Tools for policy and prevention: the Australian National Alcohol Indicators Project.
For decades, alcohol has been a leading cause of preventable death, disease, and disability in Australia, the economic costs of which measure in the billions every year (Collins & Lapsley, 2002). Yet, before the late 1990s, alcohol consumption and related harms were only occasionally attended to by a number of different organizations, in an uncoordinated and piecemeal manner across the various states and territories. Forward-thinking members of the (now disbanded) National Expert Advisory Committee on Alcohol (NEACA) managed to turn this around with their recommendation of and support for a national minimum data set on alcohol. The National Drug Research Institute (NDRI), then the National Centre for Research into the Prevention of Drug Abuse, with its prevention focus, national outlook, strong collaborative links, and research expertise, was identified as the most appropriate center to oversee the project. The Australian Commonwealth Department of Health and Ageing provided the basic financial resources, reviewed every five years. Since its inception the National Alcohol Indicators Project (NAIP) has had a strong focus on collaboration, and Melbourne's Turning Point Alcohol and Drug Research Centre has played a major collaborative role in the project, especially during the first five years.
This article was prepared for the Monitoring alcohol and other drug related harm: Building systems to support better policy international symposium and to support discussion in relation to the first two major objectives of the conference: (1.) to review international and Canadian experiences with designing, implementing and utilizing alcohol and other drug monitoring programs with a view to defining best practices and, (2.) to discuss case studies of the application of monitoring programs for a) the evaluation of significant policies relevant to the reduction of alcohol and other drug-related harm, and/or b) influencing the policy agenda.
The raw materials: Identification, access, and assembly
The primary aim of the NAIP is to track and report on trends in alcohol consumption and related harms across jurisdictions and communities with special emphasis on the wide dissemination of information. Ultimately, the objective is to use these tools to evaluate the efficacy of alcohol policy and strategies. As such, the first requirement of the NAIP was to identify and bring together relevant and reliable data. The project concentrated on obtaining access to data already regularly collected by other agencies--as opposed to collecting new or primary data (which, in any case, was precluded by the modest budget). Improvements in information technology have encouraged many administrative systems to move to electronic records management and this has vastly increased the research potential of such information.
At the outset, a number of potential sources of secondary data relevant to a national monitoring approach for alcohol were readily identifiable: mortality records; morbidity records; national alcohol consumption surveys; police reported road crashes and assault offences; and alcohol sales data (see Table 1 for details).
Access to data collected by a centralized agency (e.g., ABS mortality data) was relatively straightforward with a short data transfer period. Securing access to other noncentralized data, especially police-reported information, was a laborious process and in some cases, delays of up to 18 months occurred between the initial data request and actual data transfer. The inability to guarantee access and to directly control reporting processes is one of the limitations inherent to studies which rely heavily on data collected and controlled by others.
NAIP data requests typically occur on an annual basis. Fortunately the financial cost of obtaining official administrative information is relatively small. Most government agencies charge a minimum cost-recovery rate as opposed to the high costs associated with private data companies and population surveys.
In large part, the harm indicators selected for this project were determined by practical considerations, including access to electronic records. This selection bias admittedly produces only a partial picture of the actual impact of alcohol consumption on a population. Emergency department (ED) data is a case in point--these data are a rich source of information and hold particular potential for monitoring alcohol-related harms but, for the reasons described below, do not yet form part of the NAIP collection.
Emergency department data has a particular capacity to capture alcohol-related injuries which do not appear in official hospital admission records (and possibly include large numbers of less serious injuries). The high frequency of events and the broad spectrum of conditions captured by such records are likely to prove especially instructive where hospital admissions and/or deaths are relatively infrequent (e.g., small communities, rural areas). Unfortunately, unlike hospital separations, there is currently no systematic or standardized approach to recording ED presentations in Australia. The use of electronic record management packages (e.g., Emergency Department Information System) that potentially identify presentations by International Classification of Diseases (ICD) code is typically left to the discretion of individual hospitals, and as a result, application is generally piecemeal and many non-urban hospitals continue to use pen-and-paper records.
In addition to ED records, there are other data collections which may provide relevant and valuable information, including but not necessarily limited to: local and state alcohol, crime, and social surveys; ambulance call-outs; police drunk and disorderly reports; liquor infringement notices; reports of child abuse; sobering-up shelter admission; admissions to women's refuges; and liquor industry data (e.g., Liquor Merchants of Australia, Distilled Spirits Industry Council of Australia, Inc.). For the most part, current application of these data across all Australian jurisdictions and for comparative purposes is limited.
Shaping the measurement tools: Fashioning administrative data sets into indicators of alcohol-related harm
Data that are primarily collected for nonresearch purposes typically require a substantial amount of time and effort to shape into relevant, reliable, and consistent measurement tools. This is especially the case where data are collected independently from separate jurisdictions and where there is no nationally standardized recording system (e.g., police assault data). In order to shape these data into appropriate measurement tools or alcohol indicators, the NAIP has used a number of approaches, two of which will be discussed in detail in this section: 1. the population aetiologic fraction method and the, 2. surrogate method. Both of these methods are well established in the epidemiological research literature; however, wherever possible and appropriate, the NAIP has sought to improve upon the standard approach with tailored modifications.
This section describes each of these methods and its application by the NAIP.
Estimating alcohol-attributable mortality and morbidity using the aetiologic fraction method
Typically, routinely collected death and hospitalization data do not provide information about each individual's level or pattern of alcohol consumption. The population aetiologic fraction method provides a means of estimating the number of alcohol-attributable cases given: 1. the prevalence of drinking in the population from which the cases are derived and, 2. the relationship between consumption and specific disease or injury (i.e., relative risk or odds ratio). A multiplication of the number of people with each particular condition by the population alcohol aetiologic fraction (PAAF) specific to that condition, followed by a sum of the results, produces an estimate of the number of alcohol-attributable deaths or hospitalizations in a given population (see English et al., 1995; WHO, 2000).
The PAAF for a particular illness or injury attributable to various levels of drinking is the proportion of cases with that condition in the population that can be attributed to such drinking. For some conditions (such as alcoholic liver cirrhosis and alcohol dependence), the PAAF is one (1), because such conditions are--by definition--wholly attributable to alcohol. For other conditions (e.g., assault, road crashes, and stroke) the PAAF is less than one, because they are only partially attributable to alcohol. In these instances, the PAAF is a function of both the strength of the causal relationship between a particular level of drinking and the condition (measured as a relative risk) and the proportion of the population drinking at that particular level (i.e., drinking prevalence). There are over 40 conditions for which there exists sufficient research evidence to support a causal relationship with alcohol consumption (English et al., 1995).
PAAFs have the potential to vary widely over place and time. In part, this is because consideration of drinking prevalence within the population of interest is a critical component in their estimation. Prior to the NAIP, however, Australian estimates of alcohol-attributable mortality and morbidity typically adopted a one-size-fits-all approach to drinking prevalence estimates. Most studies assumed that a measure of drinking prevalence taken from one population at one point in time could be reliably applied across different populations and over different time periods--usually an estimate of drinking prevalence which covered the entire nation (e.g. Chikritzhs et al., 2002). As is the case for many countries, there is substantial variation in levels and patterns of alcohol consumption throughout Australia. It has been estimated, for example, that per capita alcohol consumption in the Northern Territory and some northern non-metropolitan areas of Western Australia are at least one and a half times greater than the national level (Catalano, Chikritzhs, Stockwell, Webb, & Dietze, 2001).
The NAIP addressed this substantial variation in drinking levels by using, for the first time, drinking prevalence estimates specific to each state and territory in the estimation of alcohol-attributable morbidity and mortality. In addition, since accurate documentation of trends over time was a primary goal of the NAIP, in the absence of annual surveys of drinking, per capita alcohol consumption was used to adjust prevalence estimates over time (Chikritzhs et al., 2000). This new approach was to prove efficacious when it was demonstrated that alcohol population aetiologic fractions for the Northern Territory were between 50% and 75% greater than those for Australia as a whole (Chikritzhs et al., 2000a). Using a similar approach, the most recent NAIP bulletin (No.11) estimated Indigenous alcohol attributable-deaths based on levels of alcohol consumption drawn specifically from representative Australian Indigenous populations. The use of Indigenous-specific drinking prevalence data as opposed to the standard approach of using national non-Indigenous levels of drinking prevalence dramatically increased the underlying PAAFs and the subsequent harm estimates. Compared to PAAFs based on non-Indigenous specific drinking prevalence (i.e. general population surveys), Indigenous-specific PAAFs are about 12% and 30% greater for chronic and acute alcohol-attributable conditions respectively (Chikritzhs et al., 2007a).
Throughout the late 1990s and early 2000s a variety of alcohol-attributable mortality and morbidity estimates were published by several independently funded research centers (it is interesting to note that all were funded by the Commonwealth Government). The range of estimates available did little to improve general levels of understanding concerning alcohol and harm but highlighted the regrettable absence of a consensus among researchers as to the most appropriate methodological approach. For example, Chikritzhs, Jonas, Stockwell, Heale, & Dietze (2001) estimated that there were 3,290 deaths due to hazardous and harmful alcohol consumption in 1997. For the same year, Ridolfo and Stevenson (2001) and Higgins, Cooper-Sanbury, & Williams (2000) estimated 3,411 and 3,668 such deaths respectively. Mathers, Vos, & Stevenson (1999) estimated that during 1996 there were 7,157 lives saved and 4,492 lives lost due to drinking, producing a net saving of 2,631 lives. The NAIP sought to address this disparity among findings by bringing together a consortium of Australian alcohol researchers to agree upon and establish a set of consensus recommendations in relation to the quantification of alcohol-attributable mortality and morbidity. The recommendations included guidelines for the adoption of future approaches that hopefully served as a guide to workers in the wider fields of alcohol research and to promote positive communication among independent research groups (see Chikritzhs et al., 2002).
The application of alcohol-related surrogate measures
It is rare that alcohol's role in events, which cause individuals to come to the attention of government agencies or authorities, can be reliably and directly discerned from official records. For instance, even though legal maximum breath alcohol levels for driving have been in place for decades in Australian states and territories, the reporting of driver breath alcohol concentrations for those involved in road crashes is not mandatory across all jurisdictions and is often left to the discretion of the reporting officer (especially where drivers and passengers have escaped non-fatal injury) (Chikritzhs, Stockwell, Heale, Dietze, & Webb, 2000b). Moreover, reporting systems which ostensibly make allowance for an entry identifying whether or not alcohol was involved, such as an "alcohol flag," rarely manage to achieve adequate levels of objectivity, compliance, and reliability to be given serious consideration for monitoring purposes.
For monitoring trends over time as opposed to estimating population prevalence, measures that do not necessarily capture all alcohol-related events but which reliably identify events for which alcohol is highly likely to be a major contributor--although not necessarily the only contributor--may well suffice. For example, the use of the surrogate measure "single--vehicle nighttime road crashes" (e.g. crash occurred between midnight and 1 a.m. and involved a lone driver veering off a highway into a lamppost) in the road safety research literature demonstrates how, in the absence of breath alcohol data, time of day and crash circumstances can be effectively used to identify crashes likely to be a result of intoxication (e.g., Holder & Wagenaar, 1993).
Based on the surrogate approach, the NAIP used fatally injured drunk-driver blood alcohol data to identify specific times of the day and particular days of the week which were likely to be either alcohol- or nonalcohol-related for each jurisdiction. Across the country, the most common times for alcohol-related crashes to occur were during the hours of 10 p.m. and 2 a.m. on Friday, Saturday, and Sunday nights, although longer hours were evident for Saturdays. Daytime hours between 6 a.m. and 2 p.m. on most weekdays were commonly associated with large numbers of non-alcohol-related injuries.
Using these temporal parameters, trends in alcohol-related fatalities/serious injuries versus non-alcohol-related road injuries (Chikritzhs et al., 2000b) and alcohol-related versus non-alcohol assaults were able to be made (Matthews, Chikritzhs, Catalano, Stockwell, & Donath, 2002). Thus, where the degree of alcohol involvement in an incident is uncertain or unknown, an effective alternative is to identify cases that are highly likely to be alcohol-related, given what is known about other characteristics associated with the event that are likely to be both accurate and reliable.
Being and staying policy relevant
Commentators on the transfer from research evidence to uptake of evidence-based policy (or lack thereof) have observed that "researchers and policy makers work to different imperatives" (Lin, 2003, p. 285). From the outset, the NAIP has deliberately sought to conduct research pointedly aimed at producing policy relevant outcomes that are both scientifically rigorous and readily accessible by nonresearchers. As described below this is achieved using a range of strategies.
Packaging the tools: Modes of dissemination
Rather than assume the one-size-fits-all approach to dissemination typical of academic research (i.e., peer review journal publications), the NAIP uses a range of modes for dissemination that potentially appeals to diverse audiences. One of the project's most innovative and mainstay modes of information dissemination is the four-page, color print bulletin. There are currently 11 bulletins in the NAIP series that examine trends for a range of alcohol indicators and population subgroups. The bulletins have proven a highly successful means of presenting information on alcohol consumption and harms, each release having disseminated upwards of 2000 hard copies free-of-charge with many more downloaded from the NDRI website (see www.ndri.curtin.edu.au/publications/naip.html).
Feedback indicates that the appeal of the bulletins lies in their brevity and simplicity. Limited to four pages, they use color maps and uncomplicated figures to summarize information. Each bulletin begins by focusing the reader's attention on a dot-point summary of the most crucial and salient items--all expressed in plain language. The bulletins provide a minimum of technical and methodological detail but enough to allow accurate interpretation of the data. The verbatim replication of maps, figures and summary points by print media are testimony to the utility of the approach. In many cases, the bulletins are also accompanied by technical reports that provide detailed methods and results.
There is a need for researchers to find new ways to enhance the practical impact of their work, but the obligation to maintain high quality research standards, to submit to the scrutiny of peers and the scientific community, and to contribute to the scientific body of knowledge remain fundamental. The NAIP pursues its scientific obligations as vigorously as its contribution to policy and practice, with regular submissions to peer reviewed journals, book chapters, and conference attendances.
Getting the information packaged right is a crucial first step toward bridging the information gap, but it is not sufficient to ensure uptake. Policy makers are subject to a range of competing pressures and attendance to objective research evidence is, at best, likely to be a low priority (Lin, 2003). The NAIP approach to information uptake assumes that busy policy makers, practitioners, and their informants will rarely actively seek out research evidence. Rather than waiting to be found, NAIP outputs are proactively communicated to potential audiences. This is largely achieved by strategic use of the media, plus electronic and conventional mailing lists.
For each new NAIP output (e.g., bulletin, journal article), careful consideration is given to a media strategy and how best to time the media release. Timing is especially important. Each newsworthy story competes on a daily basis for media attention. A mistimed release, perhaps one which clashes with another alcohol story, can result in minimal or no coverage. In part, the high frequency of media attention given to Bulletin 7 (drinking among underaged teenagers) was due to the fact that it was released just prior to "schoolies" week when binge drinking by holidaying school leavers typically (and predictably) draws attention from both the media and general community. In most aspects of media activity, a dedicated media officer is invaluable. The identification of salient and interesting aspects of the research, construction of succinct but engaging media releases, establishment and maintenance of important media contacts, and liaison with key stakeholders are all best dealt with by a trained media professional as such activities are usually beyond the scope and expertise of most academic researchers.
Each bulletin and media release is accompanied by an NDRI Web site posting, submissions to appropriate listserves, and coordinated distribution of hard copies by surface mail to key organizations (e.g., national and international research centers, drug and alcohol offices, health departments, treatment centers, nongovernment advocacy organizations, state and federal government ministers, professional health and medical organizations, and university libraries). Bulletins are also available for download from the Commonwealth Government's National Drug Strategy Web site (as well as the NDRI Web site).
The strong collaborative links between the NDRI and other national and international drug research centers has been a crucial factor in the expansion of NAIP activities and expertise. The NAIP Advisory Committee consists of representatives from a range of stakeholders and provides critical comment on the utility and content of draft bulletins as well as suggestions for further work. The specialized Indigenous NAIP Advisory Committee comprised of prominent members of the National Indigenous Drug and Alcohol Committee, has played a crucial role in facilitating access to Indigenous specific health information and in ensuring the wide dissemination of the first Indigenous NAIP Bulletin (No. 11).
Recognizing and responding to information needs
The NAIP follows an internally driven workplan but is also operationally flexible, allowing responses to key stakeholder information needs as they arise. Bulletin 7, for example, arose from an informal communication by the Commonwealth Government signalling a specific need for consolidated, up-to-date information on levels of consumption and harms among teenage drinkers. The media interest generated around Bulletin 7 was intense, resulting in detailed front-page coverage (including the full list of summary points lifted directly from the bulletin) by the Australian capital city's premier newspaper. Canberra is the home of Australia's federal parliament, and the article included comment from the Ministerial Council on Drug Strategy.
In November, 2002, the National Drug Strategy Aboriginal and Torres Strait Islander Peoples' Reference Group and the former National Expert Advisory Committee on Alcohol (NEACA) identified a crucial need for indicators of alcohol-related harms specific to Indigenous Australians in order to facilitate the timely and reliable measurement of the effectiveness of policies and interventions for Indigenous communities. At the request of both of these bodies, the Commonwealth Government funded the NAIP to examine whether it was feasible to construct alcohol-related harm indicators specifically applicable to Indigenous populations. A major breakthrough from the project was the identification and use of some 17 zones for reporting on alcohol-related harms with specific relevance to Indigenous communities. As a first ever Australian effort to provide standardized alcohol-attributable death rates for Indigenous people across all of Australia, Bulletin 11 prompted comment from local government, state and federal health ministers, and a range of stakeholders concerned with Indigenous health issues.
Making the most of opportunities for evaluating policy
The NAIP has a strong commitment, not only to monitoring indicators, but also to the evaluation of alcohol policy throughout Australia; probably the most notable of which was the evaluation of the Northern Territory's Living With Alcohol (LWA) program. The LWA program was introduced in 1992 and was initially funded by the imposition of a small levy on all alcoholic beverages sold in the Northern Territory containing 3% alcohol by volume or greater. The LWA Levy effectively raised the retail cost of these beverages by about 5 cents per standard drink. The Levy was removed in 1997 which in turn resulted in a fall in the real price of alcoholic beverages with more than 3% alcohol by volume. Nevertheless, LWA programs and services continued to operate until 2002 and were funded from redirected taxes collected by the Commonwealth. The public health, safety, and economic impact of the LWA program was initially evaluated by the NAIP (see Stockwell et al., 2001) and was found to have resulted in a significant cost saving to the Northern Territory during its first four years of operation. A subsequent evaluation showed that without the support of the price increase, the LWA programs and services intended to reduce alcohol-related harms would have had limited impact on reducing acute harms (e.g., road injury and violent assault) (Chikritzhs, Stockwell, & Pascal, 2005).
Results from these studies presented a strong argument for alcohol taxes combined with comprehensive programs and services designed to reduce the harms from alcohol. By capitalizing on the occurrence of a natural experiment, the NAIP was able to provide robust evidence for the efficacy of reducing the economic availability of alcohol in an Australian context. Contextually and culturally relevant research evidence is likely to stand a better chance of influencing policy than evidence derived elsewhere. It is far more difficult to disparage the significance of local evidence on the grounds that it lacks domestic relevance than it is to argue away, on the same grounds, an entire suit of concurring scholarly reviews and articles from the international literature.
The NAIP outgoing costs are small--the project operates on a modest budget with a small core research team--yet there is a range of evidence to suggest that the benefits of the project have been substantial. Resources generated by the NAIP have been used by a range of health professionals, communities, academics, government and nongovernment organizations to argue for evidence based alcohol policy in this country. As to whether such efforts will be rewarded with tangible positive change, time will tell. Perhaps more important however, are the many thousands of disseminated bulletins and the many hundreds of generated media statements and interviews, the contribution of which to increased awareness, informed community debate, and progress toward evidence-based alcohol policy is difficult to measure, but which should not be underestimated.
Catalano, P., Chikritzhs, T., Stockwell, T., Webb, M., & Dietze, P. (2001). Trends in per capita alcohol consumption in Australia, 1990/911998/99. (National Alcohol Indicators, NDRI Monograph No. 4). Perth: National Drug Research Institute, Curtin University of Technology.
Chikritzhs, T., Jonas, H., Heale, P., Dietze, P., Stockwell, T., Hanlin. K., & Webb, M. (2000). Alcohol-caused deaths and hospitalisations in Australia, 1990-1997. (National Alcohol Indicators, Technical Report No. 1). Perth: National Drug Research Institute, Curtin University of Technology.
Chikritzhs, T., Stockwell, T., Heale, P., Dietze, P., & Webb, M. (2000b). Trends in Alcohol-related Road Injury in Australia, 1990-1997. (National Alcohol Indicators, Technical Report No.2). Perth: National Drug Research Institute, Curtin University of Technology.
Chikritzhs, T., Jonas, H., Stockwell, T., Heale, P., & Dietze, P. (2001). Mortality and life years lost due to alcohol: a comparison of acute and chronic causes. Medical Journal of Australia, 19(174), 281-284.
Chikritzhs, T.. Stockwell, T., Jonas, H., Stevenson, C., Cooper-Stanbury, M., Donath, S., Single, E.. & Catalano, P. (2002). Towards a standardised methodology for estimating alcohol-caused death, injury and illness in Australia. Australian and New Zealand Journal of Public Health, 26(5), 443-450.
Chikritzhs, T., Catalano, P., Stockwell, T., Donath, S., Ngo, H., Young, D.. & Matthews, S. (2003). Australian Alcohol Indicators, 1990-2001: Patterns of alcohol use and related harms for Australian states and territories. Perth: National Drug Research Institute, Curtin University of Technology.
Chikritzhs, T., Stockwell, T., & Pascal, R. (2005).The impact of the Northern Territory's Living With Alcohol Program, 1992-2002: revisiting the evaluation. Addiction, 100, 1625-1636.
Chikritzhs, T., Pascal, R., Gray, D., Stearne, A., Saggers, S., & Jones, P. (2007a). Trends in alcohol-attributable deaths among Indigenous Australians, 1998-2004. (National Alcohol Indicators, Bulletin No. 11). Perth: National Drug Research Institute, Curtin University of Technology.
Chikritzhs, T., Gray, D., Lyons, Z., & Saggers, S.(Eds.) (2007b). Restrictions on the sale and supply of alcohol: Evidence and outcomes. (NDRI Monograph). Perth: National Drug Research Institute, Curtin University of Technology.
Clemens, S., Donath, S., Stockwell, T., & Chikritzhs, T. (2007). Alcohol consumption in Australia: National surveys from 1989 to 2004. (National Alcohol Indicators, Technical Report No.2. Perth). National Drug Research Institute, Curtin University of Technology.
Collins, D. & Lapsley H. (2002). Counting the cost: Estimates of the social costs of drug abuse ill Australia in 1998-9. (National Drug Strategy Monograph Series No. 49). Canberra: Commonwealth Department of Health and Ageing.
English, D., Holman, C., Milne, E., Winter, M., Hulse, G., Codde, J., et al. (1995). The quantification of drug caused morbidity and mortality in Australia. Canberra: Commonwealth Department of Human Services and Health.
Higgins, K., Cooper-Stanbury, M., & Williams, P. (2000). Statistics on drug use in Australia 1998. (AIHW cat no. PHE 16). Canberra: AIHW (Drug Statistics Series).
Holder, H. & Wagenaar A. (1993). Mandated server training and reduced alcohol-involved crashes: a time series analysis of the Oregon experience. Accident, Analysis and Prevention, 26(1), 89-97.
Lin, V. (2003). Improving the research and policy partnership: an agenda for research transfer and governance. In V. Lin & B. Gibson (Eds.), Evidence-based health policy: Problems and possibilities. (pp. 285297). Oxford: Oxford University Press.
Mathers, C., Vos, T., & Stevenson, C. (1999). The burden of disease and injury in Australia. Canberra: Australian Institute of Health and Welfare.
Matthews, S., Chikritzhs, T., Catalano, P., Stockwell, T., & Donath, S. (2002). Trends in alcohol-related violence in Australia, 1991/921999/00. (National Alcohol Indicators, Technical Report No.5). Perth: National Drug Research Institute, Curtin University of Technology.
Ridolfo, B. & Stevenson, C. (2001). The quantification of drug-caused mortality and morbidity in Australia, 1998. Canberra: AIHW
Stockwell, T. & Chikritzhs, T. (Eds.) (2000). International guide for monitoring alcohol consumption and alcohol-related harm (WHO/MSD IMSB/00.5). Geneva: World Health Organization.
Stockwell, T., Chikritzhs, T., Hendrie, D., Fordham, R., Ying, F., Phillips, M., Cronin J., & O'Reilly, B. (2001). The public health and safety benefits of the Northern Territory's Living With Alcohol programme. Drug and Alcohol Review, 20:167-180.
TABLE 1 Secondary data sources of alcohol-related harm indicators Data source Measure Wholesaler records of Per capita pure alcohol consumption; alcohol purchases made volume of pure alcohol consumption by liquor retailers. by beverage type Collected by liquor licensing authorities in some states/territories Individual state/territory Numbers/rates of police-reported police services violent offences (e.g. violent assault, disturbances, drunk and disorderly conduct) State/territory police Numbers/rates of police-reported services road crashes and impaired driving offences Australian Bureau of Numbers/rates of alcohol-attributable Statistics (ABS) collates deaths data from all jurisdictions Australian Institute of Numbers/rates of alcohol-attributable Health and Welfare hospital admissions (AIHW) collates data from all jurisdictions Local emergency Numbers/rates of emergency department(s) (ED). State/ department presentations territory health departments (where available) ABS and AIHW large- Self-reported alcohol consumption scale, representative national alcohol consumption surveys Data source Comment Wholesaler records of From 1990/1996, it was possible to alcohol purchases made access electronic records of annual by liquor retailers. volumes of alcohol purchases made by Collected by liquor licensed retail outlets utilized by licensing authorities in licensing departments to calculate some states/territories licensing fees. These data enabled estimation of per capita alcohol consumption, an invaluable alcohol indicator (e.g., Catalano et al., 2001). Most jurisdictions stopped collecting these data after a 1997 High Court ruled that raising taxes on alcoholic beverages, tobacco and petrol by states/ territories was unconstitutional. The ruling did not preclude the collection of wholesale alcohol purchase data by liquor licensing authorities but, for most jurisdictions, the incentive for continued collection was lost. Only Western Australia and the Northern Territory continued to collect these data. Individual state/territory Police reports of violent offences are a police services potentially rich source of information. However, without a central collation agency, data were not readily accessible. Individual agreements were reached with data custodians in each jurisdiction allowing transfer of de- identified unit records. Data typically included: time of day of offence; type of offense; sex and age of offender; and location of offence. (Matthews et al., 2002) State/territory police For several years, the Australian services Transport and Safety Bureau (ATSB), collated unit record police reports of road crashes from all jurisdictions. Driver-based data included: severity of injury, time of crash, location, sex and age. Many cases included breath alcohol level but compliance varied considerably between jurisdictions (Chikritzhs et al., 2000b). In the late 1990s, completeness of ATSB data holdings declined considerably, severely reducing utility. Australian Bureau of Includes de-identified information on Statistics (ABS) collates deceased individuals (e.g. cause, date, data from all jurisdictions sex, age, location). Use of ICD codes and established methods for estimating alcohol-attributable deaths (Chikritzhs et al., 2003) has meant that accessing these data is a NAIP priority. Australian Institute of As for mortality data, use of ICD codes Health and Welfare and application of the population (AIHW) collates data aetiologic traction method allows from all jurisdictions alcohol-attributable hospitalizations and related outcomes (e.g. bed days, PYLLs) to be estimated (Chikritzhs et al. 2003). Local emergency Systematic reporting and ICD coding of departments) (ED). State/ ED presentations is not wide-spread in territory health departments Australia, and standardized, nation-wide (where available) identification of alcohol- attributable conditions is not yet possible. Some analysts have focused on only injury-related presentations or surrogate measures (e.g. night-time injuries). Use of subjective reports of alcohol-related admissions should be treated with caution (NDRI, 2007). ABS and AIHW large- Every three years, the AIHW conducts the scale, representative National Drug Strategy Household Survey national alcohol (NDSHS)-the only substance-use-specific consumption surveys national survey. The ABS conducts the National Health Survey (NHS), which asks questions about a range of health issues including alcohol and drug use. Methods used and response rates achieved vary considerably (Clemens et al., 2007).
|Printer friendly Cite/link Email Feedback|
|Publication:||Contemporary Drug Problems|
|Date:||Sep 22, 2009|
|Previous Article:||The World Health Organization's Global Alcohol Database: opportunities for research and support for policy.|
|Next Article:||Monitoring alcohol and alcohol related problems in Sweden.|