|
Abstract
|
BACKGROUND: The ability to predict the risk of poor outcome following knee cruciate ligament reconstruction in the Australian Defence Force (ADF) population would help direct individual rehabilitation programmes and workforce planning. This study describes the epidemiology of cruciate ligament reconstruction in the ADF and demonstrates the use of Bayesian networks (BN) to predict the likelihood of fitness for ongoing service under different scenarios.
METHODS: Members who had knee cruciate ligament reconstruction through ADF were identified from billing data and matched to electronic medical records to extract demographic and clinical data. Outcome measure was medical fitness for ADF service up to 24 months after reconstruction. BN models were used to compare outcomes between (1) age groups according to military service, and (2) sexes according to body mass index (BMI).
RESULTS: From November 2012 to June 2019, a total of 1199 individuals had knee cruciate ligament reconstruction (average 169 reconstructions/year). Following reconstruction and rehabilitation, 89 (7.4%) were medically unfit for service. Scenario analysis using a tree-augmented naïve BN model showed that, compared with Navy and Air Force, Army members had a higher probability of being unfit in those aged <35 years and a lower probability in those aged ≥35 years. In both sexes, those with obese BMI had the greatest probability of being unfit.
CONCLUSION: While most ADF members were fit for ongoing military service following cruciate ligament reconstruction, service type, age, sex and BMI influenced outcome. BNs provided an interactive and intuitive method to demonstrate the impact of different variables on the outcome.
Language: en |