This project will assemble a retrospective observational cohort of people presenting to emergency department and hospital inpatient services with an acute alcohol harm and/or problematic alcohol use. The project seeks to examine the epidemiology of alcohol-related harms through the information of hospitalisation, emergency department presentations, cancer notifications, mental health ambulatory care, mortality, offending, and incarceration.
Dr Sallie-Anne Pearson, Professor,
Centre for Big Data Research in Health, UNSW Australia
Dr Adrian Dunlop, Area Director and Senior Staff Specialist,
Drug and Alcohol Clinical Services, Hunter New England Health Local Health District
Reducing the health, social, and economic burden of alcohol use is a priority in Australia and globally. Alcohol consumption is estimated to play a causal role in over 200 disease and injury conditions. Approximately 3.9% of deaths and 1.8% of hospitalisations in Australia are alcohol-related. Alcohol use negatively impacts on the community through reduced workplace productivity, traffic accidents, family problems, crime, and public disorder, with an estimated economic cost of $14 billion annually.
Recent evidence suggests increasing rates of alcohol-related problems in Australia over the past two decades despite evidence of declining population levels of consumption (particularly among young people). This increase in harms represents a significant burden on healthcare and law enforcement services. Indeed, recent estimates suggest that approximately one in ten emergency department presentations in Australia are alcohol-related, with more than 144,000 alcohol-attributable hospitalisations in Australia in 2012/13.
The aforementioned data are based on modelled estimates or on aggregated number of presentations to services. It is important to understand these events at the individual level: a significant proportion of people who experience alcohol-related problems will have recurrent problems, experience substantial morbidity and higher risk of mortality, and place significant burden on healthcare and law enforcement services. In Australia, there has been no attempt at the population level to longitudinally track people with alcohol-related problems to measure overall mortality, morbidity and other problems (e.g., offending and incarceration), despite such work for other substances (e.g., opioids).
The overall objective of this program of research is to use linked health and law enforcement data to establish and describe individuals presenting to emergency and inpatient health care services with an acute alcohol harm or problematic alcohol use; measure their health service utilisation and law enforcement engagement; and quantify morbidity, mortality, offending and incarceration among this cohort.
Specific aims include:
- Describe the cohort at their first point of contact with emergency department or inpatient hospital services within the study period for an acute alcohol harm and/or problematic alcohol use;
- Quantify healthcare service utilisation and law enforcement engagement among the cohort and assess individual and situational characteristics as predictors of frequency of engagement;
- Quantify the rate of mortality, morbidity, offending and incarceration amongst the cohort, looking at overall rates and cause-specific outcomes where possible; and
- Assess individual and situational characteristics as predictors of mortality, morbidity, offending and incarceration.
The cohort will consist of individuals with a diagnosis indicating an acute alcohol harm or problematic alcohol use presenting to inpatient services (NSW Admitted Patient Data Collection; NSW APDC) and acute services (NSW Emergency Department Data Collection; NSW EDDC) in NSW between January 1st, 2005 and December 31st, 2014. Diagnostic classification systems used by NSW APDC and NSW EDDC in this period comprise the International Classification of Diseases and Health Related Problems 10th edition Australian Modification (ICD-10-AM), International Classification of Diseases, 9th edition, Clinical Modification (ICD-9-CM) or the Systematized Nomenclature of Medicine--Clinical Terms Australian Modification (SNOMED-CT-AU). Once the cohort has been identified, their linked data from various other routinely-collected administrative data sets will be extracted, providing information on emergency department presentations, hospital separations, cancer notifications, mental health ambulatory care, mortality, offending, and incarceration.
We will describe the characteristics of the cohort and quantify healthcare service, law enforcement engagement and cause-specific mortality rate among this group. We will then develop a risk predictive model to assess the relationship between cohort characteristics and their associated risks of specific health outcomes (e.g. multi-morbidities, cancer and mortality), social consequences (e.g. offence charges and incarceration), burden of alcohol-related diseases (e.g. frequent health service attendance and economic burden).
In all analyses, multiple confounding variables will be controlled for as appropriate. We will also undertake the below analyses for the total cohort and focused specifically on those young people (e.g., aged 15-24 years old). Findings are reported in accordance with the Reporting of studies Conducted using Observational Routinely-collected health Data statement (RECORD) and Guidelines for Accurate and Transparent Health Estimates Reporting (GATHER).
We have received full ethics approval for this study as well as all datasets. We are currently in the process of planning and undertaking the various analyses outlined in the aims.
Contact the lead investigator about possible PhD project opportunities
This program of research will provide a comprehensive population-level understanding of the burden of problematic alcohol use on individuals and on healthcare and law enforcement services. It will extend knowledge of individual and situational factors that predict adverse alcohol-related outcomes, with the capacity to inform personalised intervention. The patterns of healthcare utilisation will also improve our knowledge of patient needs to enhance healthcare delivery for targeted populations. The multi-dimensional measurement of diverse events produced by this project can better reflect the scale and impact of alcohol-related problems which may be under-ascertained in the study of a single dataset.