K-level 1 to 4

participants received the standardised outpatient prosthetic rehabilitation service, as detailed in Appendix 1 (see eAddenda). An independent research assistant contacted potential participants from the Amputee Physiotherapy Service database to obtain informed verbal consent for the interview. The interview process involved coordinating telephone interviews with country physiotherapists on remote community visits, Aboriginal Health workers, nurses, and the use of telehealth. Medical records were audited for potential predictor variables and this was undertaken blind to the interviews. Box 1 outlines the predictor variable domains investigated. All potential variables were dichotomised (eg, amputation cause: atraumatic or traumatic). Receiver Operator Characteristic (ROC) curves were used to generate a threshold for dichotomous classification click here of continuous variables (eg, age). This was performed with an equal weighting for sensitivity and specificity. Table 1 in the eAddenda details the dichotomous variable classifications. Intrinsic predictor variables Amputation predictor variables Functional predictor variables • gender • age • indigenous status • metropolitan versus country • accommodation at discharge: home versus residential care • medical comorbidities: diabetes type I or II, peripheral arterial disease,

cardiac condition, renal failure, stroke, transient ischaemic attack, lower limb pathology • number of medical comorbidities, including mental health issues and musculoskeletal pathology • amputation R428 solubility dmso cause • amputation level • bilateral lower limb amputation • time to second lower limb amputation • time from amputation to prosthetic milestones: casting, fitting and definitive

prosthesis • mobility level achieved without a prosthesis: wheelchair mobility, transfers, hopping • independence with donning and doffing prosthesis, and monitoring prosthetic fit at discharge • mobility for aid use at discharge • mobility level achieved using a prosthesis at discharge: walking indoors, outdoors, stairs, slopes, grass, gravel, uneven terrain, high-level balance activities and running Full-size table Table options View in workspace Download as CSV Medical comorbidities (including mental health issues and musculoskeletal pathology) were recorded and counted for each participant. Charlson Comorbidity Index and Combined Age Charlson Comorbidity Index were calculated from medical comorbidities data.31 In the present study, amputation level was classified as transtibial or above transtibial. Bilateral lower limb amputation was defined as having undergone two major lower limb amputations. Participants were classified as able to independently perform the locomotor skill or being dependent (ie, required assistance or unable to perform). Mobility aids were either used or not used, and the aid type was not statistically weighted for its level of support.