The causes and prognoses of different types of fractures in wild koalas submitted to wildlife hospitals.

Fractures are a major problem in wild koalas of great veterinary and conservation importance as their occurrence in different locations of the body might result in varying healing success. The aim of this study was to determine the fracture types (defined by location of the fracture) occurring in wild koalas, temporal patterns, possible causes and risk factors of fracture types, and the prognosis for successfully releasing kolas with healed fracture types into the wild. Data from a total of 2031 wild koalas submitted to wildlife hospitals in South-East Queensland, Australia, over a period of 13 years were analysed. Approximately 56.7% of koalas experienced head fractures, 13.4% had torso fractures, 14.9% had limb fractures and 15% had combination fractures. A total of 84.1% of fractures were caused by vehicle collisions, 9.1% by dog attacks, 3.3% by falls from trees, 1.3% by train collisions, 0.2% by livestock trampling and 1.8% due to unknown causes. Multinominal logistic regression was used to identify risk factors (cause of fracture, age category, sex, year, three-year admission period and season of fracture event) by fracture type. The type of fracture was associated with both the cause of the fracture and the season when it occurred: for example torso fractures (compared to combination fractures) were associated with dog attacks (OR=10.98; 95% CI6.03, 20.01) and falls from trees (OR=4.79; 95% CI2.26, 10.19) relative to vehicle collisions. More submissions of koalas with head fractures due to vehicle collisions occurred in spring compared to autumn and winter, coinciding with the breeding season of koalas and increased animal movement. Prognosis for koalas with fractures was poor, with approximately 63.8% of koalas admitted dead on arrival, 34.2% euthanised, and only 2.0% of koalas able to be released. Given this data, further research into mitigation strategies to decrease the risk of fractures and to increase the observed low recovery rate should be considered.