Table 1.
Characteristics and summary of the included studies examining population suicide rates and EAS
| Study | Study design | Sample | Methods | Conclusions |
|---|---|---|---|---|
| Canetto and McIntosh, 202126 | Cohort study | Individuals dying by suicide and EAS between 1998 and 2018 over the age of 18 years in Oregon state (USA) | DWDA and suicide death rates per 100 000 calculated using WISQARS and WONDER data from the CDC and OHA annual reports for Oregon state and the USA | For women over 75 years of age, EAS rates are higher than non-assisted suicide rates. For women over 65 years of age, non-assisted suicides have increased in the past two decades. Women over 65 years of age represent 16% of non-assisted suicide deaths but 46% of deaths by EAS |
| Jones and Paton, 201522 | Cohort study | Individuals dying by suicide and EAS in Oregon, Washington, Montana, Vermont and the USA broadly, from 1990–2013 | Used non-assisted suicide rates from 1990–2013 based on age group and EAS data from Oregon and Washington. Controlled for suicide-related demographic variables, and used logistic regression to control for confounding variables | When state effects are controlled for, rates of self-initiated death increased by 8.9% where EAS is legal. When demographic variables are controlled for, self-initiated death rates increased by 11.8% where EAS is legal. Accounting for state-specific time trends, the increase in self-initiated death is 6.3%. There are higher rates in people aged over 65 years. There is no evidence to suggest a decrease in non-assisted suicide rates or increase in mean age of death by suicide |
| Nanner, 202125 | Cohort study | Belgian individuals who died by suicide between 1990 and 2015 | Used the synthetic control method to observe changes in non-assisted suicide rates in Belgium before and after legalisation of EAS (2002) compared with countries without EAS policy. Control for variables that affect suicide risk | GSCM showed an average annual suicide rate increase of 0.73 per 100 000 population (95% CI −5.7 to 7.2; P = 0.80). Placebo testing based on the SCM analysis showed equal outcomes for Belgium and the comparisons. This study failed to show a statistically significant association between the legalisation of EAS and non-assisted suicide rates |
| Steck et al, 201523 | Longitudinal cohort study | 8 527 786 Swiss men and women with a total of 24 842 suicides between 1991 and 2008 | Used the Swiss national cohort to calculate rates of EAS and non-assisted suicide (by various methods) from 1991–2008. Compared genders and age groups (15–34, 35–64 and 65–94 years), using coding provided by Federal Statistical Office | Across all age groups, increase in female self-initiated death from 13.6 to 14.3 between 1991 and 2008. In those aged over 65 years, suicide by poisoning rates doubled in men and more than tripled in women. Rate of suspected EAS increased from 2.37 to 14.98 |
| Steck et al, 201624 | Longitudinal cohort study | 5 million Swiss 2003 to 2008 | Data on EAS from 2003–2008 provided by right-to-die organisations. Calculated rates of EAS and non-assisted suicides as associated with demographic and socioeconomic variables. Compared percentages of self-initiated death with different underlying causes. Logistic regression analysis performed to examine gender differences in probability of suicide based on gender | Rates increase with age, with a greater increase for EAS than non-assisted suicide (13.8 to 30.1 v. 0.3 to 38.9). Rates of EAS similar in men and women (4.1 v. 5.0); however, men have a non-assisted suicide rate three times that of women |
| Zalman and Stack, 199621 | Population based time series analysis | All deaths by suicide in The Netherlands from 1950–1990, using 1973 and 1981 as key legal markers | National suicide rates were obtained from the national database. Specifically rates for age groups 65–74 and ≥75 years were calculated with data from the World Health Organization. Rates were compared before and after two major cases in 1973 and 1981. This study controlled for divorce rate, religiosity and economic strain. Yule Walker techniques were used to purge for autocorrelation | Legal changes surrounding euthanasia had a significant association with non-assisted suicide rates on bivariate analysis. However, once divorce, religiosity and economic strain were controlled for, the association was no longer statistically significant, neither in the population as a whole or among older people (aged 65–74 and over 75 years) |
EAS, Euthanasia and Assisted Suicide; DWDA, Death with Dignity Act; WISQARS, Web-based Injury Statistics Query and Reporting System; WONDER, Wide Ranging Online Data for Epidemiologic Research; CDC, Center for Disease Control; OHA, Oregon Heath Authority; GSCM, Generalised synthetic control method; SCM, synthetic control method.