Methodology
Estimages of avoidable deaths in Victoria 1999-2002
Aims
This analysis presents estimates of mortality attributable to tobacco smoking and other significant 'avoidable' causes for Victoria for the four-year period 1999-2002.
As well as tobacco caused deaths, we also report on deaths due to alcohol (including alcohol-related road deaths) and other drugs. Total road deaths, including those caused by alcohol or other drug use, have also been included for comparison.
Summary tables are provided for the whole of Victoria and by various geographical boundaries. These tables include figures on deaths attributable to tobacco and other causes and comparison with the overall Victorian proportions of such deaths.
Explanation of methods
The estimates presented in the tables are calculated using 'aetiological fractions' for a comprehensive list of illnesses and causes of death or injury known to be associated with smoking, alcohol or other drug use. Aetiological fractions are the proportions of cases or deaths from a condition (by age group and gender) that are attributed to tobacco (or other drug) use. For example, if there is a probability of 0.36 that a stroke is caused by smoking, then the product of this probability (the aetiological fraction) and the total number of stroke deaths in a population gives an estimate of the number of stroke deaths attributable to smoking.
The fractions used in this analysis are derived from known prevalence and risk factor data in the Australian population[1,2]. 'Partial' aetiological fractions were used in estimating deaths due to alcohol. These partial fractions[2] compare 'unsafe' (hazardous, harmful) alcohol consumption with 'responsible' (low) alcohol consumption rather than with total abstinence. The use of these partial fractions reflects the idea that 'unsafe' drinking - as compared with low alcohol consumption, which may be protective - is the major cause for concern. Aetiological fractions were applied to the underlying cause of death for every Victorian death in 1999-2002 to obtain estimates of total deaths due to tobacco, alcohol and other drugs. Details of all deaths in Victoria from 1st January 1999 to 31st December 2002 were obtained from the Australian Bureau of Statistics.
Individual deaths were grouped, according to the place of a person's usual residence, by Local Government Area, Department of Human Services Region and Primary Care Partnership catchments. Summary sheets for each area show total deaths, deaths due to smoking and other avoidable causes and comparison of the proportions of each with the state average. It should be noted that, because of the way the deaths for each cause are calculated, very small numbers in some tables do not compromise any individual's privacy. The numbers are cumulated fractions of deaths rather than actual people. For example, a reported single smoking caused death from stroke in the Shire of Nowhere may not indicate that John Smith of this Shire died of a smoking stroke but rather that Jack Brown aged 41, Fred Jones aged 58 and Mary Smith aged 37 each died of stroke in Nowhere Shire. The sum of their aetiological fractions (0.36, 0.33 and 0.32 respectively being the proportion of deaths from stroke in their age/sex groups that are due to tobacco) is 1.1 - rounded to 1 in summary tables. Therefore the one reported death is not identifying one individual in any way.
Vicky Thursfield October 2004.
Ridolfo B, Stevenson C 2001. The quantification of drug-caused mortality and morbidity in Australia, 1998. AIHW cat. no. PHE 29. Canberra: AIHW (Drug Statistics Series no. 7).
| | Quit | The Cancer Council Victoria | |