Volume 172 Issue 12

In 1998-1999 the Australian Institute of Health and Welfare carried out a national study of the burden of disease and injury in Australia,1,2 in close collaboration with the Victorian Department of Human Services, which conducted a parallel analysis concurrently in Victoria.3,4 Here, the principal investigators present the key findings. Abstract This is an overview of the first burden of disease and injury studies carried out in Australia. Methods developed for the World Bank and World Health Organization Global Burden of Disease Study were adapted and applied to Australian population health data. Depression was found to be the top-ranking cause of non-fatal disease burden in Australia, causing 8% of the total years lost due to disability in 1996. Mental disorders overall were responsible for nearly 30% of the non-fatal disease burden. The leading causes of total disease burden (disability-adjusted life years [DALYs]) were ischaemic heart disease and stroke, together causing nearly 18% of the total disease burden. Depression was the fourth leading cause of disease burden, accounting for 3.7% of the total burden. Of the 10 major risk factors to which the disease burden can be attributed, tobacco smoking causes an estimated 10% of the total disease burden in Australia, followed by physical inactivity (7%). The Australian and Victorian burden of disease and injury studies were the first comprehensive studies of this type in Australia. Both studies used methods developed for the Global Burden of Disease Study (GBD)5 to quantify the loss of health from a comprehensive set of 176 causes of disease and injury and for 10 major risk factors. They used a common metric, the disability-adjusted life year (or DALY), which combines information on both the impact of premature death and the impact of disability and other non-fatal health outcomes. One DALY can be thought of as one lost year of "healthy" life, and the burden of disease as a measurement of the gap between current health status and an ideal situation of living into old age free of disease and disability. DALYs have previously been used to provide a comprehensive assessment of the global burden of disease and injury for the World Bank,6 to inform global priority setting for health research,7 and to report on trends in population health across the world.8 The methodology used in the studies is described in Box 1. Years of life lost due to mortality (YLL) Ischaemic heart disease is by far the largest cause of years of life lost in both males and females, accounting for 20.5% of the mortality burden in 1996. Ischaemic heart disease is followed by stroke and breast cancer in females, and by lung cancer and suicide in males (Box 2A). Using a small area-based measure of socioeconomic disadvantage, we found that, in 1996, the most disadvantaged quintile of the Australian population lost 35% more years of life than the least disadvantaged quintile. For Australians aged less than 65 years, the differential burden between the lowest and highest quintiles is even greater, with a 60% excess burden in the most disadvantaged group. Years lost due to disability (YLD) Mental disorders are the leading cause of YLD, accounting for nearly 30% of the non-fatal disease burden in Australia in 1996 (Box 2B). These are followed by nervous system and sense organ disorders (16%) and chronic respiratory diseases (9%). In terms of specific conditions, depression is the leading cause of non-fatal disease burden in Australia, causing 8% of the total YLD in 1996 (Box 2B). Hearing loss and alcohol dependence and harmful use of alcohol are the second and third leading contributors to YLD for males. Dementia and osteoarthritis are the second and third leading contributors for females. In contrast to the mortality burden, the disability burden is smaller for males than for females. Neurological and sense organ disorders, mental disorders and musculoskeletal disorders are all more common in females. The male burden is higher for cardiovascular disease, diabetes, chronic respiratory diseases and cancers. Females generally have a greater incidence and prevalence of the more common non-fatal health problems, whereas males have a greater incidence of the major diseases and injuries associated with high case fatality. Thus, some of the years of the longer life span women enjoy are lived with a lower quality of life. Total disease burden (DALYs) As shown in Box 3, inclusion of non-fatal health outcomes creates a substantially different picture to that provided by traditional mortality statistics: mental disorders are now the third leading cause of overall burden (14% of total), after cardiovascular diseases (20%) and cancers (19%). Both the burden of nervous system disorders (which includes dementia) and that of chronic respiratory conditions are similar in magnitude to that of injuries. The total disease burden in males is 13% higher than that in females. Ischaemic heart disease and stroke head the list of specific causes of disease burden, together causing nearly 18% of the total (Box 2C). Chronic obstructive pulmonary disease (COPD) and lung cancer together account for another 7.3% of the total burden. Depression is the fourth leading cause of disease burden in Australia, accounting for nearly 4% of the total burden. The burden of mental disorders in Australia is dominated by affective (depression and bipolar disorder), substance use and anxiety disorders. Substance use disorders are the leading cause of mental disorders for males, accounting for 33% of their mental health DALYs (Box 4). The major cause of mental disorders for women is depression, accounting for 34% of women's mental health DALYs. The large burden from mental disorders in young adults is partly due to the high prevalence of these disorders at these ages and partly because the DALY, as an incidence-based measure, attributes the future stream of ill-health to the age at incidence (Box 5). Inclusion of the burden of suicide and ischaemic heart disease attributable to depression increases the burden of depression from 3% to 5% of total DALYs. Inclusion of the burden of cardiovascular disease attributable to diabetes increases the diabetes burden also from 3% to 5%. Depression and diabetes then become equal third leading causes of the burden of disease in Australia. Projected burden of disease in 2016 The Victorian Burden of Disease Study projected a 25% drop in the rate of all-cause DALYs in men and a 17% drop in women over the next 20 years. In Victoria, life expectancy is projected to increase by 4.6 years for men and 3.6 years for women. Mortality is forecasted to drop considerably faster than disability. The large gap in projected changes between YLL and YLD rates suggests that some of the further gains in life expectancy may occur at the expense of quality of life. Large health gains are expected in all cardiovascular diseases except cardiomyopathy. Large gains are also predicted for injuries, COPD in men and in a number of cancers (lung cancer in men; bowel cancer, stomach cancer and leukaemia in both sexes). Adverse mortality trends are driving projected increases in the burden rates from lung cancer and dementia in women and heroin overdose deaths, melanoma and diabetes in men (Box 6). Cancer will be the largest cause of disease burden in 2016, as improvements in cardiovascular health will outpace the slower improvements in cancer. Population ageing will increase the burden of neurological, sense organ and musculoskeletal disorders relative to other conditions. The burden of injuries is expected to decrease partly due to favourable trends and partly as a consequence of population ageing. The burden of dementia will increase for both sexes, and, for women in Victoria, may take over from ischaemic heart disease as the largest cause of DALYs in 2016. Diabetes, prostate cancer, hearing loss and heroin dependence in men, and lung cancer in women, are other conditions that will move up the ranking order. Stroke in men and women, and road traffic accidents, suicide and COPD in men, will drop considerably in their ranking order of projected burden in 2016. Burden attributable to risk factors Risk factors such as smoking, alcohol consumption, physical inactivity, hypertension, high blood cholesterol level, obesity and inadequate fruit and vegetable consumption are responsible for large proportions of the overall burden of disease in Australia (Box 7). The leading risk factor is tobacco smoking (responsible for 10% of the total burden), followed by physical inactivity (7%), then high blood pressure and obesity. Insufficient intake of fruits and vegetables (fewer than five servings per day) causes an estimated 3% of the total burden, and 11% of the cancer burden in Australia. The net harm associated with alcohol consumption is around 2% of the total burden, as the ill-health associated with harmful and hazardous drinking is offset by benefits from alcohol in the prevention of cardiovascular disease. The protective effect is only relevant after age 45, whereas the harmful effects of alcohol are apparent at all ages. For males, illicit drugs are responsible for a level of harm similar to that of alcohol (2.2% of total male burden of disease). Just over half this burden is due to premature mortality, the other half to YLD resulting from drug dependence or harmful use. Discussion Burden of disease analysis provides a unique perspective on health, one that integrates fatal and non-fatal outcomes, yet allows the two classes of outcomes to be examined separately. Among the top 15 causes of burden of disease and injury in Australia are four non-fatal or low-fatality diseases: depression, asthma, osteoarthritis and hearing loss. The burden of mental illnesses such as depression and alcohol dependence, and of non-fatal diseases such as osteoarthritis and hearing loss, has been seriously underestimated by traditional approaches to disease burden assessment that disregard disability and take into account only deaths. The calculation of YLL is straightforward, and the precision of the estimates is almost entirely dependent on the quality of the data on the underlying cause of death. The calculation of the disability burden (YLD) requires much more extensive epidemiological modelling, drawing on a diverse range of data sources, research findings and expert opinion. Thus, the precision of the YLD estimates is not quantifiable in the usual statistical sense. As a next step in these studies, we plan to model the uncertainty of these burden of disease estimates using simulation modelling and sensitivity analyses. Estimates of the burden attributable to each risk factor are based as far as possible on studies that controlled for other relevant risk factors, but it is likely that the complexity of the interaction between risk factors has not been fully captured. Despite this reservation, the conclusion remains that each of the risk factors analysed is responsible for much ill health, similar in magnitude to the top 10 diseases. Projections of the burden of disease to 2016 highlight the adverse trends in tobacco-related diseases for women, and diabetes and substance abuse for men, and the impact of population ageing in increasing the disability burden. These findings suggest that tobacco control, and combating physical inactivity and overweight and obesity, should be continuing priorities for public health action. The extensive epidemiological modelling for over 1200 disease and sequelae categories resulted in the identification of many gaps and deficiencies in Australian population health data (even given the high quality and extensive availability of such data in Australia compared with many other countries). Rather than avoid areas with poor or unavailable information, these studies attempt to make the best possible, internally consistent estimates within a comprehensive framework. These studies are a first step in exploring the usefulness of burden of disease methods for Australia. There are major gaps in our knowledge about the effectiveness of interventions and the associated costs. Linking of burden of disease analyses to studies of the cost-effectiveness of interventions for major health problems will allow these interventions to be judged both in terms of cost-effectiveness, and their relative impacts in reducing the burden of disease and ill-health. Until these analyses can be done, however, the results reported here provide an indication of the "unfinished" health agenda in Australia, identifying areas in which additional health gains can be made. Acknowledgements The Australian Burden of Disease Study was carried out with funding support from the Commonwealth Department of Health and Aged Care. We thank other members of the Australian and Victorian project teams who assisted in analyses, data collection and literature reviews: Bruno Ridolfo, Simon Eckermann, James Morris, Ying Chen, Derrick Bui, Nittita Prasopa-Plaizier, Bernadette Pound, Anne Magnus and Sean Tobin. References Mathers C, Vos T, Stevenson C. The burden of disease and injury in Australia. Australian Institute of Health and Welfare. Canberra: AIHW, 1999. Also at: <http://www.aihw.gov.au/publications/health/bdia.html> (accessed 12 May 2000). Mathers C, Vos T, Stevenson C. The burden of disease and injury in Australia summary report. Australian Institute of Health and Welfare. Canberra: AIHW, 1999. Also at: <http://www.aihw.gov.au/publications/health/bdiasr.html> (accessed 12 May 2000). Vos T, Begg S. The Victorian Burden of Disease Study: Mortality. Melbourne: Public Health and Development Division, Department of Human Services, 1999. Also at: <http://www.dhs.vic.gov.au/phd/9903009/index.htm> (accessed 12 May 2000). Vos T, Begg S. The Victorian Burden of Disease Study: Morbidity. Melbourne: Public Health Division, Department of Human Services, 2000. Also at: <http://www.dhs.vic.gov.au/phd/9909065/index.htm> (accessed 12 May 2000). Murray CJ, Lopez AD. The Global Burden of Disease: a comprehensive assessment of mortality and disability from diseases, injuries and risk factors in 1990 and projected to 2020. Vol 1. Cambridge, Mass: Harvard School of Public Health on behalf of the World Health Organization and the World Bank, 1996. World Bank. World Development Report 1993: investing in health. New York: Oxford University Press, 1993. Ad Hoc Committee on Health Research Relating to Future Intervention Options. Investing in health research and development. Geneva, World Health Organization, 1996. World Health Organization (WHO). The World Health Report 1999. Geneva: WHO, 1999. Stouthard M, Essink-Bot M, Bonsel G, et al. Disability weights for diseases in the Netherlands. Rotterdam: Department of Public Health, Erasmus University, 1997. DisMod [computer program], version 1.0. President and Fellows of Harvard College. All rights reserved. 1994. The Burden of Disease Unit, Harvard University. (This software can be downloaded from the WWW at: <http://www.hsph.harvard.edu/organizations/bdu/dismod/ index.htm> (accessed 12 May 2000). English DR, Holman CDJ, Milne E. The quantification of drug caused morbidity and mortality in Australia, 1995 edition. Canberra: Commonwealth Department of Human Services and Health, 1995. Beaglehole R, Dobson A, Hobbs MS, et al. CHD in Australia and New Zealand. Int J Epidemiol 1989 18 (3 Suppl 1): S145-S148. Authors' details Australian Institute of Health and Welfare, Canberra, ACT. Colin D Mathers, BSc, PhD, Principal Research Fellow, Health Division; Chris E Stevenson, BSc, MSc, Statistician, Health Division. Department of Human Services, Melbourne, Victoria. E Theo Vos, MD, MSc, Senior Epidemiologist, Policy Development and Planning Division; Stephen J Begg, MPH, Epidemiologist, Policy Development and Planning Division. Reprints will not be available from the authors. Correspondence: Dr E T Vos, Policy Development and Planning Division, Department of Human Services, 555 Collins Street, Melbourne, VIC 3000. Theo. VosATdhs.vic.gov.au 1: Australian burden of disease studies - methodology The disability-adjusted life year (DALY) For each disease or health condition, DALYs are calculated as the sum of the years of life lost due to premature mortality (YLL) in the population and the years lost due to disability (YLD) for incident cases of the health condition. The Australian studies depart from the methods used for the Global Burden of Disease Study (GBD) in the following key areas: YLL are calculated using Australian projected life expectancies; age weights are not used; adjustments for the effects of comorbidity are included;1 and disability weights developed recently by Stouthard et al are used.9 The DALY measures the future stream of healthy years of life lost due to each incident case of disease or injury. It is thus an incidence-based measure rather than a prevalence-based measure. Both the GBD and the Australian studies apply a 3% time discount rate to years of life lost in the future to estimate the net present value of years of life lost. Undiscounted DALYs were also calculated, but are not presented here. Years of life lost due to mortality (YLL) Adjustments were made to deaths registration data to correct for problems relating to coding of causes of death. Deaths registered in 1996 in which the underlying cause fell into various "ill-defined" and "unknown" categories were proportionately redistributed across other relevant causes of death.1 The YLL for a given age is determined by the average cohort (projected) life expectancy for that age, discounting future years by 3% per annum. Unlike most potential years of life lost (PYLL) measures, YLLs do not exclude deaths above a certain age, and do not give a zero value to years of life lost above that age level. Years lost due to disability (YLD) YLD estimates are made for a comprehensive set of 176 disease and injury categories involving analysis of 1260 disease stages, severity levels and sequelae. For some conditions, numbers of incident cases are available directly from disease registers or epidemiological studies, but for most conditions only prevalence data are available. In these cases, a software program, DISMOD, is used to model incidence and duration from estimates of prevalence, remission, case fatality and background mortality.10 For each incident case of a particular condition, YLD is calculated as the product of the average duration of the condition (to remission or death) and a severity weight that quantifies the equivalent loss of healthy years of life due to living with the health condition. YLD calculations draw upon many different sources of data, including population health data collections, health service data, epidemiological studies (Australian and international), and expert knowledge and judgement. For most disease and injury groups, Australian experts were consulted during the development and revision of YLD estimates. The disability weights used in DALY calculations represent societal preferences for different health states, ranging from 0 - "good or ideal health" (preferred to all other states) to 1 - "equivalent to being dead". The Australian study uses actual or derived weights from the GBD and from Stouthard et al.9 These studies used similar methods to measure average societal preferences for health states. Burden attributable to risk factors Attributable fractions were calculated for risk factors using the methods outlined by English et al,11 based on best estimates of the prevalence of exposure to each risk factor, together with best available evidence of the relative risks of diseases or injury. Projections In the Victorian study, projections of the mortality burden were made by extrapolating age-, sex- and cause-specific mortality trends from 1979 to 1996 by log-linear Poisson regression methods taking expected changes in population size into account, and constraining the total number of deaths to age- and sex-specific projections of all-cause mortality. Lacking time series of the incidence of disease, we assumed that trends in mortality are accompanied by equal changes in disease incidence, with the exception of cardiovascular diseases, for which there is evidence that decreases in case-fatality rates and lower incidence of disease have contributed equally to the drop in mortality.12 A potential source of inaccuracy is the lack of trend data in the occurrence of non-fatal conditions, which forced us to assume a stable incidence. 2: Top 10 causes of the mortality, disability and total disease burden in Australia, 1996 A: Mortality burden: percentage of total years of life lost due to mortality (YLL), by sexPersons (1348233 YLL) Percentage of total Males (752591 YLL) Percentage of total Females (595642 YLL) Percentage of total 1 Ischaemic heart disease 2 Stroke 3 Lung cancer 4 Suicide 5 Colorectal cancer 6 COPD* 7 Road traffic accidents 8 Breast cancer 9 Diabetes mellitus? 10 Dementia 20.5 8.3 6.3 5.2 4.4 4.0 3.3 2.8 2.1 1.8 1 Ischaemic heart disease 2 Lung cancer 3 Suicide 4 Stroke 5 Road traffic accidents 6 COPD* 7 Colorectal cancer 8 Prostate cancer 9 Diabetes mellitus? 10 Cirrhosis 21.0 7.3 5.9 5.6 4.5 4.2 3.9 3.0 2.1 1.7 1 Ischaemic heart disease 2 Stroke 3 Breast cancer 4 Lung cancer 5 Colorectal cancer 6 COPD* 7 Dementia 8 Diabetes mellitus? 9 Road traffic accidents 10 Ovary cancer 19.7 9.5 6.8 4.7 4.4 3.9 2.6 2.5 2.1 2.0 B: Disability burden: percentage of total years lost due to disability (YLD), by sex Persons (1162041 YLD) Percentage of total Males (578720 YLD) Percentage of total Females (583321 YLD) Percentage of total 1 Depression 2 Dementia 3 Asthma 4 Osteoarthritis 5 Adult-onset hearing loss 6 Diabetes mellitus? 7 Alcohol dependence/abuse 8 COPD* 9 Stroke 10 Ischaemic heart disease 8.0 5.6 4.8 4.8 4.1 3.8 3.5 3.3 3.3 3.1 1 Depression 2 Adult-onset hearing loss 3 Alcohol dependence/abuse 4 Dementia 5 Asthma 6 COPD* 7 Diabetes mellitus? 8 Stroke 9 Osteoarthritis 10 Ischaemic heart disease 6.2 5.7 4.9 4.4 4.3 4.2 4.1 3.9 3.9 3.9 1 Depression 2 Dementia 3 Osteoarthritis 4 Asthma 5 Generalised anxiety disorder 6 Diabetes mellitus? 7 Vision disorders 8 Stroke 9 Adult-onset hearing loss 10 COPD* 9.8 6.8 5.7 5.3 3.5 3.5 2.9 2.7 2.6 2.5 C: Total burden of disease and injury: percentage of total disability-adjusted life years (DALYs), by sex Persons (2510274 DALYs) Percentage of total Males (1331311 DALYs) Percentage of total Females (1178963 DALYs) Percentage of total 1 Ischaemic heart disease 2 Stroke 3 COPD* 4 Depression 5 Lung cancer 6 Dementia 7 Diabetes mellitus? 8 Colorectal cancer 9 Asthma 10 Osteoarthritis 12.4 5.4 3.7 3.7 3.6 3.5 3.0 2.7 2.6 2.2 1 Ischaemic heart disease 2 Stroke 3 Lung cancer 4 COPD* 5 Suicide 6 Road traffic accidents 7 Diabetes mellitus? 8 Depression 9 Colorectal cancer 10 Dementia 13.6 4.8 4.5 4.2 3.3 3.0 3.0 2.7 2.7 2.5 1 Ischaemic heart disease 2 Stroke 3 Depression 4 Dementia 5 Breast cancer 6 COPD* 7 Asthma 8 Diabetes mellitus? 9 Osteoarthritis 10 Colorectal cancer 11.1 6.1 4.8 4.7 4.6 3.2 3.1 3.0 2.9 2.7 *Chronic obstructive pulmonary disease (chronic bronchitis and emphysema). ?Includes type 1 and type 2 diabetes. Total YLL, YLD and DALYs are in parentheses. 6: Age-standardised disability-adjusted life year (DALY) rates per 1000 in Victoria in 1996 and projected to 2016, by selected causes and sex Males Females 1996 2016 Difference as percentage of 1996 rate 1996 2016 Difference as percentage of 1996 rate All causes All-cause YLL All-cause YLD 144.0 80.1 63.8 108.3 51.8 56.5 -25% -35% -12% 128.6 64.8 63.8 107.0 45.6 61.3 -17% -30% -4% Cardiovascular diseases Ischaemic heart disease Stroke Peripheral vascular disease Inflammatory heart disease 31.7 18.9 6.8 1.2 1.5 14.4 7.7 2.9 0.6 1.8 -54% -59% -58% -54% 26% 27.2 13.7 7.8 1.3 0.8 13.5 7.0 3.2 0.5 1.2 -50% -49% -59% -61% 53% Cancer Lung cancer Breast cancer Bowel cancer Prostate cancer Stomach cancer Leukaemia Melanoma 29.3 6.8 4.4 3.9 1.3 1.2 1.2 24.6 4.2 3.2 4.8 0.6 0.9 1.7 -16% -38% -26% 22% -58% -26% 41% 26.5 3.4 6.9 3.8 0.9 0.9 0.7 23.3 4.3 6.0 2.6 0.4 0.7 0.6 -12% 25% -12% -32% -53% -14% -20% Mental disorders Depression Alcohol abuse/dependence Heroin abuse/dependence 18.1 4.1 3.3 2.4 21.5 4.1 3.1 6.1 19% -- -8% 152% 18.0 6.1 1.5 0.9 17.8 6.1 1.4 0.86 -1% - -8% -3% Neurological/sense organ disorders Dementia 12.7 3.8 12.5 3.9 -2% 4% 14.8 6.3 19.0 10.6 28% 68% Chronic respiratory diseases COPD Asthma 10.6 6.3 3.3 5.9 2.6 2.8 -44% -58% -15% 9.2 4.3 3.8 7.8 4.1 3.2 -15% -5% -16% Injuries Suicide Road traffic accidents Falls 13.8 3.9 3.9 1.3 6.9 3.0 1.0 0.5 -50% -24% -74% -65% 5.6 1.2 1.5 1.1 3.0 0.8 0.6 0.7 -47% -29% -61% -42% Diabetes mellitus 4.7 5.8 24% 4.1 3.6 -13% Musculoskeletal diseases 3.8 3.8 -- 6.7 6.7 -- YLL=years of life lost due to mortality. YLD=years of life lost to disability. COPD=chronic obstructive pulmonary disease. 7: The burden of disease attributable to 10 major risk factors, Australia 1996 Percentage of total DALYs Persons Males Females Tobacco Physical inactivity High blood pressure Alcohol harm Alcohol benefit Obesity Lack of fruit and vegetables High blood cholesterol level Illicit drugs Occupation Unsafe sex 9.7 6.7 5.4 4.9 -2.8 4.3 2.7 2.6 1.8 1.7 0.9 12.1 6.0 5.1 6.6 -2.4 4.3 3.0 3.2 2.2 2.4 1.1 6.8 7.5 5.8 3.1 -3.2 4.3 2.4 1.9 1.3 1.0 0.7 DALY=Disability-adjusted life year.

Colin D Mathers · Chris E Stevenson · Stephen J Begg

In December 1998, a 37-year-old Queensland woman died from a rabies-like illness, 27 months after being bitten by a flying fox (fruit bat). Molecular techniques enabled diagnosis of infection with Australian bat lyssavirus (ABL), the second human case to be recognised and the first to be acquired from a flying fox. It must be assumed that any bat in Australia could transmit ABL; anyone bitten or scratched by a bat should immediately wash the wounds thoroughly with soap and water and promptly seek medical advice. The Australian bat lyssavirus (ABL) was first recognised in June 1996.1 It was subsequently shown not only to belong to a new genotype within the Lyssavirus genus, but also to be more closely related to classic rabies virus than any of the other five genotypes of lyssavirus.2 The first recognised human infection with ABL was in November 19961,3 (Box 1). The patient died from a rabies-like illness 20 days after first becoming unwell. She had apparently been bitten by a yellow-bellied sheathtail bat (Saccolaimus flaviventris; an insectivorous bat) about 4.5 weeks before onset of the illness (R Taylor, Public Health Physician, Rockhampton, QLD, personal communication). We describe here the features of the second recognised human infection with ABL, which had a much longer incubation period and was transmitted by a flying fox (fruit bat; Pteropus sp.). Clinical record A 37-year-old woman was admitted to Mackay Base Hospital in late November 1998 with a five-day history of fever, vomiting, anorexia, pain about the left shoulder girdle, paraesthesiae about the dorsum of her left hand and sore throat with difficulty swallowing. On examination she was acutely ill but well oriented. She was unable to fully open her mouth, was drooling saliva and had difficulty speaking. She was febrile (38 degrees C) but normotensive. Muscle tone was increased, and examination occasionally provoked painful spasms. Examination of the throat provoked spasmodic attempts to swallow. Apart from neutrophilia (12.0 x 109/L; reference range, 2.0-8.0 x 109/L), routine haematological and biochemical tests gave normal results. Twelve hours later her condition had deteriorated considerably, with increased agitation, dysphagia and dysphonia, and the muscular spasms had become more frequent and severe. She was paralysed and ventilated. At about this time, a history of a bat bite was elicited, and a diagnosis of ABL infection was considered. Cerebrospinal fluid (CSF), serum and saliva were submitted for testing. An attempt on Day 2 of hospitalisation to cease artificial ventilation was unsuccessful; when the sedative dose was reduced, she was no longer able to communicate and did not appear to understand verbal commands. Thereafter, she remained ventilator-dependent; whenever the dose of muscle relaxants was reduced, purposeless movements, such as facial grimacing and rolling eye movements, and muscular spasms, such as arching of the back, became evident. Another prominent feature of the illness was marked fluctuations of body temperature and blood pressure. On Day 4 of hospitalisation, the reference laboratory reported that a polymerase chain reaction (PCR) had detected what appeared to be a specific ABL product in the saliva. Nursing and medical staff were informed of the probable diagnosis of ABL infection, and appropriate precautions were implemented.4 The diagnosis of ABL infection was confirmed four days later. On Day 14 of hospitalisation, the patient ceased spontaneous movements and respiratory effort. Ventilator support was withdrawn; she died 19 days after onset of the illness. Post-exposure treatment (PET) was provided to seven healthcare workers because of possible percutaneous or mucous membrane exposure to the patient's saliva.4,5 The patient had attended an evening barbecue in late August 1996, 27 months before onset of the illness (Box 1). At the function, a flying fox had suddenly landed on the back of a young child. In the course of removing it, the patient was bitten at the base of her fifth left finger. Two days later, she presented to her general practitioner and was given tetanus toxoid and appropriate antibiotics. Six months later, in early March, she returned to the GP, asking about a blood test for the "bat virus". She was advised that she should instead receive PET because of a potential exposure to ABL, but decided against this. As soon as the diagnosis of ABL infection was confirmed, PET was administered to the child and to four other people exposed to the flying fox at the barbecue. Diagnostic studies No antibodies to Japanese encephalitis (JE), Murray Valley encephalitis, Kunjin or rabies viruses were detected by enzyme immunoassay tests of serum and CSF collected on Day 2 of hospitalisation. Attempts were made to culture virus by inoculating serum, CSF and saliva onto monolayers of C636, BHK-21, Vero and mouse neuroblastoma cells, but no viruses were isolated. Serum, CSF and saliva were examined for RNA of JE virus, Hendra virus (formerly known as equine morbillivirus) by in-house reverse transcriptase PCR, and for RNA of ABL by heminested reverse transcriptase PCR.6 No viral RNA was detected in serum or CSF, and neither JE nor Hendra virus RNA was detected in saliva. However, the heminested PCR for ABL in saliva produced amplicons of the expected size6 in both first- and second-round reactions (600 and 586 base pairs, respectively). Nucleotide sequencing showed that the amplicon was a lyssavirus-specific product that differed from any other ABL held at the reference laboratory. Concurrently, products from the amplification were separated electrophoretically, transferred onto a nylon membrane7 and hybridised with a digoxigenin-labelled ABL-specific probe. The probe hybridised with the control virus and with amplicons generated from the saliva, indicating that the amplicons were indeed lyssavirus-specific products. Postmortem studies Light microscopy revealed widespread and severe encephalitis affecting all parts of the brain other than the cerebellum. Inflammation and necrosis were particularly severe in the brainstem and hippocampi, where most neurones had either completely disappeared or were necrotic. There was perivascular cuffing by lymphocytes and diffuse infiltration of grey matter neuropile by microglia and lipid-laden macrophages. Occasional neurones showed neuronophagia. A few cytoplasmic inclusion bodies were seen, particularly in the hypothalamus. Light microscopy also revealed diffuse pancarditis with focal myocyte destruction and infiltration of epicardial nerve branches with mononuclear cells. There was no evidence of viral inclusions in acinar or ductal epithelial cells of the parotid and submandibular glands, but nerve bundles in each were infiltrated by mononuclear cells. Spinal cord, brainstem, cerebellum, midbrain and both cerebral hemispheres were examined by immunofluorescent antibody staining. Intense fluorescence was observed in all impression smears, indicating the widespread presence of ABL (Box 2). RNA extracted from these brain samples, as well as salivary and adrenal glands, was tested for ABL by heminested PCR; all samples were strongly positive. Portions of each tissue section were cultured with mouse neuroblastoma cells. PCR indicated successful virus isolation from brain and spinal cord after the first blind passage (Day 7 after inoculation), while specific immunofluorescent antibody staining of cells, indicating presence of ABL, was evident after the second blind passage (Day 14). Sequencing of the PCR product from the cell cultures confirmed that the isolate was the flying-fox variant of ABL. Public health responses As soon as the patient's diagnosis was announced in the media, requests for PET increased markedly throughout Queensland. Many requests were for exposures that had occurred many months, sometimes years, previously. Five hundred and eighteen courses were requested between December 1998 and February 1999 (inclusive), compared with 24 courses in the same three months the previous year, and 59 courses in the preceding three months (Box 3). Discussion The clinical presentation, duration and course of the patient's illness were virtually indistinguishable from those seen in rabies. A short, non-specific prodrome was followed by inexorable progression through distinct stages, culminating, after a relatively short illness, in coma and death. Pain and paraesthesiae about the site of the bite and signs of autonomic instability, such as hypersalivation and labile blood pressure, are also commonly seen in rabies.8 The histological features in the brain were similar to those seen in rabies, including the pathognomonic Negri-like inclusion bodies.8 Findings similar to those seen in the heart and salivary glands have been described in patients who died of rabies.9 Monoclonal antibody and molecular sequencing studies have shown distinct variants of rabies virus, each associated with a dominant mammalian reservoir.10 The first reported patient with ABL infection was infected with the virus variant associated with yellow-bellied sheathtail bats (A Gould, Senior Principal Research Officer, Australian Animal Health Laboratory, Geelong, Vic, personal communication), and, as expected, our patient was infected with the flying-fox variant. ABL has been found in all four Australian flying fox species, and, to date, all ABL-infected bats have been either unwell or dead at the time of collection.11 An extraordinary feature of the patient's illness was the very long (27 months) incubation period. The usual incubation period for rabies is 20-90 days, and 95% of cases occur within a year of exposure.8 Although rare, prolonged incubation periods have been reported,12,13 but the reason for the prolongation has not been established. Even if the patient had accepted the recommended PET, it is uncertain whether the illness would have been prevented. This is because the guidelines at the time recommended vaccine only, without rabies immunoglobulin, as treatment after a bat exposure more than three months previously.5 Rabies that occurred because PET was either delayed or did not include rabies immunoglobulin has been reported.14 The guidelines were subsequently changed to recommend rabies immunoglobulin for all bat exposures,15 but most public health authorities do not include this immunoglobulin if the potential exposure to ABL was more than 12 months previously. Media reporting of the patient's diagnosis was initially restrained but changed on her death, with some reports becoming blatantly alarmist. This media attention and the inherent concern about rabies contributed to intense public demand for PET from Queensland public health units. Although the cost of this PET was considerable, it was predominantly for "catch-up" treatments for those with historical exposures and therefore probably represents a "one-off" expense. This case reminds that ABL infection, although rare, is lethal. Any bat in Australia must be assumed to have the potential to transmit the virus, and members of the public should therefore avoid handling bats. Anyone either bitten or scratched by a bat should immediately wash the wounds thoroughly with soap and water and promptly seek medical advice, regardless of the site or severity of the exposure. Acknowledgements Many people were involved in the management of the patient and in the public health responses. We wish to thank the nursing staff of the Intensive Care Unit, Mackay Base Hospital, and the Public Health Nurses, in particular Mrs Dorothy Symons, of the Tropical Public Health Unit Network. We also thank Ms Judy Northill and Mr Alan Westacott (Queensland Health Scientific Services) and Mr David Gould (Communicable Diseases Unit, Queensland Health). References

Jeffrey N Hanna · Ian K Carney · Greg A Smith · Joseph E Deverill · John A Botha · Ina L Serafin · Bruce J Harrower · Peter F Fitzpatrick · Jeffrey W Searle

Recommendations from the Australasian Society of Thrombosis and Haemostasis Abstract The anticoagulant effect of warfarin should be kept at an international normalised ratio (INR) of about 2.5 (desirable range, 2.0-3.0), although a higher level may be better in a few clinical conditions. The risk of bleeding increases exponentially with INR and becomes clinically unacceptable once the INR exceeds 5.0. Warfarin therapy should be continued for around six weeks for symptomatic calf vein thrombosis, and for 3-6 months after proximal deep vein thrombosis (DVT) that occurs after surgery or limited medical illness. Therapy for six months or longer could be considered for DVT occurring without an obvious precipitating factor, proven recurrent venous thromboembolism (VTE), or if there are continuing risk factors. Oral anticoagulants prevent ischaemic stroke in atrial fibrillation (AF). Maximum efficacy requires an INR > 2.0, but some benefit remains at an INR of 1.5-1.9. Patients aged over 75 years are at greatest risk of intracranial bleeding during warfarin therapy for AF, and the target INR may be reduced to 2.0-2.5, or perhaps as low as 1.5-2.0, in such patients. Warfarin should be withheld if it is more likely to cause major bleeding than to protect from stroke (eg, in young people with isolated AF where the annual baseline risk of stroke is < 1%). In patients with AF, aspirin is less effective than warfarin (much less effective after such patients have had a stroke or transient cerebral ischaemia). In people with prosthetic heart valves, an INR of 2.5-3.5 is probably sufficient for bileaflet or tilting disc valves, but a higher target INR is necessary for caged ball or caged disc valves. The addition of aspirin (100 mg/day) further decreases the risk of embolism but increases the risk of gastrointestinal bleeding. Warfarin is used for preventing and treating venous or arterial thrombosis and embolism. It is a potentially hazardous drug, causing major bleeding in 1%-2% of people treated, and intracranial bleeding in about 0.1%-0.5% during each year of therapy. These risks are well recognised, but strong recent evidence that many otherwise healthy people with atrial fibrillation (AF) or venous thromboembolism (VTE) can benefit from long term warfarin therapy has led to a major increase in its use. These consensus guidelines offer advice on the selection of patients for warfarin therapy and management of such patients. The recommendations draw on proceedings of the Fifth American College of Chest Physicians Consensus Conference on Antithrombotic Therapy,1 and are consistent with the most recent Guidelines on oral anticoagulation developed for the British Society for Haematology.2 Warfarin therapy and management of complications The INR The INR (international normalised ratio) is a good indicator of effectiveness and risk of bleeding during warfarin therapy and is best kept at about 2.5, with a target range of 2.0-3.0, for most clinical indications, although higher levels may be better for certain patients (Box 1). The lower limit of this target range recognises a threshold level for effectiveness, while the upper limit is set to minimise bleeding. Starting and maintaining warfarin therapy The daily maintenance dose of warfarin differs greatly between individuals, commonly between 0.5 mg/day and 15 mg/day, and often fluctuates over time. The average maintenance dose is about 4.5 mg/day, although this is lower in the elderly. The drug is rapidly and completely absorbed and immediately blocks further hepatic synthesis of the functional vitamin K-dependent haemostasis factors (II, VII, IX, X, protein C, protein S). However, its impact on the INR is delayed until preformed coagulation factors are removed, so dose adjustment must allow for these delayed effects. The plasma half-life of warfarin is about 36 hours.3 In the past, it was customary to use a loading dose of 10 mg. However, for most situations, a reduced starting dose of 5 mg per day will achieve an INR of 2.0 in four to five days.4 INR is measured daily or every second day during the first week of treatment, with the dose of warfarin (taken in the evening) titrated against the morning's INR. It is then measured at increasing intervals depending on response. Many patients, once the dose is stable, can be well controlled with 4-6-weekly testing and dose adjustment, but others need more frequent assessment. An empirical approach to warfarin dosing can be smooth and effective but published dose-adjustment tables can help.2 Old age, reduced body weight, and impaired cardiac or liver function all predict a smaller than average dose requirement. Multiple comorbidities and a need for many drugs increase the risk of an unstable anticoagulant response. The effect of warfarin is subject to multiple interactions. These include the dietary content or extent of absorption of vitamin K, the absorption of warfarin and its effect on the liver (which are increased or decreased by many other drugs), and the clearance of blood-clotting factors.1,3 Intercurrent illness, starting or stopping therapy with other drugs (especially antibiotics and amiodarone) and changes in diet or bowel function can all influence the INR. Rechecking the INR within a few days of any change in medication or clinical condition is prudent. Bleeding is minimised by regular monitoring to avoid an excessive INR and by educating patients about how warfarin works, why their dose requirement may change, and the likely settings and symptoms of bleeding complications. Successful warfarin therapy requires a partnership with patients, who should be encouraged to have their INR checked soon after any change in their normal routine. Clinics should periodically audit their results with warfarin therapy and review exceptional cases. Between 50% and 75% of INRs are likely to fall into their designated therapeutic range.5 Two recent Australian case reports are reminders that bioequivalence has not been formally demonstrated for Coumadin and Marevan (both from Boots Healthcare Australia, Sydney, NSW), the two locally available formulations of warfarin.6 Warfarin and bleeding Major bleeding has been reported in 1.1%-8.1% of patients during each year of long-term warfarin therapy (1.1%-2.7% by anticoagulant clinics managing patients with prosthetic heart valves,7-9 1.3% in atrial fibrillation trials, and 2.8%-8.1% after a stroke or transient ischaemic attack10-12). Risk factors include old age, serious illness (cerebral, cardiac, kidney or liver disease), cerebrovascular or peripheral vascular disease, and an unstable anticoagulant effect. Forgetfulness, non-steroidal anti-inflammatory drugs and alcohol abuse may also contribute.7,9,13-15 Warfarin appears to be especially hazardous after a transient ischaemic attack or minor stroke; in one trial, 14 months of warfarin therapy with a relatively high target INR of 3.0-4.5 increased major bleeding from 0.9% to 8.1%, intracranial bleeding from 0.5% to 4.1%, and fatal intracranial bleeding from 0.2% to 2.6% (relative to low dose aspirin therapy).12 Bleeding is most likely during the first three months of treatment, and often follows trauma or unmasks a previously unsuspected comorbidity.8,13-15Age alone is not a contraindication to warfarin therapy. Although one report showed that each decade above the age of 40 raised the risk of major bleeding by almost 50%, with a maximum effect above 70 years,7 others have found that age below 70 years has no influence.8,15 The INR is the dominant determinant, whether bleeding is expressed as the absolute risk per annum (Box 2) or as relative risk. In a 1996 study, the bleeding rate was doubled as the INR increased from 2.0-2.9 to 3.0-4.4, quadrupled between 4.5-6.0, and was multiplied by five when the INR was above 7.0.15 There is a consistent increase in major bleeding (including intracranial bleeding16) when the INR exceeds 4.0-5.5.7,11,14 A 1997 trial found that each increase in INR by 0.5 multiplied the risk of major bleeding (mostly intracranial) by 1.43.12 Managing an excessively prolonged INR or bleeding caused by warfarin therapy An INR above 5.0 requires close monitoring and often needs intervention, as determined by the level of the INR and the presence or absence of bleeding (Box 3). The INR often remains elevated for some days, even if warfarin is withheld, but small amounts of vitamin K1 quickly correct the INR to safer levels. In most patients, 1-2.5 mg of oral vitamin K1 reduces the INR from 5.0-9.0 to 2.0-5.0 within 24-48 hours; this intervention is usually sufficient in the absence of bleeding.17,18 These small doses are obtained by withdrawing the desired amount from a 10 mg vial of injectable vitamin K1 and giving this orally or parenterally. When the INR is > 9.0, then 5 mg vitamin K1 may be more appropriate and can be given orally, subcutaneously or intravenously (very rarely, the last may cause a serious anaphylactoid reaction). In people with a massive accidental or self-inflicted warfarin overdose, the long half-life of warfarin means that the INR may rebound over several days as the effects of vitamin K1 wear off. In any case, the response to vitamin K1 needs to be monitored. Bleeding caused by a warfarin overdose is controlled with clotting factor replacement (Box 3), and this may also be indicated in the absence of bleeding when the risk is very high.19 Bleeding or an unstable dose-response should trigger a review of the need for warfarin. Continued treatment will require closer monitoring of the INR, both to detect the transient warfarin resistance caused by too much vitamin K1, and to avoid further overanticoagulation. Heparin treatment may be required to cover a prolonged period of warfarin resistance. Interrupting warfarin therapy for surgery When there is a need for surgery, the risk of perioperative bleeding under continued warfarin therapy must be balanced against the risk of thromboembolism if warfarin therapy is stopped.20Most surgery, including hip or knee replacement and many thoracic or abdominal operations, can proceed under continued warfarin cover without undue bleeding (provided the INR during and soon after surgery is about 1.5-2.0). Warfarin therapy is a contraindication for regional anaesthesia (eg, spinal, epidural, brachial blocks) and is unacceptable where even minor bleeding might cause critical damage (as in neurosurgery and some plastic surgery). It is also unpopular with most surgeons. However, the absolute daily risk of a serious thromboembolic event is small in most people with AF, previous systemic embolism or a prosthetic heart valve (the hazard is greatest from mitral and older-model prosthetic valves, and in patients with more than one prosthetic valve). Thus, it is safe to stop warfarin therapy for several days before and after surgery in such patients. High-dose heparin cover for these indications is rarely indicated as the risk of bleeding is usually prohibitive.20 The risk of recurrence is greatest during the first four weeks after VTE, so warfarin therapy should not be interrupted during this time if at all possible. If anticoagulants must be stopped for surgery soon after VTE, a vena cava filter can be placed to minimise the risk of life-threatening pulmonary embolism. Specific indications for warfarin therapy Deep vein thrombosis and pulmonary embolism Prevention: Heparins are now usually the preferred drugs for the prevention of perioperative VTE, but warfarin retains a limited role when the risk of thrombosis is very high. Its main role is in long-term therapy. Warfarin is no less effective than low molecular weight heparins after hip or knee replacement, and the risk of bleeding is similar or lower when therapy is started at about the time of surgery and continued at least until patients are fully mobile.21 Treatment: Anticoagulants prevent early thrombus extension and embolism and minimise late recurrence. Heparin treatment can be stopped after a minimum of five days when warfarin therapy is also being given, provided that the two drugs are overlapped for at least four days and the INR has exceeded 2.0 for two or more days.22 Increasingly, deep vein thrombosis (DVT) is now managed at home -- an approach preferred by many patients and made possible by trials which found that initial treatment with low molecular weight heparins given in a fixed dose by subcutaneous injection is no less effective or safe after DVT than standard heparin therapy. Home heparin therapy requires close monitoring to ensure compliance and a safe and effective start for warfarin therapy.23,24 Although warfarin is now usually given for 3-6 months after VTE, there is growing evidence that the optimal duration of treatment is determined by the patient's clinical presentation. Six to 12 weeks of warfarin therapy is probably enough when DVT follows surgery or transient immobilisation ("secondary" DVT), as recurrence is minimised by six weeks of treatment after symptomatic calf vein DVT,7 and by three months of treatment after proximal DVT.25,26 However, warfarin therapy for longer than six months may be required after "idiopathic" DVT, recurrent VTE, or when there is a continuing cause like cancer or an inherited or acquired "hypercoagulable" state.27-30 Whether, in these circumstances, warfarin should be given for 12 months, two years, or longer, remains under active investigation. For individuals, the choice will also be influenced greatly by risk of bleeding. Controversies in the management of DVT and VTE Calf vein thrombosis: Although calf vein DVT poses little immediate threat and is commonly believed to be clinically unimportant, it has the potential to extend and embolise. In a randomised comparison where 51 patients with symptomatic calf DVT were treated for five days with heparin only or with heparin followed by ongoing warfarin therapy, there was a recurrence during the next three months in eight of 28 patients from the first group (23%: seven clinically suspected and confirmed; five with proximal extension and one with embolism), but none in the second.31 Therefore, patients with calf vein thrombosis should be treated with warfarin unless there are contraindications. Accuracy of diagnostic tests for DVT: Venous ultrasonography has now replaced venography as the first-line diagnostic test for clinically suspected DVT. Despite its limited sensitivity to small calf vein DVTs, a negative ultrasound result almost excludes thrombosis when there is a low pretest clinical probability for DVT (a DVT score of zero on a checklist of clinical features obtained before ultrasonography, such as active cancer, immobilisation, major surgery, entire leg swelling, localised tenderness, calf swelling, pitting oedema and collateral superficial veins).32 However, for patients in whom the pretest clinical probability is moderate (DVT score of 1-2) or high (score, > 3), a negative ultrasound result does not exclude a small DVT, and they should have either early venography or further ultrasonography once or twice within the next seven days in case there is proximal extension of an undetected calf thrombus. This approach is validated by extensive clinical follow-up.33 Recurrent or idiopathic DVT or VTE: In a randomised trial of patients presenting with recurrent DVT, oral anticoagulant therapy for six months was followed by a recurrence in 21% during four years of follow-up, compared with 3% when treatment was continued. However, ongoing warfarin therapy increased the rate of major bleeding during the four years from 2.7% to 8.6%, while mortality remained unchanged.27 Similarly, in a separate trial of management after a first "idiopathic" VTE, warfarin therapy for three months was followed by recurrence in 16 of 77 patients during 10 months of follow-up, compared with only one of 76 patients in whom warfarin therapy was continued.30 However, the use of warfarin increased the annual risk of major bleeding from zero to 4%.30 These high rates of bleeding reinforce the need for careful risk assessment when considering patients for long term anticoagulant therapy after VTE. The results of these trials suggest that warfarin therapy should be continued for one year after an "idiopathic" or recurrent VTE if the risk of bleeding is acceptable, and that treatment should be extended to two years if warfarin control is straightforward and the bleeding risk remains low. Atrial fibrillation Warfarin is now widely used to prevent systemic embolism in otherwise healthy patients with atrial fibrillation (AF). In clinical trials, warfarin consistently reduced the annual risk of a first ischaemic stroke (including stroke with a residual functional deficit) by almost 70% (from 7% to 3% per annum) and mortality by 33%, at the cost of a small increase in serious bleeding (from 1.0% to 1.3% per annum).10,34,35 The prevalence of AF rises from about 3% at 65 years to more than 10% by 85 years, and AF accounts for about 1.5% of all strokes in people aged 50-59 years, and almost 25% of strokes in people aged 80-89 years. Age is therefore an important determinant of ischaemic stroke in AF (the relative risk [RR] of stroke in AF rises by 1.4 with each decade).35 Previous stroke or transient ischaemic attack (RR, 2.5), diabetes (RR, 1.7), and treated hypertension (RR, 1.6) also contribute, as do heart failure, ischaemic heart disease, a large left atrium, and left ventricular dysfunction.10 Stroke is unlikely in isolated AF but becomes more likely as additional risk factors accumulate (Box 4). This makes warfarin therapy inappropriate for young people with AF alone and no other cardiac risk factor (isolated AF), as their annual risk of stroke (< 1%) is low enough to ensure that risk of bleeding always equals or exceeds any likelihood of gain. Because of the risk of bleeding, these reports raise important questions about the best target level of INR, and about which patients with AF should be offered long-term warfarin therapy. The incidence of stroke is minimised by an INR > 2.0 and increases exponentially below this level, but some benefit remains while the INR is 1.5-1.9. When considering warfarin therapy for AF, each candidate requires a formal estimate of the relative risks of stroke (Box 4) and bleeding (Box 2). Controversies about the use of warfarin or aspirin to prevent stroke in atrial fibrillation Stroke and the INR: The risk of stroke during warfarin therapy for AF is dictated by the INR. Below 2.0, the relative risk doubles at 1.7, triples at 1.5, sextuples at 1.3, and reaches 18 times once the INR is normal, but nothing is gained by increasing the INR beyond its therapeutic threshold of 2.0.36 Results were similar when warfarin was given for secondary stroke prevention in patients with AF who had already developed a stroke or transient cerebral ischaemia.11 Again, in a randomised trial in which patients with AF plus at least one other risk factor for stroke were given either warfarin in a dose to prolong their INR (INR, 2.0-3.0; median, 2.4) or aspirin combined with a low dose of warfarin (0.5-3.0 mg/day; INR, 1.2-1.5; median, 1.3), the dose aiming for the higher INR was clearly superior.37 Aspirin or warfarin for AF? The 30% risk reduction in stroke from aspirin treatment is well below the 70% achieved with warfarin therapy.10 In a blinded analysis of clinical outcomes when the two drugs were compared, warfarin was better at preventing cardioembolic strokes and strokes of uncertain cause.38 This is consistent with the small effect observed with aspirin for secondary stroke prevention in patients with AF and who have had a stroke or TIA -- warfarin reduced the risk of recurrence by 62%, compared with only 16% for aspirin.39 It may be a useful compromise to reserve aspirin for patients with uncomplicated AF whose baseline risk of embolism is low. Warfarin, INR and aspirin in elderly patients with AF: Age above 75 years and a high INR both increase the hazard from intracranial and other major bleeding during warfarin therapy. Because there is some residual benefit at an INR of 1.5-1.9, this reduced target range may offer an acceptable exchange of safety for benefit in some elderly patients. Where the risk of bleeding is high, aspirin is less effective, but safer than warfarin. Cardioembolic stroke prevention in conditions other than AF There is evidence that cardioversion to correct a recent cardiac arrhythmia should be delayed until after three weeks of anticoagulant cover to prevent systemic embolism.35 Warfarin prevents embolic stroke and other arterial embolism, as well as VTE, after myocardial infarction (MI), and is often given for 3-6 months when MI is followed by intraventricular thrombus formation (risk factors include transmural anterior infarction and ventricular dysfunction).40 A good case also exists for long term warfarin therapy in some patients with ongoing left ventricular dysfunction.41,42Prosthetic heart valves Improved design has greatly reduced the thrombogenicity of mechanical prosthetic heart valves, but the need for effective, lifelong warfarin therapy remains because systemic embolism is still the main source of late mortality and morbidity. The risk is determined by the type of valve and its position (higher for mitral than aortic valves, greatest when both are replaced). Tissue valves, by contrast, are almost free of thromboembolic complications, except during the first three months.43The American College of Chest Physicians recommends an INR of 2.0-3.0 for recent-model bileaflet or tilting disc valves, and 2.5-3.5 for older and more thrombogenic valves that have a caged ball or disc; patients with a newly placed bioprosthetic (tissue) valve require three months of warfarin and an INR of 2.0-3.0.43 However, in our view, because the evidence is incomplete, it remains prudent to retain a target range of 2.5-3.5 for most ("low-risk") prosthetic valves while aiming higher (3.0-4.5) for older and more thrombogenic models, provided there is no contraindication (Box 1). This view is consistent with recent recommendations from the British Society for Haematology.2 Antiplatelet drugs alone are ineffective, but combining dipyridamole or aspirin (100 mg/day) with warfarin reduces the risk of systemic embolism. Meta-analysis suggests that the penalty for adding aspirin is a 2.5-times increase in major gastrointestinal bleeding,44 so the combination is perhaps best avoided, except in patients considered to be at unusually high risk of systemic thromboembolism (more than one mechanical valve, previous embolism, associated AF).43 Special circumstances for anticoagulation Antiphospholipid antibody syndrome and factor V Leiden: Two retrospective surveys of clinical outcomes in patients with antiphospholipid antibody syndrome and venous and/or arterial thrombosis suggest that warfarin therapy fails to prevent recurrent thromboses unless the INR is prolonged above 3.0.45,46 This contrasts with a more recent report of few recurrences while the INR was 2.0-3.5.47 Without better information, and until randomised trials are complete, it is not possible to make a firm recommendation about the optimal target range for this condition. The effect of aspirin alone in preventing thrombosis in the antiphospholipid antibody syndrome is unclear.45-47 There is no current evidence to suggest that patients with factor V Leiden-heterozygous abnormality should require more intense anticoagulation. It is still uncertain whether the duration of therapy should be increased in these patients, as evidence from reports about the risk of recurrent VTE is conflicting.28,29,48 Oral anticoagulants in pregnancy: Oral anticoagulants cross the placenta and should be avoided throughout pregnancy, especially during the first and third trimesters.49 Treatment at 6-12 weeks' gestation causes calcified epiphyses (chondrodysplasia punctata) and a characteristic nasal hypoplasia in offspring,50 while later exposure is associated with central nervous system abnormalities, including microcephaly.51 In one report, almost 30% of children (10 of 35) born to mothers with a prosthetic heart valve were malformed if acenocoumarol was taken through 6-12 weeks' gestation, but none of 19 developed a malformation when this drug was replaced with heparin before the sixth week.52 Continuing warfarin therapy until term also exposes infants to the risk of intracranial and other major bleeding during birth. Heparins do not cross the placenta and do not cause these problems.53-55 It is safe to breastfeed during warfarin therapy as there is minimal excretion into breast milk.56 References Hirsh J, Dalen JE, Anderson D, et al. Oral Anticoagulants. Mechanism of action, clinical effectiveness and optimal therapeutic range. Chest 1998; 114 Suppl: 445S-469S. Walker ID, Machin S, Baglin TP, et al. Guidelines on oral anticoagulation. 3rd ed. Br J Haematol 1998; 101: 374-387. Holbrook AM, Wells PS, Crowther NR. Pharmacokinetics and drug interactions with warfarin. In: Poller L, Hirsh J, editors. Oral anticoagulants. Sydney: Arnold, 1996: 30-48. Crowther MA, Ginsberg JB, Kearon C, et al. A randomized trial comparing 5 mg and 10 mg warfarin loading doses. Arch Intern Med 1999; 159: 46-48. Rose P. Audit of anticoagulant therapy. J Clin Pathol 1996; 49: 5-9. Coumadin and Marevan are not interchangeable. Aust Adverse Drug React (ADRAC) Bull 1999; 18: 6. van der Meer FJM, Rosendaal FR, Vandenbroucke JP, Briet E. Bleeding complications in oral anticoagulant therapy: an analysis of risk factors. Arch Intern Med 1993; 153: 1557-1562. Cannegieter SC, Rosendaal FR, Wintzen AR, et al. Optimal oral anticoagulant therapy in patients with mechanical heart valves. N Engl J Med 1995; 333: 11-17. Levine M, Raskob GE, Landefeld S, Kearon C. Hemorrhagic complications of anticoagulant treatment. Chest 1998; 114 Suppl: 511S-523S. Laupacis A, Boysen G, Connolly S, et al. Risk factors for stroke and efficacy of antithrombotic therapy in atrial fibrillation. Analysis of pooled data from five randomized controlled trials. Arch Intern Med 1994; 154: 1449-1457. The European Atrial Fibrillation Trial Study Group. Optimal oral anticoagulant therapy in patients with nonrheumatic atrial fibrillation and recent cerebral ischemia. N Engl J Med 1995; 333: 5-10. The Stroke Prevention in Reversible Ischemia Trial (SPIRIT) Study Group. A randomized trial of anticoagulants versus aspirin after cerebral ischemia of presumed arterial origin. Ann Neurol 1997; 42: 857-865. Landefeld S, Beyth RJ. Anticoagulant-related bleeding: clinical epidemiology, prediction and prevention. Am J Med 1993; 95: 315-328. Fihn SD, McDonnell M, Martin D, et al. Risk factors for complications of chronic anticoagulation. A multicenter study. Ann Intern Med 1993; 118: 511-520. Palareti G, Leali N, Coccheri S, et al. Bleeding complications of oral anticoagulant treatment: an inception-cohort, prospective collaborative study (ISCOAT). Lancet 1996; 348: 423-428. Hylek EM, Singer D. Risk factors for intracranial hemorrhage in outpatients taking warfarin. Ann Intern Med 1994; 120: 897-902. Weibert RT, Le DT, Kayser SR, Rapaport SI. Correction of excessive anticoagulation with low-dose oral vitamin K1. Ann Intern Med 1997; 125: 959-962. Crowther M, Donovan D, Harrison L, et al. Low-dose oral vitamin K reliably reverses over-anticoagulation due to warfarin. Thromb Haemost 1998; 79: 1116-1118. Makris M, Greaves M, Philips W, et al. Emergency oral anticoagulant reversal: the relative efficacy of infusions of fresh frozen plasma and clotting factor concentrate on correction of the coagulopathy. Thromb Haemost 1996; 77: 477-480. Kearon C, Hirsh J. Management of anticoagulation before and after elective surgery. N Engl J Med 1997; 336: 1506-1511. Clagett GP, Anderson FA, Geerts WH, et al. Prevention of venous thromboembolism. Chest 1998; 114 Suppl: 531S-560S. Hyers TM, Agnelli G, Hull RD, et al. Antithrombotic therapy for venous thromboembolic disease. Chest 1998; 114 Suppl: 561S-578S. Koopman MMW, Prandoni P, Piovella F, et al. Treatment of venous thrombosis with intravenous unfractionated heparin administered in the hospital as compared with subcutaneous low-molecular-weight heparin administered at home. N Engl J Med 1996; 334: 682-687. Levine M, Gent M, Hirsh J, et al. A comparison of low-molecular-weight heparin administered primarily at home with unfractionated heparin administered in the hospital for proximal deep-vein thrombosis. N Engl J Med 1996; 334: 677-681. Schulman S, Rhedin A-S, Lindmarker P, et al. Comparison of six weeks with six months of oral anticoagulant therapy after a first episode of venous thromboembolism. N Engl J Med 1995; 332: 1661-1665. Levine MN, Hirsh J, Gent M, et al. Optimal duration of oral anticoagulant therapy: a randomized trial comparing four weeks with three months of warfarin in patients with proximal DVT. Thromb Haemost 1995; 74: 606-611. Schulman S, Granqvist S, Holmstrom M, et al. The duration of oral anticoagulant therapy after a second episode of venous thromboembolism. N Engl J Med 1997; 336: 393-398. van den Belt AGM, Sanson B-J, Simioni P, et al. Recurrence of venous thromboembolism in patients with familial thrombophilia. Arch Intern Med 1997; 157: 2227-2232. Simioni P, Prandoni P, Lensing AWA, et al. The risk of recurrent venous thromboembolism in patients with an Arg506 to Gln mutation in the gene for factor V (Factor V Leiden). N Engl J Med 1997; 336: 399-403. Kearon C, Gent M, Hirsh J, et al. A comparison of three months of anticoagulation with extended anticoagulation for a first episode of idiopathic venous thromboembolism. N Engl J Med 1999; 340: 901-907. Lagerstedt CI, Olsson C-G, Fagher BO, et al. Need for long-term anticoagulant treatment in symptomatic calf-vein thrombosis. Lancet 1985; 2: 515-518. Wells PS, Anderson DR, Bormanis J, et al. Value of assessment of pretest probability of deep-vein thrombosis in clinical management. Lancet 1997; 350: 1795-1798. Heijboer H, Buller HR, Lensing AW, et al. A comparison of real-time compression ultrasonography with impedance plethysmography for the diagnosis of deep-vein thrombosis in symptomatic outpatients. N Engl J Med 1993; 329: 1365-1369. Singer DE. Overview of the randomized trials to prevent stroke in atrial fibrillation. Ann Epidemiol 1993; 3: 563-567. Laupacis A, Albers GW, Dalen JE, et al. Antithrombotic therapy in atrial fibrillation. Chest 1998; 114 Suppl: 579S-589S. Hylek EM, Skates SJ, Sheehan MA, Singer DE. An analysis of the lowest effective intensity of prophylactic anticoagulation for patients with nonrheumatic atrial fibrillation. N Engl J Med 1996; 335: 540-546. Stroke Prevention in Atrial Fibrillation Investigators. Adjusted-dose warfarin versus low-intensity, fixed-dose warfarin plus aspirin for high-risk patients with atrial fibrillation: Stroke Prevention in Atrial Fibrillation III randomised clinical trial. Lancet 1996; 348: 633-638. Miller VT, Pearce LA, Feinberg WM, et al. Differential effect of aspirin versus warfarin on clinical stroke types in patients with atrial fibrillation. Neurology 1996; 46: 238-240. European Atrial Fibrillation Trial Study Group. Secondary prevention in non-rheumatic atrial fibrillation after transient ischaemic attack or minor stroke. Lancet 1993; 342: 1255-1262. Cairns JA, Theroux P, Lewis HD Jr, et al. Antithrombotic agents in coronary artery disease. Chest 1998; 114 Suppl: 611S-633S. Fuster V, Gersh BJ, Giuliani ER, et al. The natural history of idiopathic dilated cardiomyopathy. Am J Cardiol 1981; 47: 525-531. Al-Khadra AS, Salem DN, Rabd WR, et al. Warfarin anticoagulation and survival: a cohort analysis from the studies of left ventricular dysfunction. J Am Coll Cardiol 1998; 31: 749-753. Stein PD, Alpert JS, Dalen JE, et al. Antithrombotic therapy in patients with mechanical and biological prosthetic heart valves. Chest 1998; 114 Suppl: 602S-610S. Cappelleri JC, Fiore LD, Brophy MT, et al. Efficacy and safety of combined anticoagulant and antiplatelet therapy versus anticoagulant monotherapy after mechanical heart-valve replacement: a metaanalysis. Am Heart J 1995; 130: 547-552. Rosove MH, Brewer PM. Antiphospholipid thrombosis: clinical course after the first thrombotic event in 70 patients. Ann Intern Med 1992; 117: 303-308. Khamashta MA, Cuadrado MJ, Mujic F, et al. The management of thrombosis in the antiphospholipid-antibody syndrome. N Engl J Med 1995; 332: 993-997. Krnic-Barrie S, O'Connor CR, Looney SW, et al. A retrospective review of 61 patients with antiphospholipid syndrome: analysis of factors influencing recurrent thrombosis. Arch Intern Med 1997; 157: 2101-2108. Eichinger S, Pabinger I, Stumpflen, et al. The risk of recurrent venous thromboembolism in patients with and without Factor V Leiden. Thromb Haemost 1997; 77: 624-628. Ginsberg J, Barron W. Pregnancy and prosthetic heart valves. Lancet 1994; 344: 1170-1172. Koren G, Pastuszak A, Ito S. Drugs in pregnancy. N Engl J Med 1998; 338: 1128-1137. Hall JG, Pauli RM, Wilson KM. Maternal and fetal sequelae of anticoagulation during pregnancy. Am J Med 1980; 68: 122-140. Iturbe-Alessio I, del Carmen Fonseca M, Mutchinik O, et al. Risks of anticoagulant therapy in pregnant women with artificial heart valves. N Engl J Med 1986; 315: 1390-1393. Ginsberg JS, Kowalchuk G, Hirsh J, et al. Heparin therapy during pregnancy. Risks to the fetus and mother. Arch Intern Med 1989; 149: 2233-2236. Fejgin MD, Lourwood DL. Low molecular weight heparins and their use in obstetrics and gynecology. Obstet Gynecol Surv 1994; 49: 424-431. Sanson B-J, Lensing AWA, Prins MH, et al. Safety of low-molecular-weight heparin in pregnancy: a systematic review. Thromb Haemost 1999; 81: 668-672. Orme ML, Lewis PJ, de Swiet M, et al. May mothers given warfarin breast-feed their infants? BMJ 1977; 1: 1564-1565. Background and evidence basis of recommendations The Australasian Society of Thrombosis and Haemostasis Consensus Guidelines for Warfarin Therapy were written on behalf of the Australasian Society of Thrombosis and Haemostasis (ASTH). The writing committee was commissioned by council and consisted of Associate Professor A S Gallus (Chairman), Dr R I Baker, Professor B H Chong, Dr P A Ockelford and Associate Professor A M Street. The guidelines were developed after extensive consultation with the membership of the ASTH, including several workshops and teleconferences. The draft recommendations were open for comment and discussion at the 1998 annual scientific meeting of the ASTH in Sydney. They draw upon review of all available evidence from published studies and from clinical experience. The aim is to provide an Australian perspective on the evidence to guide all practitioners in the safe and effective use of oral anticoagulants in hospital and the community. We are grateful for the help of Dr K McGrath, Dr M Herzberg (Quality Assurance Program in Haematology, Royal College of Pathologists of Australasia), Dr P Montanaro (Royal Australian College of General Practitioners), Dr P Steele (Australia and New Zealand Cardiac Society) and Professor J Fletcher (International Union of Angiology). Authors' details Australasian Society of Thrombosis and Haemostasis, Perth, WA. Alex S Gallus, FRACP, FRCPA, Chairman; Ross I Baker, FRACP, FRCPA; Beng H Chong, FRACP, FRCPA; Paul A Ockelford, FRACP, FRCPA; Alison M Street, FRACP, FRCPA. Reprints will not be available from the authors. Correspondence: Professor A S Gallus, Director, SouthPath, C/- Flinders Medical Centre, Bedford Park, SA 5042. 1: Range of international normalised ratio (INR) recommended for specific applications of warfarin therapy* Condition INR range Preventing DVT (high risk patients, like those who have had hip replacement) 2.0-3.0 Therapy after DVT or pulmonary embolism 2.0-3.0 Preventing systemic embolism Atrial fibrillation Valvular heart disease After myocardial infarction Tissue heart valves (first 3 months) 2.0-3.0 2.0-3.0 2.0-3.0 2.0-3.0 Bileaflet mechanical heart valve (aortic) 2.5-3.5 Mechanical prosthetic heart valve (high risk) 3.0-4.5 Preventing recurrence of myocardial infarction 3.0-4.5 Thrombosis in antiphospholipid antibody syndrome 3.0-4.5 DVT=deep vein thrombosis . *Based largely on the 5th American College of Chest Physicians Consensus Conference1 and consistent with current recommendations of the British Society for Haematology.2 2: Risk of major bleeding (% per annum) and international normalised ratio (INR) - findings of two studies23,30 INR Study 130 Study 223 < 2.0 2.0-2.9 3.0-3.9 4.0-4.9 5.0-5.9 ≥ 6 3% 2%-3% 2%-3% 4% 5% 5%-13% 0 1% 3% 4% 50% 3: Managing overdose and bleeding during warfarin therapy* Clinical setting Action INR >5.0 but < 9.0 (no bleeding) Stop warfarin, give 1-2.5mg vitamin K1, measure INR in 6-12 hours, restart warfarin at reduced dose once INR is < 5 INR ≥9.0 (no bleeding) Stop warfarin, give 5mg vitamin K1, measure INR in 6-12 hours, restart warfarin at reduced dose once INR is < 5, clotting factor replacement? if high risk of bleeding Major bleeding (any level of INR) Stop warfarin, give 5mg vitamin K1, clotting factor replacement, measure INR as required, assess need to restart warfarin INR=international normalised ratio. *Based on Makris et al, 1996.19 ?Blood products available in Australia for clotting factor replacement after warfarin overdose include fresh frozen plasma and Prothrombinex-HT (CSL Limited), a factor II, IX and X concentrate. 4: Risk of ischaemic stroke in patients with atrial fibrillation (AF), grouped by age and other risk factors* (derived from Laupacis et al10) Risk categories Patients affected per annum Lone atrial fibrillation† Age < 60 years Age 60-69 years Age 70-79 years Age ≥80 years 0 1.6% 2.1% 3.0% Age < 65 years No risk factors One or more risk factors 1.0% 4.9% Age 65-75 years No risk factors One or more risk factors 4.3% 5.7% Age >75 years No risk factors One or more risk factors 3.5% 8.1% *Hypertension, diabetes, previous stroke or transient ischaemic attack. †Atrial fibrillation without transient ischaemic attack or stroke, myocardial infarction, hypertension or heart failure.

Alex S Gallus · Ross I Baker · Beng H Chong · Paul A Ockelford

Compared with funding for other public health issues, funding for tobacco control in Australia is appallingly neglected "Tobacco smoking is the single largest preventable cause of premature death and disease in Australia."1 This is the first line of a document distributed recently with the Ministerial Council on Drug Strategy's National Tobacco Strategy 1999 to 2002-03.2 The World Health Organization, in recognition of the unabating global rise in tobacco-caused deaths -- forecast to reach 10 million a year in the decade 2020-2030 -- has elevated tobacco control to one of its priority projects,3 alongside malaria control and the essential drugs program. So, how is tobacco control faring in Australia? It is popular for Australian governments to be tough on illicit drugs, but of late they have been decidedly weak on tobacco. They have allowed smoking prevalence to languish unchanged in the 1990s, after having successfully driven it down during the preceding three decades.4 A small gain has been that several State governments have announced or have implemented bans on smoking in restaurants, but, relative to government funding of other public health issues, funding for tobacco control is appallingly neglected. Moreover, for some of the other public health issues, such as illicit drug use prevention, little evidence for policy or program effectiveness exists (Box 1). Further progress with tobacco control appears to be shackled by: ideological recoil at regulating a "legitimate product"; concern for short term political industry patronage; and ill-informed views that robust tobacco control will rapidly strangle the goose that lays the annual $5 billion golden egg of tobacco excise. In April this year, in the face of huge criticism, the Liberal Party accepted sponsorship from Philip Morris for a lunch at its annual conference, with the Federal Director, Lynton Crosby, and the Prime Minister, John Howard, shrugging off criticism, claiming that tobacco products are legal,13 just as politicians of 150 years ago might have said of opium or even slavery. The Labor Party's record is little better, having received many donations from tobacco companies over the years.14 Only the Democrats refuse tobacco sponsorship. In 1999, Prime Minister Howard said on radio "the only way you could further reduce smoking in this country is probably by making it illegal".15 Borrowing Thatcherite vernacular, he has previously declared tobacco control to be a "nanny State" activity.16 Federal Health Minister Michael Wooldridge has championed the national tobacco campaign, although he has failed to make it a priority within his portfolio. Funding for the national tobacco campaign is now at a historic low, with the National Tobacco Strategy brimming with promises, but remaining woefully underfunded.2 In 1999, after seven years of lobbying by the Australian Cancer Society, the tobacco tax system was changed, but not a cent of the $400 million windfall expected from the changed basis of excise (from per weight to per cigarette stick) has been committed to smoking prevention. For all its occasional rhetoric on the importance of tobacco control, the Federal Government, over the coming two years, is likely to derive at least a thousand times more ($10.25 billion) from tobacco excise17 than it committed to "tobacco harm minimisation" in its 1998-99 Budget ($3.8 million -- additional small allocations come from other public health programs).18 Australia and New Zealand's nine Attorneys General agreed in March this year to investigate whether to emulate their US counterparts and sue the tobacco industry. International precedents suggest that they could choose from charges ranging from conspiracy, misleading and deceptive conduct, or common law negligence. Criminal charges against individuals are allegedly also receiving active consideration.19 In the United States case, the tobacco industry hoisted the white flag in November 1998 and settled, agreeing to pay the US States $US206 billion over 25 years.20 (Population adjusted, at today's exchange rates, this would translate into $A25.7 billion.) The money from this case -- the largest settlement in world history -- will compensate the US States for the public costs of caring for uninsured, dying smokers and fund massive education programs to redress decades of industry misinformation. Since the settlement, four US civil cases have seen juries return verdicts in favour of plaintiffs.21 The latest, the Florida Engle class action, may result in an award of punitive damages totalling hundreds of billions of dollars.22 Some analysts believe it could bankrupt the industry,23 ushering in a new era of fully regulated tobacco sales with tobacco classified as a toxic substance. What has brought about these successes? Everything changed overnight for the tobacco companies when, on 8 May 1998, the Minnesota District Court ordered US tobacco companies to place on the World Wide Web what has now run to over 30 million pages of internal documents.24 Among this Niagara of revelations are thousands that are highly incriminating, including a great many authored by Australian industry operatives. The decades-long industry claims that lung cancer was only "statistically associated" rather than caused by smoking, that nicotine was not addictive, and that tobacco companies did not want children to smoke, have been confirmed as carefully orchestrated positions designed to avoid litigation. These views contrast with the industry's private admissions and its elaborate efforts to deny evidence, delay government action and deceive smokers. As a Philip Morris executive wrote in 1978: ". . . an admission by the industry that excessive cigarette smoking is bad for you is tantamount to an admission of guilt with regard to the lung cancer problem. This could open the door to legal suits in which the industry would have no defence."25 The internal documents detail a massive international conspiracy designed to muddle public thinking about tobacco and stall government action26 (Box 2). After the documents were revealed, the Australian tobacco companies have been attempting to deny that the arguments that sustained the US settlement are relevant to Australia. They are amassing a collection of press cuttings about smoking, presumably in an effort to argue that, even if the tobacco companies were determined to promote public doubt about tobacco's harms, this counted for little against what the public could read in the press. The centrepiece of this argument is that, because Australian tobacco tax more than covers the public costs of treating diseased smokers, smokers pay their way -- government claims for recompense are therefore unsupportable.38 This argument has all the moral force of a company polluting the environment, attempting to cover it up with commissioned and carefully managed science and then pleading -- but our customers pay taxes . . . use that to pay for the damage you claim, but that we dispute, occurs. A lobbying document has been circulated by Philip Morris threatening that, if the Australian and NZ Attorneys General proceed with legal action, governments risk huge legal costs and closure of Australian tobacco manufacturing with attendant job losses.38 Predictably, it fails to mention any of the recent legal decisions against the tobacco industry, nor the seachange in evidence provided by the availability of the internal documents. If the Attorneys General take courage, the industry will walk naked into court, with its manifest vulnerabilities illuminated by thousands of its own documents. If the Attorneys General proceed and succeed, we are entitled to wonder whether the money they extract will be used for the intended purpose. One year on from the settlement in the United States only three cents in the dollar were being used for tobacco control.39 Cynics might well ask if the present momentum is motivated solely by revenue raising. References Ministerial Council on Drug Strategy. Background paper. A companion document to the National Tobacco Strategy 1999 to 2002-03. A framework for action. Canberra: Commonwealth of Australia, 1999: 1. Ministerial Council on Drug Strategy. National Tobacco Strategy 1999 to 2002-03. A framework for action. Canberra: Commonwealth of Australia, 1999. Brundtland GH. WHO's International Conference on Global Tobacco Control Law: Towards a WHO Framework Convention on Tobacco Control. Speech text. January 7, 2000. <http://www.who.int/director-general/speeches/2000/ 20000107_new_delhi.html> (accessed 18 May 2000). Hill DJ, White VM, Scollo MM. Smoking behaviours of Australian adults in 1995: trends and concerns. Med J Aust 1998; 168: 209-213. Anderson J. Black spot road safety a national priority in federal budget. Media release. <http://www.dotrs.gov.au/budget/anders/ABudget12_99.htm> (accessed 22 May 2000). Lawrence C. Human Services and Health Portfolio Budget Statement 1995-96, Budget Related Paper No. 4.9. Canberra: AGPS, 1995: 31-52. Willis R. Budget Statements 1994-95. Budget Paper No. 1. Canberra: AGPS, 1994: 3.63. Wooldridge M. Health and Family Service Portfolio Budget Statement, Paper No. 1.8. Canberra: AGPS, 1997. Costello P. Budget measures 1998-99. Budget paper No. 2. Canberra: AGPS, 1998, 1-42. (Also: Commonwealth Department of Health and Family Services. Fact Sheet 1. Investment in preventive public health measures, Budget 1998-99.) Costello P. Budget measures 1999-00. Budget paper No. 2. Canberra: AGPS, 1999. (Also: Department of Health and Aged Care. Fact Sheet 4. Tough on drugs, Budget 1999-2000, May 1999.) Australian Bureau of Statistics. Causes of death, 1998. Canberra: ABS, November 1999. (Catalogue No. 3303.0.) English DR, Holman SDJ, Milne E, et al. The quantification of drug-caused morbidity and mortality in Australia, 1995 edition. Canberra: Commonwealth Department of Human Services and Health, 1995 Hardaker D. 7.30 Report, ABC Televison, 2000; 13 April. <http://www.abc.net.au/ 7.30/730subjectidx_Politics.htm> (accessed 18 May 2000). Chapman S. Tobacco Control Supersite. <http://www.health.usyd.edu.au/ tobacco/worddocs/political_donations.txt> (accessed 18 May 2000). Howard J. Comments on Jeremy Cordeaux radio program 5DN. 1999; July 22. Brough J. Minister tips tobacco sponsorship to stay. Sydney Morning Herald 1997; 3 Sept: 1. Costello P. Australian Budget, 2000-01; Budget paper No. 1. Budget strategy and outlook, Statement 5 revenue: Part II. Budget estimates of revenue, Table 6. Indirect tax. Canberra: Ausinfo. <http://www.ausinfo.gov.au/budget/papers/ bpl/html/bs5-01.htm#P57_5844> (accessed 18 May 2000). Costello P. Budget measures 1998-99, Budget paper No 2. Canberra: AGPS, 1998, 1-42. (Also: Commonwealth Department of Health and Family Services. Fact Sheet 1. Investment in preventive public health measures. Budget 1998-99.) Griffin T (South Australian Attorney General). ABC-TV. Lateline 23 March [transcript]. <http://www.abc.net.au/lateline/archives/S113020.htm> (accessed 22 May 2000). Kelder G, Davidson P, editors. The multistate Master Settlement Agreement and the future of state and local tobacco control; 1999 March 24. <http://tobacco.neu.edu/msa/index.html Executive Summary> (accessed 18 May 2000). Levin M, Menn J. $33m award to smoker who knew risks. Sydney Morning Herald 2000; March 29: 11. Williams B. Smokers win case in Fla. News and Observer (Raleigh, Va) 2000, April 8. <http://www.news-observer.com/daily/2000/04/08/nc00.html> (accessed 18 May 2000). Tobacco Control Resource Center Inc., and the Tobacco Products Liability Project, Northeastern University. <http://tobacco.neu.edu/> (accessed 18 May 2000). The Tobacco Archives. <http://www.tobaccoarchives.com> (accessed 18 May 2000). Osdene TS. Roper study proposal to Tobacco Institute, 1978, 16 February. <http://www.tobaccoinstitute.com/getallimg.asp?DOCID=TIOK0034601/4602> (accessed 18 May 2000). Cullman H. (Philip Morris) Memorandum to the files (Confidential). Interoffice correspondence. 1976; December 3. Search for -- Did:2025025286 at <http://www.pmdocs.com/default1.asp> (accessed 23 May 2000). Proposal for the organisation of the Whitecoat Project. <http://www.ash.org.uk/ papers/2501474262.pdf> (accessed 18 May 2000). Project Down Under. 1987; June 26. Search for -- Did:2021502671/2678 at <http://www.pmdocs/com/default1.asp> (accessed 23 May 2000). Boyse S (British American Tobacco). Minutes of a special meeting of the UK industry on environmental tobacco smoke. 1988, February 17. <http://www.ash.org.uk/papers/401247331.pdf> (accessed 18 May 2000). Headden S. Papers you weren't ever supposed to see. How Big Tobacco fought an all-fronts war. US News 1998; May 4. <http://www.usnews.com./ usnews/issue/980504/4toba.htm> (accessed 22 May 2000). Marlboro marketing strategy (Australia) 1990. Search for -- Did:2048571302/1384 at <http://www.pmdocs.com/default1.asp> (accessed 23 May 2000). Philip Morris (Australia) Corporate Affairs Plan 1992. Search for -- Did:2023240608/0627 at <http://www.pmdocs.com/default1.asp> (accessed 23 May 2000). Health aspects of environmental tobacco smoke: an evaluation of the scientific literature. Submission to the Health Care Committee of the National Health and Medical research Council. Sydney: November 1994, 156 pp. Staunton D (Tobacco Institute of Australia) Memorandum 1994; 7 April. Search for -- Did:2504088601/8606 at <http://www.pmdocs.com/default1.asp> (accessed 23 May 2000). Knopick P. Memorandum to William Kloepfer: TI, USA; 1980. <http://www. tobaccoinstitute.com/getallimg.asp?DOCID=TIMN0107822/7823> (accessed 18 May 2000). Francis P (Philip Morris Australia). Telex to Don Hoel (Shook, Hardy and Bacon, lawyers, Kansas). 1998; 11 December. Search for -- Did:2021588889 at <http://www.pmdocs.com/default1.asp> (accessed 23 May 2000). Chapman S. Tobacco Control Supersite. <http://www.health.usyd.edu.au/ tobacco/ozdocs.html#Plainpackaging> (accessed 18 May 2000). Philip Morris (Australia). Why US style medical cost recovery litigation is not right for Australia. Melbourne: Philip Morris, 2000; March 16: 7pp. Rees D. Tobacco money scattered by states. Richmond Times-Dispatch. Richmond, Va, 1999; 19 Nov. Authors' details Department of Public Health and Community Medicine, University of Sydney, Sydney, NSW. Simon Chapman, Associate Professor; and Associate Director, VicHealth Centre for Tobacco Control, Drummond Street, Cartlton, VIC. Reprints will not be available from the author. Correspondence: Associate Professor S Chapman, Department of Public Health and Community Medicine, University of Sydney, NSW 2006. simoncAThealth.usyd.edu.au 1: Recent federal budget commitments to major public health programs (average annual commitments for 1994-95 to 2002-03), compared with deaths from associated causes, 1998Average committed per year ($ million)No. of deaths in 199811,12Amount ($) committed per death Black-spot road safety program5 AIDS control6 Breast cancer7 Cervical cancer program638.1 49.5 51.6 10.51942 187 2558 30219619 264706 20172 34603Illicit drugsNational Drug Strategy7 National Illicit Drug Strategy8-1036.6 74.7630 63058095 118571Asthma management103.16854525Preventing falls101.711821438Tobacco (Health Australia) Tobacco Harm Minimisation 1995-199866.118224337National Tobacco Campaign Tobacco Harm Minimisation 1999-200292.018224112 2: What the tobacco companies' internal documents revealed In 1977, after years of individually denying tobacco-caused disease, the tobacco companies came together at a stately manor in Berkshire and gave birth to an international conspiracy. In the decades that followed, the companies Gagged their own scientists and signed up a phalanx of external consultants. These consultants agreed to be "managed" by industry lawyers via programs with pulp fiction titles like the Whitecoat Project27 and Project Down Under.28 Spent what a 1988 meeting described as "vast sums of money to keep the [passive smoking] controversy alive".29 Funded teenage education gestures calculated to gain political "Brownie points", while privately noting that they would not support programs that actually reduced tobacco use30 and assiduously noting the brand preferences of "new smokers" (industry-speak for teenagers).31In Australia, the tobacco companies pumped hundreds of thousands of dollars into community concerns about the government's misplaced priorities on tobacco, arguing that it should instead be addressing illicit drugs.32 Members of a group commissioned by the Tobacco Institute in 1994 to review the literature on passive smoking33 were said to be "known to us, as are their views".34 Most crucially, the word "addiction" has been the industry's bête noire. A letter between two senior US industry officials noted "[industry lawyers] remind us that the entire matter of addiction is the most potent weapon a prosecuting attorney can have in a lung cancer/cigarette case. We can't defend 'free choice' if the person was 'addicted'".35 So they invested millions in public relations efforts at denying that nicotine was addictive. There are dozens of memos between the Australian tobacco industry and the same lawyers, approving strategy, vetting scientists who might utter the profane "a" word in industry-sponsored meetings, and sanitising public documents.36 A fascinating series of memos documents the delight of Philip Morris over lobbying by the then Victorian Premier Jeff Kennett in 1992-93, designed to keep "addiction" off Australian packs by promoting instead comparatively tepid and tiny European-style warnings.37

Simon Chapman