Comparison with the General Scottish Population

Crude death rates are calculated by dividing the total number of deaths in a population by the total population size. This does not take any other population characteristics into account. Key characteristics of populations may increase the probability of death occurring, so comparison of crude death rates between different groups or populations can be misleading.

The prison population in Scotland has a different age profile to the general Scottish population. As outlined in the Deaths in Prison Custody by Age Band section, the prison population is comprised of individuals aged 16 years and over and, while the prison population has been ageing over time, it is disproportionately comprised of adults between 21 and 50 years. As the probability of death tends to increase with age, these differences in composition could have an effect on trends in the crude death rate.

A further difference between the prison population and the general Scottish population is the distribution of males and females. Across the reporting period, women comprised less than 6% of the prison population. In contrast, the average proportion of females in the general population was around 51%. As outlined in the Deaths in Prison Custody by Sex section, only 16 deaths in prison custody across the period 2012-13 to 2023-24 involved females and the maximum number occurring in any year is 3. These small numbers mean that comparisons made in this report will only include deaths involving males.

Standardised Mortality Ratios

(Supplementary Tables C1 to C6)

To compare the male prison population with the general male population in Scotland, standardisation is needed to account for differences in age structure. Standardised mortality ratios (SMRs) provide the ratio of expected mortality risk for a population, given age-specific mortality rates from a reference population.

Using SMRs allows us to calculate how many deaths would be expected in the male prison population, should the overall pattern in the male general population hold in the male prison population in Scotland. SMRs are calculated by dividing the observed number of deaths (in the male prison population) by the expected number, expressed as a percentage. SMRs are calculated here according to three broad age bands.

An SMR higher than 100 indicates a greater risk among the male prison population compared to the general male population in Scotland. Values less than 100 suggest lower than expected risk.

Confidence intervals are presented to show the level of uncertainty in the calculated ratio. If the range of values within a confidence interval contains 100, there is no statistically significant difference (at a 95% confidence level) between the male prison population and the general male population in Scotland.

Further information on the methods used here, including how the SMR is calculated, has been published by Public Health England. SMRs have been applied here as this is the method advised when comparing small populations (see Technical Briefing 3 by The Association of Public Health Authorities for further detail).

This report provides SMRs for all deaths and for two causes of death: probable suicide and drug misuse, across deaths that occurred amongst the male prison population. When providing SMRs by cause of death, findings are reported using 3-year rolling periods due to the smaller number of deaths. This helps to smooth out fluctuations over time. However, the mortality rate in any given period will be influenced by those in the preceding years. SMRs should not be used to understand how the level of risk changes over time.

The SMRs presented in this report should not be taken as indicative that the risk of death (by probable suicide, drug misuse or any other cause) changes specifically as a result  of an individual being in prison custody. It should be noted that underlying contributory or determinant factors have not been examined in the analysis. Patterns in both population groups impact on the calculated ratio for each time period presented. Therefore, consideration must be given to the patterns present in both the prison and general populations when interpreting the results. NRS produce publications on all deaths, probable suicide and drug misuse deaths and trends in the general population on their webpages.

There are two limitations to the analysis presented in this report. The first is that deaths in prison custody are grouped by financial year, whereas deaths in the general population are grouped by calendar year. This difference in time periods means results should be interpreted with caution.

The second is that the available data on deaths in the general population by age bands includes deaths involving 15-year-old males. For the time period covered by this report, the male prison population only included individuals aged 16 or over. The number of deaths involving 15 year old males in the general population is low (62 over the period 2012 to 2024, on average less than 5 per annum) but this will still have a bearing on the figures calculated in the SMRs. Again, results should be interpreted with caution.

All Deaths (standardised mortality ratios)

(Supplementary Tables C1 and C2)

Figure 15 shows the standardised mortality ratios for all male deaths in prison custody 2012-13 to 2023-24 (Scotland).

Figure 15 shows that the risk of male deaths occurring in prison custody is statistically significantly lower than in the general population in each year of the reporting period, as the confidence intervals are below 100. In 2021-22 and 2023-24, the risk is closer to that of the general population but remains statistically significantly lower. In the latest year (2023-24) the ratio calculated is at the highest level, however the upper confidence interval is calculated at 99.49, which still falls below the threshold of 100.

Supplementary table C2 shows the breakdown of each age group across each year.

Probable Suicide Deaths (standardised mortality ratios)

(Supplementary Tables C3 and C4)

Figure 16 shows the standardised mortality ratios for male probable suicide deaths in prison custody for 3-year rolling periods 2012-15 to 2021-24 (Scotland).

Figure 16 shows that the risk of male probable suicide deaths in prison custody was statistically similar to the general population across the rolling periods 2012-15 to 2018-21, as the confidence intervals cross 100. However, for the rolling periods 2019-22, 2020-23 and 2021-24 the risk of male probable suicide deaths in prison custody is statistically significantly higher than the general population (confidence intervals are higher than 100).

Supplementary table C4 shows the breakdown of each age group across each year group.

Drug Misuse Deaths (standardised mortality ratios)

(Supplementary Tables C5 and C6)

Figure 17 shows the standardised mortality ratios for male drug misuse deaths in prison custody for 3-year rolling periods 2012-15 to 2021-24 (Scotland).

Figure 17 shows that the risk of male drug misuse deaths in prison custody is statistically significantly lower than in the general population for the periods 2012-15 to 2019-2022, as the confidence intervals are below 100. However, for the periods 2020-2023 and 2021-2024 the risk of male drug misuse deaths in the prison population is statistically similar to the general male population (confidence intervals cross 100).

Again, the Standardised Mortality Ratio calculated for drug misuse deaths in the male prison population is at the highest level seen in the time series in the rolling period 2021-2024.

Supplementary table C6 shows the breakdown of each age group across each year group.