Hospital separations data

• Length of stay

In 1996-97, there were 644,407 hospital separations of people in Australia and 0.5 per cent (n=3,539) were estimated to be due to horse-related injury. Hospital separations can include multiple admissions for some patients, particularly when they are transferred to other acute care hospitals or from one ward to another. To offset the effect of multiple admissions and to have a dataset of acute admissions for horse-related hospitalisations, a selection method for estimating new incident cases was used and is described in the section "Data Issues". Using this selection method, estimated incident cases of horse-related injury hospitalisations during 1996-97 were 3,124 cases.

Estimated incident cases and their horse-related external causes are presented in Table 3. Most cases were riders of horses involved in a riding accident. These cases were admitted at a rate of 16.2 estimated incident cases of horse-related injury per 100,000 population and accounted for 96 percent of the cases (n=2,989). Horse-rider injuries from motor vehicle accidents and non-motor road-vehicle accidents were few (135 cases), with injuries due to animal-drawn vehicles and their occupants predominating in this group (53%, n=72 cases).

In Figure 4, age-specific rates of estimated incident cases of horse-related injury are presented for both sexes. It is evident that horse-related injuries are common in children and young adults of both sexes. Estimated incident case numbers are 1.4 times higher for females than for males (n=1,809 vs 1,315). From 5-34 years of age, female rates were significantly higher than male rates, with a maximum rate of 73.5 estimated incident cases of horse-related injury per 100,000 population occurring at 10-14 years of age. At ages greater than 34 years, female rates declined and were significantly lower statistically than male rates at 55 to 74 years of age. This pattern may reflect the shift in the horse riding population from young amateur female riders to older male professionals.

The predominance of young amateur female horse riders with their high exposure to the hazards associated with horse riding may account for the high injury rates in this group rather than the risk factors of female gender and youth (Buckley, Chalmers et al. 1993).

Figure 4: Age-specific rates of estimated incident cases of horse-related injury by sex, Australia 1996-97.

Note: Error bars indicate 95% confidence intervals for rates.

Crude rates of estimated incident cases of horse-related injury by state and territory of residence are presented in Figure 5. It is evident from the 95% confidence intervals, based on the Poisson distribution, that Queensland, with a rate of 29.4 estimated incident cases per 100,000 population, is significantly higher statistically than the national rate of 16.8 estimated incident cases per 100,000 population and higher than the rate in all states and the Australian Capital Territory. Rates in Victoria, South Australia, and the Australian Capital Territory (11.3, 11.4, and 5.8 estimated incident cases per 100,000 population, respectively) were significantly lower statistically than the national rate and the rate in New South Wales, Queensland, and Western Australia. The pattern of horse-related injury (highest rates in Queensland and lower rates in Victoria and South Australia) is similar to the death rate for these states (see Figure 3).

Figure 5: Crude rate of estimated incident cases of horse-related injury by state and territory of residence, Australia 1996-97.

Note: Error bars indicate 95% confidence intervals for rates. State of residence was not reported for 36 cases.

Estimated incident cases of horse-related injury by state of residence and regional distribution are presented in Table 4. Incident cases of horse-related injuries are distributed widely in Australia. Over one-third of the cases occurred in capital cities and suburbs, with the highest number in New South Wales (13.3%, n=403 cases). Queensland and New South Wales had 67 per cent (n=1023) of the rural cases, possibly as a result of riding and handling of horses recreationally and on farms and stations. Many equestrian sporting events are also held in rural areas. Queensland, with the highest rate of estimated incident cases of horse-related injury, had 70 per cent of its cases from remote regions of the state. Given the severe nature of many horse-related injuries, retrieval of injured riders would be of a particular concern in these remote regions.

In the following discussion of horse-related injuries, incident cases will be restricted to external cause category E828.2 (`ridden animal accident, animal rider injured'). This category represents 96 per cent (n=2989) of estimated incident cases of horse-related injury, is more uniform and representative of the horse rider population, and will be referred to as `horse rider' incident cases.

During 1996-97, the four most common injuries, representing 78 per cent of all the injuries reported in horse rider incident cases as principal diagnosis, are summarised by major body region in Table 5 (also see Appendix 1, Table A1.1). Fifty-nine per cent of the injuries reported were in females, with injury to the upper body region (head and upper limbs) occurring most often in both sexes. This pattern of injury is consistent with injuries from falling or being thrown by a horse (Danielsson and Westlin 1973), (Grossman, Kulund et al. 1978), and (Masters 1991). Overall, the most common injury reported was musculoskeletal, usually either a fracture, dislocation or sprain (55%, n=1634).

The most serious injury reported in both male and female cases was neurological, primarily intracranial injury which occurred in 20.4 per cent (n=609) of the incident cases. Loss of consciousness or concussion was not reported as principal diagnosis in this dataset; these clinical symptoms were reported, however, as additional diagnoses in 2.2 per cent of the cases (n=69). Brief loss of consciousness was reported in 57 per cent of these cases (n=39) and concussion in the remaining cases.

Rates of injury for the four most common injuries in each body region indicate that injury to the head and upper body was high, with the rate of intracranial injury and the rate of fracture of radius and ulna highest in females. Overall, these rates in both sexes were higher than rates for any other injury.

Based on the ABS survey of participation in sport and physical activities in 1997, males participated in horse-riding 1.8 times that of women (Australian Bureau of Statistics 1999), but the pattern of injury remained the same for both sexes, with estimated annual rates of hospitalised intracranial injury of 297.4 per 100,000 male participants and 212.7 per 100,000 female participants.

Rates of estimated incident cases of horse rider injury for three of the most common injuries in upper and lower body regions diagnosed in injured male and female horse riders are presented by age and sex in Figure 6. It is evident that male and female horse riders aged 10 to 14 years are at a high risk of injury from riding horses at this age, with rates (per 100,000 population) of upper and lower limb fractures and intracranial injuries in females significantly higher statistically in this age group than rates (per 100,000 population) in males of the same age (fracture radius ulna: 26.0±4.0 CI^* vs 8.3±2.2 CI; fracture of ankle: 3.4±1.4 CI vs 0.7±0.7 CI; and intracranial injury: 16.1±3.1 CI vs 4.8±1.7 CI, respectively). This pattern of injury may be consistent with falling or being thrown from a horse, which is common in young amateur riders (Masters 1991). With increasing age, there was a decrease in rates for all injuries in females, with female intracranial rates remaining significantly higher statistically than male rates at ages 15 to 34 years of age. At ages 35 to 64 years, male rates were generally higher than female rates of the same age, possibly due to more older male professional riders, but the rates were not significantly different statistically.

Figure 6: Principal diagnoses of estimated incident cases of horse rider injury by age and sex, Australia 1996-97, crude rates.

Length of stay

The calculation of length of stay is based on the time that injured horse riders (n=2,989) spent in hospital during their treatment and any complications associated with their injuries. This period of treatment is based on time of admission to time of discharge to previous home, or when death occurred on ward. Length of stay for treatment of horse-related injuries is calculated by dividing the number of bed-days occupied by all horse rider injury cases (irrespective of Mode of Separation) by the number of estimated incident cases. Based on this method, the average length of stay (ALOS) for horse rider injuries was 3 days.

The twelve most common horse rider injuries are listed with estimated incident cases, total bed days and average length of stay, in Table 6 and a tabulation of all injuries is presented in Appendix 1, Table A1.2. Intracranial injury, the most common injury sustained by riders, accounted for a total of 1,344 bed days, with an average length of stay of 2 days. Although intracranial injury is generally regarded as serious, for 94 per cent of the cases the average length of stay was two days (probably for observation). Only six per cent of the cases reported having an ALOS greater than three days, reflecting a more serious head injury. The maximum ALOS reported for intracranial injury was 78 days, with six cases requiring more than three weeks of treatment. Pelvic, vertebral column, and skull fractures had ALOS of more than a week. The highest ALOS (30 days) was for treatment of fracture of vertebral column with spinal cord injury (Table A2.2).