Niranjan Bidargaddi1, Geoffrey Schrader2, Graeme Tucker3, Rohan Dhillon4, Anand Ganesan5. 1. College of Medicine & Public Health, Flinders University, Adelaide, SA, Australia; Country Health South Australia, Adelaide, SA, Australia. Electronic address: niranjan.bidargaddi@flinders.edu.au. 2. College of Medicine & Public Health, Flinders University, Adelaide, SA, Australia; Country Health South Australia, Adelaide, SA, Australia. 3. Health Statistics Unit, Epidemiology Branch, SA Health, School of Medicine-University of Adelaide, SA, Australia. 4. Central Adelaide Local Area Health Network, Adelaide, SA, Australia. 5. College of Medicine & Public Health, Flinders University, Adelaide, SA, Australia.
Abstract
BACKGROUND: Increasingly, big data derived from administrative hospital records can be subject to analytics to provide clinical insights. The aim of this study was to determine the impact of psychiatric comorbidity on length of hospital stay and number of hospital admissions in cardiac patients utilising routinely collected hospitalisation records. METHODS: We routinely collected clinical and socio-demographic variables extracted from 37,580 cardiac patients, between 18 and 65 years old, admitted to South Australian hospitals between 2001/02 to 2010/11 financial years with cardiac diagnoses used to derive patient level and separation level variables used in the modelling. Multi-level models were constructed to analyse the impact of psychiatric comorbidity on both length of stay and the total number of hospitalisations, allowing for interactions between socioeconomic status and the burden of disease. Possible confounders for these models were, sex, age, indigenous status, country of birth, and rural status. RESULTS: For cardiac patients a mental health diagnosis was associated with an increase of 12.5% in the length of stay, and an increase in the number of stays by 20.0%. CONCLUSIONS: This study demonstrates the potential utility of routinely collected hospitalisation records to demonstrate the impact of psychiatric comorbidity on health service utilisation.
BACKGROUND: Increasingly, big data derived from administrative hospital records can be subject to analytics to provide clinical insights. The aim of this study was to determine the impact of psychiatric comorbidity on length of hospital stay and number of hospital admissions in cardiac patients utilising routinely collected hospitalisation records. METHODS: We routinely collected clinical and socio-demographic variables extracted from 37,580 cardiac patients, between 18 and 65 years old, admitted to South Australian hospitals between 2001/02 to 2010/11 financial years with cardiac diagnoses used to derive patient level and separation level variables used in the modelling. Multi-level models were constructed to analyse the impact of psychiatric comorbidity on both length of stay and the total number of hospitalisations, allowing for interactions between socioeconomic status and the burden of disease. Possible confounders for these models were, sex, age, indigenous status, country of birth, and rural status. RESULTS: For cardiac patients a mental health diagnosis was associated with an increase of 12.5% in the length of stay, and an increase in the number of stays by 20.0%. CONCLUSIONS: This study demonstrates the potential utility of routinely collected hospitalisation records to demonstrate the impact of psychiatric comorbidity on health service utilisation.