanism of SCI and TBI tackling injuries include force of impact, direction of impact, number of tacklers, and level(s) of impact. Maul and ruck situations have been identified as particularly dangerous for rugby union players. Spinal injuries occur through forced flexion (often involving rotation) to the ball carrier’s neck or to a player at the bottom of the ruck or maul. Serious injury may also occur due to players charging into the ruck or maul. It is often difficult to determine whether certain rugby union injuries occur in a tackle or in the ensuing ruck or maul. This is because players may be injured on impact, when there is a pile up, or on hitting the ground.

The power generated in an initial scrum engagement is potentially greater than is safe for players’ spine. Scrum related injuries occur at the engagement of the two packs or when the front rows collapse. They may also occur when a player in the front row, usually the hooker, is ‘popped’ out of the scrum. Scrum collapse may occur because of instabilities caused by a mismatch of skill, experience and strength. However, it may be intentionally collapsed despite being an illegal tactic. In Australia, collision rather than collapse is the primary mechanism of most scrum injuries.

Several interrelated factors specific to each player may influence vulnerability to serious injury. These include body size (weight and height), body type, physical preparation, age, experience, and skill. Other factors that might be considered include position played and the level of competition played. The importance of appropriate fitness, preparedness and suitable body type for playing both rugby league and rugby union in order to reduce the risk of injury has been identified extensively in the literature

149 In rugby football protective equipment can be effective for protection against minor injuries, but there is much less evidence concerning the effectiveness of protective equipment for serious injuries to the head and cervical spinal cord.

Appropriate management of suspected serious injury before hospitalisation can markedly improve recovery outcomes and prevent secondary injuries from occurring.

Dunning and Sheard (1979), looking at the sociological development of rugby football, argue that there are distinguishable stages of development in the game of rugby, each characterised by more formal, more complex rules and organisation. Each stage involved the demand for increasingly orderly and restrained behaviour as the game experienced continual societal pressures that might be considered a ‘civilising’ process.

Despite this process many studies have found increasing levels of injury accompany the development of professionalism. The inevitable development of rugby football into the modern sporting codes of rugby union and league has been greatly influenced by the pressures of professionalism and the evolving attitudes to sport of society in general. Societal trends worldwide over the last century have arguably tended towards growing competitiveness, seriousness of involvement and ‘achievement-orientation’ in most modern sports.

In recent years, while safety awareness has improved, a number of injury prevention strategies have been introduced and foul play has become a greater priority, other developing factors may have acted to adversely affect the risk of injury for rugby footballers. The increase in average game tempo, impact speed, number of tacklers, head on tackles, along with player size over the years have potentially contributed to player injuries in rugby football codes (i.e. advances in injury prevention offset by faster games and bigger players).

150 However, the growth of professionalism has also seen elite players becoming more concerned with avoiding injury in order to maintain a long-term career and meet increasing playing season demands. Additionally, the accompanying growth of media coverage has also seen the reduction in acceptability and increase in accountability for unsportsmanlike behaviour.

The need to modify the games’ rules in response to societal pressures to reduce violence, improve player safety, and increase the games entertainment value has long been evident and part of the tradition of the codes respective developments. Undoubtedly, the future popularity and ultimately survival of both codes at the elite as well as the community level depends on such adaptability as it has in the past.

151 CHAPTER THREE

METHODOLOGY

This chapter outlines the research methods used in this study to investigate rugby football spinal cord and brain injuries resulting in permanent deficit and death in NSW. Chapter Three, Section One provides the definition of these injuries followed by an overview of the study design. Section Two gives a detailed description of the data sources used and Section Three outlines the statistical analysis methods used.

SECTION ONE

Injury definitions of data included in study population

Spinal cord injury (SCI)

SCI is defined here as an injury to the cervical or upper thoracic spinal cord resulting in permanent neurological deficit that entails permanent and severe restrictions to the functioning of an injured person, that is, a persisting case of spinal injury. These injuries are graded from A to C on the American Spinal Injury Association (ASIA) impairment scale (see Appendix A for details). The ASIA impairment scale ranges from spinal injuries causing complete restriction with no preservation of sensory or motor function to incomplete restriction with motor function preserved below the neurological level, and more than half of the key muscles below this level having a muscle grade less than 3. Injuries graded A to C leave motor power below the level of spinal cord lesion insufficient for most practical uses therefore rendering the patient permanently disabled.

152 Head and brain injury

Head injury, as defined by the Sports Medicine Australia (formerly the Australian Sports Medicine Foundation), covers a wide range of injuries from lacerations to brain damage (McCrory & Dicker 1992).

Traumatic brain injury (TBI) is an insult to the brain from an external force causing permanent or temporary neurological dysfunction (including impairments of cognitive, physical, and psychosocial functions) (Khan et al. 2003).

Some apparently severe TBI may resolve precluding coverage by the NSWSIC compensation scheme. Such injuries were not included in the thesis. Head injuries described in this thesis focuses on permanently disabling TBI involving severe loss of normal functioning in accordance with the requirements for NSWSIC compensation (i.e. those involving more than 35% permanent loss of use of a body part).

Fatality

Injuries leading to death within one year of the accident date were included in this study. Injuries leading to pronounced death on the field would not have been evacuated to spinal units. However, these cases would have been reported to the NSWSIC if playing for a member club.

153 Study design

The design of the studies main component is based on the use of retrospective injury surveillance data. The availability of this data provided an invaluable opportunity and reporting imperative for descriptive and analytical longitudinal study of rugby football injuries.

The choice of design was strongly influenced by the rare and catastrophic event nature of SCI and TBI in rugby football players. While SCI and TBI are catastrophic outcomes, they are rare events occurring respectively in 1.4 (O'Connor 2001a) and 141 (O'Connor 2002b) per 100,000 of the total Australian population per year (TBI notably ten times greater than SCI).

The descriptive analytical survey using case surveillance data study design was appropriate in this study for the objective of describing the number and rate of injuries, and examining possible predictive variables for an outcome that is rare. A cohort study would have required a follow up period of 15 years to find approximately 30 cases in each code, while a case control study was inappropriate due to recall bias of both case and control subjects.

The other major component of the thesis - the descriptive analytical survey using archival film and video footage of rugby football games was primarily interested in studying the development of playing styles over the last century and was subsequently tied to the retrospective availability of footage. Similarly the comparison of early and modern player characteristics were subject to archival data availability and needed to be collected retrospectively. Attempts to systematically compile prospective video evidence of rugby play including catastrophic injury were beyond the scope of the study considering the rare nature of events.

154 The suitability of various levels of research design for the research questions posed are summarised in Table 4.

Table 4: Levels of research design and suitability to research questions posed. Study design options Suitability Randomised control trial/Experimental Participants could not ethically be randomly studies assigned into a treatment group including exposure factors that are considered to increase risk for serious injuries. Similarly, difficulties with introducing unproven potentially protective measures Cohort or Case Control studies A cohort design is not suitable for rare outcomes due to time and resource constraints. A case-control is suitable for events occurring in 2% or less of the community. However, serious sporting spinal injuries occur too rarely for concurrent enrolment of controls as each case is identified. Recall of exposure would require cases to consider exposure factors up to 16 years previous. Cross sectional/longitudinal analytical This would be the design of choice when descriptive studies examining whether a problem exists. It identifies whether a further prospective study is warranted. It also allows for the immediate description and exploration of old and new cases of the rare event, and quasi-causal analysis. Case study/field research Provides important first-line information about a disease/outcome that is usually rare or not yet reported. It describes the factors present but it can not consider these as causal factors with any assurance.

155 Study group

Various data sources have been used in an attempt to describe the frequency of serious injury events, associated risk factors and evidence of change in their incidence rate. The main sources of data examined were archival material, video footage, sports injury compensation cases and medical records.

The target population were all league and rugby players in NSW of all ages from 1984 to 1999. To establish the frequency of the serious injury events under study, the study group included all players admitted to major SCI units in NSW with permanent SCI as well as all brain injuries and fatalities reported to the NSWSIC during the study period. As all permanent injuries from the study group's data sources were included, the level of generalisation of the findings to the target population, league and rugby players of all ages is believed to be high (Figure 12).

The study group for the descriptive examination of the codes of practice included video footage of elite level games over the last century, printed rules, discussions and player statistics found in archival collections of football paraphernalia. These results will be generalised to players of all ages as the target population.

156 Figure 12: Study group chart Target population SCI and brain injury Rules and styles of play

All school, club and professional rugby league and union players of all ages in NSW

Serious permanent spinal, head injuries and fatalities in NSW resulting from Study population Development of game rules and rugby league or union football 1984-1999 style of play since inception

95% of acute rugby 70% of acute rugby league football SCI in NSW football brain injury and fatality in captured (remaining NSW captured (remaining 30% 5% transferred not covered in NSWSIC scheme, interstate, immediate non stabilised injury or alternate fatality or managed in compensation sought) paediatric facility)

NSW Spinal Unit Archival and current Video footage of archival NSW Sporting Injuries Study group acute admissions Committee compensation game rules/discussions and recent games cases

97.6% of acute SCI 67% of rugby league and 100% of brain injury and admissions in NSW 45% of rugby union SCI fatalities compensated by Spinal Units captured cases verified (remaining NSWSIC captured (remaining 2.4% 33% and 55% respectively missing medical not covered by NSWSIC) records)

Archival written material All video footage of elite NSW Spinal Unit NSW Sporting identified in NSW State games 1920s to 1990s medical records Injuries Committee Verification library archive search and identified from Screensound benefits paid case current versions of and Sports Recording records rulebooks Services databases 157 SECTION TWO

Sources of data

NSW spinal units

There have been three acute admission spinal units in NSW over the study period (Royal North Shore Hospital, Prince Henry Hospital and Prince of Wales Hospital). These units will be referred to as New South Wales spinal units (NSWSU). According to personal communications with the acting heads of these units, more than 95% of patients with spinal injuries that occur in NSW are admitted to these specialty units. The remaining 5% of cases are taken directly to the Sydney Morgue or are air lifted to spinal units in other states if they occur close to state borders (O'Connor, 2002; National Injury Surveillance Unit, 9 November 2002, pers. comm.; Dr Rutkowski,. Royal North Shore Hospital spinal unit, 16 March 1998, pers. comm.; Dr Engel, Prince Henry Hospital spinal unit, 20 March 1998, pers. comm.; Dr Epps, Sydney Children's Hospital, 8 April 2003, pers. comm.; Dr Waugh, New Children's Hospital Westmead, 20 December 2002, pers. comm.).

These estimates are supported in studies of the ASCIR data collection system which draws on reporting from spinal units across Australia (Berry et al. 2006). Admission logbooks from Royal North Shore and Prince Henry hospital spinal units were examined for all acute sporting and recreational related spinal admissions from 1984 to 1999. In 1998, the primary function of Prince Henry Hospital spinal unit focused on rehabilitation with the Prince of Wales Hospital becoming the initial transfer point for acute SCI cases within the catchment area, attested by a review of the Prince of Wales spinal unit logbooks for 1998-1999.

158 The logbook counts of SCI injury admissions for the NSWSU included those acute SCI admissions with and without neurological deficits. Excluded were admissions relating to physical assault, accidental falls from buildings, trees, or any other structure unless clearly related to a sport or recreational activity. Motor vehicle accidents, motorcycle and pedestrian accidents or any patients transferred from overseas were also excluded.

The medical records for rugby football acute SCI admissions were reviewed to determine the nature of any neurological deficit at the time of discharge from hospital and the circumstances of the injury in these activities. The data from all the spinal units were combined. The exclusion of non-acute SCI admissions prevented double counting of the prevalent SCI population frequently admitted for treatment and rehabilitation. In addition, all cases were verified using initials and birth dates to ensure there were no patients counted twice.

A coding sheet was developed to collect basic demographics, the date of the accident, the level of injury, the permanence of the condition, the details of the sport/activity and the location of accident (Sport and recreational spinal injuries data collection form – Appendix C). The author collected the data during site visits to the medical records departments of each hospital. Detailed causes of the injuries were also sought from the records when available with information supplemented from NSWSIC records (see next section – NSW Sporting Injuries Committee Insurance).

The review process of the medical records included all such patients with a spinal injury whether or not they had suffered an injury resulting in a neurological deficit. The purpose of reviewing the files of all patients was to include the ‘near misses’ of SCI – patients who were admitted to the spinal unit but discharged without any permanent disability within grades A to C on the ASIA impairment scale.

159 Due to the retrospective nature of this study 2.4% of the sought files were not available. Based on the details available from the log book admissions, these cases were not discernibly different from those located.

Data from the NSWSU were aggregated as no comparisons were made of the admitting hospital as an independent variable.

NSW Sporting Injuries Committee Insurance

NSW Sporting Injuries Committee’s (NSWSIC) Insurance Scheme data were used to validate the number of cases of injury identified from the NSWSU admissions and provide additional details about the circumstances of injury and potential risk factors. Their data were also used to assess the level/severity of brain injuries and fatalities in NSW rugby league and union football players.

This Scheme provides financial compensation to insured clubs for injured players suffering permanent impairment from injuries that occurred during sports participation. All claimants have detailed medical, psychological, occupational assessments recording the type of injury, consequent permanent condition and including the circumstances of the injury reported from eyewitness and self-reports. The medical panel of the NSWSIC assessed all cases of compensation for permanent SCI and brain injuries for the permanency and percentage loss of function.

Financial compensation is only processed once a player’s permanent impairment has stabilised. All head injuries require detailed brain injury assessments that include cognitive assessments of long term functioning. Only permanent and disabling brain injuries that had clearly been attributed to injury associated with sporting activities, and that complied with the NSWSIC definition as permanent loss of mental capacity and/or permanent physical impairment, involving severe loss of normal functioning in accordance with the requirements for NSWSIC compensation (i.e., those involving more than 35% permanent loss of use of a body part), were

160 included in this study. Information collected by the NSWSIC included eyewitness reports, case notes, pathology and radiology reports, neuropsychological assessments and post mortem findings. The NSWSIC eyewitness reports were often detailed and provided numerous quotes that were used in the results to illustrate in detail common accounts of serious football injury events.

Rugby league case files from the NSWSIC were reviewed from 1984 to 1999. All rugby union case files were reviewed from 1984 to 1989 only because virtually all (99%) of rugby union clubs sought private insurance arrangements in 1990. All case files for these periods were available for analysis. The NSWSIC case files for all fatalities, brain, and spinal injuries were examined for demographics, date and location of accident, level of injury, resulting condition, causal mechanism, team, grade, position and time of game. The data were collected using a coding sheet designed for this purpose and used by one reviewer and the author during site visits to the offices of the NSWSIC where files are held (NSWSIC benefits paid cases data collection form- Appendix C).

It is estimated that 70% of rugby football brain injuries and fatalities in NSW from 1984 to 1999 were compensated by the NSWSIC. This estimate is based on the proportion NSWSIC membership contributed to the total populations of rugby football players in NSW over this period (NSWSIC Membership figures 1984-1999 – Appendix D). It allows for a liberal estimate of 10% loss to non stabilised injuries or cases where alternate compensation was sought (Personal Communication – Jon Anderson CEO NSWSIC March 2003)

Cases from the NSWSU were matched for date of birth, accident date, and patients’ initials. Only the initials and dates of births of patients recorded by the NSWSU were used to facilitate validation with NSWSIC cases, preserving the confidentiality of the medical information acquired. This method identified 67% of rugby league cases and 45% of the rugby union cases treated by NSWSU for acute SCI between 1984 and 1999. The remaining 33% and 55% respectively were

161 playing in teams not covered by the NSWSIC. Two cases where NSWSIC paid benefits to rugby league claimants in 1996 and 1997 that were not identified in the NSWSU cases were included in the analysis due to their associated brain injury or fatality leading to management in non-spinal unit facilities. Table 5 provides a summary of the availability of injury data from each source and code.

Table 5: Availability of data by source, injury and code

Spinal cord injuries Brain injuries and fatalities Sources Rugby league Rugby union Rugby league Rugby union NSWSU 1984-1999 1984-1999 NA NA NSWSIC 1980-1999 1980-1989 1980-1999 1980-1989

Rules of play

Identification of the literature

The literature was reviewed for risk factors associated with rugby football injuries, the history of both codes, their sociological development and changes in the rules of play. Literature was also searched for relevant sporting injury and public health articles. An online search using Medline, the National Sport Information Centre literature database, University of New South Wales, and NSW State library literature databases was made using the following keywords: Rugby, Union, League, Football, Sport (domain search grouping), Injury, Injuries, Spine, Spinal, Head, Brain, Fatal, Fatality, Fatalities (injury search grouping) Australia, NSW, New South Wales (geographical search grouping) Injury Prevention, and Public Health (discipline search grouping). There were no language or year restrictions made in these searches. Articles identified during the search were examined for other references not included in online databases.

Online searches of football related sites, including official club homepages were examined for player size data. Sports archive collections held at the Mitchell Library, State Library and NSW Library were reviewed. The review focused on accessing all available past versions of game rules, player size data, annual club

162 reports, game/tour souvenir booklets, and ‘how to play’ guides. The player details for teams from pre 1960 were obtained from the sporting archive collections at the NSW State Library in football magazines, game pamphlets and club reports. Comparisons using teams from 1999 to 2000 were used as teams from the post 1989 period had limited available player statistics. Modern player statistics for 1999 to 2000 teams were ultimately sourced from rugby football websites of similar teams, i.e. elite club, state and international level players whose details were gathered pre 1960.

Archival film and video footage

The Screensound, formerly known as National Film and Television Archives, archival video database was searched for footage of rugby league and union games throughout the 20th century. Keywords rugby, football, league, and union were used to elicit relevant titles. Excluded from the search were television news segments, as these contain only seconds of actual playing footage. Any archival footage that was football related but clearly did not contain actual playing footage was similarly excluded. Most (68%) of the pre 1970s footage came from Cinesound and Movie-tone newsreels. Footage of post 1970s games was obtained from Sports Recording Services, which hold footage for many elite level games in both union and league in the 1990s.

A list of 304 titles was generated from the above mentioned sources. Videos had to be ordered from storage in the national archive in Canberra and were often on or not located on the first attempt. Therefore, videos not obtained from the initial order were requested on three occasions over six months in an attempt to access all titles. Of a possible 304 titles, 208 titles were obtained and reviewed comprising close to 30 hours of footage and 4350 analysed incidents of contact between players. Additional footage of forty games from 1990 to 2000 was obtained from Sports Recording Services.

163 Five countries have dominated as rugby code playing nations over the century. Games reviewed were all between the same five countries facilitating comparisons over time. Teams in the games reviewed over the entire period were all elite internationals or Australian state teams playing rugby league or union (Australia, England, South Africa, New Zealand, New South Wales, Queensland) at the same level of competition. Reviews of rugby league and union were made separately due to their different playing styles and rule development.

The representativeness of this source of material must be considered as footage used mostly comprised of highlights focusing on the types of play considered to be the most exciting parts of the game for audience appeal. Highlights are more likely to include tackling plays in which ‘tries’ are scored and where position advancing runs are made or critically defended.

However, the fact that the highlight footage used in the study mostly covered scoring and exciting plays was of value as these activities are integral to the game and involve high intensity play that is often associated with injury (Garraway et al. 1991; Kew et al, 1991; Roux et al. 1987; Silver, 2001; Silver, 2002; Orchard & Seward, 1994; Williams, 2002).

A data collection tool was developed to describe and compare the styles of play observed from the game footage obtained (Video incident review form and coding manual – Appendix E). Details were recorded of each incident that involved contact between players such as a tackle, scrum, lineout, or ruck and the number of players involved in the incident. Scrums were categorised as normal, wheeling, popping, or collapsing (Appendix E) while tackles were categorised in terms of the level and direction of the first three impacts to an individual ball carrier (see below for brief description).

164 Ball carrying style Two hands = Ball held in front of and away from body with two hands

Swinging = Ball held to side with swinging movement

Waist = Ball held at waist level

Chest = Ball held at chest level

Level of impact 1st tacklers level of impact – Recording 1st tacklers first committed point of impact.

2nd tacklers level of impact – Recording 2nd tacklers first committed point of impact.

3rd tacklers level of impact – Recording 3rd tacklers first committed point of impact.

(Refer to Figure 1 in Appendix E for definitions of level of impact scale i.e. legs/waist/chest/head-neck levels marked on human figure.)

Direction of impact 1st tacklers direction of impact – Recording direction of 1st impact.

2nd tacklers direction of impact – Recording direction of 2nd impact

3rd tacklers direction of impact – Recording direction of 3rd impact

(Refer to Figure 2 in Appendix E for definitions of direction of impact scale i.e clock face.)

One research assistant was trained by the author (i.e. the thesis candidate) to define the type of involvement, the level/direction of impacts, the number of players involved in a contact incident and the method of data entry onto the recording tool (Appendix E). The author and the research assistant completed three pilot sessions allowing discussion of definitions and fine-tuning modifications to the recording tool. The research assistant who was blind to the research hypotheses completed all final footage reviews. Intra-rater and inter-rater reliability was tested for twenty- seven observations. Spearman’s Rho correlations ranged between 0.854 to 0.905 for inter-rater reliability and 0.921 to 0.902 for intra-rater reliability (Table 6 and 7). Intra-rater reliability was tested between the initial and second ratings, 90 days

165 apart by the research assistant, in two randomly selected games containing 27 recorded contact incidents. Inter-rater reliability between the research assistant and the author was tested using randomly selected games containing 38 recorded contact incidents.

Table 6: Intra-rater reliability of recording tool a

Number of tacklers Correlation Coefficient .917 Sig. P<0.001 Ball carrying style Correlation Coefficient .902 Sig. P<0.001 Level of first impact Correlation Coefficient .921 Sig. P<0.001 a Spearman's rho (n= 27)

Table 7: Inter-rater reliability of recording tool a

Number of tacklers Correlation Coefficient .854 Sig. P<0.001 Ball carrying style Correlation Coefficient .860 Sig. P<0.001 Level of first impact Correlation Coefficient .905 Sig. P<0.001 a Spearman's rho (n= 38)

The content validity of the recording tool was judged suitable by a sports research team comprised of Professor James Lawson, Dr Stephen Wilson and the author. The recording tool considered ball carrying style, level and direction of impact, and number of players involved in each player contact incident observed. Details regarding scrums were also recorded. Based on findings in the literature, the factors assessed were judged to critically influence the risk of serious injury. However, one clearly identifiable weakness in the recording tool was the inability to quantify player velocity when tackled.

166 The final coding sheet was examined for face validity and judged appropriate by the research team, Professor James Lawson, Dr Stephen Wilson and the author. The coding sheet reflected the defined concepts under study with the appropriate number of categories for each contact incident. Several pilot tests identified the need to reduce the number of ranks used by some parameters, where this detail could not reliably be determined.

Only one other analogous study (Norton et al. 1999) could be found that allowed comparison of methodology with the archival film and video footage component of the present study. However, this study (Norton et al. 1999) was located only after completing the archival film and video footage analysis in the present study. The comparable study evaluated video recordings of elite Australian Football League (AFL) football games to derive and compare data over a 36 year period on major components of the games structure. This included total playing/non playing time, time spent in specific activities, ball velocity, and player height and mass.

These descriptive data were used to assess the physiological implications and risk of injury in the modern AFL game. Emphasis was placed on assessing the physiological demands on players and how this has changed over several decades to assist coaching strategies. This is opposed to the present study which is primarily concerned with the implications of changing playing style for serious injury risk.

Four complete elite Victorian Football League/AFL games, one from each decade 1960 to 1990, were assessed by a judge who timed the duration of specified activities. An estimation was also made of the distance travelled by the ball relative to time elapsed, in order to estimate the velocity of the ball which would reflect game speed. Official records from AFL House were used to derive player size data. Descriptive statistics were used to analyse most of the data (Norton et al. 1999).

167 The main differences in approach taken in the AFL study (Norton et al. 1999) was that it utilised several measures involving time that allowed inferences to be made about game speed and physiological demands on players that were not possible in this study. These measures were of critical importance to the AFL study (Norton et al. 1999) because its main objective was to assess changing physiological demands of the game where as this study focuses on issues directly relevant to serious injury risk.

The AFL study (Norton et al. 1999) limited its analysis to a handful of complete games instead of using a random sample of numerous complete and incomplete games. While the analysis of complete games is ideal, it was not always possible in this study which analysed hundreds of hours of archival footage stretching back to the 1930s.

Several methodological limitations were acknowledged by the AFL study authors this included that only four videos were analysed leading to bias because weather and tactics were specific to games, the judges rating system was subjective, and that ball velocity was a restricted measure of game speed (Norton et al. 1999).

Interestingly, the authors of the AFL study arrived at similar conclusions to the present study regarding the significance of changes in playing style and player size from the data collected i.e. increased game speed and player size increases the risk of injury (see literature review for details – Chapter Two, Sections Six to Eight).

168 SECTION THREE

Statistical analysis

All statistical analysis was performed using SPSS for Windows, Version 10.0.7 (SPSS Inc, Chicago, USA), Microsoft EXCEL, Version 2000 (Microsoft Corporation, New York, USA), and EPINFO, Version 6 (Centers for Disease Control and Prevention, USA).

Incidence rate analysis

For SCI, the numerator was the NSWSU SCI cases, stratified by code of play. The denominator was the number of participants in rugby league and union football in NSW each year of the study period, and was estimated from figures provided by the NSW rugby league and rugby union associations (personal communication; NSW Rugby Union & NSW Rugby League, December 2002). Records of participants provided by the associations from the 1980s in both codes were inconsistently collected so participant figures have been extrapolated where absolute numbers were not available.

For brain injuries and fatalities, the numerator was the NSWSIC compensated brain injury and fatality cases. The denominator was the membership of rugby league and union football players in the NSWSIC Insurance Scheme each year of the study period and was ascertained from figures published in the NSWSIC annual reports (NSWSIC annual reports, 1984-1999). The NSWSIC membership of rugby league players accounts for upwards of 80% of the player population estimated by the NSW Rugby League association.

169 The participant populations for the miscellaneous sports addressed in the overview section of the results, i.e. water-skiing, cycling, surfing, snow skiing and trail- bike/pro-motorbike sports for NSW from 1984 to 1994, could not be reliably estimated. The most reliable sport and recreational participant figures for NSW are available from the Australian Bureau of Statistics (ABS) (ABS Cat No. 4177.0; ABS Cat No. 6285.0) and the Exercise, Recreation and Sport Survey (Dale & Ford 2002) funded by the Australian Sports Commission and State/Territory Departments of Sport and Recreation (Appendix G). However, these sources do not go back past 1996, only include adult participants, and often only specify conglomerated category activities, for example an ice/snow related category which covers snow skiing, and a power boating category which covers water skiing.

Standard reporting of the rate of cumulative incidence rate of injury in sporting injuries literature is usually stated per 1,000 or 10,000 players. This study reports the cumulative incidence rate as per 100,000 due to the infrequent number of serious injuries in the study. As the numbers of rugby league and union players have been relatively static over the study period, the approximation of the annual incidence rate of injury per 100,000 participants could also be expressed as the case rate per 100,000 player years at risk.

Trend analysis and code comparisons

Injury rate trends in this study were determined using Chi-square tests and by comparing the statistical significance of proportional differences in incidence rates between grouped years within sports and over time between sports. The initial quartile period from 1984 to 1987 was used as a baseline for comparison with subsequent periods. This initial period served as an ideal baseline as it occurred before the introduction of many safety initiatives in rugby and had the highest number of rugby football spinal unit admissions and serious permanent injuries. The incidence rates in these periods were compared for significant proportional difference. Survival analysis was not endeavoured, as it required unavailable exposure data.

170

The relative risk of SCI between codes and 95% confidence intervals around the relative risks were calculated. For code comparisons, the relative risk calculations used injured and non-injured players as the outcomes with each rugby code as the exposures.

Risk factor analysis

Pearson Chi-square tests (Fishers exact test was used for 2X2 tables i.e. when looking for associations between dichotomous variables.) and goodness of fit tests were used for examining the associations of injuries, risk factors and codes. For comparisons between country and metropolitan teams, the relative risk and confidence intervals were calculated using all SCI, brain injuries and fatalities as the outcome and non-injured players as the study group with each team region as the exposure. The denominators were estimated from figures provided by the NSW rugby league and rugby union associations (NSW Rugby Union & NSW Rugby League, December 2002, pers. comm.).

Archival video and player size analysis

In the archival video analysis, a Pearson Chi-square test (Fishers exact test for 2X2 tables) was used for the associations between playing styles, period and code and was suitable to the nature of the data which was asymmetric, discrete and not normally distributed.

The player size analysis required comparisons between continuous variables including player height and weight, and ages were tested using t-test and ANOVA for significant differences between pre 1970 and post 1970 groupings. Data were checked to ensure that all the assumptions underlying the tests used for comparison (i.e. T-Test and ANOVA) were satisfied. This included random sampling, normal distribution of dependent variable, and equality of variance.

171 Teams were chosen randomly from available data sources that covered elite level teams only. All players from the 1920s to 1950s were from elite club teams or representatives in state and international test teams. Comparable elite club, state and international test teams were chosen from 1999 to 2000 so that teams were similar levels of competition and country of origin.

To facilitate comprehension of material, tables presented in the findings are summaries drawn from the completed statistical analysis. Much of the analysis was completed for both rugby codes together and individually, requiring many pages of results. Please refer to Appendix H for the entire set of detailed tables.

When assessing the various assumptions required for using the parametric tests it was found that the distribution of the dependent variables varied from normal for weight, approximating normal for height and skewed for age. However, analysis was primarily interested in weight and height differences not age. The homogeneity of variance assumption was assessed with the Levene test of equality of variance. Calculations for t-tests not assuming equal variances were used when required. ANOVAs were abandoned if variance in the dependent variable were not homogenous in the groups being compared (the non parametric test equivalent, the Kruskall Wallis test, was then employed). A formal sample size assessment was not undertaken as an opportunity sample was used, as the true size of the total population, that is all football film/video recordings ever made in the 20th century, was unknown.

172 Data limitations

Research design

The disadvantages of a descriptive analytical study include: reduced ability to assess causality due to the temporal relationship between the participant becoming a case and the number of players at risk of injury, and the inability to establish retrospectively the accurate player person-time at risk.

Incidence rate calculations

A most informative measure of an incidence rate is the incidence density rate which utilises the person-time at risk. This accounts for dynamic populations where the person time at risk varies over the population. This is best expressed as the rate of injury per 10,000 player hours. This study was only able to approximate the calculation of at risk player hours expressed per 10,000 player population.

Statistical analysis

Chi Square tests assume that events are independent. The study design poses potential threats to this underlying assumption in some instances where people may be counted two or more times in the control group not obviously in the incident group.

Coverage

The NSWSIC data between 1990 and 1999 describes rugby league players only, as NSW rugby union adopted private insurance arrangements in 1990, and hence union injuries from 1990 were not recorded by the NSWSIC for union players. A comprehensive analysis of the incidence rate or mechanism of rugby union head injuries was not possible as the NSWSIC was the only source of head injury data and was devoid of union player data after 1990.

173 The retrospective method of the NSWSIC compensation scheme meant that several injuries that occurred within the study period may not have been processed if the injury had not stabilised and were hence omitted from analysis. Players of both codes may also have been omitted from analysis if compensation claims to the NSWSIC were not made. This may have occurred because players may have been involved in other legal proceeding to attain financial compensation for injury.

It is also possible that some SCI were omitted from analysis if the injury caused immediate fatality and were therefore not evacuated to the spinal units. Additionally, analysis may have excluded very young children who suffered SCI and were evacuated to a paediatric facility instead of one of the study units.

In Australia, it is possible to identify each hospitalised adult who has suffered an SCI from trauma. This is due to the small number of specialist treatment units nationally and the well established hospital protocol of referring cases with SCI to one of these units. However, the coverage of children is not complete, as some cases will be treated in paediatric hospitals rather than spinal units (O'Connor 2002c).

Based on routine hospital separations data 1986-1998, it is estimated that about half of the new cases of SCI that occur each year amongst 0-14 year olds nationally, are covered by ASCIR. The National Injury Surveillance unit estimates that for the year 1998-1999, three under 15 year old SCI cases treated in paediatric facilities were unreported by the Australian Spinal Cord Injury Register, accounting for around 2% of total SCI in Australia (O'Connor 2002c).

It must also be considered that evacuation procedures for country injury locations may bias results underestimating the incident of country injuries. This is because accidents occurring in NSW country regions close to state borders may be taken to interstate facilities and not be counted in the survey.

174 A search of the NSW state libraries newspaper articles database found no mention of SCI received in an under 14 year old while playing rugby league or union in NSW. The ABS do not release data on injuries sustained by less than ten people per year. Therefore, ABS data does not allow the detection of this rare event. Personal communication with the National Injury Surveillance Unit (29 Nov 2002), RNSH and PHH Spinal Unit directors (1998; 5 December 2002), Director of Sydney Children's hospital (8 April 2003) and Staff Specialist at New Children's Hospital Westmead (20 December 2002) reported no cases in their respective tenures.

Risk factor detail

Examination of medical records was the only source of detail on the circumstances surrounding injury. However, medical records were not designed for the purpose of identifying risk factors, and consistent information on important risk factors was not available in every case. The NSWSIC benefits paid case files provide much detail on the circumstances surrounding injuries, such as eye witness testimonies. However, they did not record important risk factor data consistently for every compensated case.

Consistent collection of risk factors using systematic prospective surveillance of large player populations is required to validate and explore conclusions drawn from this thesis’ consideration of risk factors.

Film and video analysis

The available material will be biased due to the selection of the games filmed and plays focused on by the journalists reporting on the codes. All games analysed were at the elite and international level by virtue of the accessibility of video footage. Moreover, much of the video footage was essentially highlights from games therefore, not all game's rules could be assessed.

175 The use of video analysis in biomechanics research has steadily progressed in recent years with technology now allowing the systematic measurement of data such as the velocity and gait of players. The potentially poor quality of archival footage made it impossible to quantify the assessment of player velocity as had been done in other prospective studies (McIntosh et al. 2000). Hence, the video analysis in the thesis, while sometimes quantified, was essentially qualitative and can only be suggestive not empirical.

CHAPTER THREE SUMMARY

The design of the studies main component is based on the use of retrospective injury surveillance data. The availability of this data provided an invaluable opportunity and reporting imperative for descriptive and analytical longitudinal study of rugby football injuries. The choice of design was strongly influenced by the rare and catastrophic event nature of SCI and brain injury in rugby football players.

Various data sources and statistical techniques have been used in an attempt to describe the frequency of serious injury events, associated risk factors and evidence of change in their incidence rate. The main sources of data examined were; a retrospective analysis of clinical and compensation case file records during 1984-1999; a review of game rules, weights and heights of players, over the past 100 years; a review of film and video footage of rugby football games spanning most of the 20th century.

Data limitations in this study include the disadvantages of utilising a descriptive analytical study design, the inability to calculate person time at risk exposure, and limitations of retrospective data sources utilised in terms of consistency of record detail.

176 CHAPTER FOUR

FINDINGS

The first section of this chapter provides an overview of the number of injuries and fatalities associated with all sport and recreational activities identified from the NSWSIC (1980 to 1999) and NSWSU (1984 to 1994) data sources (see Chapter three – methodology, section two). Section two then focuses its attention on rugby football with a descriptive analysis of the rate of serious rugby football injuries in NSW 1984 to 1999 and associated risk factors. Section three presents the findings of the archival game footage and player size analysis which provides comparisons of changing playing style and player size over the 20th century.

SECTION ONE

OVERVIEW OF SERIOUS SPORT AND RECREATIONAL INJURIES IN NSW

Spinal cord, brain and fatal injuries compensated by the NSW Sporting Injuries Committee Insurance Scheme 1980 to 1999

Whilst the NSWSIC was unable to provide explicit risk categories for the sports it insures due to confidentiality policies, it was able to provide the premium rates it charges that purportedly reflect assessed injury risk (NSWSIC premium rates – Appendix D). Rugby league, rugby union, motor cycling, amateur horse race riding and polo attracted the highest premium rates.

Of all the other sport covered by the NSW Sporting Injuries Committee between 1980 and 1999, rugby league and union codes accounted for the greatest proportion (77.9%; 95%CI 70.0% - 84.3%) of compensated brain, neck and fatal injuries (Table 8). The remaining 22.1% (95%CI 15.4% - 30.0%) of injuries were contributed by 18 other sport and recreational activities.

177 While rugby league and union accounted for the greatest number of serious injuries, the incidence rate per 100,000 participants per year of serious injuries in other sports with relatively small participant populations was sometimes higher.

The annual incidence rate of brain injury between 1980 and 1999 for rugby league participants was 1.83 (95%CI 1.2 - 2.6) per 100,000 participants per year and 1.68 (95%CI 0.5 - 4.3) per 100,000 for rugby union participants per year (Table 8). The annual incidence rate of brain injury for other sports with lower participant populations was as high as 156.99 (95%CI 4.0 – 871.5) per 100,000 cycling participants per year; 147.71 (95%CI 3.7 – 820.2) per 100,000 Taekwondo participants per year; and 100.00 (95%CI 2.5 – 555.9) per 100,000 Hangliding participants per year.

However, it can be observed that the small participant populations lead to large confidence intervals around the estimates of injury incidence making significant differences with other sports hard to detect. The only significant differences in the incidence of brain injury between other sports and the rugby codes (i.e. where confidence intervals do not overlap) were for cycling and Taekwondo which were found to have a greater incidence of brain injuries than rugby league.

The annual incidence rate of SCI was highest in rugby union with 5.44 (95%CI 2.9 – 9.3) per 100,000 rugby union participants per year compared with 1.71 (95%CI 1.1 – 2.5) SCI per 100,000 rugby league participants. Both codes had a higher rate of SCI than soccer with 0.13 (95%CI 0.0-0.8) SCI per 100,000 soccer participants per year (Table 8) which was the only other sport responsible for a compensable SCI.

The annual incidence rate of fatalities for rugby league participants was 1.47 (95% CI 0.9 – 2.2) per 100,000 participants per year and 2.93 (95%CI 1.2 – 6.0) per 100,000 for rugby union participants per year (Table 8).

178 The annual incidence rate of fatalities for other sports with lower participant populations was as high as 77.22 (95%CI 2.0 – 429.5) fatalities per 100,000 for water polo participants per year and 10.91 (95%CI 1.3 – 39.4) per 100,000 flat horse racing participants per year. However, the only activity with a significantly higher fatality rate than the rugby codes was spear fishing with 55.73 (95% CI 6.7 – 200.5) fatalities per 100,000 spear fishing participants per year (Table 8).

179 Table 8: NSW Sporting Injuries Committee compensated brain, neck injuries, and fatalities 1980 to 1999a Activity Participants Brain Spine Fatalb N (per 100,000 participants per N (per 100,000 participants N (per 100,000 participants year, 95%CI)c per year, 95%CI)c per year, 95%CI)c Rugby league 1,636,301 30 (1.83, 95%CI 1.2 - 2.6) 28 (1.71, 95%CI 1.1 – 2.5) 24 (1.47, 95%CI 0.9 – 2.2) Rugby uniond 238,776 4 (1.68, 95%CI 0.5 – 4.3) 13 (5.44, 95%CI 2.9 – 9.3) 7 (2.93, 95%CI 1.2 – 6.0) Soccer 742,358 2 (0.27, 95%CI 0.03 – 1.0) 1 (0.13, 95%CI 0.03 – 0.8) 0 Touch football 206,309 0 0 2 (0.97, 95%CI 0.1 – 3.5) Spear fishing 3,600 0 0 2 (55.73, 95%CI 6.7 – 200.5) Fishing 146,390 1 (0.68, 95%CI 0.02 - 3.8) 0 3 (2.05, 95%CI 0.4 – 6.0) Swimming 20,745 0 0 2 (9.64, 95%CI 1.2 – 34.8) Flat horse racing 2,818 1 (35.49, 95%CI 0.9 - 197.6) 0 1 (35.49, 95%CI 0.9 – 197.6) Horse racing (amateur) 1,272 1 (78.62, 95%CI 2.0 – 437.2) 00 Polo 6,674 1 (14.98, 95%CI 0.4 – 83.5) 00 Pony 252,952 1 (0.40, 95%CI 0.1 – 2.2) 00 Motorcycling 18,332 0 0 2 (10.91, 95%CI 1.3 – 39.4) AFL 61,703 1 (1.62, 95%CI 0.04 – 9.0) 0 1 (1.62, 95%CI 0.04 – 9.0) Netball 36,362 2 (5.50, 95%CI 0.7 – 19.9) 00 Basketball 17,769 1 (5.63, 95%CI 0.1 – 31.4) 00 Cycling 637 1 (156.99, 95%CI 4.0 – 871.5) 00 Hangliding 1,000 1 (100.00, 95%CI 2.5 – 555.9) 00 Hockey 32,901 1 (3.04, 95%CI 0.7 – 16.9) 00 Taekwondo 677 1 (147.71, 95%CI 3.7 – 820.2) 00 Water polo 1,295 0 0 1 (77.22, 95%CI 2.0 – 429.5) Total 3,432,460 49 42 45 a Source: NSW Sporting Injuries Committee database 1980 to 1999. b Head and spine injuries resulting in death are accounted for in both injury and fatality figures. c Annual incidence rate per 100 000 NSWSIC member players. d Rugby union cases were reviewed up until 1990.

180 Sport and recreational admissions to NSW Spinal units 1984 to1994

The broad category of recreational diving accounted for 31.1% (95%CI 26.9% – 35.5%) of all admissions to spinal units (Table 9). If the unspecified rugby football admissions are combined for both codes admissions, rugby football accounted for 25.3% (95%CI 21.4% – 29.5%) of all admissions to spinal units. Equestrian activities accounted for 11.1% (95%CI 8.4% - 14.4%) of all admissions, followed by water skiing (3.9%, 95%CI 2.3% – 6.0%), bicycle riding (3.6%, 95%CI 2.1% – 5.8%), and surfing 3.0% (95%CI 1.6% – 5.0%). These admissions included resolving SCI injuries and permanent neurological deficits.

The incidence rate of admissions for recreational diving was estimated as 0.2 (95%CI 0.2 – 0.3) SCI admissions per 100,000 estimated persons in NSW per year (Table 9). The incidence rate of admissions for rugby league was estimated as significantly higher at 5.1 (95%CI 3.8 – 6.7) SCI admissions per 100,000 estimated participants per year, as was rugby union at 8.2 (95%CI 6.0 – 10.8) SCI admissions per 100,000 estimated participants per year. Equestrian activities had similar figures to rugby league with 5.6 (95%CI 4.2 – 7.3) SCI admissions per 100,000 estimated participants per year.

The incidence rate for the 69 admissions contributed by water skiing, cycling, surfing, skiing and trail bike/pro-motorbike activities and 82 admissions contributed by a composite of infrequent sporting activities categorised as ‘other’ could not be calculated as the entire participant populations in NSW from 1984 to 1994 could not be reliably estimated.

The most reliable sport and recreational participant figures for NSW are available from the Australian Bureau of Statistics Catalogue No. 4177.0 (1998; 1999a; 2000a) and the Exercise, Recreation and Sport Survey (Dale and Ford, 2002) funded by the Australian Sports Commission and State/Territory Departments of Sport and Recreation (see Appendix G). However, these sources are not available

181 prior to 1996. Furthermore these sources only include adult participants, and often only specify conglomerated categories activities, for example, the ice/snow related category which covers snow skiing and the power boating category which includes water skiing.

Table 9: Acute sport and recreational admissions to the main spinal injury units in NSW 1984 to 1994a Activity N (% logbook admissions, 95%CI) Per 100,000 participants per year (95%CI)b Diving 145 (31.1%, 26.9% - 35.5%) 0.2 (0.2 - 0.3) Equestrian 52 (11.1%, 8.4% - 14.4%) 5.6 (4.2 - 7.3) Rugby league 49 (10.5%, 7.9% - 13.7%) 5.1 (3.8 - 6.7) Rugby union 49 (10.5%, 7.9% - 13.7%) 8.2 (6.0 - 10.8) Unspecified rugby football 20 (4.3%, 2.6% - 6.6%) Unable to calculatec Water skiing 18 (3.9%, 2.3% - 6.0%) Unable to calculated Bicycle 17 (3.6%, 2.1% - 5.8%) Unable to calculated Surfing 14 (3.0%, 1.6% - 5.0%) Unable to calculated Snow skiing 11 (2.3%, 1.2% - 4.2%) Unable to calculated Trailbike/pro - motorbike 9 (1.9%, 0.9% - 3.6%) Unable to calculated Othere 82 (17.6%, 14.2% - 21.4%) Unable to calculated Total 466 a Source: Logbooks NSWSU b Annual incidence rate per 100 000 estimated players. c Unable to estimate population as code not specified/may include unofficially organised games. d Unable to reliably estimate entire populations for these sports in NSW over study period. e Australian rules, Basketball, Bodyboarding, Bushwalking, Cricket, Dune Buggy, Fishing, Flying Fox, Glider Plane, Soccer, Golf, Gymnastics, Hockey, Judo, Kayaking, Motorboat, Netball, Para/hang gliding, Parachuting, Rock Climbing/Abseiling, Rodeo, Scuba Diving, Slippery Dip, Snowboard, Softball/Baseball, Swimming, Swing, Tobogganing, Touch Football, Trampoline, Water Related Activities, Waterslide, Wrestling

182 SECTION TWO

SERIOUS RUGBY FOOTBALL INJURIES

NSWSU and NSWSIC rugby football SCI cases 1984 to 1999

Spinal unit admissions

The number of spinal unit admissions associated with permanent and non- permanent injuries for both codes vary over the study period from one to sixteen per year (Table 10).

Table 10: Spinal unit admissions and cervical spinal cord injury for rugby union and league footballers in New South Wales 1984 – 1999a Spinal Unit Admissions Cervical spinal cord injury only League Union Unspecified Leagueb Union Unspecified 1984 4 6314 0 1985 5 4032 0 1986 1 3313 0 1987 4 7236 0 Sub total 14 20 8 8 15 0 1988 3 4022 0 1989 7 5 0 1 2 0 1990 5 6 0 3 0 0 1991 7 2 0 2 0 0 Sub total 22 17 0 8 4 0 1992 9 1 6 4 0 1c 1993 1 8 2 0 2 0 1994 3 3 4 1 1 0 1995 6 2 3 0 2 1d Sub total 19 14 15 5 5 2 1996 4 5 0 3 3 0 1997 0 1 2 1 1 0 1998 6 3 2 2 1 0 1999 3 1 0 0 0 0 Sub total 13 10 4 6 5 0 Total 68 61 27 27 29 2 a Source: NSWSU medical records b Includes two cases from NSWSIC case records (1996,1997) c Backyard football game d Schoolyard touch football game

183 On average, there were 4.3 (standard deviation {SD} ±3.7) spinal unit admissions per year in rugby league and 3.8 (SD ±3.5) spinal unit admissions per year in rugby union in NSW. The estimated annual incidence rate of admissions for league and union players was 4.9 (95% CI 3.8 – 6.2) spinal unit admissions per 100,000 estimated players per year for rugby league and 7.0 (95% CI 5.3 – 9.0) spinal unit admissions per 100,000 estimated players per year for rugby union (Table 11).

Table 11: Summary of spinal unit admissions and cervical spinal cord injury for rugby union and league footballers in New South Wales 1984 to 1999a Spinal Admissions Spinal Cord Injury League Union League Union Average number of 86775 54625 86775 54625 participants per year Total injuries 1984 - 1999 68 61 27 29 4.3 3.8 1.7 1.8 Mean cases per year (SD ±3.7) (SD ±3.5) (SD ±2.6) (SD ±2.9) 4.9 7.0 1.9 3.3 Incidence rate per 100,000 (95%CI (95%CI (95%CI (95%CI estimated players per year 3.8 - 6.2) 5.3 - 9.0) 1.3 - 2.8) 2.2 - 4.8) a Source: NSWSU medical records and NSWSIC case records

The relative risk of a spinal unit admission for rugby union players was 1.22 times greater than for rugby league players (95%CI 1.01 - 1.45, X2= 4.08, P<0.05) (Table 12). The incidence rate of spinal unit admissions did not significantly change over the 16-year study period of 1984 to 1999 for league (Trend X2= 0.10, P>0.05) or union (Trend X2= 3.58, P>0.05). Though not statistically significant (Trend X2= 3.58, P>0.05), spinal unit admissions for rugby union have gradually decreased from 9.2 per 100,000 estimated players per year (95%CI 5.6 - 14.1) in 1984 to 1987 to 4.6 per 100,000 estimated players per year (95%CI 2.2 - 8.4) in 1996 to 1999 (Table 12).

184 Table 12: Summary of trends in spinal unit admissions and the incidence rate of cervical spinal cord injuries for rugby league and union players in New South Wales 1984 –1999a League Union Admissions N (per 100,000 estimated players N (per 100,000 estimated players per year, 95%CI)b per year, 95%CI)b 1984-87 14 (4.0, 95%CI 2.2 - 6.8) 20 (9.2, 95%CI 5.6 - 14.1) 1988-91 22 (6.5, 95%CI 4.1 - 9.8) 17 (7.8, 95%CI 4.5 - 12.5) 1992-95 19 (5.2, 95%CI 3.1 - 8.1) 14 (6.4, 95%CI 3.5 - 10.8) 1996-99 13 (3.9, 95%CI 2.1 - 6.6) 10 (4.6, 95%CI 2.2 - 8.4) Trend X2= 0.10, DF=1, P>0.05 Trend X2= 3.58, DF=1, P>0.05 1984-1999 RR (Union Vs League)= 1.22 (95%CI 1.01 - 1.45), X2= 4.08, P<0.05 Spinal N (per 100,000 estimated players N (per 100,000 estimated players b b Cord Injury per year, 95%CI) per year, 95%CI) 1984-87 8 (2.3, 95%CI 1.0 - 4.5) 15 (6.9, 95%CI 3.8 - 11.3) 1988-91 8 (2.4, 95%CI 1.0 - 4.6) 4 (1.8, 95%CI 0.5 - 4.7) 1992-95 5 (1.4, 95%CI 0.4 - 3.2) 5 (2.3, 95%CI 0.7 - 5.3) 1996-99 6 (1.8, 95%CI 0.7 - 3.9) 5 (2.3, 95%CI 0.7 - 5.3) Trend X2= 0.59, DF=1, P>0.05 Trend X2= 5.81, DF=1, P<0.05 1984-1999 RR (Union Vs League)= 1.34 (95%CI 1.01 - 1.67), X2= 4.09, P<0.05 a Source: NSWSU medical records and NSWSIC case records b Annual incidence rate per 100, 000 estimated players

185 Permanent cervical spinal cord injuries

Fifty-six rugby football players (27 league and 29 union) with permanent cervical spinal cord injury were identified from the NSWSIC and NSWSU medical records between 1984 and 1999 (Table 10). These permanent injuries are a subset of the total spinal unit admissions that include resolving injuries. On average, there were 1.7 (SD ±2.6) permanent SCI injury per year in rugby league and 1.8 (SD ±2.0) permanent SCI per year in rugby union in NSW. Standard deviations greater than the average were the result of a wide variation in the yearly number of SCI. In rugby league the year-to-year number of SCI from 1984 to 1999 ranged from zero to four, and in rugby union from zero to six.

The estimated annual incidence rate of SCI for rugby league players was 1.9 (95% CI 1.3 - 2.8) per 100,000 estimated players per year and 3.3 (95%CI 2.2 - 4.8) per 100,000 estimated rugby union players per year (Table 11). The relative risk of a SCI was 1.34 times greater for rugby union players than for rugby league players between 1984 and 1999 (95%CI 1.01 - 1.67, X2= 4.09, P<0.05) (Table 12).

Analysing the relative risk of spinal unit admissions and permanent SCI between rugby union and league players for each quartile period found that rugby union players were 1.52 times more likely to be admitted to spinal units (95% 1.09 - 1.91, X2= 5.85, P<0.05) and 1.69 (95% 1.16 - 2.10, X2= 6.86, P<0.05) times more likely to sustain permanent SCI compared with rugby league players in the period (1984 to 1987). The relative risk of spinal unit admissions and permanent SCI between rugby union and league players in subsequent periods were not statistically significant.

The incidence rate of permanent SCI in Rugby league ranged from 2.4 (95%CI 1.0 - 4.6) per 100,000 estimated players per year during the period 1988 to 1991 to 1.4 (95%CI 0.4 - 3.2) per 100,000 estimated players per year during the period 1992 to 1995 (Table 12). The incidence rate of permanent SCI in rugby league did not significantly decline (Trend X2= 0.59, P>0.05) over the entire study period 1984 to 1999.

186 Rugby union did experience a significant decline in trend in incidence rate of permanent SCI over the study period. There was a peak of 6.9 (95%CI 3.8 - 11.3) SCI per 100,000 estimated players per year in 1984 to 1987 declining to 1.8 (95%CI 0.5 - 4.7) per 100,000 estimated players per year in 1988 to 1991. This remained low at 2.3 (95%CI 0.7 - 5.3) per 100,000 estimated players per year for the rest of the study period (Trend X2= 5.81, P<0.05) (Table 12).

In rugby league SCI, tests of proportional difference indicated no significant differences between the incidence rate in the baseline period and each subsequent period. In rugby union SCI, significant differences were found between the incidence rate in the baseline period and all subsequent quartile periods (Table 13).

Table 13: Tests of proportional difference between the incidence ratea in the baseline period and each subsequent period for rugby union SCIb 1984 to 1987 with 1984 to 1987 with 1984 to 1987 with 1988 to 1991 1992 to1995 1997 to 1999 P value P<0.05 P<0.05 P<0.05 a Annual incidence rate per 100, 000 estimated players b Source: NSWSU medical records and NSWSIC case records

For all permanent cervical spinal cord injuries, damage was most common (X2= 25.43, P<0.0001) to the C4 to C7 (90%) level of the cervical spinal cord for both codes of rugby football (Table 14).

Table 14 : Site of cervical spinal cord injuries in rugby football players in NSW from 1984 to 1999a Injury Level Number (%) C1/2 - C2/3 3 (5, 95%CI 1.1% - 14.9%) C3 - C3/4 3 (5, 95%CI 1.1% - 14.9%) C4 - C4/5 22 (39, 95%CI 26.5% - 53.3%) C5 - C5/6 17 (30, 95%CI 18.8% - 44.1%) C6 - C6/7 11 (20, 95%CI 10.2% - 32.4%) Total 56 (100) a Source: NSWSU medical records and NSWSIC case records

187 NSWSIC rugby football brain injuries and fatalities 1984 to 1999

Brain injuries

Twenty-seven rugby league football players with permanent brain injury were identified from the NSWSIC medical records for the period 1984 to 1999 (Table 15). On average, there were 1.7 (SD ±1.3) permanent brain injuries per year for rugby league players. The NSWSIC data for NSW rugby union players for 1984 to 1989 identified three cases of permanent brain injury between 1984 and 1987. For this period, on average there were 0.75 (SD ±1.5) permanent brain injuries per year for rugby union players.

Table 15: Summary of brain injuries and fatalities for rugby union and league footballers in New South Wales 1984–1999a Brain injury Fatalities League Unionb League Unionb Average participants per year 81137 23392 81137 23392 27 3 19 5 Total injuries (1984-1999) (1984-1987) (1984-1999) (1984-1987) 1.7 0.75 1.2 1.25 Mean cases per year (SD ±1.3) (SD ±1.5) (SD ±1.0) (SD ±0.5) Annual incidence rate per 2.1 3.2 1.5 5.3 100,000 players per year (95%CI (95%CI (95%CI (95%CI (1984-1999) b 1.4 - 3.0) 0.7 - 9.4) 0.9 - 2.3) 1.7 - 12.5) a Source: NSWSIC case records b Rugby union figures only for the period 1984 to 1987

The annual incidence rate of brain injuries for rugby league players was 2.1 (95%CI 1.4 - 3.0) per 100,000 players per year and 3.2 (95%CI 0.7 - 9.4) per 100,000 rugby union players per year (Table 15). Between 1984 and 1987, rugby union players were associated with 1.49 times more brain injuries compared to rugby league players but this was not statistically significant (95%CI 0.42 - 5.28, X2= 0.34, P> 0.05), possibly due to the limited period of comparison (Table 16).

188 There was no significant reduction (Trend X2= 0.01, P> 0.05) in the incidence rate of brain injuries identified by the NSWSIC case files over the period 1984 to 1999 for rugby league (Table 16).

Table 16: Rugby league brain injuries and fatalities from NSW Sporting Injuries Committee benefits paid cases 1984 to 1999a League Unionb Brain injuries N (Annual incidence rate per N (Annual incidence rate per 100,000 participants, 95%CI) 100,000 participants, 95%CI) 1984-87 7 (2.2, 95%CI 0.9 - 4.4) 3 (3.2, 95%CI 0.7 - 9.4) 1988-91 8 (2.5, 95%CI 1.1 - 4.9) NA 1992-95 4 (1.2, 95%CI 0.3 - 3.1) NA 1996-99 8 (2.5, 95%CI 1.1 - 4.9) NA 1984-1999 Trend X2= 0.01, DF=1, P>0.05 NA

1984-1987 RR (Union Vs League)= 1.489 (95%CI 0.42 - 5.28), X2= 0.34. P>0.05 Fatalities N (Annual incidence rate per N (Annual incidence rate per 100,000 participants, 95%CI) 100,000 participants, 95%CI) 1984-87 7 (2.2, 95%CI 0.9 - 4.4) 5 (5.3, 95%CI 1.7-12.5) 1988-91 3 (0.9, 95%CI 0.2-2.7) NA 1992-95 3 (0.9, 95%CI 0.2-2.6) NA 1996-99 6 (1.9, 95%CI 0.7-4.1) NA 1984-1999 Trend X2= 0.07, DF=1, P>0.05 NA

1984-1987 RR (Union Vs League)= 2.48 (95%CI 0.83 - 7.40), X2= 2.58, P>0.05 a Source: NSWSIC case records b Rugby union cases were analysed up until 1990.

Tests of proportional difference in incidence rates between the baseline period 1984 to 1987, and subsequent four-year periods for brain injuries in rugby league players, indicated no significant differences between the incidence rate in the baseline period and each subsequent period.

189 Fatalities

For the period 1984 to 1999, 19 fatalities were reported for rugby league players with an average of 1.2 (SD ±1.0) fatalities per year (Table 15). For the period 1984 to 1987, five fatalities were reported for rugby union players with an average of 1.3 (SD ±0.5) fatalities per year (Table 15). Rugby league had 1.5 (95%CI 0.9 - 2.3) fatalities per 100,000 member players per year for the period 1984 to 1999, and rugby union had 5.3 (95%CI 1.7 - 12.5) fatalities per 100,000 member players per year for the period 1984 to 1987.

Between 1984 and 1987, rugby union players were associated with 2.48 times more fatalities than rugby league players, however, this was not found to be statistically significant (95%CI 0.83 - 7.40, X2= 2.58, P>0.05) possibly due to the limited period available for comparison (Table 16).

There was no significant (Trend X2= 0.07, P>0.05) change in the incidence rate of fatalities identified by the NSWSIC case files over the period 1984 to 1999 for rugby league (Table 16). Similarly, there were no significant proportional differences in the incidence rate of rugby league fatalities between the baseline period 1984 to 1987 and each subsequent four-year period.

190 Risk factors (Extrinsic)

The following sub section analyses details gathered from the NSWSU and NSWSIC records concerning extrinsic risk factors associated with serious injury.

Types and style of play that increase the risk of serious injury

Spinal cord injury

The analysis of the NSWSU data 1984 to 1999 verified by NSWSIC records indicated that 78% (95%CI 57.7% - 91.4%) of rugby league SCI were tackle related with 90% (95%CI 66.9% - 98.7%) of these to the ball carrier (Table 17). The remaining 22% of SCI in rugby league were contributed by scrums 19% (95% CI 6.3% - 38.1%) and other activities. The number of scrum related SCI in rugby league dropped from four during the 1980s to one during the 1990s, while tackle related SCI stayed consistent with ten during the 1980s and eleven during the 1990s.

Conversely, in rugby union scrums caused significantly (X2=4.44, P<0.05) more SCI in rugby union, 45% (95% CI 26.5% - 64.3%), than in rugby league. Of all rugby union SCI, tackles accounted for 35% (95% CI 17.9% - 54.3%) of which, 78% (95% CI 40.0% - 97.2%) were sustained to the ball carrier.

It was difficult to distinguish whether an injury occurred in a tackle or in an ensuing maul or ruck, however it was clear that 17% (95%CI 5.85% - 35.78%) of SCI in union were associated with mauls and rucks. Scrum related SCI in rugby union declined in number by more than half in the 1990s (N=4) when compared to the 1980s (N=9). However, the proportion of all SCI accounted for by scrums was not significantly (X2=0.00, P>0.05) associated with these periods dropping from 47% (95%CI 24.5% - 71.1%) during the 1980s to 40% (95%CI 12.2% - 73.8%) during the 1990s.

191 Table 17: Cause of SCI, spinal admission, brain injury and fatalities by rugby football code in NSW 1984 to 1999a Rugby league Rugby unionc Scrum 5 (19%) 13 (45%) Ruck or maul 0 5 (17%) SCI Tackled 17 (63%) 7 (24%) Tackling 2 (7%) 2 (7%) Unspecified collision or tackle 3 (11%) 2 (7%) Tackle related total 21 (78%) 10 (35%) Total 27 (100%) 29 (100%) Spinal Scrum 11 (27%) 17 (50%) admission Ruck or maul 0 5 (16%) not resulting Tackled 14 (34%) 3 (9%) in permanent Tackling 9 (22%) 2 (6%) SCI Unspecified collision/tackle 7 (17%) 6 (19%) Tackle related total 24(59%) 6 (19%) Total 41 (100) 32 (100%) Brain Not known 4 (15%) 0 injury Tackled 14 (52%) 2 (67%) Tackling 7 (26%) 0 Unspecified collision or tackle 2 (7%) 1 (33%) Tackle related total 22 (81%) 2 (67%) Total 27 (100%) 3 (100%) Scrum 0 1 (20%) Fatality Ruck or maul 0 1 (20%)

Exertionb 8 (42%) 1 (20%)

Unspecified collision 1 (5%) 0

Not known 1 (5%) 0 Tackled 4 (21%) 2 (40%) Tackling 5 (25%) 0 Tackle related total 9 (47%) 2 (40%) Total 19 (100%) 5 (100%) a Source: NSWSU medical records and NSWSIC case records. b One rugby league case was exertion after heavy tackle and one after scrum collapse. c Rugby union brain injuries and fatalities cases were analysed up until 1990.

The causes of rugby football spinal admissions that did not result in permanent spinal cord injury were very similar to those that did result in SCI (Table 17). Accordingly, there was no significant relationship found between injury cause and neurological status on discharge from spinal units in rugby league (X2 = 1.73, P>0.05) or rugby union players (X2 = 0.43, P>0.05).

192 Brain injury

Brain injury data collected by the NSWSIC from 1984 to 1999 were all associated with either tackles or collisions (Table 17). Rugby league players contributed 90% of the available brain injury data. These rugby league players reported 67% (95% CI 43.0% - 85.4%) of brain injuries were sustained by the ball carrier and 33% (95% CI 14.6% - 57.0%) were sustained by the tackler. There was a significant (X2=4.77, P<0.05) difference in the proportion of non-fatal brain injuries associated with being tackled (76%, 95% CI 50.1% - 93.2%) compared with doing the tackling (24%, 95% CI 6.8% - 49.9%) in rugby league. Rugby union reported 67% (95% CI 9.4% - 99.2%) of head injuries sustained by the ball carrier and 33% (95% CI 0.8% - 90.6%) as an unspecified collision.

Fatalities

The fatality data collected by the NSWSIC for 1984 to 1999 were all associated with cardiac arrhythmia and the sequelae of severe spinal cord or brain injuries. Of the 19 fatalities occurring in rugby league, 63% (95%CI 38.4% - 83.7%) were associated with heart failure, 67% (95%CI 34.9% - 90.1%) of these accounted for by exertion and the remainder by tackles (33%, 95% CI 9.9% - 65.1%) of which 75% (95% CI 19.4% - 99.4) were to the ball carrier (Table 17). Two cases of exertion were found to be associated with endogenous morbid pathological process accompanying pre-existing cardio vascular disease. One case of exertion was found to be exacerbated by the ingestion of the anabolic steroid oxymesterone.

Brain injuries were associated with 32% (95%CI 12.6% - 56.6%) of rugby league fatalities, of which 33% (95%CI 4.3% - 77.7%) received spinal injuries in addition to brain injuries. Fatal brain injuries were no more likely (X2=1.0 P>0.05) to be associated with being tackled (25%, 95%CI 0.6% - 80.6%) than with doing the tackling (75%, 95%CI 19.4% - 99.4%). Of the five fatalities occurring in rugby union, 20% (95%CI 0.5% - 71.6%) were associated with heart failure accounted for by exertion. The remaining fatalities were caused by brain injuries from being

193 tackled (40%, 95% CI 5.3% - 85.3%) or SCI associated with a scrum or maul (40%, 95%CI 5.3% - 85.3%).

Three deaths associated with head injuries in rugby league occurred when the player’s head struck the ground when tackled. The two rugby union tackle related fatalities, between 1984-1989, were both caused by brain injuries and were a result of a penalised head high tackle and a heavy tackle to the chest causing the player to hit his head against the ground.

Tackles Multiple tacklers

Evidence of an increased risk of injury from multiple tackling was explored in the NSWSU/NSWSIC data. Many tackling injuries involved multiple tacklers, where more than one opponent player was involved in the tackle of a single player. Commonly, this results in the ball carrier being unable to break their fall using their hands and arms resulting in their head taking full impact with the ground.

The difficulties in distinguishing between tackling, ruck and maul phases of play in rugby union means that it is difficult to accurately determine the role of multiple opponents in directly causing serious injury. However, the tackle mechanism could be clearly determined in rugby league, where 29% (95%CI 11.3% - 52.2%) of SCI and 32% (95%CI 13.9% - 54.9%) of brain injuries associated with tackles involved multiple opponents. Of the fatalities associated with tackles in rugby league a large proportion, 44% (95%CI 13.7% - 78.8%), involved multiple opponents with the remaining 56% (95%CI 21.2% - 86.3%) associated with single opponents.

NSWSIC case study 1 – tackled player injury (multiple tacklers) – brain injury A 27-year-old man, playing amateur club rugby league, was running forward carrying the ball and was tackled by two opponents. With his arms held in to his body by both tacklers, his head struck heavily on the ground with immediate loss of consciousness. He suffered a right subdural haematoma and a period of raised intracranial pressure that caused an 85% loss of mental capacity including severe memory impairment, 50% visual loss, and a left leg paresis.

194 NSWSIC case study 2 – tackled player injury (multiple tacklers) - SCI A 15-year-old boy wearing headgear, playing for a local team, was tackled by three players when running with the ball. The third player's knee struck the ball carrier in the neck causing the lining of the vascular artery to peel leading to a blockage to the brain resulting in a stroke. The stroke caused 50% loss of mental capacity, 20% loss of speech and 35% loss of use of right arm and left leg.

Level of impact

High tackles risk of injury to the ball carrier

Head high tackles resulting in the opponent player being penalised were associated with 5% (95% CI 0.1% - 22.8%) of tackle-related brain injuries in rugby league and none in rugby union. No SCI in either code involved penalised head high tackles.

Relationship between ball carrying style and tackling level

An examination of archival game footage for an observable relationship between ball carrying style and tackling level throughout early and modern games for both codes identified that opponent players consistently attempt to target the ball rather than the ball carrier. Failure to adhere to this practice sees the ball passed to a team-mate at the last possible moment by the ball carrier, keeping the ball in play and resulting in opposing players being drawn to the location of the former ball carrier creating space for the attacking team.

The ball carrying style and level of tackle tallied (Table 18) from an examination of football video footage in both codes from the 1930s to the 1990s indicated a significant (X2=273.4, P<0.001) association between ball carrying style and tackling level. Higher tackles to the more susceptible anatomical location of the upper chest were associated with carrying the ball high on the chest (76%, 73.94 - 78.64) or in front of body with two hands (19%, 17.08 - 21.43). Tackles to the head and neck were associated with carrying the ball high on the chest (59%, 52.60 - 64.68) or in front of body with two hands (32%, 26.09 - 37.52).

195 Elaboration tables used to control for time period and code played, found the relationship between the two variables (i.e. level of first impact and ball carrying style) were the same across all partial tables as in the original table. This excludes the possibility of spurious or conditional relationships existing with time period or code played.

Table 18: Level of first impact by ball carrying style of all incidents analyseda Ball Level of first impact % (95%CI) carrying style Legs Waist Chest Head/Neck 30 41 19 32 Two Hands (26.51% - 32.58%) (38.14% - 44.41%) (17.08% - 21.43) (26.09% - 37.52%) 2 3 1 3 Swinging (1.36% - 3.42%) (1.75% - 3.88%) (0.82% - 2.18%) (1.29% - 5.78%) 14 5 3 7 Waist (12.02% - 16.70%) (4.00% - 6.92%) (2.21% - 4.17%) (4.01% - 10.37%) 54 51 76 59 Chest (50.74% - 57.36%) (47.58% - 53.95%) (73.94% - 78.64%) (52.60% - 64.68%) Total 100 100 100 100 a Source: Screensound Australia and Sports Recording Services rugby football video collections Pearson Chi-square =273.4, P<0.001, N= 3443

Approaching tackles low with head down

In the rugby league NSWSIC case reports, 50% (95% CI 1.3% - 98.7%) of tackling- related SCI, 78% (95%CI 29.0% - 96.3%) of tackling-related brain injuries and 60% (95%CI 14.7% - 94.7%) of tackling-related fatalities were associated with a low tackle, commonly with head down, by the injured player. Injury would occur through contact of the tackler’s head against a body part of the opponent, most often the knee or hip. Tackled players in rugby league were at risk of injury when running with their head down into tackles; as this was associated with 18% (95%CI 3.8% - 43.4%) of SCI in tackled players and 14% (95%CI 1.8% - 42.8%) of brain injuries in tackled players.

In rugby union NSWSU and NSWSIC case reports, 10% (95%CI 0.3% - 44.5%) of tackling-related SCIs were associated with low tackles. None of the brain injuries or fatalities in the rugby union NSWSIC case reports were associated with tackling low.

196 NSWSIC case study 3 – tackling player injury A 20-year-old man playing elite Rugby league for a professional team attempted to tackle an opposition player very low in a bent over position with his head down not looking up, and was struck in the right temple by the knee of the ball carrier. He fell to the ground, managed to stand up and attempted to continue but collapsed minutes later. He sustained bilateral subdural haematomas and cerebral oedema. This caused 70% loss of mental capacity including memory impairment, partial quadriparesis, 75% visual loss and post-traumatic epilepsy.

Rucks and mauls

Examination of the five NSWSU/NSWSIC cases of rugby union ruck and maul injuries did not always clarify whether injuries were occurring at the initial impact of players or in an ensuing pile-up which often involved collapsing players. However, all cases with sufficient detail of the incident involved collapsing players but did not necessarily implicate it as the cause of injury.

NSWSIC case study 4 – collapsing maul injury A 20-year-old amateur country rugby union player fell over during a maul, catching his head under his body, which collapsed with other players falling on top. This resulted in complete C5/6 quadriplegia with permanent immobility.

Scrums

Examination of the NSWSIC/NSWSU cases of SCI associated with scrums in rugby league found all five injured players were playing in the hooker position. In 40% (95%CI 5.3% - 85.3%) of cases, injury clearly occurred during the initial contact during the engagement phase of the scrum. Another 40% of cases were injured when the scrum collapsed with subsequent scrum movement. The remaining case (20%) was injured from being ‘popped’ in the scrum when the opposition pack started to break and his team-mates maintained the drive. There were no rugby league school team players injured in scrums.

197 Of the thirteen cases of SCI from scrums in rugby union, 62% (95%CI 31.58% - 86.14%) were associated with scrum collapse. Of these, at least one (12.5%, 95%CI 0.3% – 52.7%) collapse occurred instantly on initial contact of packs, which may have actually been caused by an engagement injury, and another case (12.5%, 95%CI 0.3% – 52.7%) was associated with a ‘popping’ incident.

Another 23% (95%CI 5.04% - 53.81%) of cases were evidently injured during the initial contact of the engagement phase of the scrum. The remaining 15% (95%CI 1.92% - 45.45%) of cases were injured as the result of being popped in the scrum when team-mates started to break and the opposition maintained the drive. Where position information was available for scrum SCI cases (10/13), 60% (95%CI 26.24% - 87.85%) of rugby union players injured in scrums were playing in the hooker position and 40% (95%CI 12.16% - 73.76%) were playing in the prop position.

In rugby union, three school team players aged between 16 years and 19 years sustained SCI associated with scrum-related incidents, one of whom was inexperienced in the hooker position as it was his first game in a forward's position. He was deemed to have an unsuitable body type due to his tall thin build by the attending doctor after the incident. His head struck the opposing hooker's shoulder when the scrum was being formed as he was slower packing down for scrum and was caught with his head up.

NSWSIC case study 5 – scrum collapse injury rugby union A 27-year-old rugby union loose head prop was injured in the second half of a game during a scrum collapse receiving flexion and rotation of head to the left causing permanent C5/6 quadriplegia. Contributing factors to the scrum having collapsed included the wet conditions, the higher number of scrums and subsequent player fatigue.

NSWSIC case study 6 – scrum collapse injury rugby league A 22-year-old rugby league hooker was injured in the second half of a game during a scrum collapse causing permanent complete C5 quadriplegia. Testimony from team-mate "The scrum was packed in the normal fashion and the opposition team won the ball. On seeing that the ball was lost, I broke my bind with the hooker to

198 tackle the opposing halfback. The opposing pack was still interlocked and pushing forward and consequently came over on top of hooker forcing him to the ground in an awkward position".

NSWSIC case study 7 – scrum engagement injury A 32-year-old rugby union hooker was injured in the first half of the game during a scrum engagement receiving hyperflexion causing permanent C5/6 quadriplegia. He was not straightened up when the scrum engaged. His own side started to engage before the opposition was ready. The player slipped in wet conditions and his head was forced down onto his chest.

NSWSIC case study 8 – scrum ‘popping’ injury A 26-year-old rugby union hooker injured in second half of the game during a scrum drive receiving hyperflexion causing permanent C4/5 quadriparesis. His chin was forced onto his chest when the front rows lifted. A testimony from the patient said: "As soon as I saw the halfback pick up the ball my loose head prop let go of me and stood to the side of the scrum. At that instant, the opposing tight head prop grabbed me. The opposing team then started to put the second drive on. My second rower and tight head prop still had a tight grip on me and started to withstand the drive. I was squeezed between both packs and my feet were lifted off the ground pushing my head forward until my chin was forced hard up against my chest."

Illegal play

Examination of the NSWSIC case reports indicated the danger of players raising their knees or elbows in anticipation of impact, whether intentionally or inadvertently. Of all the rugby league tackle-related permanent brain injuries between 1984 and 1999, 31.8% (95%CI 13.9% – 54.9%) of cases were caused by contact with the knee or elbow.

There were five cases of illegal play in rugby league involving tackling head injuries; two spear-tackle associated injuries; one high-tackle associated injury, one injury from an intentional kick to the head, and one penalised head high tackle involving a raised elbow that resulted in a fatality.

There was at least one fatal case of an SCI in rugby union where a testimony alleges that scrums were not being set in accordance with the rules.

199 Injury management strategies for suspected serious injuries

There were eight NSWSIC cases of brain injury (30%, 95%CI 13.8% - 50.2%) in rugby league, that included two fatalities, where the injured continued to play after the injury incident. Of these, two players received severe blows to the head on two occasions, one player during the same game, and the other player in games one month apart. In rugby union, there was one case of fatal brain injury in which the player had sustained enough serious head injuries in past matches to warrant advice from a doctor that he stopped playing.

NSWSIC case study 9 – multiple concussive injuries A 29-year-old playing country rugby league in a forward position received a concussive blow in a tackle in the second half. He was then repositioned to the wing (a relatively protected position) and received a second head injury. He subsequently collapsed and died due to bilateral subdural haematomas and massive cerebral oedema despite resuscitation attempts. It is possible he suffered from the ‘second impact syndrome’.

NSWSIC case study 10 – no treatment for head injury A 27-year-old league player was tackled hard by two opponents in a trial match receiving a fatal head injury. The patient was hungover from drinking the previous evening, and did not take part in warming up or stretching. He did not experience the effects immediately and received no treatment. The injury resulted in death from Hind brain infarction with traumatic dissection of vertebral artery with thrombosis and acute pneumonia.

NSWSIC case study 11 – no treatment for injury A 37-year-old rugby league player was struck in the head by an opponent’s knee as affecting tackle receiving a head injury with blood clot. The player continued playing in the rest of game, receiving no treatment. He was assessed as having lost 85% of mental capacity and use of left arm, 70% left leg, 100% sexual function, 50% right leg, right arm and both eyes and developed epilepsy.

200 Protective equipment for serious injury

Only two brain injured rugby league players in the NSWSIC cases were reported to be wearing soft headgear. The effectiveness of protective equipment against serious injury cannot be evaluated without more consistently collected data on the use of protective devices among injured and non-injured players.

Risk factors (Intrinsic)

The following sub section analyses details gathered from the NSWSU and NSWSIC records concerning intrinsic risk factors associated with serious injury.

Weight/height and body type

As far as player size can be inferred from grade and level of competition played, injured players were often reported to be in the top grades of play and higher levels of competition (see subsequent section on level of competition).

There was one case identified where unsuitable body type was the possible contributing factor to serious injury. This case involved a scrum engagement injury resulting in complete C5 quadriplegia for a 19-year-old rugby union player in the hooker position. The attending physician's medical record reported the player as having a tall thin build unsuited to the hooker position. This was the player’s first game in the forward position so inexperience may have contributed to the injury. The injured player’s head struck the opposing hookers shoulder because he was slower at packing down to form the scrum and was caught with his head up. NSWSIC records verified eleven rugby league players and one rugby union player had fatal heart attacks from exertion and/or existing disease.

201 Physical preparation

As described, above, several fatal heart attacks were attributed to exertion which is potentially related to player preparation. The NSWSIC rugby league cases reported one fatal brain injury where the player was unsuitably prepared to play; he was found to be hung over from drinking the previous evening and did not take part in warm up or stretching sessions before a trial match. The effect of alcohol further contributed to injury with the player being unaware of his injury and receiving no immediate treatment. The NSWSIC analysis revealed one case of fatal myocarditis and one of fatal head injury (previously reported in Section two) where players were playing their second game of the day (both league). In at least one case of rugby union SCI, a high number of scrums and subsequent player fatigue was indicated as a contributing factor to scrum collapse.

Skills, experience and technique

The NSWSU/NSWSIC cases reported that in rugby league there were three SCI in players who were reported to be running head down when entering a tackle in which they were injured. Playing out of position, and therefore possible inexperience, in the front row was suggested as the contributing factor to injury in one rugby union and two rugby league players sustaining SCI in scrums.

Player position

Only a third of injured players had position information recorded and there may have been a bias associated with reporting front row players, commonly thought to carry greater risk of injury. Examination of the NSWSU/NSWSIC records suggested that in rugby league, forwards were most likely (79%, 95%CI 54.44% - 93.95%) to receive serious injury where player position was recorded. Rugby union forwards, particularly front row players, were most likely (92%, 95%CI 61.52% - 99.79%) to receive serious injury due to the risks associated with scrums.

202 Hookers appeared to be vulnerable to SCI sustained in scrums. Where position was stated, hookers were injured in 60% (95%CI 26.24% - 87.85%) of rugby league SCI and 42% (95%CI 15.17% - 72.33%) of rugby union SCI.

Age

Of the 102 rugby football players who sustained a serious injury or fatality between 1984 and 1999, the average age was 24 years (SD ±6.7, Median = 23 years) with 52% of injured players aged between 20 to 29 years (Table 19). Ages ranged from 13 years to 46 years old. Over 50% of injuries were to players under 24 years of age, however, there were few injuries to players below 16 years old or over 40 years old. Of the two players injured under 15 years of age one was a 13-year-old rugby league player who collapsed from exertion while playing touch football during training and died from cardiac failure due to hypertrophic obstructive cardiomyopathy (further details in section on NSWSIC rugby football fatalities). The other was a 14-year-old rugby league player injured when tackled, developing a right subdural haematoma that resulted in considerable loss of mental capacity.

Table 19: Descriptive age analysis of seriously injured rugby football players in NSW 1984 to 1999a Age group Percent 14 and under 2 15-19 26 20-24 30 25-29 22 30-34 14 35-39 4 40 and over 2 a Source: NSWSU medical records and NSWSIC case records

Country Vs Metropolitan competitions

Team membership information was available for 63% (22/35) of all rugby union serious injuries and fatalities, and 61% (41/67) of all rugby league NSWSU/NSWSIC injury and fatality cases (Table 20). The proportion of participants in metropolitan and country teams in rugby league and in rugby union

203 over the study period was reflected in the proportions of all serious injuries and fatalities accounted for by metropolitan and country teams. Rugby league participants have a fairly even spilt between country (52%) and metropolitan (48%) teams and a similar distribution of serious injuries and fatalities. There are a greater number of metropolitan team (62%) participants in rugby union reflected in 64% of serious injuries and fatalities occurring in players from metropolitan teams (Table 20).

The incidence rate of all serious injuries and fatalities for rugby league split by team region was 3.2 (95%CI 0.2 – 4.8) per 100,000 estimated metropolitan team players, compared to 2.7 (95%CI 1.7 – 4.3) per 100,000 estimated country team players in NSW over the study period 1984 to 1999 (RR= 1.14, 95%CI 0.62 – 2.10, X2 = 0.17, P>0.05). The incidence rate of all serious injuries and fatalities for rugby union was 2.6 (95%CI 1.4 – 4.3) per 100,000 estimated metropolitan team players. This was compared to 2.4 (95%CI 0.1 – 4.7) per 100,000 estimated country team players in NSW over the study period 1984 to 1999 (RR= 1.07, 95%CI 0.45 – 2.56, X2 = 0.03, P>0.05). However it must be considered that evacuation procedures for country injury locations may bias results underestimating the incident of country injuries as regions close to state borders may be taken to interstate facilities and not be counted in the survey.

Table 20: Team region of all injured players by code (where information available) All SCI, brain injuries and fatalities 1984-1999 Rugby league teams Rugby union teams Country team injuries/fatalities – N, % (per 20 (49%) 8 (36%) 100,000 estimated players per year, 95%CI)b(2.7, 95%CI 1.7 – 4.3) (2.4, 95%CI 0.1 – 4.7) City team injuries/fatalities – N, % (per 21 (51%) 14 (64%) 100,000 estimated players per year, 95%CI)b(3.2, 95%CI 0.2 – 4.8) (2.6, 95%CI 1.4 – 4.3) Estimated players 1984-1999 c Rugby league teams Rugby union teams Country teams participants 721970 (52%) 332123 (38%) City teams participants 666434 (48%) 541886 (62%) Total 1388404 (100%) 874009 (100%) a Source: Injuries - NSWSU medical records/NSWSIC case records. b Annual incidence rate per 100, 000 estimated players c Membership - personal communication - NSWRL & NSWRU

204 Level of competition

The level of competition was recorded in 32% of rugby union and 34% of rugby league injuries, and was commonly (78%) 1st/A grade or reserve 1st/A grade even if only at social club level. However, this may reflect a reporting bias, as the team grades of all injured players were not known. At least two brain injuries (one fatal) and eight spinal unit admissions were attributed to elite rugby league players over the study period although 6/8 spinal unit admissions had substantially resolving injuries so as not to fall under this studies definition of permanent serious injury.

Time in game and season

Forty percent of all serious injuries and fatalities reported by the NSWSU/NSWSIC recorded the time in game the incident occurred. There was no significant difference (X2=4.8 P>0.05) in the number of serious injuries and fatalities occurring in the second half of the game (58%, 95%CI 40.8% - 73.0%) compared with the first half for both codes combined. Brain injuries (41.6%, 95%CI 15.2% - 72.3%, X2=6.2 P>0.05), SCI (63.6%, 95%CI 40.7% - 82.8%, X2=4.2 P>0.05) and fatalities (66.6%, 95%CI 22.3% - 95.7%, X2=5.6 P>0.05) occurred no more frequently in the first half compared with the second half for both codes combined. Brain injuries were no more likely (X2=5.9 P>0.05) to occur in the first half than SCI or fatalities.

The date of injury was recorded for 98% of all serious injury and fatality cases. As would be expected, injuries were found to mostly occur (89%) between April and August, covering the main playing season for most teams in Australia (Table 21). No single month from April to August was found to be significantly more likely associated with serious injuries and fatalities (X2=10.8 P>0.05).

Table 21: Distribution of all rugby football serious injuries and fatalities over months of the year Month Feb Mar Apr May Jun Jul Aug Sept Percent 2 8 22 14 20 18 15 1 a Source: Injuries - NSWSU medical records/NSWSIC case records.

205 Position on field

Of the 26 cases where position was recorded, commonly (81%, 95%CI 60.7% - 93.5%) game play preceding injury was close to the try line. However, players’ position on the field at the time of accident was not consistently recorded in enough cases to assess association with various injuries.

Training

Few cases (2/102) of serious injury or mortality were reported to have occurred during training. Both codes recorded one such serious injury during training. In rugby league, a 13-year-old collapsed from exertion while playing touch football in training. He died from cardiac failure due to hypertrophic obstructive cardiomyopathy. In rugby union, a 24-year-old received SCI when tackled in a ruck during training leaving him an incomplete C5 quadriplegic..

206 SECTION THREE DEVELOPMENT OF RUGBY FOOTBALL

Style of football played in modern times (post 1989) compared to that played in earlier periods (pre 1958)

A number of the elements in the style of play were examined as a potential risk of injury. This section examines styles of play over the last century for observable changes using archival footage spanning back to the 1930s.

The observational period was split into upper and lower 20th percentiles for comparisons of early and modern play i.e. chronologically, the first fifth of available incidents are compared with the last fifth. This avoided an arbitrary split in the middle of the period (or at some other stage), as there was no clear point to mark the development of earlier styles of play into modern ones. The terms ‘earlier’ and ‘modern’ will be used in the following section to refer to the two periods compared: pre 1958 and post 1989.

The many thousands of incidents were viewed (the basic unit of analysis) to allow for a large enough sample size to test for significant associations. Re-analysing the data for two half chronological periods, pre 1970 and post 1970 led to the same conclusions being drawn from the results.

Around 80% of all football video footage observed involved tackles. Ball carrying style, number of tacklers, and the level/angle of initial impact were recorded for these incidents. The level and angle of subsequent impacts, as occurs with multiple tacklers, was also recorded (see Appendix E for coding manual definition of terms and recording system).

Number of tacklers

The game footage observed found tackling incidents that involved more than one opponent tackling player (33.7, 95%CI 30.2% - 37.3%) were more frequently recorded in the modern games viewed (P<0.001) (Table 22). An elaboration

207 analysis of level of impact by game period controlling for rugby code found the same relationship across all partial tables. However, the association for rugby union was much weaker and was not found to be significant (P>0.05). This suggests an intervening relationship with rugby code for the association between number of tacklers and game period. (See Appendix H for complete tables.)

Table 22: Number of tacklers by perioda Tacklers Pre 1958 tackling incidents N = 692 Post 1989 tackling incidents N = 700 Observed % (95%CI) Observed % (95%CI) One player 86.4 (83.6% - 88.8%) 66.3 (62.6% - 69.7%) Multiple 13.6 (11.1% - 16.3%) 33.7 (30.2% - 37.3%) Total 100 100 a Source: Screensound Australia and Sports Recording Services football video collections Fisher’s exact test P<0.001, N= 1392

Level of impact

The video analysis observations found head/neck tackling incidents were less likely (P<0.001) in modern games (3.6%, 95%CI 2.3% - 5.2%) than in earlier games (13.1%, 95%CI 10.6% - 16.0%) (Table 23). Players were most likely (P<0.001) to be tackled at chest level (46.9%, 95%CI 43.1% - 50.7%) in modern games and at waist level in early games (41.6%, 95%CI 37.7% - 45.4%). These relationships held for the level of second impact (P<0.001). The use of Chi-square to test the association of third impact level with period of game was precluded. This is because an unacceptably high percentage of expected cell counts less than five found for that contingency table was found and it was not valid to collapse any categories.

An elaboration analysis of level of impact by game period was conducted to control for rugby code. It found the relationship in the partial table for rugby union was weaker than in rugby league table although both displayed the same relationship as the original table. This suggests an intervening relationship with rugby code in the association between level of tackle and game period. The use of Chi-square tests were precluded for the elaborated tables of the direction of the second and third impact because of the high percentage of expected cell counts less than five. (See Appendix H for complete tables.)

208 Table 23: Level of initial tackle by perioda Level of 1st Pre 1958 tackling incidents Post 1989 tackling incidents impact N = 645 N = 699 Observed % (95%CI) Observed % (95%CI) Legs 22.5 (19.3% - 25.9%) 29.9 (26.5% - 33.4%) Waist 41.6 (37.7% - 45.4%) 19.6 (16.7% - 22.7%) Chest 22.8 (19.6% - 26.2%) 46.9 (43.1% - 50.7%) Head/Neck 13.1 (10.6% - 16.0%) 3.6 (2.3% - 5.2%) Total 100 100 a Source: Screensound Australia and Sports Recording Services video football video collections Pearson Chi-square P<0.001, N= 1344

Direction of impact

Because of the dangers associated with head on impacts, early and modern games were examined for direction of impact to the ball carrier. The video analysis results indicate that the proportion of impacts to the front half of the ball carrier (41.1, 95%CI 37.4% - 44.8%) was greater (P<0.001) in modern games compared to earlier games (21.5%, 95%CI 18.3%- 24.8%) (Table 24). There was no relationship between the directions of second impacts (P>0.05) with period of game. There were insufficient cell counts to calculate the significance of the relationship between third impacts with period of game.

An elaboration analysis of the direction of first and second impact by game period controlling for rugby code found the same relationship across all partial tables and the same relationship in the partial table as the original table. This suggests a direct relationship between direction of tackle and game period. The low cell counts for elaborated tables controlling for rugby code in the association between the direction of the third impact and game period precluded the use of Chi-square tests (see Appendix H for complete tables).

Table 24: Direction of initial tacklers impact by perioda Angle of Pre 1958 tackling incidents N = 643 Post 1989 tackling incidents N = 696 1st impact Observed % (95%CI) Observed % (95%CI) Front half 21.5 (18.3%- 24.8%) 41.1 (37.4% - 44.8%) Sides 19.9 (16.8% - 23.2%) 12.2 (9.8% - 14.8%) Back half 58.6 (54.7% - 62.4%) 46.7 (42.9% - 50.4%) Total 100 100 a Source: Screensound Australia and Sports Recording Services video football video collections Pearson Chi-square P<0.001, N= 1339 209 Impact speed

Comparisons of qualitative observations of rugby football film and video footage of early and modern games suggest that modern game players, in both codes, tend to build up running momentum before impacting opposition players. Whereas early game players appear to run less before impacting other players. This can most obviously be observed by comparing the depth (amount of vertical field position covered) of both defending and attacking lines as they prepare to engage opposition players. Footage of games in earlier times tends to show players, who were not in the immediate vicinity of the ball carrier, to be stationary or moving at slow speeds. Forwards were often bunched together for long periods after scrummaging leaving fast running plays to the backs. This was in contrast to modern play where forwards hastily return to offensive and defensive positions after scrums in both rugby league and union. Head on impacts are likely to have a greater relative velocity between players compared to impacts where the ball carrier is being chased (see Chapter Two - Literature Review, section six). As head on impacts have been found to be more likely in modern games (Table 24), a greater relative velocity between players on modern games might also be inferred.

Ball carrying style

Video footage found modern players carried the ball to their chest before being tackled more commonly (83.6%, 95%CI 80.6% - 86.3%, P<0.001) than in earlier players (34.2%, 95%CI 30.5% - 37.9%). Earlier players were most likely to hold the ball in front of the body with two hands (51.3%, 95%CI 47.4% - 55.1%, P<0.001) (Table 25).

An elaboration analysis of ball carrying style by game period controlling for rugby code found the same relationship across all partial tables as in the original table. This suggests a direct relationship between ball carrying style and game period (see Appendix H for complete tables).

210 Table 25: Ball carrying style by perioda Ball carrying style Pre 1958 tackling Post 1989 tackling incidents incidents N = 655 N = 697 Observed % (95%CI) Observed % (95%CI) Two Hands 51.3 (47.4% - 55.1%) 13.6 (11.1% - 16.4%) Swinging 3.1 (1.8% - 4.6%) 0.6 (0.1% - 1.4%) Waist 11.5 (9.1% - 14.1%) 2.2 (1.2% - 3.5%) Chest 34.2 (30.5% - 37.9%) 83.6 (80.6% - 86.3%) Total 100 100 a Source: Screensound Australia and Sports Recording Services video football video collections Pearson Chi-square P<0.001, N= 1352

Dangerous scrums and tackling technique

The examination of video archives was not able to establish any statistical difference in the proportion of dangerous scrums observed (including wheeling, collapsing or popping) between early and modern games (P>0.05). Nevertheless, potentially dangerous scrums did account for 25.3% (95%CI 21. 9% - 28.9%) of all scrums observed for the entire period. Of these dangerous scrums, 48% (95%CI 40.3% - 56.5%) were wheeling, 49% (95%CI 40.9% - 57.2%) collapsing and 3% (95%CI 0.7% - 6.5%) popping.

211 Comparisons of rugby football players’ age, height and weight, pre 1950 with post 1998

Players in elite international test teams (Australia, England, New Zealand, South Africa) from the 1920s to 1950s were contrasted with comparable teams from 1999-2000 for average age, height, and weight.

Elite league players have increased in height from an average of 177.8cm (SD ± 5.8) to 182.7cm (SD ±6.7) over the century (P<0.001) (Table 26). Their average weight has increased from 80.5kg (SD ± 8.2) to 94.2kg (SD ±9.3) (P<0.001). Similarly, elite union players have increased in height from an average of 182.8cm (SD ±14.2) to 186.3cm (SD ±7.5) (P<0.05) and in weight from an average of 85.0 kg (SD ±10.8) to 100.2kg (SD ±12.5) (P<0.001) (Table 26).

Table 26: Separated codes comparing age, height and weight for early and modern playersa CODE FACTOR PERIOD N Mean SD T-test League AGE Early 106 24.8 3.2 Sig. Modern 518 25.2 3.7 P>0.05 HEIGHT Early 122 177.8 5.8 Sig. Modern 520 182.7 6.7 P<0.001 WEIGHT Early 169 80.5 8.2 Sig. Modern 520 94.2 9.3 P<0.001 Union AGE Early 88 25.8 3.5 Sig. Modern 101 27.6 3.3 P<0.001 HEIGHT Early 88 182.8 14.2 Sig. Modern 101 186.3 7.5 P<0.05 WEIGHT Early 88 85.0 10.8 Sig. Modern 101 100.2 12.5 P<0.001 a Source: Early player data - ES.Marks and Davis sporting archive collections. Modern player data – Rugby football websites

Rugby union players were on average significantly (P<0.05) older, taller and heavier than league players in both early and modern teams (Table 27).

212 Table 27: Comparisons of age, height and weight by codes and perioda PERIOD FACTOR CODE N Mean SD T-test Early AGE League 106 24.8 3.2 Sig. Union 88 25.8 3.5 P<0.05 HEIGHT League 122 177.8 5.8 Sig. Union 88 182.8 14.2 P<0.01 WEIGHT League 169 80.5 8.2 Sig. Union 88 85.0 10.8 P<0.01 Modern AGE League 518 25.2 3.7 Sig. Union 101 27.6 3.3 P<0.001 HEIGHT League 520 182.7 6.7 Sig. Union 101 186.3 7.5 P<0.001 WEIGHT League 520 94.2 9.3 Sig. Union 101 100.2 12.5 P<0.001

Position details were available for the sampled players allowing comparisons between forwards and backs. For both codes backs were, on average, a year or two younger than forwards (P<0.001), 4 to 9cm shorter (P<0.001), and around 10kg heavier (P<0.001)(Table 28).

Table 28: Comparisons of age, height and weight by player position and codea CODE FACTOR POSITION N Mean SD T-test League AGE Forward 328 25.7 3.5 Sig. Back 255 24.5 3.4 P<0.001 HEIGHT Forward 323 183.7 6.7 Sig. Back 262 180.1 6.3 P<0.001 WEIGHT Forward 346 95.9 9.9 Sig. Back 276 86.1 8.7 P<0.001 Union AGE Forward 100 27.9 3.4 Sig. Back 89 25.5 3.2 P<0.001 HEIGHT Forward 100 189.2 12.6 Sig. Back 89 179.5 6.3 P<0.001 WEIGHT Forward 100 102.1 10.9 Sig. Back 89 83.0 9.5 P<0.001 a Source: Early player data - ES.Marks and Davis sporting archive collections. Modern player data – rugby football websites

Early rugby union forwards are on average over two years older than backs from the same period (P<0.001) as are modern rugby union forwards and backs (P<0.01) (Table 29). Early rugby league forwards are on average over one year older than backs from the same period (P<0.05) as are modern rugby league forwards and backs (P<0.001).

213 Early rugby union forwards are on average over 8cm taller than backs from the same period (P<0.001) as are modern rugby union forwards and backs (P<0.001) (Table 29). Early rugby league forwards are on average over 3cm taller than backs from the same period (P<0.001) as are modern rugby league forwards and backs (P<0.001). Early rugby union forwards are on average over 16kg heavier than backs from the same period (P<0.001) as are modern rugby union forwards and backs (P<0.001). Early rugby league forwards are on average over 9kg heavier than backs from the same period (P<0.001) as are modern rugby league forwards and backs (P<0.001).

Table 29: Comparisons of age, height and weight by player position, code and perioda CODE FACTOR POSITION PERIOD N Mean SD T-test League AGE Forward Early 54 25.3 2.9 Sig. Back 42 23.9 2.9 P<0.05 HEIGHT Forward 50 180.3 4.4 Sig. Back 48 176.1 5.4 P<0.001 WEIGHT Forward 73 85.3 6.7 Sig. Back 62 76.3 6.6 P<0.001 Union AGE Forward 44 27.1 3.6 Sig. Back 44 24.5 3.0 P<0.001 HEIGHT Forward 44 188.5 17.3 Sig. Back 44 177.1 6.3 P<0.001 WEIGHT Forward 44 93.0 7.4 Sig. Back 44 76.9 7.1 P<0.001 CODE FACTOR POSITION PERIOD N Mean SD T-test League AGE Forward Modern 274 25.8 3.6 Sig. Back 213 24.7 3.5 P<0.001 HEIGHT Forward 273 184.3 6.8 Sig. Back 214 181.0 6.1 P<0.001 WEIGHT Forward 273 98.7 8.6 Sig. Back 214 88.9 7.1 P<0.001 Union AGE Forward 56 28.5 3.2 Sig. Back 45 26.5 3.1 P<0.01 HEIGHT Forward 56 189.7 7.0 Sig. Back 45 181.9 5.4 P<0.001 WEIGHT Forward 56 109.2 7.2 Sig. Back 45 88.9 7.5 P<0.001 a Source: Early player data - ES.Marks and Davis sporting archive collections. Modern player data – Rugby football websites

214 When teams were examined by country, most of the observed differences remained consistent for Australia, South Africa and New Zealand but not for English teams. English players’ average weight significantly increased from around 82kg (SD ±8.4) to 90kg (SD ±9.5) (P<0.0001) but their age and height remained around 24 years (Early SD=3.0, Modern SD= 4.8) and 178cm tall (Early SD ±5.4, Modern SD ±9.5). Comparisons between players (all sampled from elite level club teams or representatives in state and international test teams) of all the countries revealed that English players were, on average, usually slightly younger, and smaller than players from comparable teams in other countries.

215 CHAPTER FOUR SUMMARY

The estimated annual incidence rate of SCI for rugby league players was 1.9 per 100,000 estimated players per year, and 3.3 per 100,000 estimated rugby union players per year over the study period 1984 to 1999.

There was no significant change in the incidence of rugby league related serious spinal cord injuries (1984-1999), fatalities (1984-1999) or serious head injuries (1984-1999) (P>0.05). There was a small but significant decline in rugby union related serious spinal cord injuries (1984-1999, P<0.05). However, the relative risk of spinal cord injury was 1.34 times greater for rugby union compared to rugby league over the entire study period (P<0.05).

The annual incidence rate of brain injuries for rugby league players was 2.1 per 100,000 players per year (1984-1999) and 3.2 per 100,000 rugby union players per year (1984-1987). There was no significant reduction (P> 0.05) in the incidence rate of brain injuries identified by the NSWSIC case files over the period 1984 to 1999 for rugby league.

The analysis indicated that 78% of rugby league SCI were tackle related with 90% of these to the ball carrier. The remaining 22% of SCI in rugby league were contributed by scrums 19% and other activities. The number of scrum related SCI in rugby league dropped from four during the 1980s to one during the 1990s, while tackle related SCI stayed consistent with ten during the 1980s and eleven during the 1990s.

Conversely, in rugby union scrums caused significantly (P<0.05) more SCI in rugby union, 45%, than in rugby league. Of all rugby union SCI, tackles accounted for 35% of which, 78% were sustained to the ball carrier.

216 It was often difficult to distinguish whether an injury occurred in a tackle or in an ensuing maul or ruck, however it was clear that 17% of SCI in union were associated with mauls and rucks. Scrum related SCI in rugby union declined in number by more than half in the 1990s when compared to the 1980s. However, the proportion of all SCI accounted for by scrums was not significantly (P>0.05) associated with these periods dropping from 47% during the 1980s to 40% during the 1990s.

Brain injury cases were all associated with either tackles or collisions. Rugby league players contributed 90% of the available brain injury data. These rugby league players reported 67% of brain injuries were sustained by the ball carrier and 33% were sustained by the tackler. There was a significant (P<0.05) difference in the proportion of non-fatal brain injuries associated with being tackled (76%) compared with doing the tackling (24%) in rugby league. Rugby union reported 67% of head injuries sustained by the ball carrier and 33% as an unspecified collision.

The fatality data collected were all associated with cardiac arrhythmia and the sequelae of severe spinal cord or brain injuries. Of the 19 fatalities occurring in rugby league, 63% were associated with heart failure, 67% of these accounted for by exertion and the remainder by injuries in tackles (33%). Two cases of exertion were found to be associated with endogenous morbid pathological process accompanying pre-existing cardio vascular disease. One case of exertion was found to be exacerbated by the ingestion of the anabolic steroid oxymesterone.

Of the five fatalities occurring in rugby union, 20% were associated with heart failure accounted for by exertion. The remaining fatalities were caused by brain injuries from being tackled (40%) or SCI associated with a scrum or maul (40%).

217 Findings suggest increased risk of serious injury associated with multiple tacklers, high tackles, tackling with head down, illegal and dangerous play, inappropriate body type, inexperienced players, unfit/unprepared players, and the front row positions in the scrum. The force of impact between participants is a key causal risk factor for serious injury that intermediates many other risk factors.

Elite rugby league and union players from 1999-2000 teams were significantly heavier (P<0.05) and taller (P<0.05) than players pre 1950. Players in modern elite games post 1989 were more likely to be tackled by multiple opponents (P = 0.000), tackled head on (P<0.05), at chest level (P<0.05) and at greater speeds than their earlier counterparts pre 1958. However, modern players appear to be no more aggressive or perpetrate greater foul play than their predecessors.

218 CHAPTER FIVE DISCUSSION

Relative contribution of various organised sports and recreational activities to serious injury and fatalities in NSW

The inherent risk of serious injuries associated with the rugby football codes and other high risk sport and recreational activities has been widely acknowledged in several epidemiological studies of sports injury (Berry et al. 2006; Boufos et al. 2006; Carmody et al. 2005; Haylen, 2004; O'Connor & Cripps 1999; O'Connor, 2000a; O’Connor, 2002; O'Connor, 2002a; O'Connor & Cripps 1996; Watt & Finch 1996; Cripps, 2000; Armour et al. 1997; Scher 1998; Taylor & Coolican 1988; Taylor & Coolican 1987; Silver 1992; Yeo 1998a; Wilson et al. 1996; Rotem et al. 1998). The rugby codes often have more participants than other high risk sports and therefore not only have a relatively high incidence of serious injuries but also account for the largest number of seriously injured participants compared to other high risk organised sports. The rugby codes consequently pose the largest burden on society in respect to serious and permanent injuries of all organised sports in NSW.

NSWSIC 1980-1999

The highest risk sports covered by the NSWSIC were rugby league, rugby union, motor cycling, amateur horse race riding and polo, as indicated by their premium rates (see Appendix D). Of all compensated brain, neck and fatal injury between 1980 and 1999 contributed by every sport covered by the NSWSIC, rugby league and union codes accounted for the greatest proportion of compensated cases. However, a comprehensive analysis of rugby union head injuries and fatalities over the entire study period was not possible as the NSWSIC was the only source of consistent data and was devoid of union player records after 1990.

219 Risk exposure considered in terms of actual playing time at risk was not calculable retrospectively over the study period. When participant populations were accounted for, the incidence of brain injury in various sports was only higher than the rugby codes for cycling and Taekwondo due to their small participant numbers. None of the Australian studies reviewed compared the risk of serious brain injury between organised sports. Several that assessed the relative risks of concussion between organised sports found that the rugby codes, particularly rugby league, were the highest risk sports for sustaining concussion injuries (NHMRC 1994; Northern Sydney Area Health Service 1997; Neumann et al. 1998; Gibbs 1994; Seward et al. 1993; Gabbett, 2000; MacDougal & Osbourne 1992).

The NSWSIC data also demonstrated that the rugby codes had a higher incidence rate of SCI than soccer, which was the only other sport responsible for a compensable SCI. This is consistent with other Australian SCI studies (Yeo 1993; O’Connor & Cripps 1999; O’Connor, 2000a; O'Connor, 2001a; O'Connor, 2002) which have found rugby league and union to be the greatest cause of SCI among organised sports. Spear fishing was the only sport covered by the NSWSIC with a significantly higher annual incidence rate of fatalities than the rugby codes. This was due to spear fishing's very small participation numbers, although this still indicates the inherent high risk of fatality associated with this activity. There is a dearth of literature providing epidemiological study of spear fishing and rock fishing injury and fatality requiring attention from future research.

220 NSWSU admissions 1984-1994

Recreational diving had the greatest proportion of admissions to the spinal units (including both resolving SCI injuries and permanent neurological deficits). While recreational diving accounts for the greatest frequency of sport and recreational related spinal unit admissions in NSW, the participant population is potentially drawn from the entire state population. Unlike the rugby football codes and other high risk organised sports, recreational diving is an unorganised activity which often occurs in unsupervised areas such as private pools and natural water bodies. This detrimentally affects the activities’ amenability to injury prevention strategies outside of public education campaigns.

One quarter of admissions to the two spinal units between 1984 and 1994 were contributed by the rugby football codes. If the unspecified rugby football admissions are combined for both codes admissions, rugby football had the second greatest proportion of admissions to spinal units. This was followed by equestrian activities, water skiing, bicycle riding and surfing. However, where participant population estimations were possible, equestrian activities, rugby league and rugby union had the highest incidence of spinal unit admission. As would be expected, these findings about the relative contributions of various sport and recreational activities to spinal unit admissions in NSW, are consistent with other studies of SCI in sport and recreational activities in Australia (Yeo 1993; O'Connor & Cripps 1999; O'Connor, 2000a; O'Connor, 2002a; O'Connor & Cripps 1996) which are also based on data gathered from major spinal units.

221 Rugby football SCI, Brain injury and fatality 1984-1999

This study arguably constitutes a census of spinal unit admissions and SCI for rugby football players in NSW 1984–1999. This process confirmed that 95% (O'Connor, 2002; National Injury Surveillance Unit, 9 November 2002, pers. comm.; Dr Rutkowski,. Royal North Shore Hospital spinal unit, 16 March 1998, pers. comm.; Dr Engel, Prince Henry Hospital spinal unit, 20 March 1998, pers. comm.; Dr Epps, Sydney Children's Hospital, 8 April 2003, pers. comm.; Dr Waugh, New Children's Hospital Westmead, 20 December 2002, pers. comm.) of acute severe spinal cord injury cases among NSW rugby football players over the study period were included in the analysis.

While the spinal unit admissions did not all necessarily result in permanent neurological deficit (i.e. SCI), the causes of rugby football spinal admissions that did not result in permanent spinal cord injury were comparable to those that did result in SCI. The similar mechanisms involved with these ‘near miss’ injuries mean that they offer a useful indicator of the risk of such injuries associated with rugby union and league football. The 57 SCI cases between 1984 to 1999 associated with both codes, displayed similar mechanisms of injury to that shown in earlier studies in terms of the internal process of injury involved (i.e. flexion/extension/ rotation causing damage to cervical vertebrae and spinal cord) (Taylor & Coolican 1987; Silver 1984; Silver & Stewart 1994; Wilson et al. 1996; Rotem et al. 1998). As is noted in these studies, damage was most common to the C4/5 level of spinal cord in both codes reflecting the mostly severe nature of injuries observed.

The relative risk of spinal unit admission and SCI over the entire study period was significantly greater for rugby union compared with rugby league players. This reflects the high risk of injury associated with rugby union scrums, rucks and mauls compared to rugby league where these aspects of the game are not emphasised or do not exist.

222 Continued variance in these relatively small numbers makes conclusive interpretation of these figures difficult, however, current evidence suggests that rugby union, unlike rugby league, has sustained a significant reduction in SCI over the entire study period. The small but significant decrease in the incidence of these SCI in rugby union found in this study between 1984-99 in NSW, signifies the effectiveness of numerous rule changes and safety programs (Rotem et al. 1998; Rotem & Davidson, 2001; Silver 1992; Yeo 1998a) aimed at improving player safety particularly in the scrum, ruck and maul since the 1980s in Australia. It is worth noting however that NSWSIC/NSWSU data indicated that while scrum related SCI in rugby union declined in number by more than half in the 1990s, when compared to the 1980s, the proportion of all SCI accounted for by scrums did not significantly decline over the study period. This suggests that reductions in scrum injuries alone do not explain the overall decrease in SCI observed in rugby union players over the study period. The conclusion that rugby union SCI has significantly reduced over the study period must also be tempered with the potential bias introduced by the loss of a data source in the latter years of the study. The NSWSIC lost many of its rugby union subscribers to private insurance arrangements in the mid nineties. However, this data source acted primarily only as a supplementary source of info and case cross check to spinal unit data for the SCI analysis component of the study.

The analysis of NSWSIC and NSWSU cases indicates that the incidence of rugby league related SCI in NSW did not significantly decrease from 1984 to 1999 and remained relatively constant. Rugby league's failure to demonstrate a reduction in SCI over the study period, unlike rugby union, may be partly attributable to the significantly greater relative risk of SCI for rugby union compared with rugby league players at the beginning of the study period. This left greater room for reducing rugby union SCI than for rugby league over the study period. Additionally, variations in rules and playing style in rugby league mean that, unlike in rugby union, the tackle rather than the scrum, ruck and maul is the major source of SCI and thus the league code was less amenable to the SCI prevention strategies utilised by rugby union.

223 The evidence for rugby union brain injuries and fatalities is incomplete due to limited NSWSIC coverage, however the incidence of rugby league related brain injuries and fatalities clearly did not decrease in NSW from 1984 to 1999 despite the introduction of several serious injury prevention strategies.

The NSWSU/NSWSIC data analysed indicate permanent spinal cord and head injury in both codes of football, while infrequent, still consistently occur. Anecdotal newspaper reports on the frequency of serious football injuries in the years 2000 to 2002 suggest they may have been a particularly bad years for both codes with regards to serious SCI and brain injury. A recently published study of football injuries based on the ASCIR monitoring of NSW spinal units, found only a small, but non-significant decline in the incidence rate of SCI in rugby union and rugby league from 1986 to 2003 (Berry et al. 2006).

Risk factor analysis

The association of many, often interrelated, extrinsic and intrinsic risk factors with the serious injuries and fatalities observed in the NSWSIC/NSWSU analysis is a complex relationship. The ability to assign causality to these risk factors was strictly limited by the descriptive nature of this study. However, a form of triangulation is achieved through assessing various interrelated risk factors from diverse disciplines, theoretical frameworks and bodies of evidence. These include both conclusions drawn from the original data collected in this study and from the literature reviewed.

Concurrence in these sources support the validity of the conceptual framework that has been developed. This conceptual framework seeks to explain the complex relationship between risk factors and identify the fundamentally important risk factors for serious injury. This in turn is used as the basis for evaluating and developing serious injury prevention strategies.

224 Interrelationship of risk factors

Although the Haddon Matrix has been utilised as a research tool by injury epidemiologists in a range of injury prevention settings (Conroy & Fowler 2000; Short 1999; Runyan 1998; Chorba 1991), the application in this study to serious rugby football injuries has provided a framework for conceptualising risk factors that permits consideration of temporal issues in rugby football injury prevention, and likens these injuries to epidemiological concepts of disease (see Chapter Two, Section Five – Risk factors for serious injury in rugby football). This framework has provided a systematic point of application for research and the development of prevention strategies as has been its use in other studies (Lett et al. 2002).

This approach has led to the development of several conclusions about the importance of various risk factors for serious injury. Key to these conclusions are that assessment of all risk factors associated with serious injury in this study and in the literature suggest that the most fundamental risk factors for serious rugby football injury are speed of game, force and level of impact.

The interrelationships of risk factors for acute and chronic injuries in elite AFL players have been conceptualised in a theoretical flow chart model by Norton et al. (2001). Adapting this idea to risk factors for serious injury and fatality in rugby football allows the hypothesis of a novel theoretical model (Figure 13).

This model illustrates the concepts discussed and conclusions drawn about the relative importance of various risk factors for serious injury and fatalities in rugby football. The model splits extrinsic and intrinsic risk factors and accounts for the temporal dimension of injury prevention (i.e. pre and post injury event). The arrows show direction of influence between the risk factors and how they ultimately influence serious injuries or fatalities and the outcome for the players.

225 Providing an overview of the complex web of risk factors makes it possible to glean a novel perspective on the central factors that influence the occurrence of serious or fatal injuries in rugby football. This perspective is critical to formulating a strategic approach to serious injury prevention in rugby football and determining where resources are best allocated.

It is important to note the interrelated nature of many of the factors. The force, direction and level of impact all contribute directly to the likelihood of injury. These factors are in turn influenced by game speed, playing style, player size and fitness. Another example is the relationships between age, grade, experience, skill, body size and type. These factors all interact to affect the likelihood of injury occurring, its severity and subsequent long-term outcome for the injured player. It is hypothesised that many risk factors relate to a few key factors that are most directly responsible for injury and fatality risk. According to this hypothesis the key factors that underpin the determinants of injury and fatalities are the speed of play and the force of impact (Figure 13).

226 Figure 13 – Interrelationships of risk factors for serious injury and fatality in rugby football

227 Extrinsic risk factors

Scrums, rucks, mauls and tackles

The analysis of NSWSU/NSWSIC cases confirms other studies’ findings that rugby union players appear to be at a greater relative risk of spinal cord injury than rugby league players mainly due to the dangers of intense scrum, ruck and maul activity exclusive to rugby union (Taylor & Coolican 1987; Armour et al. 1997; Silver 1984). However, as awareness has grown of the dangers of these plays, preventative measures aimed at ‘de-powering’ and controlling these engagements have successfully reduced these injuries.

Scrums are capable of producing much greater forces than needed to damage a player’s spine (Wetzler et al. 1996; Bauze & Ardran 1979). Engagement probably poses the greatest danger of serious injury due to the impulsive force created, but subsequent danger in collapse or popping still carries significant danger. This finding was made in Taylor & Coolican’s study of Australian footballers (1987) and is borne out of the current analysis of NSWSU/NSWSIC scrum injury cases.

While some injuries were superficially attributed to scrum collapse, further investigation of player testimonies attested that injury occurred before collapse. Instability in the complex forces working in an active scrum are dangerous as they increase the likelihood of scrum collapse or ‘popping’ which have often been the cause of spinal cord injury. Scrums require sufficient training to allow safe participation, as there are many safety skills required. Inexperienced players and those with inappropriate body types such as long thin neck, may be at greater risk of serious injury in scrum activity. Players in the front row, principally the hooker, are at the greatest risk of serious injury in a scrum.

228 As previously noted, the NSWSU/NSWSIC analysis and previous studies have found that scrums account for a smaller proportion of SCI in rugby league than in rugby union, which appears to be a result of variations in rules in which less emphasis is placed on the use of scrums. As this aspect of the game is less emphasised, the forces expended in them are greatly reduced, as are their number and duration.

The findings of NSWSU/NSWSIC analysis substantiate other studies’ (Taylor & Coolican 1987; Milburn 1993) findings of the increased risk of serious injury posed to hookers and inexperienced players in scrums. Illegal tactics such as intentionally collapsing a scrum, popping a player or ‘crotch binding’ are also indicated as particularly dangerous.

The current sequential scrum-engagement technique used at junior levels with increased control by referees has proven somewhat effective in reducing serious injuries in this phase of play for junior players. While changes to under-19 rules introduced in 1985 and altered laws for adults were mostly designed to prevent scrum collapse, it has been argued (Taylor & Coolican 1988) that these are inadequate, because scrum collapse is less significant than the danger of impact forces generated by the engagement of two packs. A more effective strategy would be to ‘depower’ initial impact forces by front rows packing separately and then being followed in succession by the second and back rows.

Rucks and mauls pose dangers for injury in rugby union, as they are an ill-defined area of the game, which is difficult to control by the referee. Charging into a ruck or maul with shoulders below hips has been banned as particularly dangerous but is often difficult to enforce. The analysis of NSWSU/NSWSIC cases often found it difficult, as in other studies (Silver 1984; Silver & Stewart 1994), to determine whether rugby union players were injured in the initial impact of a tackle or in an ensuing ruck or maul.

229 In the analysis of NSWSU/NSWSIC cases, tackling plays in both rugby league and union account for virtually all serious head injuries, and in rugby league account for the vast majority of spinal cord injuries. Most tackle related injuries were to the tackled player but considerable proportions of injuries were also to the tackler. Rugby league players are at a greater risk of serious tackling injuries (particularly to the head) compared to rugby union players because of greater emphasis on this aspect of the game. However, the style of play in rugby union, particularly at the elite level, is increasingly emphasising tackling aspects of play.

In rugby league, many serious tackling injuries and fatalities involve multiple tacklers, where more than one opponent player was involved in the tackle of a single player. Commonly, this results in the ball carrier being unable to break their fall using their hands and arms resulting in their head taking full impact with the ground. In rugby union, the difficulties in distinguishing between tackling, ruck and maul phases of play means that it is harder to accurately determine the role of multiple opponents indirectly causing serious injury. However, the dynamics of how injury might occur are likely to be similar i.e. multiple tackling opponents hinder the ability of the ball carrier to prepare for and withstand impacts with players and the ground.

The NSWSIC/NSWSU also confirmed the danger of high tackles as has been noted in the literature reviewed (Dunning & Sheard 1979; Australian Rugby Football Union 1995; Australian Rugby League 1995; Taylor & Coolican 1987; Silver 1984; Edgar 1995; Scher 1978; Scher 1983a; Scher 1991b). Of interest then to the development of injury prevention strategies is the finding from the archival video analysis that ball carrying style is associated with tackling level.

230 Training and or rule changes concerning ball carrying style must have important effects on tackling style and level carrying implications for the occurrence of serious injuries. This understanding (relationship of ball carrying style to impact level) may prove important for the development of rule modifications, or coaching strategies designed at encouraging tackling impacts at middle body target zones (chest to waist) which appear to carry the least risk of injury to the tackler and ball carrier.

Conversely the danger of serious injury for tacklers approaching impacts too low with their heads down has also been identified in the NSWSIC/NSWSU cases analysed and the literature reviewed (Thomas et al. 1999; Scher 1981; Silver 1992).

Illegal play

The NSWSU/NSWSIC analysis confirmed that illegal play continues to account for a considerable proportion of serious head and neck injuries in Australian rugby codes. Illegal play such as head high tackling and spear tackles are of particular danger. In the fast-paced full contact environment of a football game a fine distinction lies between playing admirably hard and being excessively aggressive. While players seek to physically dominate their opponents, it is hoped that they do not wish to purposefully harm them. The effectiveness of rules designed for player safety depends on referee enforcement and appropriate penalty deterrents for transgressions. At the elite level, foul play has received increased attention and policing (NHMRC 1994; Larkins 1995; Australian Rugby Football Union 1995; Australian Rugby League 1995; Robilliard, M. – Australian Rugby Football Union 1996, pers. comm. 22 November; Meredith, M. – NSW Rugby League 1996, pers. comm. 10 December; Dr Hazzard, H. – National Rugby League 2000, pers. comm. 18 July). However, it is much more difficult to maintain this level of vigilance at the community level where most footballers play.

231 Intensity of match play – speed of play and force of impacts

Very few players in the NSWSIC/NSWSU analysis or the literature reviewed (Norton et al. 1999; Taylor & Coolican 1987) were seriously injured during training despite generally greater exposure time than actual match play. This reflects the importance of match intensity on the risk of serious injury occurring. Furthermore the importance of match intensity is indicated by the increased risk of injury associated with higher levels of competition where play is more intense and impacts forces are greater (see discussion to follow on intrinsic risk factors). Match intensity influences injury rates through its effect on speed of play and force of impacts which are fundamental risk factors for serious injury as observed in the NSWSIC/NSWSU data and in the literature reviewed (Garraway et al. 1991; Hall 1995; Wilson et al. 1999; Norton et al. 1999; Silver 1984; Silver, 2001; Taylor & Coolican 1987; Roux et al. 1987).

Stage of season

The importance of match intensity can also arguably be demonstrated by the predominance of serious injuries very early and late in the season in the literature reviewed (Alsop et al. 2000; Garaway & Macleod 1995; Garraway et al. 1991; Armour et al. 1997; Kew et al. 1991; Williams & McKibbin 1987; Quarrie et al. 2002; Taylor & Coolican 1987; Gabbett, 2000). Early and late stages of the season correspond respectively to early try out matches where final team positions are sought and end of season final matches, both of which tend to elicit intense play from participants. However this relationship was not demonstrated in the NSWSIC/NSWSU cases where most serious injuries occurred relatively evenly spread over mid to late season. This difference in findings might be explained by the fact that this study focused solely on serious injuries of an explicit type whereas the literature reviewed incorporated minor injuries into their analysis.

232 Position on field

Commonly, where NSWSIC/NSWSU records detailed the player’s position on the field at the time of injury, the game play preceding serious injury was close to the try line in a critical defensive/offensive situation. This possibly reflects the intense nature of play at these stages of the game and their implication for serious injury risk. However, the position on the field at the time of the accident was not consistently recorded in enough cases to assess its association with various injuries. A study of all injuries during a Rugby World Cup pre-qualifying tournament found slightly more injuries occurred in the defensive half of the field of play than in the offensive half (Wekesa et al. 1996). A study of minor injuries in rugby league found that slightly more injuries occurred in the injured players own half of the playing field reflecting the increased force and aggression used when players are defending their own half (Gibbs 1994). This factor is also of interest in that it indicates to rule makers and particularly to referees the importance of paying exceptional attention to player safety issues at these points in the game.

Number of players

The effect that different numbers of players on the field would have on injury rates has not been directly studied, but may be inferred from comparing game variations adopting different numbers of players. Rugby league has two less players than in rugby union and experiences a relatively lower risk of some serious injuries. However, many other variations in game rules confound this comparison.

More suitable evidence is gained within rugby union where there is an international competition running with variations to the 15-man-a-side rules (also with significantly shorter games) i.e. super sevens. Under these rule variations, many more tries are scored as increased room between players makes break through runs more likely. Players appear more likely to be tackled while being chased rather than in head on impacts. While reducing the intensity and number of impacts, the reduction in player numbers emphasises other more skilled aspects of the game such as running, passing and kicking with more exciting, open and

233 continuous plays. However, there are no published studies comparing rates of injury between persons playing under different variations of game rules which requires further study.

Acute injury management strategies for suspected serious injuries

There were a substantial number of brain injury NSWSIC/NSWSU cases in both codes, including several fatalities, where the injured player continued to play after the injury incident, often receiving multiple, severe impacts. This occurred in cases within the same match, over two matches played on the same day, and in two games played in the same month. The latter case, which was fatal had sustained enough serious head injuries in past matches to warrant advice from a doctor that he stopped playing. These cases indicate the importance of establishing improved acute injury management protocols to ensure that all injured players are adequately assessed before being allowed to return to play, and that these recommendations are enforced by teams and playing associations.

Consensus on suspected spinal cord injury management contrasts with disagreements over the appropriate assessment and management of suspected brain injury evident in the literature reviewed (NHMRC 1994; Sydney Morning Herald 1995; Clarke 1998; McCrory & Berkovic 1998; McCrory, 2001a; McCrory, 2001b; Wilberger 1993; Gronwall & Wrightson 1975; Maddocks & Saling 1995; Geffen et al. 1998; Cantu & Voy 1995; Saunders & Harbaugh 1984; Annegers et al. 1980; Salcido & Costich 1992). Acute management of suspected spinal cord and head injuries is extremely important to recovery outcomes. The benefit of any doubt in suspected serious injury management should be given to the side of player safety. Particular danger has been observed in the analysis of NSWSIC brain injury and fatality cases from injured players returning to play before they have properly recovered from injury.

234 Several experts have argued that mandatory exclusions from play are excessive and individual assessments of injured players are needed (McCrory & Berkovic 1998; McCrory, 2001a; McCrory, 2001b). International recognition of the importance of standardised injury assessment and planned injury management responses suggest that players with suspected injuries (including apparently mild head injuries) should be excluded from further play until suitable evidence demonstrates that they are completely recovered from injury and are fit to resume to play. At the community level, the inability to always provide suitable medical evaluation of injuries put the onus on the player and coach to ensure that they are fit to play. In this regard, mandatory exclusion periods may give the benefit of the doubt to player safety.

Protective equipment for serious injury

The thesis study design did not allow any evaluation of the effectiveness of protective equipment for rugby football players. While the literature indicates the effectiveness of protective equipment for serious injury prevention in some sports e.g. helmets for bicyclists (McDermott 1992; Attewell et al. 2001), their effectiveness for preventing serious injuries in rugby football has not yet been demonstrated (Wilson 1998; MacDougal & Osbourne 1992; McIntosh & McCrory, 2000; Gerrard 1998; Bishop 1996; Milburn 1993; McCrory, 2001c; Geffen et al. 1998).

An interesting comparison can be made with American football in which a long history of attention to injury prevention has seen the standard adoption of extensive protective equipment for players. However, this led to a cycle in which ever more equipment was needed to ensure player safety and many aspects of the game had to significantly change including player selection and general playing tactics (Hrysomallis & Morrison 1997).

235 Several studies have proposed conflicting hypotheses on whether protective equipment is likely to prove useful for preventing serious injuries in the rugby football codes, some arguing that such equipment can potentially reduce these injuries (Norton et al. 1999; Quarrie & Chalmers, 2001) while others arguing that such equipment may in fact increase the risk of serious injury (Bishop 1996; Geffen et al. 1998; Garraway et al. 2000).

The only epidemiological studies conducted have not been able to provide evidence that existing protective equipment used is an effective means of injury protection in rugby football for anything other than minor injuries (Milburn 1993; McCrory, 2001c; Wilson 1998; McIntosh & McCrory, 2000). However, continued research is required to clarify the role of protective equipment in the prevention of serious injury in rugby football, and to possibly develop more effective means of providing equipment protection to players. Player attitudes and behaviour regarding the use of protective equipment is fundamental to its effectiveness or failure to prevent serious injury in rugby football. Further research of these issues must accompany the evaluation of current and developed protective equipment.

Intrinsic risk factors

Anthropometric and physiological characteristics

A key hypothesis developed in this study regards the role of anthropometric and physiological characteristics as protective factors against serious injury. The hypothesis is derived from a synthesis of conceptual perspectives including: epidemiological evidence provided by this study and others comparing the rate of injury at diverse levels of the game where anthropometric and physiological characteristics vary greatly; fundamental biomechanical principles; biomechanical studies of the relationship between impact force, anthropometric and physiological characteristics and serious injury.

236 The importance of appropriate physical preparation in reducing the risk of injury has been identified extensively in the literature (Sherry & Wilson 1998; Geffen in Sherry 1998; Williams, 2002; Yeo 1998a; Yeo 1998b; Upton et al. 1996; Silver & Stewart 1994; Milburn 1993). The NSWSIC/NSWSU analysis revealed several cases where players were seriously injured and deemed not suitably physically prepared (e.g. not warmed up, hungover from previous night, unfit and fatigued). However, speed of play (Roux et al. 1987) and forces of engagement (Taylor & Coolican 1987) might be the most important aetiological factor in the majority of rugby injuries.

Of note, is the fact that players’ fitness and strength influence the velocity and force they are able to exert in an impact. It also influences their ability to keep up with play and position themselves to be able to build momentum for head-on impacts. Several studies in rugby football and other sports have found little evidence to suggest that fitness, flexibility and strength provided significant protection from serous injuries (Lee et al. 2001; Silver, 2001; Silver, 2002). Many studies have also found a greater risk of injury at higher levels of competition where players are strongest and fittest (Taylor & Coolican 1987; Davidson 1987; Bottini et al. 2000; Burry & Gowland & 1981; Armour et al. 1997; Kew et al. 1991; Jakoet & Noakes 1998; Seward et al. 1993; Silver 1984; Silver, 2001; Silver, 2002; Silver & Stewart 1994; Quarrie et al. 2001; Lee et al. 2001; Lee & Garraway 1996).

Traditionally, larger players have been thought to be better equipped to play rugby football and withstand serious injury. Certainly, an impact between a smaller and larger opponent will give the advantage of inertia to the larger player, indicating the importance of appropriately matching players for size. However, this advantage is lost when opponent players are of a similar size which is usually the case. Larger players are likely to produce a greater force in impacts. As the greater the force exerted in an impact, the greater the potential risk of an injury occurring, it might be expected that games with larger players have an increased risk of serious injury for all players offset only in a limited fashion by the protective factor this large size provides. This hypothesis is borne out by the NSWSIC/NSWSU analysis and the

237 literature reviewed which indicate that the risk for serious injury increases for larger players as level of competition and age increases.

Body type plays a large role in determining players’ weight, height and strength. Other characteristics might also affect vulnerability to serious injury. The clearest example of this is the increased risk of spinal injury to players with long, thin necks (particularly while scrummaging) as demonstrated in the NSWSIC/NSWSU analysis and the literature reviewed (Milburn 1993; Yeo 1993; Australian Torts Reports 1987; NHMRC 1994). Recognition of this risk has led to more appropriate selection of players and a growing emphasis on strengthening exercises for the neck and warming-up exercises before games, particularly in non-elite football. These measures have formed the basis of football spinal cord injury prevention education campaigns in Australia (Yeo 1998a). An in-depth analysis of the kinetics of rugby union scrummaging by Milburn (1993) supported the evaluation of body type in front-row forwards, arguing that appropriate conditioning of neck musculature has been shown to reduce the incidence of spinal injuries and so should be given a high priority irrespective of possible law and refereeing changes.

Nonetheless, the authors of one of Australia’s most comprehensive investigation into football spinal injuries have argued that there is little evidence that a long, slender neck intrinsically increases the risk of neck injury (Taylor & Coolican 1988). Similarly, a strong neck musculature has not been shown to reliably protect against these injuries. Children whose bones and muscles are yet to fully develop, still seem to be at a much lower risk of spinal cord injury despite having less appropriate body types than older players. While the NSWSIC/NSWSU analysis did reveal one case of a junior player receiving a SCI who was deemed to have an inappropriate body type, the many other injured players were apparently of suitable physique.

238 The contention that children are at a lower risk of SCI despite less appropriate body types is also supported by the NSWSIC/NSWSU data and by the literature reviewed (Taylor & Coolican 1988; Lee & Garraway 1996; Armour et al. 1997; Kew et al. 1991; Jakoet & Noakes 1998; Seward et al. 1993; Silver, 2001; Silver, 2002; Silver & Stewart 1994; Quarrie et al. 2001).

The statistics in the NSWSIC/NSWSU analysis indicates that the seriously injured football players studied were generally young men with players under 14 and over 40 years old, and rarely, if ever, seriously injured. The reason for this apparent paradox is simple and consistent with the hypothesis regarding the anthropometric and physiological characteristics previously stated. The reason is that less force is exerted in impacts among junior players than their senior counterparts and they therefore experience a smaller risk of serious injury. This is because of the very nature of children's size, differences to game rules that affect impact forces at these lower levels of the game, and reduced match intensity. This again indicates the fundamental importance of game speed and impact forces in determining the risk of injury.

This evidence suggests that future reductions in the rate of serious injury would most likely come from minimising the forces involved in impacts in addition to ensuring that players are of suitable body type and physically prepared. It must be emphasised that this concept does not suggest that attention to suitable body type and physical preparation should be abandoned as an injury prevention strategy. Rather an argument might be made that a purely narrow focus on these strategies ignores fundamental risk factors for serious injury and limits the ability for injury prevention stratagem to affect further reductions in serious injury.

239 Player position

Player position information collected in NSWSU/NSWSIC records was sometimes incomplete and there may have been a bias associated with reporting front row players, commonly thought to carry greater risk of injury as a result of their greater physical involvement in the game, both in attack and in defence. Available data conferred with the literature reviewed (Orchard & Seward 1994; Secin et al. 1999; Wetzler, et al. 1996; Taylor & Coolican 1987; Silver 1984; Silver 1994; Armour et al. 1997; Williams & McKibbin 1987; Wetzler et al. 1998; Gissane et al. 1997) that in both codes forwards were most likely to receive serious injury. Hookers’ appeared to be particularly vulnerable to SCI sustained in scrums.

In rugby union, back line players not involved in scrums tend to be injured more by tackling injuries, mainly while carrying the ball (Taylor & Coolican 1987; Silver 1984; Silver 1994; Burry & Gowland 1981; Kew et al. 1991). As in rugby union, rugby league backs are more likely to be injured in tackles because they are not involved in scrum activity but are responsible for critical defensive tackling and attacking runs.

However, NSWSU/NSWSIC data and the literature reviewed (Taylor & Coolican 1987) suggest that the risk of SCI for back line players in rugby league may be greater than that in rugby union because of the greater risk of tackling injuries for rugby league. Of interest is the current trend in rugby union towards tackling play which may see shifts in injury patterns.

Several studies have found an association between anthropometric characteristics and player position (Nicholas 1997; Quarrie et al. 1996; Carlson et al. 1994; Leeet al. 1997). Front row players are usually more endomorphic than other forwards, and forwards are generally larger than backs. However, coaches for elite rugby football teams in both codes are increasingly using larger mesomorphic players rather than endomorphic or ectomorphic players. Associated with the favouring of mesomorphic players, is the decline in the distinction between the traditional body types associated with various playing positions (Huxley, 2003).

240 Skills, experience and technique

NSWSIC/NSWSU findings suggest, that as in other studies (Silver 1992; Burry & Gowland 1981; Scher 1977), tackling injuries are often a result, at least in part, of dangerous or poor technique that exposes either player to greater risk of injury.

Many other studies of rugby football spinal cord injury around the world have concluded that in rugby football the cause of serious injury is largely a function of technique (Silver 1992; Burry & Gowland 1981; Milburn 1993; Scher 1977). The importance of skills training in tackling technique to minimise the risk of injury has been long recognised by football trainers, and it often forms a fundamental part of training sessions (particularly for younger players).

The ability to perform a tackle and participate safely in a scrum, ruck, or maul is a highly adept skill that comes only with training, practice, and experience. Particular danger exists if teams are not matched for strength, age, and experience. Players’ inexperience was identified as a contributing factor to injury in several NSWSU/NSWSIC cases, often when players were out of their normal playing position

241 Level of competition

As discussed, despite the importance of size, preparedness, skill and experience for injury prevention, NSWSIC/NSWSU data and a review of other surveys of acute rugby injuries at all levels of competition, suggests that the risk of injury increases with age, grade, and competitive level (Taylor & Coolican 1987; Davidson 1987; Bottini et al. 2000; Burry & Gowland 1981; Armour et al. 1997; Kew et al. 1991; Jakoet & Noakes 1998; Seward et al. 1993; Silver 1984; Silver, 2001; Silver, 2002; Silver & Stewart 1994; Quarrie et al. 2001; Lee et al. 2001; Lee & Garraway 1996). Many of these studies have found a greater number of injuries among more skilled players because of greater size; a more aggressive approach to the game; and, generally, higher speeds involved in impacts.

This contradicts somewhat the commonly accepted wisdom that increased player size, preparation, and skill is the most effective protection against injury. The analysis of NSWSU/NSWSIC cases demonstrated virtually no serious injuries to children under 16-year-olds. Where information was available, many injured players, while mostly in community level teams, played in the higher grades. Several elite players were seriously injured during the study period. However, a considerable risk remains for community level players, and their larger numbers lead to a greater frequency of injury. Rule variations aimed at improving player safety also exist at junior levels.

242 Fatalities and pre-existing conditions

The fatalities analysed were most often caused by cardiac arrhythmia, sequelae of severe SCI or brain injuries. This reflects the extreme nature of the serious SCI and brain injury under study but underestimates the deaths caused by these serious injuries. This is because the study only included cases of fatality sustained within one year of receiving a related injury. The severe reduction in lifespan often associated with serious SCI and brain injury means that many seriously injured players would die well before a natural lifetime as a result of their injuries.

The deaths caused by cardiac arrhythmia were sometimes associated with endogenous morbid pathological process accompanying pre-existing cardio vascular disease, or found to be exacerbated by the ingestion of anabolic steroids. The ability to reduce the incidence of fatalities caused by cardiac arrhythmia may depend on the adoption of pre-participation physical evaluation screening protocols, as have been used with some success in the United States (Lyznicki et al. 2000). The feasibility of the use of web-based screening systems to facilitate efficiency should be investigated.

Country vs Metropolitan competitions

In the NSWSIC/NSWSU cases analysed, the proportions of all serious injuries and fatalities accounted for by metropolitan and country teams did not differ from expected values as it reflected the proportion of participants in metropolitan and country teams in rugby league and in rugby union over the study period. However, country rugby football competitions often do not have the same injury management infrastructure available to metropolitan teams. The location where injuries occur is also likely to be remote from the metropolitan centres that contain the main spinal units equipped to treat acute SCI and brain injuries. Resources and procedures available to country team players must be improved to ensure equity with those provided to metropolitan teams.

243 Historical, sociological, playing style, player characteristics and developments in rugby football

Although archival video and literature analysis is qualitative in nature, the findings from these sources suggest that in modern play, multiple tacklers are more common, and impacts are at greater speeds and more likely head on. These factors coincide with some of the fundamental risk factors for serious injury assessed by this study. This finding is part of a possible explanation as to why injury rates persist, despite increasing injury prevention efforts, as has been suggested elsewhere in other sports (Norton et al. 1999).

Multiple tacklers are likely to collide with the ball carrier from different directions and levels. This increases considerably the sheer forces experienced by the ball carrier and makes it more difficult to prepare for impact with opposing players, as well as the ground when falling.

The dangers associated with head on impacts in the amplification of impact forces have been identified. The finding that modern games were more likely to have head on impacts between players might be expected due to improved player fitness, competitive determination, and rule changes (e.g. rugby league players are now required to fall back 10 metres from the ball after a completed tackle). As discussed earlier, the ability to maximise momentum for impacts is related to the fitness, strength and determination of a player, all of which are arguably greater for modern players. In addition, modern rule modifications (e.g. the 10 metre rule in rugby league) and developments in playing style (e.g. the reduced distinction between backs and forwards in rugby union) arguably act to maximise players running speed when tackling.

Dangerous (now banned) tactics and foul play occur less frequently in modern games. The expectation that less head/neck level tackles are occurring in modern games, particularly at the elite level as increasing penalties and monitoring processes are implemented is confirmed by the archival video analysis.

244 Proportions of observed waist high tackles were higher in earlier games. Conversely, proportions of chest high tackles in modern games were higher than in earlier games. These associations were found to be statistically significant. When league and union were split (elaborated) in analysis, the reduction in the proportion of head/neck high tackles observed for union games reduced less dramatically than for league. This might arguably be attributed to rugby league having traditionally greater media exposure and consequently earlier formalisation of a judicial processing for tackling infringements.

Modern players are more likely to carry the ball to their chest. As discussed, tacklers from both periods tend to target the ball in their tackles thus modern players are most likely tackled at chest level whereas earlier players were more likely tackled at waist level.

The demands of professionalism mean that modern elite players (and their community level emulators) are more physically prepared and possibly committed to football careers than their earlier counterparts. However, with shifting loyalty, improved foul play policing, demands to play more games and the effects of societal ‘civilising’ processes, players may be no more aggressive than those from earlier last century. The analysis comparing elite rugby football player statistics from the first half of the 20th century with current players, found modern players are larger, fitter, train more intensively and are possibly stronger than their former counterparts. These developments increase the forces exerted in impacts. However, because strength and fitness have only been shown to provide limited protection against serious injury, the increased forces in impacts may actually be increasing the overall risk of serious injury.

The development of healthier lifestyles and increasing achievement orientation among the general public in modern times, might suggest that increases in average elite player size and fitness could also be inferred somewhat to community level players but this requires further study.

245 Implications of study findings for injury prevention infrastructure in Australia and methodological approaches to rugby football injury prevention

The use of a ‘public health approach’ has proven indispensable to the assessment of serious rugby football injuries and their associated risk factors, as well as the development of injury prevention strategies and their evaluation. However, its continued success depends on one of its basic tenets – maintenance of the cycle of research, intervention, and evaluation. The conflicting evidence sometimes presented within this study regarding the successes and limitations of injury prevention strategies are not necessarily contradictions. They reflect the reality that this approach is effective but cannot function as a static and/or intermittent system. Its success will always rest on the continuity of its fundamental cycle. In practical terms this means that vigilance in injury prevention must be maintained even if successes are demonstrated. Complacency, which is arguably the natural state of human affairs, must be avoided to ensure player safety is maintained and that further injury reduction can be achieved. To ensure this, these activities must be integrated into a reliable infrastructure that maintains a consistent approach.

The importance of unifying sporting injury prevention infrastructure in Australia has been a continuing theme in the injury prevention literature for a number of years, with several authors drawing attention to vastly improved systems utilised in other comparable countries such as New Zealand and the United States (Orchard & Finch 2002). Rugby football injury prevention activities i.e. injury surveillance, injury prevention strategy development, effective implementation and evaluation of interventions, need to capitalise on the existing infrastructure of the rugby football associations and be fully integrated into them. Sophisticated linkage of these systems to a central sporting injury monitoring body, which seeks to facilitate and monitor best practices in a consistent manner is required.

246 Activities of the rugby football association's injury prevention programs and reporting requirements should be directed by this central sporting injury monitoring body. If the creation of such a central body is not feasible in the short term, this role may be taken by existing government affiliated bodies such as sporting injury compensation bodies (e.g. NSWSIC) as is currently the case in New Zealand.

Authors have argued that the failure to develop a reliable sporting injury prevention infrastructure in Australia is because of the difficulties of successfully persuading policy makers and the community that sporting injuries are preventable compared to other injuries such as those that occur on the road (Orchard & Finch, 2002). However, policy makers and the community were also initially reluctant to acknowledge the benefits of many injury prevention strategies for road injuries such as breath testing and seatbelts that have now been thoroughly accepted as effective. Road injury prevention also benefited from an extensive network of relevant government and affiliated organisations that facilitated the development of an integrated injury prevention infrastructure (Lawson 1991).

Lett et al. (2002) argue that injury specialists have failed to successfully persuade policy makers and the community that injuries are preventable partly due to the lack of a unified understanding of injury control. The authors suggest that the two most important models utilised in injury control, Haddon's Matrix and the Public Health Approach, should be combined to provide a unified framework for understanding injury prevention. The complementary contribution of these two models offers a more coherent, and comprehensive explanation of injury prevention. The Public Health Approach lacks a methodical point of application, which is provided by the Haddon Matrix. Conversely, the Haddon Matrix lacks a systematic action plan.

247 The integration of these models complementarily combines the theoretical framework of Haddon's matrix with the Public Health Approach's systematic implementation strategies. There are no published accounts of the Haddon Matrix being utilised to assess risk factors for serious injuries and fatalities in rugby football. This study has endeavoured to provide such a unified theoretical framework for understanding serious injury and fatality prevention issues in rugby football.

248 CHAPTER FIVE SUMMARY

The rugby codes pose the largest burden on society in respect to serious and permanent injuries of all organised sports in NSW.

Continued variance in these relatively small numbers makes conclusive interpretation of these figures difficult, however, current evidence suggests that rugby union, unlike rugby league, has sustained a significant reduction in SCI over the entire study period. The small but significant decrease in the incidence of these SCI in rugby union found in this study between 1984-99 in NSW, signifies the effectiveness of numerous rule changes and safety programs aimed at improving player safety particularly in the scrum, ruck and maul since the 1980s in Australia. The incidence of rugby league related brain injuries and fatalities did not decrease in NSW from 1984 to 1999.

A framework for conceptualising risk factors adapted for this study has provided a systematic point of application for research and the development of prevention strategies. Providing an overview of the complex web of risk factors makes it possible to glean a novel perspective on the central factors that influence the occurrence of serious or fatal injuries in rugby football. This approach has led to the development of several conclusions about the importance of various risk factors for serious injury. Key to these conclusions are that assessment of all risk factors associated with serious injury in this study and in the literature suggest that the most fundamental risk factors for serious rugby football injury are speed of game, force and level of impact.

This evidence suggests that future reductions in the rate of serious injury would most likely come from minimising the forces involved in impacts in addition to ensuring that players are of suitable body type and physically prepared.

249 The use of a ‘public health approach’ has proven indispensable to the assessment of serious rugby football injuries and their associated risk factors, as well as the development of injury prevention strategies and their evaluation. However, its continued success depends on one of its basic tenets – maintenance of the cycle of research, intervention, and evaluation. Complacency, which is arguably the natural state of human affairs, must be avoided to ensure player safety is maintained and that further injury reduction can be achieved. To ensure this, these activities must be integrated into a reliable infrastructure that maintains a consistent approach.

The importance of unifying sporting injury prevention infrastructure in Australia has been a continuing theme in the injury prevention literature for a number of years, with several authors drawing attention to vastly improved systems utilised in other comparable countries such as New Zealand and the United States.

Rugby football injury prevention activities need to capitalise on the existing infrastructure of the rugby football associations and be fully integrated into them. Sophisticated linkage of these systems to a central sporting injury monitoring body, which seeks to facilitate and monitor best practices in a consistent manner is required. Activities of the rugby football association's injury prevention programs and reporting requirements should be directed by this central sporting injury monitoring body or existing government affiliated bodies.

The continued annual occurrence of catastrophically serious injuries leading to permanent brain damage and quadriplegia associated with rugby league and union in NSW suggest relatively novel approaches to the development of preventive strategies.

250 CHAPTER SIX CONCLUSION

Summary

Serious injuries in the NSW rugby football codes persist despite increasing awareness of the importance of injury prevention and management over the two decades covered in the study period.

Overall, while safety measures have showed success, the continued and persistent occurrence of serious injuries and fatalities indicate an ongoing need to reduce these tragic injuries in both codes.

Force and nature of impact between participants have been identified as fundamental factors that affect risk of serious injury. The understanding of this causal link is under utilised in current prevention approaches and should be used as the basis for investigations into future preventive strategies. However, further evidence for the effectiveness of specific strategies is required before full endorsement.

Analysis of the development of the football codes over the last century has demonstrated that the limits of human physical endurance are increasingly tested as players become fitter and stronger. Developments in style of play, such as the greater likelihood of multiple tacklers and head on impacts at speed are potentially putting players at increasing risk of injury.

The need to modify the games’ rules in response to societal pressures to reduce violence, improve player safety and increase the games entertainment value has long been evident. The history of both rugby union and league football is filled with continual developments of the laws of the game that have proved to benefit the sports and their players, although these have sometimes met with disapproval at the time of their institution.

There is a clear need to provide central support, advocacy and a unified strategy for the myriad infrastructure of organizations that play a part in sports safety and injury prevention.

251 The main findings of this thesis indicate that serious injuries in the NSW rugby football codes persist despite increasing awareness of the importance of injury prevention and management over the two decades covered in the study period (see Chapter 1, Research Objectives, points 1-3). Overall, while safety measures have showed success, the continued and persistent occurrence of serious injuries and fatalities experienced by rugby football players in NSW indicate an ongoing need to reduce these tragic injuries in both codes.

The description of the risk factors associated with serious injury and evaluation of the effectiveness of injury prevention strategies (see Chapter 1, Research Objectives, points 4 & 6-9), provided evidence that consistently indicated that the force and nature of impact between participants are fundamental factors that affects risk of serious injury. The understanding of this causal link should be used as the basis for future preventive strategies.

Fundamental injury-prevention strategies such as appropriate body-type selection, physical preparation, and skills training remain critically important and must be maintained. Nonetheless, their limited capability for protecting players from injury, along with the increasing risk of injury for more skilled and physically prepared players, repeatedly suggests that efforts should be made to also develop strategies that will decrease the forces experienced in impacts.

Protective equipment is not likely to be useful in reducing the incidence of spinal cord injury for rugby football players and its value as protection from head injury for players is limited (McIntosh & McCrory, 2000; Wilson 1998; MacDougal & Osbourne 1992). The potential adverse effects of using such equipment must be weighed against their limited protective value for relatively rare injuries. Further evidence of their effectiveness is required before full endorsement. Research and development into protective equipment should continue as these investigations may eventually substantially increase and prove their efficacy.

252 Analysis of the development of the football codes over the last century (see Chapter 1, Research Objectives point 5) has demonstrated the limits of human physical endurance are increasingly tested as players become fitter and stronger. Certainly testing the limits of physical capability should not extend to testing the body’s capability for increased risk of serious debilitating injury but rather of the superior limits of skill and agility.

The need to modify the games’ rules in response to societal pressures to reduce violence, improve player safety and increase the games entertainment value has long been evident. The history of both rugby union and league football is filled with continual developments of the laws of the game that have proved to benefit the sports and their players, although these have sometimes met with disapproval at the time of their institution. It is worthy to consider therefore, that sustained efforts to improve the games should not be viewed as an attempt to break with the games tradition but rather as a continuance of their natural progression. Undoubtedly, the future popularity and ultimately survival of both codes at the community level, as well as at the elite level, depends on such adaptability.

It is disappointing that both rugby union and league attempts to set up much needed injury surveillance programs and safety review committees have not been sustained and demonstrated effective ongoing activity to prevent injuries. This is essential for the effectiveness, or otherwise, of preventative measures to be evaluated with robust epidemiological data as has been frequently advocated.

Sporting organizations should progressively modify existing injury surveillance systems to incorporate the collection of data needed to effectively evaluate safety developments. This requires players to have their playing time monitored for injury and the collection of detailed risk factor data. These systems may also seek to utilise video technology to improve the analysis of injury incidents. 253 Similarly government attempts to establish centralised bodies responsible for sporting safety advocacy, such as Sportsafe, have not been sustained and require immediate action. There is a clear need to provide central support, advocacy and a unified strategy for the myriad infrastructure of organizations that play a part in sports safety and injury prevention. This is also necessary for areas of weakness to be identified and addressed on an ongoing basis .

254 RECOMMENDATIONS

Summary

Greater utility should be gained from understanding the fundamental importance force and nature of impact play in the risk of serious football injuries, informing prevention strategies including rule modifications. Potential rule modifications designed to improve safety should seek to emphasise the skilled aspects of the games rather than those requiring brute force

Systematic evaluation is required of rule modifications that would reduce the forces of impact experienced and minimise dangerous types of impacts e.g. reducing number of players and restriction of the tackling zone.

Continued emphasis on the importance of player selection, training and preparation. For example develop a system for screening players for deficient tackling and scrummaging technique and institute a continuous process of education where the latest safety developments coming from research are effectively implemented through regular training campaigns.

Rugby associations should impose strict requirements on their community level club members regarding the punishment of illegal play.

Reducing risk exposure to known high-risk situations in rugby union such as rucks, mauls and scrums might be achieved by more formal guidelines for referee control at senior levels of competition.

Suspected head injury management needs standardised and reliable, repeated measures tests, to quantify evidence in considerations regarding returning to play after suspected injury. As loss and return to normal functioning can often be difficult to measure, standardised baseline monitoring to determine normal cognitive functioning levels for all players should be a part of the registration process.

Justification for the introduction of specific rule modifications and other pervasive preventive measures requires evidence from the highest levels of research design. Large scale, long term injury surveillance systems are needed to provide sufficiently robust data to evaluate and develop safety strategies, in addition to monitoring injury rates.

255 The analysis in this thesis examined the cumulative incidence of serious injuries and fatalities, but was not able to accurately measure incidence density rates based on player hours at risk. The conclusions drawn from this study’s analysis, while corroborating other research findings, would benefit from observational and experimental studies not possible with the current research. While continual examination and revision of the Rules is needed in the interests of player safety, justification for the introduction of specific rule modifications and other pervasive preventive measures requires evidence from the highest levels of research design.

Observational cohort and case-control studies need to be set up within existing large-scale rugby football injury surveillance systems to properly evaluate risk factors and the effectiveness of preventive measures. Controlled experimental games might employ video analysis techniques, as utilised here, which could be developed to systematically evaluate the effects of such rule changes. The efficacy of these principles might, as have enhanced referee scrum-management principles in rugby union, be demonstrated at junior levels and could eventually be extended into the more senior levels of the game.

The inherent danger of catastrophic injury in rugby football and the potential to reduce these injuries must be brought into public focus and to the attention of administrators of both codes. This will reduce complacency at the administrative level and ensure that player safety considerations are given the highest priority.

The evidence analysed, consistently indicates that the fundamental factor that affects serious injury risk is the force of impact. Potential rule modifications designed to improve safety should seek to emphasise the skilled aspects of the games rather than those requiring brute force. Recommendations made here that are already being implemented should be viewed as providing a positive evaluation of their utility.

256 Decreasing forces in impacts

Systematic evaluation is required of rule modifications that would reduce the forces of impact experienced in scrums, such as successive scrum engagement (i.e. front-row engage, followed in succession by second row) or reducing players in participating the scrum (as is currently practiced at very junior levels). Studies have considered player contributions to force in scrummaging have concluded that players not essential to the stability of the scrum still contribute considerable force to it (Milburn 1993). These players are the most appropriate players to remove from scrums. The breakaways could stand unconnected to the sides of a scrum as they provide considerable force to a scrum push and contribute to destabilising forces (while the lock adds little force but acts to stabilise his team mates).

The forces experienced in the average tackle might be reduced in several ways requiring evaluation. The number of players on field could be reduced. This may make chasing tackles more likely (rather than head on) and emphasise running, passing and kicking more than the force of player contact but requires trials before being recommended.

Variations in player numbers in different competitions such as the rugby union sevens tournaments provide the opportunity to evaluate whether reducing player numbers on the field decreases the risk of players incurring serious injury. As yet, there is no published evidence making this comparison. It might be useful to use experimental designs where style of play and injury risk is monitored and evaluated between matched players of various team sizes. For example, to assess the effect that reducing various numbers of players has on game dynamics such as number of tackles, their level, angle and speed as well as other factors that may relate to injury risk.

257 The number of tacklers allowed to tackle a ball carrier could be reduced to two or even one, diminishing the increasing danger posed by multiple tacklers in both codes.

Restriction of the tackling zone of an initial impact to below the shoulders might decrease the likelihood of accidental high tackles and would have the effect of making it more difficult for multiple players to tackle one player. An additional restriction might be that initial tackling impact must also be above the hips so as to improve the safety for tacklers who are often injured when tackling too low and struck by a knee. Uniforms could carry a standard bright stripe at the shoulder and hip level to highlight the restricted tackling target zone.

Player selection and training

Appropriate body type, experience/skill and player matching is crucial for safe scrummaging and tackling. This requires a formalised process of player selection and training.

Training packages including instructional videos, as have been used in the past, should be continually sent to all coaches in official competitions to inform them of their duty of care regarding player safety in terms of selection and technique training. This needs to be a continuous process of education where the latest safety developments coming from research are effectively implemented. This could help to ensure that a standard consensual approach is adopted nationally. Similarly, such a campaign could be used for referees to ensure a standardised safety conscience approach is used to manage all games. Many safety strategies are dependent on referees for their efficacy. Such education processes might seek to utilise emerging technologies such as CD-ROM videos to facilitate distribution and interactivity.

258

Schoolboy rugby football players should be matched for size, skill and experience rather than by age.

Some players are at a higher risk of sudden death, requiring appropriate screening procedures to ensure all players are appropriately fit to play rugby football. The feasibility of mandatory screening of football players for certain congenital health problems, e.g. Ischaemic conditions, as part of team registration needs to be assessed. However, mass-screening procedures for asymptomatic players, who are not clearly from high-risk groups is probably not a viable proposition in Australia considering the rarity of such cases. This may be just advisable for identifiably high-risk groups such as obese persons, those with familial history of coronary disease, or among the Aboriginal population.

In the scrum, the inherent danger of certain positions, such as the hooker in a scrum, requires greater coaching attention to these players such as a more formalised safety training schedule particularly for less experienced players.

Improve and emphasise awareness of skills training for tackling techniques. Develop and recommend specific ball carrying and tackling techniques based on collected evidence, e.g. carry ball to waist, don’t tackle lower than waist, keep head up and eyes open, keep head to the side and lead with shoulder when tackling, learn to anticipate how and in which direction the ball carrier will fall. Referees and coaches need to continue to emphasise the importance of these principles.

Develop a system for screening players for deficient tackling and scrummaging technique. Train coaches and trainers to explicitly recognise playing technique problems that might increase the risk for injury. This might be assisted with visually, illustrative, instructional videos or CD- ROMs. 259 Illegal play

Continued efforts, as have been successfully displayed by the respective rugby associations, are needed to prevent illegal play incidents. Rules concerning foul play are only valuable if correctly implemented. The effectiveness of this process must be persistently evaluated. The level of vigilance maintained at the elite level is difficult to sustain at the community level. However, this is where most footballers play and therefore requires special attention. Rugby associations should impose strict requirements on their community level club members regarding the punishment of offenders.

Risk exposure

Decreasing the injury risk exposure (i.e. on field playing time) of players will reduce the overall rate of injuries. Both codes might seek to reduce the actual playing hour demands on players (but should not seek to diminish training demands). Reducing risk exposure to known high-risk situations in rugby union such as rucks, mauls and scrums might be achieved to some degree by referees strictly limiting the duration of these dangerous types of play. Currently referees do try to minimise the duration of these plays but there may be some opportunity for more formal guidelines at senior levels.

Injury management

Suspected head injury management needs standardised and reliable, repeated measures tests, to quantify evidence in considerations regarding returning to play after suspected injury. Two tiers to injury management include the immediate response to injury and the follow up assessments to determine fitness to play in future games. Players ideally should be taken from play (and play stopped if necessary), appropriately assessed by

260 qualified personnel, evacuated for medical attention if needed, and not allowed to return to play until fitness to play is demonstrated.

As loss and return to normal functioning can often be difficult to measure, standardised baseline monitoring to determine normal cognitive functioning levels for all players should be a part of the registration process. Where the resources for these processes do not exist (such as in some community level competitions), mandatory exclusion periods for even minor concussion should be considered.

Players, coaches and officials should be aware of SCI risk, immediately stop a game if an SCI is suspected, immobilise the injured player, and arrange immediate paramedic evacuation.

Doctors and sports trainers need to be adequately trained and qualified, and subject to ongoing education.

Protective equipment

Continued investigations into the use of protective equipment are necessary, despite evidence showing that it currently has very limited usefulness for preventing serious injury. It may be possible to establish, feasibly with further technological developments, that protective equipment provides a significant protective factor against serious injury thus warranting its widespread adoption.

In order to properly evaluate the effectiveness of protective equipment use, large scale observational level studies are needed utilising data from rugby football injury surveillance systems. The systems need to consistently collect detailed information about type of equipment usage in both injured and uninjured players.

261 Referees need to ensure that soft headgear is not used to manipulate an opponent. Players and coaches need to support safety behaviour discouraging targeting the headgear of opposing players or the use of the helmeted head as an implement. Attitudes to usage must be thoroughly assessed, particularly regarding current low levels of acceptance.

Injury surveillance

Large scale, long term injury surveillance systems are needed to provide sufficiently robust data to evaluate and develop safety strategies, in addition to monitoring injury rates. Injury surveillance systems set up by the respective rugby codes need to be developed further and effectively maintained. They should aim to cover larger proportions of their player populations and ideally should seek to be able to collect detailed information regarding the mechanism of injury and associated risk factors.

The roles of existing safety review committees (e.g. NRL Head and Neck Committee, ARFU Safety Committee) in overseeing the continual examination and revision of game rules in the interests of player safety (making systematic use of epidemiological data provided by injury surveillance systems) need to be developed. There is little evidence that these committees have maintained activity and demonstrated effectiveness to date. Annual injury update publications (including peer reviewed journals) could apprise the rugby football community and the general public of developments in trends and injury prevention, developing awareness of prevention efforts.

262 REFERENCES

Alexander E Jr. 1990. Don't drink and dive. Surgical Neurology 34(3): 159.

Almekinders LC. 1999. Anti-inflammatory treatment of muscular injuries in sport. An update of recent studies. Sports Medicine 28(6): 383-8.

Alsop JC, Chalmers DJ, Williams SM. 2000. Temporal patterns of injury during a rugby season. Journal of Science and Medicine in Sport 3(2): 97-109.

American Spinal Injury Association (ASIA) 1992. International standards for neurological and functional classification of spinal cord injury. Rev ed. ASIA/IMSOP, Chicago.

Andersen MB, Williams JM. 1988. A model of stress and athletic injury: Prediction and prevention. Journal Sport Exercise Psychology 10: 294-306.

Annegers JF, Grabow JD, Kurland LT, Laws ER Jr. 1980. The incidence, causes, and secular trends of head trauma in Olmstead County, Minnesota, 1935-1974. Neurology 30: 912-919.

Anonymous (“Old international”). 1910. How to play rugby: Spalding’s Athletic Library. American Sports Publishing Co, New York. (E.S. Marks Sporting Collection, Mitchell Library, NSW State Library).

Archival player size information. Davis Sporting Collection No.2 (Mitchell Library, NSW State Library) (QLD annual Box 42) (RL yearbook 1958 Box 43) (New Zealand Football League Souvenir official programmes Box 43) (The rugby annual, ARU 1956 Box 43) E.S. Marks Sporting Collection (Redcaps rugby league annual 1926) ( tour in England 1926-27) ( Football Club: The official programme 1935) ( cricket and athletic club football programme 1932).

Armour KS, Clatworthy BJ, Bean AR, Wells JE, Clarke AM. 1997. Spinal injuries in New Zealand rugby and rugby league – a twenty-year survey. New Zealand Medical Journal 110: 462-5.

Athanasou J, Brown D, Murphy G. 1996. Vocational achievements following spinal cord injury in Australia. Disability and Rehabilitation 18(4): 191-196.

Atkinson L, Merry G. 2001. Advances in neurotrauma in Australia 1970-2000. World Journal of Surgery 25(9): 1224-9.

Attewell RG, Glase K, McFadden M. 2001. Bicycle helmet efficacy: a meta- analysis. Accident Analysis & Prevention 33(3): 345-52.

263

Australian Bureau of Statistics (ABS). 1995. National health survey: injuries, Australia 1995. Cat.no.4384.0, ABS, Canberra

Australian Bureau of Statistics (ABS). 1997. Australia – sport and recreation: a statistical overview, 1997. Cat.no.4156.0, ABS, Canberra.

Australian Bureau of Statistics (ABS). 1998. Australia – participation in sport and physical activity, 1997/98. Cat.no.4177.0, ABS, Canberra.

Australian Bureau of Statistics (ABS). 1999a. Australian social trends 1999: culture & leisure - sport: sporting Australians. Cat.no.4384.0, ABS, Canberra.

Australian Bureau of Statistics (ABS). 1999b. Causes of death, Australia 1999. Cat.no.3303.0, ABS, Canberra.

Australian Bureau of Statistics (ABS). 1999c. Australia – participation in sport and physical activity, 1998/99. Cat.no.4177.0, ABS, Canberra.

Australian Bureau of Statistics (ABS). 2000a. Australia – participation in sport and physical activity, 1999/00. Cat.no.4177.0, ABS, Canberra.

Australian Bureau of Statistics (ABS). 2000b. Estimated resident population: states and territories of Australia. Cat.no.3201, ABS, Canberra.

Australian Bureau of Statistics (ABS). 2001a. The social impacts of sport and physical recreation: an annotated bibliography. Prepared on behalf of the Recreation and Sport Industry Statistical Group by the National Centre for Culture and Recreation Statistics. ABS, Canberra.

Australian Bureau of Statistics (ABS). 2001b. Involvement in organised sport and physical activity, Australia. Cat.no.6285.0, ABS, Canberra.

Australian Bureau of Statistics (ABS). 2002a. Australia now year book Australia 2002: culture and recreation sport and recreation. Cat.no.1301.0, ABS, Canberra.

Australian Bureau of Statistics (ABS). 2002b. Australia now year book: transport special article – a history of road fatalities in Australia. Cat.no.1301.0, ABS, Canberra.

Australian Injury Prevention Network. (AIPN). 2001. Strategy – January 2001 to December 2005. AIPN, Brisbane.

Australian Institute of Health and Welfare & Commonwealth Department of Health and Family Services. (AIHW & DHFS). 1997. First report on national health priority areas 1996. Cat.no.PHE 1, AIHW Canberra.

264 Australian Institute of Health and Welfare & Department of Health and Family Services (AIHW&DHFS). 1998. National health priority areas report, injury prevention and control. Cat.no.PHE 3, AIHW, Canberra.

Australian Institute of Health and Welfare (AIHW) 2001. Australian hospital statistics 1999–00. cat.no.HSE 14 (Health Services Series no. 17), AIHW, Canberra.

Australian Institute of Health and Welfare (AIHW). 2000. Australia’s health 2000: the seventh biennial health report of the Australian Institute of health and welfare. AIHW, Canberra.

Australian Public Schools Amateur Athletic Association. 1907. Australian Public Schools Amateur Athletic Association rules. E.S. Marks Sporting Collection (Mitchell Library, State Library NSW).

Australian Rugby League (ARL). 1995. The international laws of the game. ARL, Sydney.

Australian Rugby Football Union (ARFU). 1995. Laws of the game. ARFU, Sydney.

Australian Sports Commission (ASC). 1997. National participation framework. ASC, Canberra.

Australian Sports Commission (ASC). 2006. Budget brings unprecedented level of funding to Australian sport. Ausport 3(1): 3.

Australian Sports Injury Data Working Party. 1997. Australian sports injury data dictionary: guidelines for injury data collection and classification for the prevention and control of injury in sport and recreation. SportSafe Australia (Australian Sports Commission) and Sports Medicine Australia, Canberra.

Australian Standards. 1992. Swimming safety Part 1- fencing for swimming pools; Part 2, Location of fencing for private swimming pools. Australian Standards, Canberra.

Australian Torts Reports. 1987. Watson vs Haines (1987). Australian Torts Reports: 80-094. CCH Australia Ltd

Australian Transport Safety Bureau. 2000. Road Fatalities Australia; population data – estimated resident population, Australia. Cat.no.3201.0, Department of Transport and Regional Services, Canberra.

Australian Transport Safety Bureau. 2001. Road fatalities – Australia 2001 statistical summary. Department of Transport and Regional Services, Canberra.

265 Australian Transport Safety Bureau. 2005. Road Deaths Australia Monthly Bulletin August 2005; 1-7.

Bader RS, Goldberg L, Sahn DJ. 2004. Risk of sudden cardiac death in young athletes: which screening strategies are appropriate? Pediatric Clinics of North America 51(5):1421-41.

Baguley I, Slewa-Younan S, Lazarus R, Green A. 2000. Long-term mortality trends in patients with traumatic brain injury. Brain Injury 14(6): 505-12.

Bailes JE, Cantu RC. 2001. Head injury in athletes. Neurosurgery 48(1): 26-46.

Bailes JE, Herman JM, Quigley MR, Cerullo LJ, Meyer PR Jr. 1990. Diving injuries of the cervical spine. Surgical Neurology 34(3): 155-8.

Barth JT, Freeman JR, Winters JE. 2000. Management of sports-related concussions. Dental Clinics of North America 44(1): 67-83.

Bauman A, Owen N. 1999. Physical activity of adult Australians: epidemiological evidence and potential strategies for health gain. Journal Science and Medicine in Sport 2(1): 30-41.

Bauze RJ, Ardran GM. 1979. Experimental production of forward dislocation in the human cervical spine. British Journal Bone Joint Surgery 60: 239-245.

BBC Sports News online. 2002. The price of professionalism. http://news.bbc.co.uk/sport2/hi/rugby_union /default.stm, last updated Friday 20 August, accessed 12 November 2002.

Bedbrook G. 1980. Discussion to papers of Steinbruck K, Green BA, Kiwerski J, Gaspar VG, Griffiths EL, Frankel HL. et al. Paraplegia 18: 133-138.

Bedbrook G. 1992. Fifty years on, fundamentals in spinal cord injury care are still important. Paraplegia 30: 10-13.

Bergman SB, Yarkony GM. 1997. Spinal cord injury rehabilitation. 2. Medical complications. Archives of physical medicine and rehabilitation 78(3): S53-8.

Berry JG, Harrison JE, Yeo JD, Cripps RA, Stephenson SC. 2006. Cervical spinal cord injury in rugby union and rugby league: are incidence rates declining in NSW?. Australian & New Zealand Journal of Public Health 30(3):268-74.

Better Health Commission. 1986. Looking forward to better health: the task forces and working groups. Volume 2. AGPS, Canberra.

Bishop PJ. 1996. Factors related to quadriplegia in football and the implications for intervention strategies. American Journal of Sports Medicine 24(2):235-9.

266

Bixby-Hammett D, Brooks WH. 1990. Common injuries in horseback riding. A review. Sports Medicine 9(1): 36-47.

Bland JM, Altman DG. 2000. Statistics Notes: The odds ratio. British Medical Journal 320: 1468.

Blanksby BA, Wearne FK, Elliott BC. 1996. Safe depths for teaching children to dive. Australian Journal of Science & Medicine in Sport 28(3): 79-85.

Blanksby BA, Wearne FK, Elliott BC, Blitvich JD. 1997. Aetiology and occurrence of diving injuries. A review of diving safety. Sports Medicine 23(4):228-46.

Blitvich JD, McElroy GK, Blanksby BA, Douglas GA. 1999. Characteristics of 'low risk' and 'high risk' dives by young adults: risk reduction in spinalcord injury. Spinal Cord 37(8): 553-9.

Blitvich JD, McElroy GK, Blanksby BA. 2000. Risk reduction in diving spinal cord injury: teaching safe diving skills. Journal of Science & Medicine in Sport 3(2): 120- 31.

Blum C, Shield J. 2000. Toddler drowning in domestic swimming pools. Injury Prevention 6(4): 288-90

Blumer CE, Quine S. 1996. Surveillance of traumatic spinal cord injury in Australia: the identification of information needs. Spinal Cord 34(11): 639-43.

Bond G, Randall C, Richard A, Rodgers BM. 1995. Pediatric equestrian injuries: assessing the impact of helmet use. Pediatrics 95(4): 487-9.

Bottini E, Poggi EJ, Luzuriaga F, Secin FP. 2000. Incidence and nature of the most common rugby injuries sustained in Argentina (1991-1997). British Journal of Sports Medicine 34(2): 94-7.

Boufos S, Dennis R, Finch F. 2006. A profile of hospitalisations and deaths due to sport and leisure injuries in New South Wales, 2000-2004 NSW Injury Risk Management Research Centre, Sydney.

Brandstater ME, Bontke CF, Cobble ND, Horn LJ. 1991. Rehabilitation in brain disorders – specific disorders. Archives of Physical Medicine & Rehabilitation 72(4- S): S332-40.

Bratton RL. 1997. Preparticipation screening of children for sports. Current recommendations. Sports Medicine 24(5): 300-7.

267 Brentnall, R. 1995. The ‘Safeplay’ code for NSW junior rugby league. Report to the NSW Sporting Injuries Committee. NSWSIC, Sydney.

Brewer J, Davis J. 1995. Applied physiology of rugby league. Sports Medicine 20(3): 129-35.

Browning KH, Donley BG. 2000. Evaluation and management of common running injuries. Cleveland Clinic Journal of Medicine 67(7): 511-20.

Burke DC. 1987. Planning a system of care for head injuries. Brain Injury 1(2):189-98.

Burry HC, Calcinai CJ. 1988. The need to make rugby safer. British Medical Journal 296: 149-150.

Burry HC, Gowland H. 1981. Cervical injury football – a New Zealand survey. British Journal of Sports Medicine 15(56): 15-19.

Cantu RC, and Voy R. 1995. Second impact syndrome: a risk in any contact sport. The Physician and Sports Medicine 23(6): 27-34.

Cantu RC, Bailes JE, Wilberger JE Jr. 1998. Guidelines for return to contact or collision sport after a cervical spine injury. Clinics in Sports Medicine 17(1): 137-46.

Cantu RC. Mueller FO. 2000. Catastrophic football injuries: 1977-1998. Neurosurgery 47(3): 673-7.

Carey V, Chapman S. Gaffney D. 1994. Children's lives or garden aesthetics? A case study in public health advocacy. Australian Journal of Public Health 18(1): 25-32.

Carlson BR, Carter JE, Patterson P. 1994. Physique and motor performance characteristics of US national rugby players. Journal of Sports Science 2(4): 403-12.

Carmody DJ, Taylor TKF, Parker DA, Coolican MRJ, Gumming RG. 2005. Spinal cord injuries in Australian footballers 1997-2002. Medical Journal of Australia 182(11):561-564.

Carvell JE, Fuller DJ, Duthie RB, Cockin J. 1983. Rugby football injuries to the cervical spine. British Medical Journal 286(6358): 4950.

Cass DT, Ross F, Lam LT. 1996. Childhood drowning in New South Wales 1990- 1995: a population-based study. Medical Journal of Australia 165(11-12): 610-2.

Chalmers DJ. 1998. Mouthguards. Protection for the mouth in rugby union. Sports Medicine 25(5): 339-49.

Chapman PJ. 1996. Mouthguard protection in sports. Australian Dental Journal 41(3): 212.

268 Chorba TL. 1991. Assessing technologies for preventing injuries in motor vehicle crashes. International Journal of Technology Assessment in Health Care 7(3): 296- 314.

Clarke KS. 1998. Epidemiology of athletic head injury. Clinical Journal of Sports Medicine 17(1): 1-12.

Close PJ, Scally PM, Laing BA. 1993. Cervical spine fractures due to spear tackles in two rugby league players. Injury 24(3): 189-90.

Cobble ND, Bontke CF, Brandstater ME, Horn LJ. 1991. Rehabilitation in brain disorders – intervention strategies. Archives of Physical Medicine & Rehabilitation 72(4-S): S324-31.

Commonwealth of Australia. 1994. Better health outcomes for Australians. AGPS, Canberra.

Condie C, Rivara FP, Bergman AB. 1993. Strategies of a successful campaign to promote the use of equestrian helmets. Public Health Reports 108(1): 121-6.

Connell, J. Zuel, B. 1994. Football Helmets proposed. Sydney Morning Herald 3 June 1994: p3.

Conroy C, Fowler J. 2000. The Haddon matrix: applying an epidemiologic research tool as a framework for death investigation. American Journal of Forensic Medicine & Pathology 21(4): 339-42.

Cooke JB. 1923. The future of the game: The Rugby Football League. A paper read at the conference of Rugby Football League, held at Llandudno, 23rd June 1923 (Davis Sporting Collection No.2 Box 42, Mitchell Library, NSW State Library).

Cordain L, Gotshall RW, Eaton SB. 1998. Physical activity, energy expenditure and fitness: an evolutionary perspective. International Journal of Sports Medicine 9: 328-335.

Cripps RA. 2000. Horse-related injury in Australia. Australian Injury Prevention Bulletin 24.

Cripps RA. 2004. Spinal cord injury, Australia 2002-03. Injury Research and Statistics Series. AIHW, Adelaide.

Cripps RA, O’Connor P. 1998. Spinal cord injury Australia 1996-97. Australian Injury Prevention Bulletin 18.

269 Daily Telegraph. 1998. Football player evacuated. 10th April 1998: p15.

Daily Telegraph. 2000. Junior players mismatched. 7th June 2000: p6.

Dale T, Ford I. 2002. Participation in exercise, recreation and sport 2001. Australian Sports Commission, Belconnen; ACT.

Daly RM, Bass SL, Finch CF. 2001. Balancing the risk of injury to gymnasts: how effective are the counter measures? British Journal of Sports Medicine 35(1): 8-18.

Dane S, Can S, Karsan O. 2002. Relations of Body Mass Index, body fat, and power of various muscles to sport injuries. Perceptual & Motor Skills. 95(1): 329-34.

Davidson RM. 1987. Schoolboy rugby injuries, 1969-1986. Medical Journal of Australia 147: 119-120.

Davis RM, Pless B. 2001. British Medical Journal bans "accidents". British Medical Journal 322(7298): 1320-1.

De Loes M. 1997. Exposure data. Why are they needed? Sports Medicine 24(3): 172-175.

Dietzen CJ, Topping BR. 1999. Rugby football. Physical Medicine & Rehabilitation Clinics of North America 10(1): 159-75.

Dixon AE, Shulman S. 1995. Sudden death during sports activities. New England Journal of Medicine 333(26): 1784-1785.

Dorsch MM, Woodward AJ, Somers RL. 1987. Do bicycle safety helmets reduce severity of head injury in real crashes? Accident Analysis & Prevention 19(3): 183- 90.

Dunning E, Sheard K. 1979. Barbarians, gentlemen and players: a sociological development of rugby football. Australian National University Press, Canberra.

Edgar M. 1995. Tackling rugby injuries. Lancet 345: 1452-53.

Edixhoven P, Sinha SC, Dandy DJ. 1981. Horse injuries. Injury 12(4): 283-287.

Edmond KM, Attia JR, Deste CA, Condon JT. 2001. Drowning and near-drowning in Northern Territory children. Medical Journal of Australia 175(11-12): 605-8.

Egger G. 1991. Sports injuries in Australia: causes, costs and prevention. Health Promotion Journal of Australia 1(2): 28-33.

Egger GJ, Vogels N, Westerterp KR. 2001. Estimating historical changes in physical activity. Medical Journal of Australia 175: 635-636.

270 Eime RM, Finch CF, Sherman CA, Garnham AP. 2002. Are squash players protecting their eyes? Injury Prevention 8(3): 239-41.

Eime RM, Finch CF. 2002. Have the attitudes of Australian squash players towards protective eyewear changed over the past decade? British Journal of Sports Medicine 36(6): 442-5.

Faccioni A. 1994. Speed development for team sport athletes. Sports Coach 17(2): 32-34.

Fadale PD, Hulstyn MJ. 1997. Common athletic knee injuries. Clinics in Sports Medicine 16(3): 479-99.

Fick DS. 1995. Management of concussion in collision sports. Guidelines for the sidelines. Postgraduate Medicine 97(2): 53-6, 59-60.

Finch CF. 1995. 'Sports injury prevention', in Ozanne-Smith J, Williams, F. (eds), Injury research and prevention: a text. Monash University Accident Research Centre: Clayton.

Finch CF. 1997. An overview of some definitional issues for sports injury surveillance. Sports Medicine 24(3): 157-63.

Finch CF, Donohue S, Garnham A. 2002. Safety attitudes and beliefs of junior Australian football players. Injury Prevention 8(2): 151-4.

Finch CF, Elliott BC, McGrath AC. 1999. Measures to prevent cricket injuries: an overview. Sports Medicine 28(4): 263-72.

Finch CF, Hennessy M. 2000. The safety practices of sporting clubs/centres in the city of Hume. Journal of Science & Medicine in Sport 3(1): 9-16.

Finch CF, Kelsall HL. 1998. The effectiveness of ski bindings and their professional adjustment for preventing alpine skiing injuries. Sports Medicine 25(6): 407-16.

Finch CF, Kenihan MA. 2001. A profile of patients attending sports medicine clinics. British Journal of Sports Medicine 35(4): 251-6.

Finch CF, McGrath A. 1997. Sportsafe Australia: a national sports safety framework. Australian Sports Commission, Canberra.

Finch CF, McIntosh AS, McCrory P. 2001. What do under 15 year old schoolboy rugby union players think about protective headgear? British Journal of Sports Medicine 35(2): 89-94.

Finch CF, Mitchell DJ. 2002. A comparison of two injury surveillance systems within sports medicine clinics. Journal of Science & Medicine in Sport 5(4): 321-35.

271

Finch CF, Owen N. 2001. Injury prevention and the promotion of physical activity: what is the nexus? Journal of Science & Medicine in Sport 4(1): 77-87.

Finch CF, Ozanne-Smith J, Williams F. 1995. The feasibility of improved data collection methodologies for sports injuries. National Sports Research Centre, Monash: Victoria.

Finch C, Valuri G, Ozanne-Smith J. 1998. Sport and active recreation injuries in Australia: evidence from emergency department presentations. British Journal of Sports Medicine 32(3): 220-5.

Finch CF, Valuri G, Ozanne-Smith J. 1999. Injury surveillance during medical coverage of sporting events – development and testing of a standardised data collection form. Journal of Science & Medicine in Sport 2(1): 42-56.

Fingerhut LA, Cox CS, Warner M. 1998. International comparative analysis of injury mortality. Findings from the ICE on injury statistics. International Collaborative Effort on Injury Statistics. Advance Data (303): 1-20.

Finvers KA, Strother RT, Mohtadi N. 1996. The effect of bicycling helmets in preventing significant bicycle-related injuries in children. Clinical Journal of Sport Medicine 6(2): 102-7.

Fisher KJ, Balanda KP. 1997. Caregiver factors and pool fencing: an exploratory analysis. Injury Prevention 3(4): 257-61.

Fradkin AJ, Finch CF, Sherman CA. 2001. Warm up practices of golfers: are they adequate? British Journal of Sports Medicine 35(2): 125-7.

Frankel HL, Hancock DO, Hyslop G. 1969. The value of postural reduction in the initial management of closed injuries of the spine with paraplegia and tetraplegia. Paraplegia 7(3): 179-192.

Franklin BA, Fletcher GF, Gordon NF, Noakes TD, Ades PA, Balady GJ. 1997. Cardiovascular evaluation of the athlete. Issues regarding performance, screening and sudden cardiac death. Sports Medicine 24(2): 97-119.

Fricker P. 1999. The young athlete. Australian Family Physician 28(6): 543-7.

Futterman LG, Myerburg R. 1998. Sudden death in athletes: an update. Sports Medicine 26(5): 335-50.

Gabbett TJ. 2000. Incidence, site, and nature of injuries in amateur rugby league over three consecutive seasons. British Journal of Sports Medicine 34(2): 98-103. Garlick D. 2001. Sports medicine education. British Journal of Sports Medicine 35(1): 78.

272

Garrard J. 1992. Promoting health and evaluating change. Australian Health Review 15(2): 213-24.

Garraway WM, Lee AJ, Hutton SJ, Russell EB, Macleod DA. 2000. Impact of professionalism on injuries in rugby union. British Journal of Sports Medicine 34(5): 348-51.

Garraway WM, Lee AJ, Macleod DA, Telfer JW, Deary IJ, Murray GD. 1999. Factors influencing tackle injuries in rugby union football. British Journal of Sports Medicine 33(1): 37-41.

Garraway WM, MacLeod DA, Sharp JC. 1991. Rugby injuries – the need for case registers. British Medical Journal 303(6810): 1082-3.

Garraway WM, Macleod DAD. 1995. Epidemiology of rugby football injuries. Lancet 345: 1485-87.

Gaspar VG, Silva RME. 1980. Spinal cord lesions due to water sports and occupations: our experience in 20 years. Paraplegia 18: 106-108.

Geffen L, Geffen G, Geffen S, Hinton-Bayre AD. 1998. 'Head injuries', in Sherry E, Wilson S. (ed) Oxford Handbook of sports medicine. Oxford University Press, New York.

Geffen S, Gibbs N, Geffen L. 1997. Thoracic spinal fracture in a rugby league footballer. Clinical Journal of Sports Medicine 7(2): 144-6.

Gemberling C. 1996. Preparticipation sports evaluation: an overview. Nurse Practitioner Forum 7(3): 125-35.

Gerrard DF. 1998. The use of padding in rugby union – an overview. Sports Medicine 25(5): 329-32.

Gibbs N. 1994. Common rugby league injuries. Recommendations for treatment and preventative measures. Sports Medicine 18(6): 438-50.

Gierup J, Larsson M, Lennquist S. 1976. Incidence and nature of horse-riding injuries, a one-year prospective study. Acta Chir Scand 142(1): 57-61.

Gilmore H. 1998. Jailed for footy tackle. The Daily Telegraph 3 August 1998: p4.

Gissane C, Jennings DC, Cumine AJ, Stephenson SE, White JA. 1997. Differences in the incidence of injury between rugby league forwards and backs. Australian Journal of Science and Medicine in Sport 29(4): 91-4.

273 Glaun R. Egnal A. Allen J. Noakes TD. 1984. Are high schools adequately prepared to cope with serious rugby injuries? South African Medical Journal 66(20): 768-70.

Gotsch K, Annest JL, Holmgreen MS, Gilchrist J. 2002. Nonfatal sports and recreation-related injuries treated in emergency departments – United States, July 2000–June 2001. Journal of the American Medical Association 288(16): 1977–79.

Grayson E. 1996. Medicolegal aspects of deliberate foul play in rugby union. British Journal of Sports Medicine 30(3): 191-2.

Gregory B, van Valkenburgh J. 1990. The athlete's knee. Journal of Post Anesthesia Nursing 5(6): 414-7.

Griffiths J. 1982. The Book of English international rugby, 1871-1982. Willow Books, London.

Gronwall D, Wrightson P. 1975. Cumulative effect of concussion. Lancet 2: 995-7.

Grossman JAI, Kulund DN, Miller CW, Winn HR, Hodge RH. 1978. Equestrian Injuries: Results of a prospective study. JAMA 240: 1881-1882.

Hall JC, Burke DC. 1978. Diving injury resulting in tetraplegia. Medical Journal of Australia 1: 171.

Hall SJ. 1995. Basic biomechanics. 2nd ed. Brown & Benchmark, Madison.

Hamilton LH, Hamilton WG, MelzterJD. 1989. Personality, stress, and injuries in professional ballet dancers. American Journal of Sports Medicine 17: 263-267.

Hamilton MG, Tranmer BI. 1993. Nervous system injuries in horse back riding accidents. Journal of Trauma-Injury Infection & Critical Care 34(2): 227-32.

Hankel HL, Montero FA, Penny PT. 1980. Spinal cord injuries due to diving. Paraplegia 18: 118-22.

Harrison J. 1999. Update on sports injury surveillance. Injury Issues Monitor 17: 3.

Hartkopp A, Bronnum-Hansen H, Seidenschnur AM. 1997. Survival and cause of death after traumatic spinal cord injury: a long-term epidemiological survey from Denmark. Spinal Cord 35: 76-85.

Haylen PT. 2004. Spinal injuries in rugby union, 1970–2003: lessons and responsibilities. Medical Journal of Australia 181(1): 48-50.

Heads I. 2000. 100 years of rugby league: a millennium special; parts 7-10. The Daily Telegraph May 2000; Saturday supplement.

274

Hinton-Bayre AD, Geffen GM, Geffen LB, McFarland KA, Friis P. 1999. Concussion in contact sports: reliable change indices of impairment and recovery. Journal of Clinical & Experimental Neuropsychology 21(1): 70-86.

Hoskins TW. 1978. Rugby injuries to the cervical cord. British Medical Journal 30: 1783.

Hoy WE, McFarlane R, Pugsley DJ, Norman R, Mathews JD. 1996. Markers for cardiovascular and renal morbidity: expectations for an intervention programme in an Australian Aboriginal community. Clinical & Experimental Pharmacology & Physiology 23(8): S33-7.

Hrysomallis, C. Morrison, WE. 1997. Sports injury surveillance and protective equipment. Sports Medicine 24(3): 181-3.

Hulkower S, Fagan B, Watts J, Ketterman E, Fox BA. 2005. Clinical inquiries: Do preparticipation clinical exams reduce morbidity and mortality for athletes? Journal of Family Practice 54(7):628-32.

Huxley, J. Faster, taller, stronger: the formula. Sydney Morning Herald July 26th 2003; p1, p10

Ingemarson H, Grevsten S, Thoren L. 1989. Lethal horse riding injuries. J Trauma 29: 25-30.

International Rugby Football board. 1995. Australian Rugby Football Union Handbook 1995. Laws of the game of rugby football. Australian Rugby Football Union, Sydney.

Jacobson GA. Blizzard L. Dwyer T. 1998. Bicycle injuries: road trauma is not the only concern. Australian & New Zealand Journal of Public Health 22(4): 451-5.

Jakoet I, Noakes TD. 1998. A high rate of injury during the 1995 Rugby World Cup. South African Medical Journal 88(1): 45-7.

Jennett B. 1996. Epidemiology of head injury. Journal of neurology, neurosurgery and psychiatry 60: 362-369.

Junge A, Dvorak J, Chomiak J, Peterson L, Graf-Baumann T. 2000. Medical history and physical findings in football players of different ages and skill levels. American Journal of Sports Medicine 28(5): S16-S21. Junge A. 2000. The influence of psychological factors on sports injuries: review of the literature. American Journal of Sports Medicine 28(5): S10-S15.

275 Katoh S, Shingu H, Ikata T, Iwatsubo E. 1996. Sports-related spinal cord injury in Japan (from the nationwide spinal cord injury registry between 1990 and 1992). Spinal Cord 34(7): 416-21.

Kew T, Noakes TD, Kettles AN, Goedeke RE, Newton DA, Scher AT. 1991. A retrospective study of spinal cord injuries in Cape Province rugby players, 1963- 1989 – incidence, mechanisms & prevention. South African Medical Journal 80: 127-133.

Kewalramani LS, Taylor RG. 1975. Injuries to the cervical spine from diving accidents. Journal of Trauma 15(2): 130-141.

Khan F, Ian J Baguley IJ, Cameron ID. 2003. Rehabilitation after traumatic brain injury. Medical Journal of Australia 178(6): 290-295.

Kinsella G, Ford B, Moran C. 1989. Survival of social relationships following head injury. International Disability Studies 11(1): 9-14.

Kluger Y, Jarosz D, Paul DB, Townsend RN, Diamond DL. 1994. Diving injuries: a preventable catastrophe. Journal of Trauma 36(3): 349-51.

Kolakowsky-Hayner SA, Gourley EV, Kreutzer JS, Marwitz JH, Cifu DX, Mckinley WO. 1999. Pre-injury substance abuse among persons with brain injury and persons with spinal cord injury. Brain Injury 13(8): 571-81.

Kriss TC, Kriss VM. 1997. Equine-related neurosurgical trauma: a prospective series of 30 patients. Journal of Trauma-Injury Infection & Critical Care 43(1): 97-9.

La Vecchia C, Levi F, Lucchini F, Negri E. 1994. Worldwide pattern of mortality from motor vehicle accidents, 1950-1990. Social and Preventive medicine 39(3): 150-78.

Larkins PA. 1995. Preventing sports injury remains a challenge; sport is a part of the Australian way of life, but the injury rate could be lower. Medical Journal of Australia 163: 230-231.

Lavis M, Rose J, Jenkinson T. 2001. Sports doctors' resuscitation skills under examination: do they take it seriously? British Journal of Sports Medicine 35: 128- 130.

Lawson J. 1991. (Ed.) Public health Australia: an introduction. McGraw Hill, Sydney.

Lawson J, Bauman AE. 2001. (eds) Public Health Australia: an introduction. (2nd ed). McGraw Hill, Sydney.

276 Lawson J, Rotem T, Wilson S. 1995. Catastrophic injuries to the eyes and testicles in footballers. Medical Journal of Australia 163(5): 242-244.

Lee AJ, Garraway WM, Arneil DW. 2001. Influences of pre-season training, fitness and existing injury on subsequent rugby injury. British Journal of Sports Medicine 35: 412-417.

Lee AJ, Garraway WM. 1996. Epidemiological comparison of injuries in school and senior club rugby. British Journal of Sports Medicine 30(3): 213-7.

Lee AJ, Myers JL, Garraway WM. 1997. Influence of players' physique on rugby football injuries. British Journal of Sports Medicine 31(2): 135-8.

Lett R, Kobusingye O, Sethi D. 2002. A unified framework for injury control: the public health approach and Haddon's matrix combined. Injury Control and Safety Promotion 9(3): 199-205.

Ley P. 1991. Isolation fencing and drownings in backyard pools. Medical Journal of Australia 154(10): 711-2.

Lloyd RG. 1987. Riding and other equestrian injuries: considerable severity. British Journal of Sports Medicine 21(1): 22-4.

Locke S. 1999. Exercise-related chronic lower leg pain. Australian Family Physician 28(6): 569-73.

Lower T. 1994. Rugby league surveillance report: report to the New South Wales Sporting Injuries Committee Research and Injury Prevention Programme. NSWSIC, Sydney.

Lynch SA, Renstrom PA. 1999. Groin injuries in sport: treatment strategies. Sports Medicine 28(2): 137-44.

Lysens RJ, Ostyn MS, Vanden S, Auweele Y. 1989. The accident-prone and overuse-prone profiles of the young athlete. American Journal of Sports Medicine 77: 612-619.

Lyznicki JM, Nielsen NH, Schneider JF. 2000. Cardiovascular screening of student athletes. American Family Physician 62(4): 765-74.

MacDougal GA, Osbourne NA. 1992. Head and facial lacerations in rugby union. Sport Health 10(2): 23-25.

Macklin K. 1962. The history of rugby league football. Anchor Press, Essex. Maddocks DL, Saling MM. 1995. Is cerebral concussion a transient phenomenon? Medical Journal of Australia 162(3): 167.

277 Mathers C, Vos T, Stevenson C. 1999. The burden of disease and injury in Australia. AIHW, Canberra.

McCrea M, Kelly JP, Kluge J, Ackley B, Randolph C. 1997. Standardised assessment of concussion in football players. Neurology 48(3): 586-8.

McCrory PR. 1997. Were you knocked out? A team physician's approach to initial concussion management. Medicine & Science in Sports & Exercise 29(7): S207- 12.

McCrory P. 1998. 'Neurological injuries in Australian and rugby football' in Jordan BD,Tsaris P, and Warren RF (eds) Sports Neurology, 2nd Ed. Lippincott-Raven, Philadelphia: pp. 441–451.

McCrory P. 2001a. When to retire after concussion. British Journal of Sports Medicine 35(6): 380-2.

McCrory P. 2001b. Does second impact syndrome exist? Clinical Journal of Sport Medicine 11(3): 144-9.

McCrory P. 2001c. Do mouthguards prevent concussion? British Journal of Sports Medicine 35: 81-82.

McCrory P. 2002. Concussion: What advice should we give to athletes post concussion? British Journal of Sports Medicine 36: 316-318.

McCrory PR, Ariens T, Berkovic SF. 2000. The nature and duration of acute concussive symptoms in Australian football. Clinical Journal of Sports Medicine 10(4):235-8.

McCrory PR, Berkovic SF. 1998. Second impact syndrome. Neurology 50(3): 677-83.

McCrory PR, Bladin PF, Berkovic SF. 1997. Retrospective study of concussive convulsions in elite Australian rules and rugby league footballers: phenomenology, aetiology, and outcome. British Medical Journal 314(7075): 171-4.

McCrory P, Dicker G. 1992. Head & Brain Injury in Sport: Australian Sports Medicine Federation Draft Policy. Sport Health 10(4): 36-39.

McCrory P, Seward H. 1992. Comparison of concussion rates in different codes of football (abst). Proceedings of the Australian Sports Medicine Federation National Scientific Conference, Canberra 1992.

McDermott FT, Lane JC, Brazenor GA, Debney EA. 1993. The effectiveness of bicyclist helmets: a study of 1710 casualties. Journal of Trauma-Injury Infection & Critical Care 34(6): 834-45.

278 McDermott FT. 1985. Prevention of road accidents in Australia. Pediatrician 12(1): 41-5.

McDermott FT. 1992. Helmet efficacy in the prevention of bicyclist head injuries: Royal Australasian College of Surgeons initiatives in the introduction of compulsory safety helmet wearing in Victoria, Australia. World Journal of Surgery 16(3): 379- 83.

McIntosh A, Dowdell B, Svensson N. 1998. Pedal cycle helmet effectiveness: a field study of pedal cycle accidents. Accident Analysis & Prevention 30(2): 161-8.

McIntosh A, McCrory P, Comerford J. 2000. The dynamics of concussive head impacts in rugby and Australian rules football. Medicine and Science in Sports and Exercise 32(12): 1980-84.

McIntosh AS, McCrory P. 2000. Impact energy attenuation performance of football headgear. British Journal of Sports Medicine 34(5): 337-41.

McIntosh A, McCrory P. 2001. Effectiveness of headgear in a pilot study of under 15 rugby union football. British Journal of Sports Medicine 35: 167-169.

Meinertz T, Hofmann T, Zehender M. 1991. Can we predict sudden cardiac death? Drugs 41: S2; 9-15.

Meldahl RV, Marshall RC, Scheinmann MC. 1988. Identification of persons at risk for sudden cardiac death. Medical Clinics of North America 72(5): 1015-31.

Milburn PD. 1987. A comparison of the mechanics of hip and crotch binding techniques in rugby union scrummaging. Australian Journal of Science and Medicine in Sports 19: 3-9.

Milburn PD. 1990. The kinetics of rugby union scrummaging. Journal Sports Science 8(1): 47-60.

Milburn PD. 1993. Biomechanics of rugby union scrummaging: Technical and safety issues. Sports Medicine 16(3): 168-79.

Mitchell R. 2002. NSW Water Safety Taskforce. New South Wales Public Health Bulletin. 13(4): 80-1.

279 Modern player size information. Rugby Football club website addresses. [On-line] Available at: www.ARU.com.au; www.NRl.com.au; www.sarugby.net; www.webrugby.com; www.rugbyleaguer.co.uk; www.rugby.com.au; www.nzrugby.co.nz; www.rfu.com. Accessed on 20/6/2000

Moller J, Elkington J. 1997. National health priority areas – injury prevention and control. Commonwealth Department of Health and Family Services, AIHW, Canberra.

Mueller FO. 1998. Fatalities from head and cervical spine injuries occurring in tackle football: 50 years' experience. Clinics in Sports Medicine 17(1): 169-82.

National Health and Medical Research Council. 1994. Football injuries of the head and neck. AGPS, Canberra.

National Health and Medical Research Council. 1999. Paradigm shift-injury: from problem to solution. New research directions. AGPS, Canberra.

National Health Priority Areas Report. 1997. Injury Prevention and Control, AGPS, Canberra.

Neumann DC, McCurdie IM, Wade AJ. 1998. A survey of injuries sustained in the game of touch. Journal of Science & Medicine in Sport 1(4): 228-35.

Nicholas CW. 1997. Anthropometric and physiological characteristics of rugby union football players. Sports Medicine 23(6): 375-96.

Noakes T, Jakoet I. 1995. Spinal cord injuries in rugby union players. How much longer must we wait for proper epidemiological studies? British Medical Journal 310(6991): 1345-6.

Noguchi T. 1994. A survey of spinal cord injuries resulting from sport. Paraplegia 32(3): 170-3.

Northern Sydney Area Health Service (NSAHS). 1997. NSW youth sports injury survey. NSAHS, Sydney.

Norton KI, Scwerdt S, Lange K. 2001. Evidence for the aetiology of injuries in Austrlian football. British Journal of Sports Medicine 35: 418-423.

Norton KI, Craig NP, Olds TS. 1999. The evolution of Australian football. Journal of Science & Medicine in Sport 2(4): 389-404.

NSW Rugby Football League. 1913. Laws of the game. Harold Murray, Sydney. Davis Sporting Collection No.2, Box 56 (Mitchell Library, NSW State Library)

280 NSW Rugby Union (NSWRU). 1926. Laws of the Game of football. NSWRU, Sydney. Davis Sporting Collection No.2, Box 56 (Mitchell Library, NSW State Library)

NSW Sport Injuries Insurance Scheme. Annual Reports of the NSW Sporting Injuries Committee 1984-95. NSWGSIIS, Sydney.

Nutbeam D, Wise M, Bauman A, Harris E, Leeder S. 1993. Goals and targets for Australia’s health in the year 2000 and beyond. AGPS, Canberra.

O'Connor FG, Johnson JD, Chapin M, Oriscello RG, Taylor DC. 2005. A pilot study of clinical agreement in cardiovascular pre participation examinations: how good is the standard of care? Clinical Journal of Sport Medicine 15(3):177-9.

O'Connor P. 2000a. Development and utilisation of the Australian Spinal Cord Injury Register. Spinal Cord 38(10): 597-603.

O'Connor P. 2000b. Spinal cord injury Australia 1998/99. Australian Injury Prevention Bulletin 22.

O’Connor P. 2001a. Spinal cord injury Australia, 1999-00. Australian Institute of Health and Welfare, Canberra.

O'Connor P. 2001b. Work related spinal cord injury, Australia 1986-97. Injury Prevention 7(1): 29-34.

O'Connor P. 2002a. Injury to the spinal cord in motor vehicle traffic crashes. Accident Analysis and Prevention 34(4): 477-85.

O’Connor P. 2002b. Hospitalisation due to traumatic brain injury (TBI), Australia 1997–98. Injury Research and Statistics Series. AIHW, Adelaide.

O’Connor P. 2002c. Spinal cord injury Australia, 2000-01. AIHW, Canberra.

O'Connor PJ. 2005. Survival after spinal cord injury in Australia. Archives of Physical Medicine & Rehabilitation. 86(1):37-47.

O'Connor P, Cripps RA. 1996. The Australian Spinal Cord Injury Surveillance System: Annual report 1995/96. National injury Surveillance Unit, Adelaide.

O'Connor P, Cripps R. 1997. Spinal Cord Injury, Australia 1995/96. Injury Prevention Bulletin 16.

O'Connor P, Cripps R. 1999. Spinal Cord Injury, Australia 1997/98 Injury Prevention Bulletin 21.

281 Odor JM, Watkins RG, Dillin WH, Dennis S, Saberi M. 1990. Incidence of cervical spinal stenosis professional and rookie football players. American Journal of Sports Medicine 18(5): 507-509.

Ohry A, Rozin R. 1982. Spinal cord injuries resulting from sport – the Israeli experience. Paraplegia 20: 334-338.

Olney D, Marsden A. 1986. The effect of head restraints and seat belts on the incidence of neck injury in car accidents. Injury 17(6): 5-7.

Orchard J, Chan M, Garlick D, Long F, Best J, McIntosh A. 2002. Second year report on the three year rugby injury surveillance project unpublished report. NSW Sporting Injuries Committee, Sydney.

Orchard J, Marsden J, Lord S, Garlick D. 1997. Preseason hamstring muscle weakness associated with hamstring muscle injury in Australian footballers. American Journal of Sports Medicine 25(1): 81-5.

Orchard J, Seward H. 1994. Football injuries. Sports Coach 17(2): 28-31.

Orchard JW, Finch CF. 2002. Australia needs to follow New Zealand’s lead on sports injuries. Medical Journal of Australia. 177: 38-39.

Orchard JW, Read JW, Neophyton J, Garlick D. 1998. Groin pain associated with ultrasound finding of inguinal canal posterior wall deficiency in Australian Rules footballers. British Journal of Sports Medicine 32(2): 134-9.

Ozanne-Smith J, Finch C, Routley V. 1994. 'Sport and recreational related injury' in Harrison JE, Cripps RA. (eds) Injury in Australia, an epidemiological review. AGPS, Canberra. (p. 45-95)

Paffenbarger RS Jr, Hyde RT, Wing AL, Steinmetz CH. 1984. A natural history of athleticism and cardiovascular health. JAMA 252(4): 491-5.

Paix BR. 1999. Rider injury rates and emergency medical services at equestrian events. British Journal of Sports Medicine 33(1): 46-8.

Parkkari J, Kujala UM, Kannus P. 2001. Is it possible to prevent sports injuries? Review of controlled clinical trials and recommendations for future work. Sports Medicine 31(14): 985-95.

Phelan A. Sport rule makers liable as crippled players win case. Sydney Morning Herald 20 October 1998; p1.

Pitt WR, Balanda KP. 1991. Childhood drowning and near-drowning in Brisbane: the contribution of domestic pools. Medical Journal of Australia 154: 661-665.

282 Pitt WR, Balanda KP. 1998. Toddler drownings in domestic swimming pools in Queensland since uniform fencing requirements. Medical Journal of Australia. 169(10): 557-8.

Pitt WR, Cass DT. 2001. Preventing children drowning in Australia. Medical Journal of Australia. 175(11-12): 603-4.

Porter CD. 1999. Football injuries. Physical Medicine & Rehabilitation 10(1): 95- 115.

Pounder DJ. 1984. "The grave yawns for the horseman." Equestrian deaths in South Australia 1973-1983. Medical Journal of Australia 141(10): 632-5.

Powell KE, Balie SN. 1994. The public health burdens of sedentary living habits: theoretical but realistic estimates. Medical Science in Sports and Exercise 26: 851- 856.

Press JM, Davis PD, Wiesner SL, Heinemann A, Semik P, Addison RG. 1995. The national jockey injury study: an analysis of injuries to professional horse-racing jockeys. Clinical Journal of Sports Medicine 5(4): 236-40.

Public Health Association of Australia. 1999. Injury – a major public health problem. Policy Statements 1999. Pirie, Canberra. (p.83-84)

Quarrie KL, Alsop JC, Waller AE, Bird YN, Marshall SW, Chalmers DJ. 2001. The New Zealand rugby injury and performance project. A prospective cohort study of risk factors for injury in rugby union football. British Journal of Sports Medicine 35: 157-166.

Quarrie KL, Cantu RC, Chalmers DJ. 2002. Rugby union injuries to the cervical spine and spinal cord. Sports Medicine 32(10): 633-53.

Quarrie KL, Chalmers DJ. 2001. Impact of professionalism on injuries in rugby union. British Journal of Sports Medicine 35: 450-453.

Quarrie KL, Handcock P, Toomey MJ. 1996. The New Zealand rugby injury and performance project. Anthropometric and physical performance comparisons between positional categories of senior A rugby players. British Journal of Sports Medicine 30(1): 53-6.

Rae K, Britt H, Orchard J, Finch C. 2005.Classifying sports medicine diagnoses: a comparison of the International classification of diseases 10-Australian modification (ICD-10-AM) and the Orchard sports injury classification system (OSICS-8). British Journal of Sports Medicine 39(12):907-11.

Raymond CA. 1988. Summer's drought reinforces diving's dangers. JAMA 260(9): 1199-1200.

283

Rosenfeld JV, McDermott FT, Laidlaw JD, Cordner SM, Tremayne AB. 2000. The preventability of death in road traffic fatalities with head injury in Victoria, Australia. The Consultative Committee on Road Traffic Fatalities. Journal of Clinical Neuroscience 7(6): 507-14.

Rotem TR, Davidson R. 2001. Epidemiology of Schoolboy football injuries. International Sports Med Journal 2(2). [Online] Available at: www.esportmed.com/ismj

Rotem TR, Lawson JS, Wilson SF, Engel S, Rutkowski SB, Aisbett CW. 1998. Severe cervical spinal cord injuries related to rugby union and league football in New South Wales, 1984-1996. Medical Journal of Australia 168(8): 379-381.

Roux CE, Goede R, Visser GR, VanZyl WA, Noakes TD. 1987. The epidemiology of schoolboy rugby injuries. South African Medical Journal 71(5): 307-13.

Rowland TW. 1986. Preparticipation sports examination of the child and adolescent athlete: changing views of an old ritual. Pediatrician 13(1): 3-9.

Runyan CW. 1998. Using the Haddon matrix: introducing the third dimension. Injury Prevention 4(4): 302-7.

Ryan RJ. 1983. The history of rugby league football in Australia. University Microfilms International, Ann Arbour, Michigan.

Salcido R, Costich JF. 1992. Recurrent traumatic brain injury. Brain Injury 6: 293-8.

Saunders RL, Harbaugh RE. 1984. The second impact in catastrophic contact- sports head trauma. Journal of the American Medical Association 252(4): 538-539.

Scher AT. 1977. Rugby injuries to the cervical spinal cord. South African Medical Journal 51: 473-5.

Scher AT. 1978. The high rugby tackle: an avoidable cause of cervical spinal injury? South African Medical Journal 53(25): 1015-8.

Scher AT. 1981. Vertex impact and cervical dislocation in rugby players. South African Medical Journal 58: 5927-8.

Scher AT. 1982. 'Crashing' the rugby scrum: an avoidable cause of cervical spinal injury: case reports. South African Medical Journal 61(24): 919-20.

Scher AT. 1983a. Serious cervical spine injury in the older rugby player: an indication for routine radiological examination. South African Medical Journal 64(4): 138-40.

284 Scher AT. 1983b. Rugby injuries to the cervical spine sustained during rucks and mauls: case reports. South African Medical journal 64(12): 456-8.

Scher AT. 1983c. The 'double tackle', another cause of serious cervical spinal injury in rugby players: case reports. South African Medical Journal 64(15): 592-4.

Scher AT. 1990a. Premature onset of degenerative disease of the cervical spine in rugby players. South African Medical Journal 77(11): 557-8.

Scher AT. 1990b. Cervical vertebral dislocation in a rugby player with congenital vertebral fusion. British Journal of Sports Medicine 24(3): 167-8.

Scher AT. 1991a. Spinal cord concussion in rugby players. American Journal of Sports Medicine 19(5): 485-488.

Scher AT. 1991b. Paralysis due to the high tackle: a black spot in South African rugby. South African Medical Journal 79(10): 614-5.

Scher AT. 1991c. Catastrophic rugby injuries of the spinal cord: changing patterns of injury. British Journal of Sports Medicine 25(1): 57-60.

Scher AT. 1995. Bodysurfing injuries of the spinal cord. South African Medical Journal 85(10): 1022-4.

Scher AT. 1998. Rugby injuries to the cervical spine and spinal cord: a 10-year review. Clinical Journal of Sports Medicine 17(1): 195-206.

Secin FP, Poggi EJ, Luzuriaga F, Laffaye HA. 1999. Disabling injuries of the cervical spine in Argentine rugby over the last 20 years. British Journal of Sports Medicine 33(1): 33-6.

Selecki BR, Berry G, Kwok B, Mandryk JA, Ring IT, Sewell MF. 1986. Experience with spinal injuries in New South Wales. Australian & New Zealand Journal of Surgery 56(7): 567-76.

Seward H, Orchard J, Hazard H, Collinson D. 1993. Football injuries in Australia at the elite level. Medical Journal of Australia 159: 298-301.

Sharp JCM, Murray GD, Macleod DAD. 2001. A unique insight into the incidence of rugby injuries using referee replacement reports. British Journal of Sports Medicine 35(1): 34.

Sherman CA, Finch CF. 2000. Preventing injuries to competitive and recreational adult golfers: what is the evidence? Journal of Science & Medicine in Sport 3(1): 65-78.

285 Sherry E. 1998. 'Epidemiology of Sporting injuries' in Sherry E, Wilson S. (ed) Oxford Handbook of sports medicine. Oxford University Press, New York.

Short D. 1999. Using science to prevent injuries: dissecting an event using the Haddon Matrix. Journal of Emergency Medical Services 24(9): 68-70, 72-4.

Silver JR. 1979. Rugby injuries to the cervical cord. British Medical Journal 32(1): 192-193.

Silver JR. 1984. Injuries of the spine sustained in rugby. British Medical Journal 288(6410): 37-43.

Silver JR. 1992. Injuries of the spine sustained during rugby. British Journal of Sports Medicine 26(4): 253-8.

Silver JR. 2001. Professionalism and injuries in rugby union. British Journal of Sports Medicine 35: 138-140.

Silver JR. 2002. The impact of the 21st century on rugby injuries. Spinal Cord 40(11): 552-9.

Silver JR, Stewart D. 1994. The prevention of spinal injuries in rugby football. Paraplegia 32(7): 442-53.

Smartplay. 2002. Smartplay. [On-line] Available at: www.smartplay.com.au/www/act/Article/Pub/ArticleDetail.asp?lngArticleID=1. Accessed 10/10/02.

Smith GS. 2001. Public health approaches to occupational injury prevention: do they work? Injury Prevention 7: 3-10.

Soden RJ, Walsh J, Middleton JW, Craven ML, Rutkowski SB, Yeo JD. 2000. Causes of death after spinal cord injury. Spinal Cord 38(10):604-10.

Soopramanien A. 1994. Epidemiology of spinal injuries in Romania. Paraplegia 32(11): 715-22.

Sorli JM. 2000. Equestrian injuries: a five year review of hospital admissions in British Columbia, Canada. Injury Prevention 6(1): 59-61.

Sosin DM, Sniezek JE, Thurman DJ. 1996. Incidence of mild and moderate brain injury in the United States, 1991. Brain Injury 10: 47-54.

Spinesafe. 1997. Annual report 1995-1996. Moorong Spinal Unit, Sydney.

Sport and Recreation Ministers' Council (Australia) 1997, Active Australia: A National Participation Framework, Australian Sports Commission, Canberra.

286

Sports Medicine Australia. 2002. About Sports Medicine Australia [On-line] Available at: www.sma.org.au. Accessed 10/10/02.

Standards Australia. 1995. A list of Australian standards: recreation. Standards Australia, Sydney.

Steinbruck K, Paeslack V. 1980. Analysis of 139 spinal cord injuries due to accidents in water sports. Papers read at the Annual Scientific Meeting of the International Medical society of Paraplegia held in Mulhouse, France, in July 1979 (Part II). Paraplegia 18: 86-93.

Stevenson MR, Rimajova M, Edgecombe D, Vickery K. 2003. Childhood drowning: barriers surrounding private swimming pools. Pediatrics. 111(2): E115-9.

Strategic Injury Prevention Partnership. 2001. National injury prevention plan: priorities for 2001–2003. Department of Health and Aged Care, Canberra.

Surf Life Saving Australia. 2000. Annual Report 2000. SLSA, Sydney.

Swimming Pools Act 1992 – Swimming Pools Regulation (No.2)

Swimming Pools Act 1992 – Swimming Pools Regulation (No.49)

Sydney Morning Herald. The impact of the ten metre rule. 16 November 1993; p48.

Sydney Morning Herald. Football foul play alleged. 29 September 1994; p6.

Sydney Morning Herald. It's a knockout. Those hard hits all part and parcel of the game. 25 February 1995; p68.

Sydney Morning Herald. Rules blamed for crippling in rugby. 20 April 1998; p5.

Taimela S, Osterman L, Kujala U. 1990. Motor ability and personality with reference to soccer injuries. Journal of Sports Medicine and Physical Fitness 30: 194-201.

Tator CH, Rowed DW, Schwartz ML. 1982. (eds): Sunnybrook cord injury scales for assessment of neurological injury and neurological recovery in early management of acute spinal cord injury. Raven Press, New York: 7.

Taylor TKF, Coolican MRJ. 1987. Spinal-cord injuries in Australian footballers, 1960-1985. Medical Journal of Australia 147: 112-118.

Taylor TKF, Coolican MRJ. 1988. Rugby must be safer: preventive programmes and rule changes. Medical Journal of Australia 149: 224.

287 Thomas BE, McCullen GM, Yuan HA. 1999. Cervical spine injuries in football players. American Journal of Orthopaedic Surgery 7(5): 338-47

Thomas S, Acton C, Nixon J, Battistutta D, Pitt WR, Clark R. 1994. Effectiveness of bicycle helmets in preventing head injury in children: case-control study. British Medical Journal 308(6922): 173-6.

Thompson DC, Nunn ME, Thompson RS, Rivara FP. 1996. Effectiveness of bicycle safety helmets in preventing serious facial injury. Journal of the American Medical Association 276(24): 1974-5.

Thompson DC, Rivara FP, Thompson R. 2000. Helmets for preventing head and facial injuries in bicyclists. Cochrane Database of Systematic Reviews. (2):CD001855.

Thompson NJ, Morris RD. 1994. Predicting injury risk in adolescent football players: The importance of psychological variables. Journal Pediatric Psychology 19: 415-429.

Thompson PD, Klocke FJ, Levine BD, van Camp SP. 1994. 26th Bethesda Conference recommendations for determining eligibility for competition in athletes with cardiovascular abnormalities. Task Force 5: coronary artery disease. Medicine & Science in Sports & Exercise 26(10):S271-5.

Thomson NJ. 1991. Recent trends in Aboriginal mortality. Medical Journal of Australia 154: 235-9.

Torg JS, Ramsey-Ermrhein JA. 1997. Management guidelines for participation in collision activities with congenital, developmental or post –injury lesions involving the cervical spine. Clinical Journal of Sports Medicine 16(3): 501-30.

Torg JS, Sennet B, Pavlov H, Leventhal MR, Glasgow SG. 1993. Spear tackler's spine: an entity precluding participation in tackle football and collision activities that expose the cervical spine to axial energy inputs. American Journal of Sports Medicine 21(5): 640-649. Trinca GW. 1987. The influence of alcohol countermeasures in changing drink driving attitudes. Asia-Pacific Journal of Public Health 1(1): 42-6.

Truswell AS. 1997. National policies promoting better nutrition, physical fitness and sports for all in Australia. World Review of Nutrition & Dietetics 81: 148-59.

Upton PAH, Roux CE, Noakes TD. 1996. Inadequate pre-season preparation of school boy rugby players: a survey of players at 25 Cape Province high schools. South African Medical Journal 86: 531-3.

Van Camp SP, Bloor CM, Mueller FO, Cantu RC, Olson HG. 1995. Nontraumatic sports death in high school and college athletes. Medicine & Science in Sports & Exercise 27(5): 641-7.

288 Van Mechelen W, Hlobil H, and Kemper HCG. 1992. Incidence, severity, etiology and prevention of sports injuries; a review of concepts. Sports Medicine 14(2): 82- 99.

Van Mechlen W. 1997a. Sports injury surveillance. 'One size fits all?' Sports Medicine 24(3): 164-168.

Van Mechlen W. 1997b. The severity of sports injuries. Sports Medicine 24(3): 176- 179.

Wakelam HBT.1937. Rugby football: how to succeed. Whitefriars Press, London.

Walsh J, DeRavin JW. 1995. Long-term care; disability and ageing. The Institute of Actuaries of Australia, Session Meeting, October/November, Sydney August 1995.

Walsh J. 1988. Costs of spinal cord injury in Australia. Paraplegia 26: 380-388.

Waters RL, Adkins RH, Yakura JS. 1991. Definition of complete spinal cord injury. Paraplegia 9: 583-581.

Watson W, Ozanne-Smith J. 1998. The cost of injury in Victoria. Monash University Accident Research Centre, Melbourne.

Watt GM, Finch CF. 1996. Preventing equestrian injuries. Locking the stable door. Sports Medicine 22(3): 187-97.

Watt GM, Ozanne-Smith J. 1994. Non-fatal injuries to young Victorians, 1986–1991. Medical Journal of Australia 160: 790–4. Weaver NL, Marshall SW, Miller MD. 2002. Preventing sports injuries: opportunities for intervention in youth athletics. Patient Education & Counseling 46(3): 199-204.

Wekesa M, Asembo JM, Njororai WW. 1996. Injury surveillance in a rugby tournament. British Journal of Sports Medicine 30(1): 61-3. Welcher JB, Szabo TJ. 2001. Relationships between seat properties and human subject kinematics in rear impact tests. Accident Analysis & Prevention 33(3): 289- 304.

Wetzler MJ, Akpata T, Albert T, Foster TE, Levy AS. 1996. A retrospective study of cervical spine injuries in American rugby, 1970 to 1994. American Journal of Sports Medicine 24(4): 454-8.

Wetzler MJ, Akpata T, Laughlin W. 1998. Occurrence of cervical spinal injuries during the rugby scrum. American Journal of Sports Medicine 26(2):177-80.

Whitlock MR. 1999. Injuries to riders in the cross country phase of eventing, the importance of protective equipment. British Journal of Sports Medicine 33(3): 212- 4.

289 Wigglesworth EC. 1987. Spinal injuries and football. Medical Journal of Australia 147: 109-110.

Wigglesworth EC. 2001. Towards an Australian Institute of Trauma Research: learning the lessons of history. ANZ Journal of Surgery 71(12): 765.

Wilberger JE Jr. 1998. Athletic spinal cord and spine injuries. Guidelines for initial management. Clinics in Sports Medicine 17(1): 111-20.

Wilberger JE. 1993. Minor head injuries in American football. Prevention of long term sequelae. Sports Medicine 15(5): 338-43.

Williams JPR, McKibbin B. 1978. Cervical spine injuries in rugby union football. British Medical Journal 2(6154):1747

Williams JPR. 2002. Editorial – Rugby Union. Spinal Cord 40(11):551-551.

Williams P, McKibbin B. 1987. Unstable cervical spine injuries in rugby – a 20-year review. Injury 18(5): 329-332.

Williamson A, Schmertmann M. 2002. Patterns of drowning and near drowning in NSW. New South Wales Public Health Bulletin 13(4): 78-80.

Wilson BD, Quarrie KL, Milburn PD, Chalmers DJ. 1999. The nature and circumstances of tackle injuries in rugby union. Journal of Science & Medicine in Sport 2(2): 153-62. Wilson BD. 1998. Protective headgear in rugby union. Sports Medicine 25(5): 333-7.

Wilson SF, Atkin PA, Rotem T, Lawson JS. 1996. Spinal cord injuries have fallen in rugby union in New South Wales. British Medical Journal 313(7071):1550.

Wintenmute GJ, Wright MA. 1991. The attitude-practice gap revisited: risk reduction beliefs and behaviours among owners of residential swimming pools. Paediatrics 88(6): 1168-71.

Woodward A, Dorsch MM, Simpson D. 1984. Head injuries in country and city. A study of hospital separations in South Australia. Medical Journal of Australia 141(1): 13-7.

Woolford S, Bourdon P, Craig N, Stanef T. 1993. Body composition and its effect on athletes performance. Sports Coach 16(4): 24-27.

Woolgar, J. 1999. England : the official R.F.U. history. Virgin, London.

Wright TM, Hood RW, Burstein AH. 1982. Analysis of material failures. Orthopedic Clinics of North America 13(1): 33-44.

290 Yarkony GM, Formal CS, Cawley MF. 1997. Spinal cord injury. Assessment and management during acute care. Archives of Physical Medicine and Rehabilitation 78(3): S48-52.

Yeo JD. 1979. Five year review of spinal cord injuries in motorcyclists. Medical Journal of Australia 2(1):381.

Yeo JD. 1980. Spinal cord injuries in motorcyclists. The Journal of the Western Pacific Orthopaedic Association 17(1): 56-58.

Yeo JD, Walsh J. 1987. Prevention of spinal cord injuries in Australia. Paraplegia 25(3): 221-224.

Yeo JD. 1993. Prevention of spinal cord injuries in an Australian study. Paraplegia 31(12): 759-763.

Yeo JD. 1998a. Rugby and spinal injury: what can be done? Medical Journal of Australia 168(8):372-3.

Yeo JD. 1998b. Spine injuries. in Sherry E, Wilson S. (ed) Oxford Handbook of sports medicine. Oxford University Press, New York.

Yeo JD, Walsh J, Rutkowski S, Soden R, Craven M, Middleton J. 1998. Mortality following spinal cord injury. Spinal Cord 36(5):329-36.

Young MC, Fricker PA, Thomson NJ, Lee KA. 1999. Sudden death due to ischaemic heart disease in young Aboriginal sportsmen in the Northern Territory, 1982-1996. Medical Journal of Australia 170(9): 425-8.

291 APPENDICES Appendix A

ASIA Impairment scale

i ASIA impairment category

To measure the change in the degree of impairment of cases in response to the combined effects of treatment in the acute care facility and during rehabilitation, ASIA impairment category for each case was recorded at admission and at discharge. This measure of impairment was derived from the Frankel cord injury scale (see references below). The categories relevant to an assessment of persisting cases of SCI are presented below:

A = Complete. No sensory or motor function is preserved at and below the spinal lesion. B = Incomplete. Sensory but not motor function is preserved below the neurological level. C = Incomplete. Motor function is preserved below the neurological level, and the majority of key muscles below the neurological level have a muscle grade less than 3 and usually some sensory sparing. D = Incomplete. Motor function is preserved below the neurological level, and the majority of key muscles below the neurological level have a muscle grade greater than or equal to 3 and almost normal sensation below the level. E = Normal motor power and sensation

References

Frankel HL, Hancock DO, Hyslop G et al. The value of postural reduction in the initial management of closed injuries of the spine with paraplegia and tetraplegia. Paraplegia 1969; 7(3): 179-192.

Tator CH, Rowed DW, Schwartz ML. (eds): Sunnybrook cord injury scales for assessment of neurological injury and neurological recovery in early management of acute spinal cord injury. New York: Raven Press, 1982: 7.

ii Appendix B

Major definitions used for the pilot Australian Rugby Union

injury survey in 2000-2002

iii The major definitions used for the pilot Australian Rugby Union injury survey in 200-2002 are as follows (Orchard, Chan, Garlick, Long, Best, and McIntosh, 2002).

"Definition of an injury: An injury or medical condition that caused a player to miss a match. Definition of cohort: - that is, the group of players to be followed during the season. These were taken as the 15 players in the first grade team for Round 1. In 2001, teams were encouraged to add to their cohort any player who played first grade during the regular season thereafter. Some teams accepted this offer and their cohort size increased during the year. Consistency was obtained in that all players only joined the survey by playing a first grade match for that team during the year. Definition of season: The season surveyed for each team started at Round 1 of the regular season and continued until the last match (finals were included if the team played in them). In order to equally compare injury rates from teams with seasons of different lengths, a “team season” was defined as 300 player weeks (that is, 15 players being followed for 20 weeks). Definition of player status for a round: In any given round, a player in the cohort was recorded as having a ‘status’ of one of the following: (1) playing in first grade (2) playing another type of match (e.g. second grade or representative match) (3) not playing due to an injury (4) not playing due to a team bye (5) not playing due to other reasons (e.g. suspended, personal reasons, retirement). By definition, every player in the selected cohort was playing in first grade during round 1. Definition of missing data: Where the status of a player (in the cohort) is unknown for a given round during the season. Definition of injury recovery: An injury was considered to be recovered when the player returned to playing a match. Any further episode after return to play was considered a new injury. Definition of injury onset: Onset was divided into (1) injuries occurring in first grade matches (2) injuries occurring in other matches (3) injuries occurring at training (4) injuries with a gradual onset (5) injuries occurring outside rugby, including illness. Injury seasonal incidence: Injury seasonal incidence was measured in units of injuries per team per season, which has been defined as injuries occurring per 300 player weeks. All injuries were included in the numerator. All players were considered at risk for each week of the survey. The final round for each team was not considered to be exposure time due to the definition of an injury (that it must cause a player to miss a game), as there was no future round for the player to miss. Injury match incidence. Injury match incidence: Injury match incidence was measured in units of injuries per 1000 player hours. Only injuries that occurred in first grade matches were included in the numerator and only players participating in first grade matches in each week were included in the denominator. A match was considered to last 80 minutes for Elite, City and Country matches and 70 minutes for Schools matches. All players were considered to have played the full game (as exact time of participation was not available). The final round for each team was not considered to be exposure time due to the definition of an injury (that it must cause a player to miss a game), as there was no future round for the player to miss. Injury pre valence Injury prevalence was defined as matches missed per team per season (or matches missed through injury per 300 player weeks)."

iv

Appendix C

Medical record/Case file data collection form

v Sport and recreational spinal injuries

ID# ______/_____

Sex M/F

DOB ______

Sport ______

Accident date ______

Level ______

Resulting Cond.

______

Mechanism ______Source ______

Other

Position______

Grade______

Location of accident______

Home address______

vi NSWSIC benefits paid cases

ID# ______/_____

Sex M/F

DOB ______

Sport ______

Accident date ______

Injury Type ______Level ______

Resulting Cond.

______

Mechanism ______Source ______

Testimony

______

Position______

Grade______

Location of accident______

Home address______

vii

Appendix D

NSW Sporting Injuries Committee membership

and premium rates

viii Membership of NSW Sporting Injuries Committee

Membership covered by the NSWSIC for the rugby football codes and horse related sports 1984-1999 is shown in the tables A-C below. Non-contestants, officials, etc. have not been included in these tables. Figures were recorded or estimated at the 30 June for each year and as some figures were estimations, the latest corrections available were used.

It has been estimated that in NSW there were around 106,000 participants in Rugby league in each year of the review period. The NSWSIC scheme covered nearly 80% of this population with around 84,00 participant members per year. Rugby union left the scheme in 1990 to make private insurance arrangements. The total participants in NSW of Rugby union has been estimated at around 31,000 per year over the review period. The SIIC scheme covered over 75% of this population with around 23,500 participant members per year until 1990.

Rugby League

Table A: Information as to the number of individuals included in the accommodation of Rugby League organizations accommodated by the Insurance-Scheme 1984-99 Year Adults Non-adults Total 1984 23,621 60,908 84,529 1985 24,616 54,123 78,739 1986 22,628 57,591 80,219 1987 24,013 57,502 81,515 1988 23,647 56,319 79,966 1989 23,655 56,319 79,974 1990 24,319 54,308 78,627 1991 25,724 55,038 80,762 1992 27,616 56,040 83,656 1993 31,481 50,442 81,923 1994 25,565 59,245 84,810 1995 22,966 59,454 82,420 1996 21,629 62,509 84,138 1997 20,782 63,915 84,697 1998 18,245 58,172 76,147 1999 19,163 56,630 75,793 Source : NSWSIC annual reports 1984 to 2000

ix Rugby Union

Table B: Information as to the numbers of individuals included in the accommodation of Rugby Union organizations accommodated by the Insurance Scheme 1984-99 Year Adult Non-adult Total 1984 13,760 9,597 23,357 1985 13,600 9,990 23,590 1986 14,000 10,600 24,600 1987 12,180 9,840 22,020 1988 12,180 11,330 23,510 1989 12,180 11,330 23,510 1990 ------1991 3,360 ------3,360 1992 ------1993 ------25 25 1994 ------25 25 1995 ------25 25 1996 ------25 25 1997 ------40 40 1998 395 28 423 1999 785 53 838 Source : NSWSIC annual reports 1984 to 2000

Other sports of interest

Table C: Information as to the number of individuals included in the accommodation of sporting organizations accommodated by the Insurance Scheme 1984-99 Year Horse racing Mini- harness Pony riding Polo 1984 144 ------15,634 350 1985 150 152 15,000 350 1986 95 200 18,000 350 1987 110 190 27,621 324 1988 98 200 27,621 315 1989 100 200 27,621 310 1990 200 ------15,678 310 1991 100 ------17,736 310 1992 100 ------17,731 310 1993 100 52 ------320 1994 80 338 ------320 1995 65 286 ------340 1996 59 260 ------340 1997 20 260 ------340 1998 17 160 ------330 1999 17 120 ------360 Source : NSWSIC annual reports 1984 to 2000

x NSW Sporting Injuries Committee premium

Table D - NSW Sporting Injuries Committee risk category premium rates Sport $ Premium $ Premium (incl GST) (incl GST) Adult Junior Motor-cycling 40.92 9.46 Rugby Union 37.95 9.46 Horse-race Riding (Amateur) 22.72 4.73 Polo 22.72 4.73 Rugby League 17.38 3.63 Hang Gliding 15.18 3.19 Australian Rules 7.65 1.54 Basketball 3.85 0.77 Netball 3.85 0.77 Indoor Soccer 3.19 0.77 Soccer 3.19 0.77 Touch Football 3.19 0.77 Cricket 1.98 0.38 Source: Personal communication- NSWSIC 5/3/2003

xi

Appendix E

Video incident review form and Coding manual

xii

xiii Rugby Football Video Analysis Coding Manual

Coding sheets provided to record tackling and scrum mechanisms occurring in video archives footage of rugby union and league games over the last 80 years.

Explanation of variables

Code - League or Union Game description - Teams involved, location, club match/grand final/e.t.c Date - Date of game Source - Archival footage source e.g. Cinesound review e.t.c Video time - Time code start/ stop / total time of each footage segment Total Incidents - Total number of tackles/scrums for each footage segment

Number in tackle - For each tackling incident how many tacklers were involved (include only fully committed tacklers)

Ball carrying style - Two hands = Ball held in front of and away from body with two hands Swinging = Ball held to side with swinging movement Waist = Ball held at waist level Chest = Ball held at chest level

Note: Most interested in carrying style at point of impact, however, if ball runner clearly changes style just before impact multiple values may be indicated as follows; number as occurred and indicate “PI” for style at point of impact. If no multiple values only one box needs to be marked.

1st tacklers level of impact - Recording 1st tacklers first committed point of impact. (legs/waist/chest/head-neck) 2nd tacklers level of impact - Recording 2nd tacklers first committed point of impact. 3rd tacklers level of impact - Recording 3rd tacklers first committed point of impact. (Refer to Figure 1)

1st tacklers direction of impact – Recording direction of 1st impact. (clock face) 2nd tacklers direction of impact – Recording direction of 2nd impact 3rd tacklers direction of impact – Recording direction of 3rd impact (Refer to Figure 2)

Scrum - Indicate whether a scrum is; Normal Wheeling (turning around) Collapsed Popping (involves incident of player being popped out of scrum) Note: May be multiple values numbered in order of occurrence.

General coding notes

If no multiple values then a simple cross/dot/tick will suffice (however keep consistent style) Strike a vertical line through non-applicable sections e.g. through sections 4, 5 and 6 if there is only one tackler An “S” in the “Number of tackles” box indicates incident was a scrum. “L” indicates a lineout. “R” indicates ruck or maul in which level and angle of impacts are often difficult to determine. All incidents that cannot be clearly defined must still be recorded by indicating a question mark (?) in the undefined section(s) (this is very important to establish rate of undeterminable incidents)

xiv Figure 1 – Impact level

Level 1

Level 2

Level 3

Level 4

Figure 2 – Direction of impact 12

9 3

6

xv

Appendix F

Glossary

xvi Rugby football Glossary

Rugby football/rugby codes – reference to both rugby union and rugby league football codes. Note the term football in the absence of mention of rugby is commonly known as a reference to soccer.

Summary

Tackle - Where one or more players may attack a single opponent

Scrum - Group of players who join and push together against opponents – under the direction of a referee. The scrum is a set play in which the two opposing positioned forward packs come together to contest possession of the ball. It is essentially used to restart play after minor infringements.

Maul and Ruck - Loose scrum like plays where players join and push together without creating a formal scrum.

Play terms

Law 18 - Tackle “A tackle occurs when a player carrying the ball in the field-of-play is held by one or more opponents so that while he is so held he is brought to the ground or the ball comes in contact with the ground. If the ball carrier is on one knee, or both knees, or is sitting on the ground, or is on top of another player who is on the ground, the ball carrier is deemed to have been brought to the ground.”

Law 20 - Scrummage “A scrummage, which can only take place in the field-of-play, is formed by players from each team closing up in readiness to allow the ball to be put on the ground between them but is not to be formed within five metres of the touch line. The middle player in each front row is the hooker, and the players on either side of him are props.”

Law 21 - Ruck “A ruck, which can only take place in the Field-of-play, is formed when the ball is on the ground and one or more players from each team are on their feet and in physical contact, closing around the ball between them. If the ball in a ruck is on or over the goal line the ruck is ended. Rucking is the act of a player who is in a ruck using his feet to retrieve or retain the ball without contravening Law 26 (Foul play law).”

Law 22 - Maul “A maul, which can only take place in the field-of-play, is formed by one or more players from each team on their feet and in physical contact closing round a player who is in possession of the ball. A maul ends when the ball is on the ground or the ball or a player carrying it emerges from the maul or when a scrummage is ordered. If the ball in a maul is on or over the goal line the maul is ended.”

(Australian Rugby Union Football Union Handbook 1995)

xvii Other play terms

“Scrum Popping” This occurs at the time of scrum engagement in the front row forward player in the “tight head “position. The opposing front row forward in the “loose head” position in combination with the opposing hooker lever the “tight head” front row forward upwards and out of the scrum. The resultant forced flexion of the cervical spine followed by rapid extension of the cervical spine occurs as the head explodes from the scrum as a cork would “pop” from a bottle thus the term “popping”. It is believed scrum popping can occur accidentally.

“Scrum Wheeling” This refers to a turning action (hence wheeling) which changes the angle of a scrum. This is sometimes used as a technique to manoeuvre key players further away from important areas of play.

“Crotch binding” Holding around the upper leg and groin of fellow team mate in a scrum.

"Head high tackle" Tackling above shoulders

"Head Bin" Off field area assigned for treatment of facial or head injuries.

"Sin Bin" Off field area assigned for penalising players who commit a foul or demonstrate poor sportsmanship. Official timekeeper used to monitor players time penalty.

Important rugby union positions

Hooker The hooker is a forward whose main function is to compete with the opposing hooker for possession of the ball in the scrum. A hooker is positioned in the middle of the front row of the scrum and attempts to drag the ball out with his feet to gain possession for his team.

Loose head prop Forward front row player positioned on the left side of hooker in scrum. Closest to feed in of ball.

Tight head prop Forward front row player positioned on the right side of hooker in scrum. Places head between heads of opposing hooker and loose head prop.

Halfback The halfback is considered to be the link between the forwards and the backs. This player must pass the ball out of the scrum to the backs.

(Eddy, A. – NSW Rugby Union 1996. pers. comm., 26 November)

(Meredith, M – NSW Rugby League 1996. pers. comm., 10 December)

xviii Medical Glossary

Arachnoid- thin membrane of the brain and spinal cord that lies between the dura mater and the pia mater

Arrhythmias- irregular heartbeat

Bilateral- towards, at, or belonging to both sides e.g. of brain.

Brain Oedema- swelling resulting from increased water content occurring as a result of injury to the brain.

Brain Haemorrhage- localized bleeding resulting from injury to the blood vessels in or around the brain. There are 4 types of haemorrhage: extradural, subdural, subarachnoid, and intracerebral.

Cardiac arrhythmia/failure/arrest- cessation of the pumping activity of the heart resulting in failure of circulation of the blood throughout the body. Cardio vascular disease- term covering diseases of the cardiovascular system. Include congenital, valvular, myopericardial or coronary heart disease, high blood pressure and arrhythmias. Concussive convulsions- observable within seconds of insult are to be distinguished from later epileptic seizures. Such concussive convulsions are rare, transient, and do not necessarily lead to the development of epilepsy.

Cerebral Oedema- swelling in the brain due to an increase in its water content, associated with such lesions as tumour, abscess, concussion or haemorrhage.

Cerebral haemorrhage- haemorrhage arising within the cerebral parenchyma.

Cerebral vascular accident- sudden rupture or blockage of a blood vessel within the brain causing serious bleeding or local obstruction to circulatory flow in the brain, resulting in a stroke.

Closed head injury- trauma to the head, which does not fracture or penetrate the skull but severely shakes the brain and may result in brain damage. This can occur as a result of an auto accident, sports injury, fall, assault, work related, accident at home, or from a bullet wound.

xix Coma- a state of unconsciousness and unresponsiveness that results from disturbance or damage to areas of the brain.

Cognitive- awareness with perception, reasoning and judgement, intuition, and memory; The mental process by which knowledge is acquired.

Cognitive Deficit- difficulties in reasoning, judgment, intuition and memory, lack of awareness and insight.

Concussion/contusion-a mild injury or bruise to the brain which may result in a brief period of loss of consciousness. This may cause memory loss, difficulty concentrating headaches, nausea, vomiting and dizziness.

Congenital abnormality- an abnormality present at birth. The abnormality may have been inherited genetically from the parents, or occurred as a result of damage or infection of the uterus, or may have occurred at the time of birth. Dura matter- the outermost meninx of the brain, thick, dense and inelastic. It lines the interior of the skull and is two layered, meningeal and endosteal. Endosteal- 1. relating to the endosteum (The layer of vascular connective tissue lining the medullary cavities of bone) 2 : located within bone or cartilage. Epilepsy- any of various disorders marked by disturbed electrical rhythms of the central nervous system and typically manifested by convulsive attacks usually with clouding of consciousness. Evacuation- the act of removing. Extradural- situated on the outside of or lying outside dura matter Extradural haematoma- haematoma between the dura matter and the skull. Glasgow coma scale- scale which evaluates the patient's level of awareness, which indirectly indicates the extent of neurologic injury. The scale rates three categories of patient responses; eye opening, best verbal response, and best motor response. The lowest score is 3 and is indicative of no response, the highest score is 15, indicates the patient is alert and aware of his or her surroundings.

Haematoma- a localized collection of blood, usually clotted, caused by bleeding from a ruptured blood vessel.

xx Haemorrhage- bleeding, the escape of blood from any part of the vascular system. It may be venous, arterial or capillary from blood vessels into tissues, into or from the body.

Hypertrophy- increase in the size of an organ

Hypertrophic obstructive cardiomyopathy- congenital condition in which there is asymmetric hypertrophy of the left ventricle and septum so that there is an outflow obstruction.

Infarct- an area of necrosis in a tissue or organ resulting from obstruction of the local circulation by a thrombus or embolus

Infarction- the process of forming an infarct

Intracranial haemorrhage- bleeding into the cranium.

Intracerebral haemorrhage- bleeding into the brain from a ruptured vessel, and is one of mechanisms that can cause a stroke.

Intracerebral haematoma- bleeding in and around brain tissue leads to a build-up of blood within the brain itself, these hematomas usually result from penetrating wounds or blood vessels that rupture.

Lesion- pathological disturbance, such as an injury, an infection, or a tumour.

Loss of consciousness- lack of awareness and having perception

Meningeal- of, relating to, or affecting the meninges (plural of meninx) Meninx- any of the three membranes that envelop the brain and spinal cord and include the arachnoid, dura mater, and pia mater Mild head injury- loss of consciousness if at all of 20 minutes or less, post traumatic amnesia of less than 24 hours, non-focal negative neurologic exam, and normal Neuro diagnostic studies. Symptoms are highly variable and can emerge anywhere from 24-hours to two weeks post injury. These symptoms can range from headache to dizziness to emotional, physical, cognitive or intellectual. Myocarditis- heart inflammation Myopericardial- heart inflammation

xxi Necrosis- death of living tissue ; specifically death of a portion of tissue differentially affected by local injury (as loss of blood supply, corrosion, burning, or the local lesion of a disease)

Oedema- presence of excessive amounts of fluid in the intercellular tissue spaces of the body, due to increased transudation of fluid from the capillaries.

Parenchyma- functional tissue of an organ

Paraparesis- latest term for paraplegia.

Paraplegia- condition in which the legs are paralysed.

Pia mater- the delicate and highly vascular membrane of connective tissue investing the brain and spinal cord, lying internal to the arachnoid and dura mater, dipping down between the convolutions of the brain, and sending an ingrowth into the anterior fissure of the spinal cord

Physical Aggression- a forceful physical action performed as self-protective or inappropriate to the particular situation.

Pneumonia- lung disease

Poor judgment and safety awareness- the inability to integrate all information when making decisions, inability to identify hazards in the immediate area. Posttraumatic epilepsy- may develop within days to months following traumatic brain injury with prolonged periods of unconsciousness. Occurs in close to a quarter of hospital admissions for sports related head injury. Quadriparesis- latest term for quadriplegia.

Quadriplegia- condition in which both arms and legs are paralysed.

Raised intracranial pressure- following a brain injury there is often a build-up of pressure within the skull, which compresses delicate brain tissue and may lead to further brain injury. The brain its membranes and cerebrospinal fluid are all encased in the skull, therefore resulting in no space to accommodate the accumulation of blood or swelling hence -the pressure builds up.

xxii Second impact syndrome- variant of malignant brain oedema syndrome observed following head trauma when still suffering symptoms of a previous head injury

Stroke- sudden diminution or loss of consciousness, sensation, and voluntary motion caused by rupture or obstruction (as by a clot) of a blood vessel of the brain -- called also apoplexy, brain attack, cerebral accident, cerebrovascular accident

Subarachnoid haemorrhage- the type of brain haemorrhage in which blood from a ruptured blood vessel spreads over the surface of the brain. The most common cause is a ruptured aneurysm.

Subdural - beneath the dura matter. Subdural haematoma- a blood clot that forms between the dura and the brain tissue. If this bleeding occurs quickly it is called an acute subdural hematoma. If it occurs slowly over weeks it is called a chronic subdural hematoma. The clot may cause increased pressure and may need to be surgically removed. Thrombosis - blood clot Vascular- blood vessel-related Vertebral- vertebra-related

Source: Sherry E, Wilson S. 1998. (ed) Oxford Handbook of sports medicine. Oxford University Press, New York.

xxiii

Appendix G

Estimates of the participant populations for miscellaneous sports

xxiv

The Exercise, Recreation and Sport Survey 2001 Sport Total '000 Cycling 400.8 Waterskiing/powerboating 56.8 Surf 172.7 Ice/snow 89.7 Motor 55.2 Source: Tim Dale And Ian Ford 2002. Participation in exercise, recreation and sport 2001 Australian Sports Commission Belconnen ACT

Participation in Sport and Physical Activities, Australia, 1997-2000 Total '000 Total '000 Total '000 Sport 1997-1998 1998-1999 1999-2000 Cycling 181.1 190.5168.3 Surf 101.9 121.3 107.6 Ice/snow 81.4 99.482.3 Source: Australian Bureau of Statistics. Participation in Sport and Physical Activities, Australia, 1998; 1999c; 2000a (cat. no. 4177.0).

xxv

Appendix H

Statistical Analysis

xxvi

Archival Rugby football Video analysis

Frequencies DATECAT6

Frequency Percent Valid Percent

1957 and before 843 19.4 49.8

Valid 1990 and after 850 19.5 50.2

Total 1693 38.9 100.0

Missing System 2658 61.1

Total 4351 100.0

Crosstabs Number of tacklers by date split Crosstab

DATECAT6 Total 1990 and after 1957 and before

Count 598 464 1062 1 Tackler % within DATECAT6 86.4% 66.3% 76.3% INCCAT Count 94 236 330 Multiple Tacklers % within DATECAT6 13.6% 33.7% 23.7%

Count 692 700 1392 Total % within DATECAT6 100.0% 100.0% 100.0%

Chi-Square Tests

Value df Exact Sig. (2-sided) Exact Sig. (1-sided)

Pearson Chi-Square 77.967(b) 1

Continuity Correction(a) 76.858 1

Likelihood Ratio 80.051 1

Fisher's Exact Test .000 .000

Linear-by-Linear Association 77.911 1

N of Valid Cases 1392

a Computed only for a 2x2 table

b 0 cells (.0%) have expected count less than 5. The minimum expected count is 164.05.

xxvii Ball carrying style by Date split Crosstab

DATECAT6 Total 1990 and after 1957 and before

Count 336 95 431 Two Hands % within DATECAT6 51.3% 13.6% 31.9%

Count 20 4 24 Swinging % within DATECAT6 3.1% .6% 1.8% BALLCL Count 75 15 90 Waist % within DATECAT6 11.5% 2.2% 6.7%

Count 224 583 807 Chest % within DATECAT6 34.2% 83.6% 59.7%

Count 655 697 1352 Total % within DATECAT6 100.0% 100.0% 100.0%

Chi-Square Tests

Value df Asymp. Sig. (2-sided)

Pearson Chi-Square 344.157(a) 3 .000

Likelihood Ratio 362.287 3 .000

Linear-by-Linear Association 289.693 1 .000

N of Valid Cases 1352

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 11.63.

Angle of 1st impact by Date split Crosstab

DATECAT6 Total 1990 and after 1957 and before

Count 138 286 424 Front-half % within DATECAT6 21.5% 41.1% 31.7%

Count 128 85 213 ANGLE1CL Sides % within DATECAT6 19.9% 12.2% 15.9%

Count 377 325 702 Back-half % within DATECAT6 58.6% 46.7% 52.4%

Count 643 696 1339 Total % within DATECAT6 100.0% 100.0% 100.0%

xxviii Chi-Square Tests

Value df Asymp. Sig. (2-sided)

Pearson Chi-Square 62.193(a) 2 .000

Likelihood Ratio 63.262 2 .000

Linear-by-Linear Association 41.711 1 .000

N of Valid Cases 1339

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 102.28.

Angle of 2nd impact by Date split Crosstab

DATECAT6 Total 1990 and after 1957 and before

Count 24 63 87 Front-half % within DATECAT6 28.6% 27.8% 28.0%

Count 30 95 125 ANGLE2CL Sides % within DATECAT6 35.7% 41.9% 40.2%

Count 30 69 99 Back-half % within DATECAT6 35.7% 30.4% 31.8%

Count 84 227 311 Total % within DATECAT6 100.0% 100.0% 100.0%

Chi-Square Tests

Value df Asymp. Sig. (2-sided)

Pearson Chi-Square 1.134(a) 2 .567

Likelihood Ratio 1.135 2 .567

Linear-by-Linear Association .207 1 .649

N of Valid Cases 311

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 23.50.

xxix Angle of 3rd impact by Date split Crosstab

DATECAT6 Total 1990 and after 1957 and before

Count 1 8 9 Front-half % within DATECAT6 11.1% 36.4% 29.0%

Count 5 10 15 ANGLE3CL Sides % within DATECAT6 55.6% 45.5% 48.4%

Count 3 4 7 Back-half % within DATECAT6 33.3% 18.2% 22.6%

Count 9 22 31 Total % within DATECAT6 100.0% 100.0% 100.0%

Chi-Square Tests

a 4 cells (66.7%) have expected count less than 5. The minimum expected count is 2.03.

Level of 1st Impact by Date Split Crosstab

DATECAT6 Total 1990 and after 1957 and before

Count 145 209 354 Legs % within DATECAT6 22.5% 29.9% 26.3%

Count 268 137 405 Waist % within DATECAT6 41.6% 19.6% 30.1% LEVEL1CL Count 147 328 475 Chest % within DATECAT6 22.8% 46.9% 35.3%

Count 85 25 110 Head/neck % within DATECAT6 13.2% 3.6% 8.2%

Count 645 699 1344 Total % within DATECAT6 100.0% 100.0% 100.0%

Chi-Square Tests

Value df Asymp. Sig. (2-sided)

Pearson Chi-Square 153.720(a) 3 .000

Likelihood Ratio 157.935 3 .000

Linear-by-Linear Association .236 1 .627

N of Valid Cases 1344

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 52.79.

xxx Level of 2nd Impact by Date Split Crosstab

DATECAT6 Total 1990 and after 1957 and before

Count 9 19 28 Legs % within DATECAT6 10.7% 8.1% 8.8%

Count 12 18 30 Waist % within DATECAT6 14.3% 7.6% 9.4% LEVEL2CL Count 42 189 231 Chest % within DATECAT6 50.0% 80.1% 72.2%

Count 21 10 31 Head/neck % within DATECAT6 25.0% 4.2% 9.7%

Count 84 236 320 Total % within DATECAT6 100.0% 100.0% 100.0%

Chi-Square Tests

Value df Asymp. Sig. (2-sided)

Pearson Chi-Square 38.767(a) 3 .000

Likelihood Ratio 34.836 3 .000

Linear-by-Linear Association .931 1 .335

N of Valid Cases 320

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 7.35.

Level of 3rd Impact by Date Split Crosstab

DATECAT6 Total 1990 and after 1957 and before

Count 5 5 Legs % within DATECAT6 25.0% 17.9%

Count 2 2 4 Waist % within DATECAT6 25.0% 10.0% 14.3% LEVEL3CL Count 6 12 18 Chest % within DATECAT6 75.0% 60.0% 64.3%

Count 1 1 Head/neck % within DATECAT6 5.0% 3.6%

Count 8 20 28 Total % within DATECAT6 100.0% 100.0% 100.0%

xxxi Chi-Square Tests

6 cells (75.0%) have expected count less than 5. The minimum expected count is .29.

Ball carrying style by 1st level of impact BALLCL * LEVEL1CL Crosstabulation

LEVEL1CL Total Head/neck Legs Waist Chest

Count 265 403 249 85 1002

Two Hands % within BALLCL 26.4% 40.2% 24.9% 8.5% 100.0%

% within LEVEL1CL 29.5% 41.2% 19.2% 31.6% 29.1%

Count 20 26 18 8 72

Swinging % within BALLCL 27.8% 36.1% 25.0% 11.1% 100.0%

% within LEVEL1CL 2.2% 2.7% 1.4% 3.0% 2.1% BALLCL Count 128 52 40 18 238

Waist % within BALLCL 53.8% 21.8% 16.8% 7.6% 100.0%

% within LEVEL1CL 14.2% 5.3% 3.1% 6.7% 6.9%

Count 486 496 991 158 2131

Chest % within BALLCL 22.8% 23.3% 46.5% 7.4% 100.0%

% within LEVEL1CL 54.1% 50.8% 76.3% 58.7% 61.9%

Count 899 977 1298 269 3443

Total % within BALLCL 26.1% 28.4% 37.7% 7.8% 100.0%

% within LEVEL1CL 100.0% 100.0% 100.0% 100.0% 100.0%

Chi-Square Tests

Value df Asymp. Sig. (2-sided)

Pearson Chi-Square 273.350(a) 9 .000

Likelihood Ratio 263.528 9 .000

Linear-by-Linear Association 43.027 1 .000

N of Valid Cases 3443

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 5.63.

xxxii Split by Rugby league and union games Number of tacklers by date split Crosstab

DATECAT6 Total 1990 and after CODE 1957 and before

Count 191 233 424 1 Tackler % within DATECAT6 87.6% 83.2% 85.1% INCCAT Count 27 47 74 Union Multiple Tacklers % within DATECAT6 12.4% 16.8% 14.9%

Count 218 280 498 Total % within DATECAT6 100.0% 100.0% 100.0%

Count 407 231 638 1 Tackler % within DATECAT6 85.9% 55.0% 71.4% INCCAT Count 67 189 256 League Multiple Tacklers % within DATECAT6 14.1% 45.0% 28.6%

Count 474 420 894 Total % within DATECAT6 100.0% 100.0% 100.0%

Chi-Square Tests

CODE Value df Exact Sig. (2-sided) Exact Sig. (1-sided)

Pearson Chi-Square 1.876(b) 1

Continuity Correction(a) 1.544 1

Likelihood Ratio 1.901 1 Union Fisher's Exact Test .204 .107

Linear-by-Linear Association 1.872 1

N of Valid Cases 498

Pearson Chi-Square 103.809(c) 1

Continuity Correction(a) 102.304 1

Likelihood Ratio 106.493 1 League Fisher's Exact Test .000 .000

Linear-by-Linear Association 103.693 1

N of Valid Cases 894

a Computed only for a 2x2 table

b 0 cells (.0%) have expected count less than 5. The minimum expected count is 32.39.

c 0 cells (.0%) have expected count less than 5. The minimum expected count is 120.27.

xxxiii Ball carrying style by Date split Crosstab

DATECAT6 Total 1990 and after CODE 1957 and before

Count 117 64 181 Two Hands % within DATECAT6 55.5% 22.9% 36.9%

Count 5 2 7 Swinging % within DATECAT6 2.4% .7% 1.4% BALLCL Count 20 5 25 Union Waist % within DATECAT6 9.5% 1.8% 5.1%

Count 69 208 277 Chest % within DATECAT6 32.7% 74.6% 56.5%

Count 211 279 490 Total % within DATECAT6 100.0% 100.0% 100.0%

Count 219 31 250 Two Hands % within DATECAT6 49.3% 7.4% 29.0%

Count 15 2 17 Swinging % within DATECAT6 3.4% .5% 2.0% BALLCL Count 55 10 65 League Waist % within DATECAT6 12.4% 2.4% 7.5%

Count 155 375 530 Chest % within DATECAT6 34.9% 89.7% 61.5%

Count 444 418 862 Total % within DATECAT6 100.0% 100.0% 100.0%

Chi-Square Tests

CODE Value df Asymp. Sig. (2-sided)

Pearson Chi-Square 273.256(b) 3 .000

Likelihood Ratio 298.071 3 .000 League Linear-by-Linear Association 240.021 1 .000

N of Valid Cases 862

Union -2 cells (25.0%) have expected count less than 5. The minimum expected count is 3.01.

League - 0 cells (.0%) have expected count less than 5. The minimum expected count is 8.24.

xxxiv Angle of 1st impact by Date split Crosstab

DATECAT6 Total 1990 and after CODE 1957 and before

Count 43 112 155 Front-half % within DATECAT6 21.1% 40.3% 32.2%

Count 37 26 63 ANGLE1CL Sides % within DATECAT6 18.1% 9.4% 13.1% Union Count 124 140 264 Back-half % within DATECAT6 60.8% 50.4% 54.8%

Count 204 278 482 Total % within DATECAT6 100.0% 100.0% 100.0%

Count 95 174 269 Front-half % within DATECAT6 21.6% 41.6% 31.4%

Count 91 59 150 ANGLE1CL Sides % within DATECAT6 20.7% 14.1% 17.5% League Count 253 185 438 Back-half % within DATECAT6 57.6% 44.3% 51.1%

Count 439 418 857 Total % within DATECAT6 100.0% 100.0% 100.0%

Chi-Square Tests

CODE Value df Asymp. Sig. (2-sided)

Pearson Chi-Square 22.782(a) 2 .000

Likelihood Ratio 23.316 2 .000 Union Linear-by-Linear Association 12.603 1 .000

N of Valid Cases 482

Pearson Chi-Square 40.094(b) 2 .000

Likelihood Ratio 40.511 2 .000 League Linear-by-Linear Association 30.277 1 .000

N of Valid Cases 857

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 26.66.

b 0 cells (.0%) have expected count less than 5. The minimum expected count is 73.16.

xxxv Angle of 2nd impact by Date split Crosstab

DATECAT6 Total 1990 and after CODE 1957 and before

Count 6 15 21 Front-half % within DATECAT6 24.0% 31.9% 29.2%

Count 7 10 17 ANGLE2CL Sides % within DATECAT6 28.0% 21.3% 23.6% Union Count 12 22 34 Back-half % within DATECAT6 48.0% 46.8% 47.2%

Count 25 47 72 Total % within DATECAT6 100.0% 100.0% 100.0%

Count 18 48 66 Front-half % within DATECAT6 30.5% 26.7% 27.6%

Count 23 85 108 ANGLE2CL Sides % within DATECAT6 39.0% 47.2% 45.2% League Count 18 47 65 Back-half % within DATECAT6 30.5% 26.1% 27.2%

Count 59 180 239 Total % within DATECAT6 100.0% 100.0% 100.0%

Chi-Square Tests

CODE Value df Asymp. Sig. (2-sided)

Pearson Chi-Square .668(a) 2 .716

Likelihood Ratio .671 2 .715 Union Linear-by-Linear Association .182 1 .669

N of Valid Cases 72

Pearson Chi-Square 1.221(b) 2 .543

Likelihood Ratio 1.230 2 .541 League Linear-by-Linear Association .002 1 .960

N of Valid Cases 239

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 5.90.

b 0 cells (.0%) have expected count less than 5. The minimum expected count is 16.05.

xxxvi Angle of 3rd impact by Date split Crosstab

DATECAT6 Total 1990 and after CODE 1957 and before

Count 3 3 Front-half % within DATECAT6 42.9% 30.0% ANGLE3CL Count 3 4 7 Union Sides % within DATECAT6 100.0% 57.1% 70.0%

Count 3 7 10 Total % within DATECAT6 100.0% 100.0% 100.0%

Count 1 5 6 Front-half % within DATECAT6 16.7% 33.3% 28.6%

Count 2 6 8 ANGLE3CL Sides % within DATECAT6 33.3% 40.0% 38.1% League Count 3 4 7 Back-half % within DATECAT6 50.0% 26.7% 33.3%

Count 6 15 21 Total % within DATECAT6 100.0% 100.0% 100.0%

Chi-Square Tests

Union - 4 cells (100.0%) have expected count less than 5. The minimum expected count is .90.

League - 4 cells (66.7%) have expected count less than 5. The minimum expected count is 1.71.

xxxvii Level of 1st Impact by Date Split Crosstab

DATECAT6 Total 1990 and after CODE 1957 and before

Count 36 71 107 Legs % within DATECAT6 17.4% 25.4% 22.0%

Count 106 63 169 Waist % within DATECAT6 51.2% 22.5% 34.7% LEVEL1CL Count 49 130 179 Union Chest % within DATECAT6 23.7% 46.4% 36.8%

Count 16 16 32 Head/neck % within DATECAT6 7.7% 5.7% 6.6%

Count 207 280 487 Total % within DATECAT6 100.0% 100.0% 100.0%

Count 109 138 247 Legs % within DATECAT6 24.9% 32.9% 28.8%

Count 162 74 236 Waist % within DATECAT6 37.0% 17.7% 27.5% LEVEL1CL Count 98 198 296 League Chest % within DATECAT6 22.4% 47.3% 34.5%

Count 69 9 78 Head/neck % within DATECAT6 15.8% 2.1% 9.1%

Count 438 419 857 Total % within DATECAT6 100.0% 100.0% 100.0%

Chi-Square Tests

CODE Value df Asymp. Sig. (2-sided)

Pearson Chi-Square 49.206(a) 3 .000

Likelihood Ratio 49.760 3 .000 Union Linear-by-Linear Association 1.781 1 .182

N of Valid Cases 487

Pearson Chi-Square 115.792(b) 3 .000

Likelihood Ratio 123.410 3 .000 League Linear-by-Linear Association 2.450 1 .117

N of Valid Cases 857

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 13.60.

b 0 cells (.0%) have expected count less than 5. The minimum expected count is 38.14.

xxxviii Level of 2nd Impact by Date Split Crosstab

DATECAT6 Total 1990 and after CODE 1957 and before

Count 2 5 7 Legs % within DATECAT6 8.0% 10.6% 9.7%

Count 1 5 6 Waist % within DATECAT6 4.0% 10.6% 8.3% LEVEL2CL Count 14 33 47 Union Chest % within DATECAT6 56.0% 70.2% 65.3%

Count 8 4 12 Head/neck % within DATECAT6 32.0% 8.5% 16.7%

Count 25 47 72 Total % within DATECAT6 100.0% 100.0% 100.0%

Count 7 14 21 Legs % within DATECAT6 11.9% 7.4% 8.5%

Count 11 13 24 Waist % within DATECAT6 18.6% 6.9% 9.7% LEVEL2CL Count 28 156 184 League Chest % within DATECAT6 47.5% 82.5% 74.2%

Count 13 6 19 Head/neck % within DATECAT6 22.0% 3.2% 7.7%

Count 59 189 248 Total % within DATECAT6 100.0% 100.0% 100.0%

Chi-Square Tests

Union - 5 cells (62.5%) have expected count less than 5. The minimum expected count is 2.08.

League - 2 cells (25.0%) have expected count less than 5. The minimum expected count is 4.52.

xxxix Level of 3rd Impact by Date Split Crosstab

DATECAT6 Total 1990 and after CODE 1957 and before

Count 2 2 4 Waist % within DATECAT6 66.7% 40.0% 50.0% LEVEL3CL Count 1 3 4 Union Chest % within DATECAT6 33.3% 60.0% 50.0%

Count 3 5 8 Total % within DATECAT6 100.0% 100.0% 100.0%

Count 5 5 Legs % within DATECAT6 33.3% 25.0%

Count 5 9 14 LEVEL3CL Chest % within DATECAT6 100.0% 60.0% 70.0% League Count 1 1 Head/neck % within DATECAT6 6.7% 5.0%

Count 5 15 20 Total % within DATECAT6 100.0% 100.0% 100.0%

Chi-Square Tests

Union 4 cells (100.0%) have expected count less than 5. The minimum expected count is 1.50.

League - 5 cells (83.3%) have expected count less than 5. The minimum expected count is .25.

Ball carrying style by Level of 1st impact BALLCL * LEVEL1CL Crosstabulation

LEVEL1CL Total Head/neck CODE Legs Waist Chest

Union BALLCL Count 108 190 108 23 429

Two Hands % within BALLCL 25.2% 44.3% 25.2% 5.4% 100.0%

% within LEVEL1CL 43.9% 52.6% 30.5% 31.1% 41.4%

Count 6 7 7 2 22

Swinging % within BALLCL 27.3% 31.8% 31.8% 9.1% 100.0%

% within LEVEL1CL 2.4% 1.9% 2.0% 2.7% 2.1%

Count 22 26 13 6 67

Waist % within BALLCL 32.8% 38.8% 19.4% 9.0% 100.0%

% within LEVEL1CL 8.9% 7.2% 3.7% 8.1% 6.5%

Chest Count 110 138 226 43 517

xl % within BALLCL 21.3% 26.7% 43.7% 8.3% 100.0%

% within LEVEL1CL 44.7% 38.2% 63.8% 58.1% 50.0%

Count 246 361 354 74 1035

Total % within BALLCL 23.8% 34.9% 34.2% 7.1% 100.0%

% within LEVEL1CL 100.0% 100.0% 100.0% 100.0% 100.0%

Count 157 213 141 62 573

Two Hands % within BALLCL 27.4% 37.2% 24.6% 10.8% 100.0%

% within LEVEL1CL 24.0% 34.6% 14.9% 31.8% 23.8%

Count 14 19 11 6 50

Swinging % within BALLCL 28.0% 38.0% 22.0% 12.0% 100.0%

% within LEVEL1CL 2.1% 3.1% 1.2% 3.1% 2.1% BALLCL Count 106 26 27 12 171

League Waist % within BALLCL 62.0% 15.2% 15.8% 7.0% 100.0%

% within LEVEL1CL 16.2% 4.2% 2.9% 6.2% 7.1%

Count 376 358 765 115 1614

Chest % within BALLCL 23.3% 22.2% 47.4% 7.1% 100.0%

% within LEVEL1CL 57.6% 58.1% 81.0% 59.0% 67.0%

Count 653 616 944 195 2408

Total % within BALLCL 27.1% 25.6% 39.2% 8.1% 100.0%

% within LEVEL1CL 100.0% 100.0% 100.0% 100.0% 100.0%

Chi-Square Tests

CODE Value df Asymp. Sig. (2-sided)

Pearson Chi-Square 56.869(a) 9 .000

Likelihood Ratio 57.498 9 .000 Union Linear-by-Linear Association 22.144 1 .000

N of Valid Cases 1035

Pearson Chi-Square 227.791(b) 9 .000

Likelihood Ratio 215.315 9 .000 League Linear-by-Linear Association 21.581 1 .000

N of Valid Cases 2408

a 2 cells (12.5%) have expected count less than 5. The minimum expected count is 1.57.

b 1 cells (6.3%) have expected count less than 5. The minimum expected count is 4.05.

xli