Is sports injury a large enough population problem?

Government agencies will only act if there is strong evidence of a big (population-level) problem that needs to be addressed. Prevention of injuries, including via policy action, cannot occur without good quality, accurate, reliable and timely data about the size of the problem.14 15 In general, government health departments look to routine population-wide mortality or morbidity (generally from hospitals) data collections, to answer these questions.15 Unfortunately, the different data sources they use rank the importance of sports injuries differently. Data regarding deaths do not feature sport prominently and hospital admissions data rank sports injuries very much lower than other injury causes. On the other hand, sports injuries are a major source of emergency department presentations, and are the leading cause of child injury emergency department visits.16 Because sport does not rank highly as a cause of injury mortality/major morbidity this limits its ranking as a priority by government health departments, as well as international bodies such as the WHO and the Global Burden of Disease and Injury estimates.17

Routine health sector injury surveillance data are usually coded to the International Classification of Diseases (ICD-9 or 10) and injury cases are identified through a principal diagnosis of injury and/or an external cause code. The ICD-10-AM (Australian modification), used in several jurisdictions, includes >200 activity codes for identifying the specific types of sport/leisure activity of the injured person at the time the injury event occurred,18 as well as place codes to broadly identify the place of occurrence of event at the time the person was injured (including sports/athletics area).19 Of course, these codes are only useful if they are applied with high quality and completeness. There is evidence that there are a relatively large number of instances in which activity codes are not provided for injury hospitalisations and so it has been estimated that ICD-10-AM-activity coded data may still underestimate the true incidence of cases by 6.0–22.9%.18

Notwithstanding this sub-optimal case ascertainment, such data can still be used to estimate trends in population rates of hospitalised sports injury, especially if an assumption is made of non-differential case-ascertainment over the period of interest. In New South Wales (NSW), Australia, an analysis of routinely-collected hospital admissions data for the period 1 July 2003–30 June 2008 estimated the overall sports injury hospitalisation rate to be 195.5/100 000 residents and for there to a non-significant annual decline over the 5-year period.20 Unfortunately, however, injury hospitalisations are only the tip of the iceberg because the vast majority of sports injuries do not require attention by medical professionals, let alone in a hospital setting.21 This was assessed directly in NSW in the midpoint of the interval used for the above trend analysis of hospitalisations. People who reported sustaining a sports injury were asked to indicate where they received treatment for this injury, and specifically targeted questions were analysed in the annual NSW Population Health Survey.22 Only 2.8% of the sports-injured respondents were admitted to hospital for that injury. Added to this, 6.1% stated they were treated in a hospital emergency department, totalling only 8.9% of population-level sports injuries treated in hospital settings.

These findings have significant implications for the ability to answer the first major question in table 1. For most sports injury cases that are not treated at a hospital, even when they are recorded in routine databases, case ascertainment is incomplete. This is a significant problem because government health departments generally only rely on routinely-collected data from hospitals. Another implication is that despite several years of sports agency attention in NSW, there has been no significant decline in sports injury hospitalisation rates.20 The reasons for this are unclear but are likely to be related to the implementation and dissemination of injury prevention measures, or the lack thereof. Perhaps there has been wide scale non-adoption/non-translation of evidence-based approaches. Perhaps the current interventions (most of which have been derived from more clinically-orientated approaches) are not suited for wide-scale delivery and adoption; implementation approaches could be too diffuse or diluted; or the right population groups not targeted. Without concurrent evidence about barriers against, and facilitators for, intervention delivery and uptake it is not possible to identify whether or not these factors have made a difference.23 There may also have been changes in participation rates over the same period of time, but without good population level data on this that cannot be determined either.

Which parts of the community are most vulnerable?

Most government health departments recognise population-health differentials and that there can be significant social/cultural/economic determinants of ill health and increased injury rates that are exacerbated by factors such as where and how people live. Not much is known about this in relation to sports injury but one Australian study showed that sports injury rates were highest in the second and third highest quintiles of social disadvantage and lowest in those with greatest social disadvantage, perhaps reflecting a reduced rate of participation in sport in that group. In contrast, age-standardised hospitalisation rates were significantly higher in remote and very remote communities, though the full reasons for this are unknown.

Epidemiological studies consistently show that the population groups with high injury frequency and rates are males (of all ages) and children/younger people aged 10–40 years, with a peak injury hospitalisation incidence around 15–25 years.20 24 In Australia, population-level sport injury hospitalisations are most commonly associated with the football codes, other team ball sports (eg, netball and basketball), forms of active transportation for sport/recreation (eg, cycling, skating/blading), equestrian sports and ice/snow sports.24 25 These figures are consistent with what is known about population-wide sport participation levels.

There are several implications of these findings including the fact that sports injury is a young–middle-aged person's issue, a population group with relatively low overall use of hospital service delivery. Team ball sports and transport-orientated activities predominate as settings for the injury events. However, government health departments have no jurisdiction over these activities, as they play no role in the delivery of sport.

Although there are likely to be population disparities in sports injury rates, at least when measured by hospitalisations and by location of residence, the reasons for this are unknown. For example, is the solution to do with sport/leisure infrastructure provision, participation in sport/leisure pursuits, availability and provision of medical and allied health services or even opportunities for formal sport versus leisure? In any case, government health departments do not traditionally play a major role in implementing/supporting most of these solutions and this means that it is very hard for them to identify a role for themselves alone.

Why should the government address the issue?

Even when there are data to illustrate that sports injury is a significant public health issue, and is associated with socio-economic disparity, there is still a need to demonstrate why government health departments specifically need to address it. Why should it be their responsibility rather than that of individuals, communities, industry or even other government sectors? In the main, this come from demonstrating that there would be significant costs to government health departments for not acting and significant benefits if they did.

Table 2 shows previous ratings of the broad level of evidence for SIP, as presented in Australian Commonwealth Health Department reports.26 27 While developed in the Australian context, these ratings draw on the best international evidence and well represent the international status. The last column in the table gives an up-to-date assessment of the knowledge base and highlights continuing significant knowledge gaps. Most notably, these relate to incomplete knowledge about the size of the public health sports injury burden, very little evidence about the existence of effective prevention solutions, a total lack of implementation activities across the population and no information at all about cost-effectiveness or cost–benefit scenarios associated with investments in the interventions at the population-level. The clear conclusions from these knowledge gaps is that there is no information available to government health departments to justify why they should invest in public health approaches to SIP.

Summary of the current status

In Australia, as elsewhere, while the past decade has seen a large increase in knowledge about injury risk factors and potential solutions to reducing these risks, most interventions remain largely untrialled; the effectiveness of implemented solutions is not known and there is no information about the cost-effectiveness of potential interventions or the likely benefits of their introduction. Table 3 summarises the status of answers to the policy questions posed in table 1. As can be seen, there are very clear gaps in the knowledge base and so it is not surprising that that there is currently no co-ordinated governmental response to the sports injury problem in Australia. This paper has illustrated the key arguments with data from Australia but the implications and conclusions have broad applicability globally.