TY - JOUR
T1 - Emergency department mental health presentations by people born in refugee source countries: an epidemiological logistic regression study in a Medicare Local region in Australia
AU - Enticott, Joanne
AU - Cheng, I-Hao
AU - Russell, Grant Mervyn
AU - Szwarc, Josef
AU - Braitberg, George
AU - Peek, Anne
AU - Meadows, Graham Nicholas
PY - 2015
Y1 - 2015
N2 - This study investigated if people born in refugee source countries are disproportionately represented among those receiving a diagnosis of mental illness within emergency departments (EDs). The setting was the Cities of Greater Dandenong and Casey, the resettlement region for one-twelfth of Australia?s refugees. An epidemiological, secondary data analysis compared mental illness diagnoses received in EDs by refugee and non-refugee populations. Data was the Victorian Emergency Minimum Dataset in the 2008?09 financial year. Univariate and multivariate logistic regression created predictive models for mental illness using five variables: age, sex, refugee background, interpreter use and preferred language. Collinearity, model fit and model stability were examined. Multivariate analysis showed age and sex to be the only significant risk factors for mental illness diagnosis in EDs. `Refugee status?, `interpreter use? and `preferred language? were not associated with a mental health diagnosis following risk adjustment for the effects of age and sex. The disappearance of the univariate association after adjustment for age and sex is a salutary lesson for Medicare Locals and other health planners regarding the importance of adjusting analyses of health service data for demographic characteristics.
AB - This study investigated if people born in refugee source countries are disproportionately represented among those receiving a diagnosis of mental illness within emergency departments (EDs). The setting was the Cities of Greater Dandenong and Casey, the resettlement region for one-twelfth of Australia?s refugees. An epidemiological, secondary data analysis compared mental illness diagnoses received in EDs by refugee and non-refugee populations. Data was the Victorian Emergency Minimum Dataset in the 2008?09 financial year. Univariate and multivariate logistic regression created predictive models for mental illness using five variables: age, sex, refugee background, interpreter use and preferred language. Collinearity, model fit and model stability were examined. Multivariate analysis showed age and sex to be the only significant risk factors for mental illness diagnosis in EDs. `Refugee status?, `interpreter use? and `preferred language? were not associated with a mental health diagnosis following risk adjustment for the effects of age and sex. The disappearance of the univariate association after adjustment for age and sex is a salutary lesson for Medicare Locals and other health planners regarding the importance of adjusting analyses of health service data for demographic characteristics.
UR - http://www.publish.csiro.au/?act=view_file&file_id=PY13153.pdf
U2 - 10.1071/PY13153
DO - 10.1071/PY13153
M3 - Article
SN - 1448-7527
VL - 21
SP - 286
EP - 292
JO - Australian Journal of Primary Health
JF - Australian Journal of Primary Health
IS - 3
ER -