Using data to answer questions of public health importance for ACT Health, with an emphasis on routinely-collected linked data

Abstract

My field placement was with the Epidemiology Section in the Population Health Protection and Prevention Division at ACT Health. Within this placement, I have completed four projects for this thesis: an analysis of Emergency Department (ED) data; a gastroenteritis outbreak investigation; an evaluation of a population health survey and, for my main project, a study of unplanned hospital readmissions. One of the motivations for undertaking these projects was to promote better use of the routinely-collected linked data to answer questions of public health importance for ACT Health. My data analysis project was an analysis of frequent ED use in the Australian Capital Territory (ACT). This is the first study to quantify and characterise ED frequent users in the ACT. The results support existing evidence that frequent users tend to be older, female, and/or single, and commonly present with pain-related conditions. The data also showed that compared to non-frequent ED users, frequent users were more likely to be referred by police, corrective or community services; arrive by ambulance, not wait to be assessed, or leave at their own risk. In addition, we investigated visit intervals, rarely reported on in other studies. This study found around one third of frequent users returned within 7 days, with 41% of their visits having the same diagnosis as the last visit. Early identification and follow-up in the community for frequent users will assist in the development of targeted strategies to improve health service delivery to this vulnerable group. Unexpected return to hospital has negative impacts on families and healthcare systems. We examined which conditions have the highest rates of readmission and contribute most to 30-day unplanned readmissions in the ACT, and which patient characteristics are associated with readmissions. The study identified a 30-day unplanned readmission rate of 6.2%, with admission rates highest for alcohol-related liver disease (19.2%), and heart valve disorders (17.4%). Older age and comorbidities are strong predictors for 30-day unplanned readmissions. For some conditions the rates were relatively high, suggesting areas to target for reducing readmissions. Therefore, when developing preventative strategies and post-discharge plans, particular consideration should be given to patients at older age or with underlying comorbidities. As part of the ACT Health Survey Program (HSP), the ACT General Health Survey (GHS) is a computer-assisted telephone interviewing survey conducted every year among ACT residents. My evaluation of the GHS found that it is a useful tool to monitor trends of overweight, obesity, nutrition and physical activity for adults and children in the ACT. The data collected are used to provide evidence to understand and analyse overweight and obesity patterns in the ACT and create awareness of unhealthy lifestyles. However, improvements could be made in a few areas, including: developing a proper evaluation plan and a data quality statement, increasing the sample size and the proportion of young people in the sample population. I also carried out an outbreak investigation of foodborne gastroenteritis that occurred among staff and public members at a large national institution in Canberra. I conducted two studies for this outbreak – a retrospective cohort study and a case control study. The epidemiological, environmental and laboratory evidence suggested the outbreak was caused by C. perfringens toxin Type A, with the likely vehicles of transmission being butter chicken and rice. The findings of this investigation suggest that a breakdown in temperature control and good food handling practices may have resulted in C. perfringens bacterium growing rapidly and producing a toxin which caused the illness. This project also indicated that the value of a second epidemiological study was questionable given the limited time and resources available. Show more