Figures use the underlying cause of death
A range of information is available from the death certificate, and figures for deaths from a particular cause can be produced in a number of ways: more on such points is available via the 'Death Certificates and Coding the Causes of Death' page of the National Records of Scotland (NRS) website.
The figures for alcohol deaths (using both the old and the new definitions) are produced on the 'underlying cause' basis, so they are the numbers of deaths for which the disease or injury which initiated the chain of morbid events leading directly to death was one of those which are listed in the relevant definition.
Causes of death which are not covered by the National Statistics (NS) definitions
The figures for alcohol deaths do not include all deaths which may be caused by alcohol - for example, they do not include deaths:
- as a result of road accidents, falls, fires, suicide or violence involving people who had been drinking; or
- from some medical conditions which are considered partly attributable to alcohol, such as certain forms of cancer.
The reasons for this include the need to be able to provide reasonably consistent trends over time and for different parts of the United Kingdom (UK). The NS definitions used include only those causes of death which are regarded as being most directly due to alcohol consumption and for which figures can be obtained from the statistics of registered deaths, due to lack of consistent statistical information about (e.g.) accidental deaths, suicides and homicides which are directly due to the consumption of alcohol.
Including appropriate proportions of deaths from causes such as road accidents and certain forms of cancer would produce considerably higher figures for alcohol-related deaths. Further information about this is available from (e.g.) the Information Services Division (ISD) Scotland paper 'Alcohol attributable mortality and morbidity: alcohol population attributable fractions for Scotland', which was published in June 2009, and the Scottish Public Health Observatory (ScotPHO) report ‘Hospital admissions, deaths and overall burden of disease attributable to alcohol consumption in Scotland’, which was published in February 2018.
Moving averages and likely ranges of year-to-year statistical variability
The charts that are included with the tables show that the number of alcohol deaths (using either definition) registered in Scotland may fluctuate noticeably from one year to the next. Therefore, in addition, the charts provide 5-year moving average values, which should be a better guide to the underlying level of such deaths and any long-term trend.
The charts also show the likely range of values around the moving average. This likely range of statistical variability in the figures is estimated by assuming that the numbers represent the outcome of a Poisson process, with the underlying rate of occurrence in each year being the same as the value of the 5-year moving average which is centred on that year. ‘Upper’ and ‘lower’ boundaries of an approximate ‘95% confidence interval’ around the moving average are calculated by adding/subtracting twice the standard deviation (for a Poisson distribution, the mean and the variance are the same, so the standard deviation is simply the square root of the moving average). In the case of alcohol-related deaths (i.e. the old NS definition), in the period from 1981 to 2014 (inclusive), only two of the 34 years had figures which were outwith this range - and this is broadly in line with what one would predict based on statistical theory (as only about 5% of observations would be expected to fall outwith an approximate 95% confidence interval).
The numbers of alcohol deaths for areas within Scotland may be subject to large percentage year-to-year fluctuations. Therefore, the tables give 5-year moving averages as well as figures for individual years for (e.g.) NHS Board areas, as the moving averages should provide a better guide to the underlying level, and any long-term trend, than the figure for any given year, or the change between one year and the next.
A separate section of the NRS website provides more information about fluctuations in and possible unreliability of death statistics for small areas, for small sub-groups of the population, or for short periods.
The introduction of the new NS definition: alcohol-specific deaths
In Spring 2017, the Office for National Statistics (ONS) initiated discussions with NRS, Scottish Government, ISD and some other government bodies regarding the need to review the NS definition of alcohol-related deaths that ONS, NRS and others used to produce statistics. This led to ONS consulting a range of interested parties on the NS definition of alcohol deaths that should be used in future. The relevant documents, which were all prepared by ONS (with input from NRS and others, who contributed as appropriate when necessary), are as follows:
- ‘Alcohol mortality definition review’ (published 28 June 2017) – the consultation document
- ‘Response to consultation on the National Statistics definition of alcohol-related deaths’ (published 6 October 2017) – summarised the responses to the consultation and set out what would be done in consequence
- ‘The impact of using the new definition of alcohol-specific deaths’ (published 27 October 2017) – compared the numbers of deaths counted by the old and the new definitions, and the resulting age-standardised death rates, for the UK as a whole and for each of its four countries. It also provided several time-series of figures, on the basis of the new definition, for 2001 to 2015. Table 1 of this publication compares the causes of death which are covered by the two definitions.
- ‘Alcohol-specific deaths in the UK: registered in 2016’ (published 7 November 2017) – the first edition of ONS’s annual ‘alcohol deaths’ publication to use the new definition, it included comparisons of the age-standardised death rates for each sex for the four countries of the UK.