International suicide rates: recent trends and implications for Australia

Impact of national suicide prevention strategies on suicide rates

Up to Publications

prev page| TOC |next page

Figures 65 through 68 shows the rates of suicide for all ages and adolescents (15-24 years) in countries with national suicide prevention strategy and the years in which the national suicide prevention strategy was implemented. The implementation of strategies was preceded by substantial increases in suicide rates among males and static rates among females in Finland, Australia, and Norway (adolescence and all ages). Sweden was witnessing declining rates prior to the introduction of their plan in males and females. In the years following the introduction of a strategy, reductions in suicide among males occurred in Finland and Australia, and increased or stabilised in Norway and Sweden. Rates in young females increased in Norway and Sweden following the implementation of their national strategies.

To examine the impact of national suicide prevention plans on suicide mortality, the average rates and trends in the five years before and after the implementation of the strategy were compared.² Analysis considered changes in rates and trends for all ages and for the 15-24 year age group. For example, in Australia, the average rate for the 1990 through 1994 was compared to the average rate for 1995-1999. Average rates are calculated based on the aggregated numbers and populations for the 5-year period. Trends were calculated via linear regression and indicate the direction and magnitude of the slope created by rates in the 5-year period. The direction (upward or downward) is indicated by the presence of a symbol + or – while the value of the slope indicates the magnitude or the trend. T-values indicated whether the change in average suicide rates or slopes are significant (p<0.05) or non-significant (n.s.).

Adolescents
All ages
Limitations

Adolescents

Table 15 presents a summary of analysis for the impact of national suicide prevention strategies on suicide rates among adolescents. In Norway, there was a significant change in suicide mortality trend among adolescent females following the introduction of the national strategy. Mortality trends changed from a negative slope (indicating declining rates) before the plan to a positive slope (suggesting increasing rates) after. Similar but non-significant changes in slope (from negative to positive) were revealed in Australian females, Norwegian males, and Swedish males and females. In Finland, male and female rates appeared to continue the upward trend that was evident in the years preceding the national strategy implementation. This analysis suggests that national suicide prevention strategies have had little or no impact on reducing suicide rates among the young.

Table 15. Changes in average suicide rates and trends following the implementation of national suicide prevention strategies, 15-24 years

Table 15 is separated into 4 smaller tables in this HTML version for accessibility reasons. It is presented as one table in the PDF version.Top of page

Australia (strategy implemented in 1995)*

	Average rate and (95% CI) for 5 years before	Average rate and (95% CI) for 5 years after	Trend [slope and (SE)] for 5 years before	Trend [slope and (SE)] for 5 years after	Difference between slopes [t-value and (p)]
Females	4.87 (2.82 – 06.94)	5.75 (4.08 – 07.43)	-0.37 (.31)	0.01 (.31)	-1.68 (ns)
Males	25.96 (24.61 – 2.30)	25.57 (22.76 – 28.38)	-0.07 (.54)	-0.31 (1.13)	0.13 (ns)

*National Youth Suicide Prevention Strategy.

Finland (strategy implemented in 1992)

	Average rate and (95% CI) for 5 years before	Average rate and (95% CI) for 5 years after	Trend [slope and (SE)] for 5 years before	Trend [slope and (SE)] for 5 years after	Difference between slopes [t-value and (p)]
Females	7.98 (6.42 – 09.53)	6.90 (4.97 – 08.83)	0.28 (.61)	0.21 (.78)	0.06 (ns)
Males	44.20 (38.77 – 49.63)	36.85 (32.34 – 41.35)	1.99 (1.90)	0.11 (1.84)	0.22 (ns)

Norway (strategy implemented in 1994)

	Average rate and (95% CI) for 5 years before	Average rate and (95% CI) for 5 years after	Trend [slope and (SE)] for 5 years before	Trend [slope and (SE)] for 5 years after	Difference between slopes [t-value and (p)]
Females	5.62 (4.47 – 05.86)	5.60 (4.44 – 06.77)	-0.16 (.27)	0.58 (.34)	-3.31 (.02)
Males	25.12 (22.48 – 27.75)	21.70 (20.40 – 23.00)	-0.20 (1.07)	0.34 (.49)	-0.33 (ns)

Sweden (strategy implemented in 1995)

	Average rate and (95% CI) for 5 years before	Average rate and (95% CI) for 5 years after	Trend [slope and (SE)] for 5 years before	Trend [slope and (SE)] for 5 years after	Difference between slopes [t-value and (p)]
Females	5.42 (4.24 – 06.59)	5.63 (4.90 – 06.37)	-0.18 (.46)	0.31 (.23)	-1.48 (ns)
Males	13.36 (11.02 – 15.71)	12.50 (11.05 – 13.95)	-0.79 (.84)	0.14 (.59)	-0.74 (ns)

All ages

Table 16 provides a summary of the all-ages analyses involving the implementation of national suicide prevention strategies. Analysis of trends before and after the adoption of strategies yields varying results for different countries. Among Australian females, the suicide mortality trend increased after the implementation of the national youth strategy. In Finland, changes from positive slopes preceding the implementation to negative slopes following were revealed. In Norway and Sweden (among males and females), downward trends were evident in the five years before the national suicide prevention strategies were introduced. Subsequently, downward slopes remained, but of smaller magnitude, suggesting a possible stabilising effect. Thus, based on these findings, with the possible exception of Finland, reductions in suicide mortality were not associated with the implementation of national strategies. Top of page

Table 16. Changes in average suicide rates and trends following the implementation of national suicide prevention strategies, all ages

Table 16 is separated into 4 smaller tables in this HTML version for accessibility reasons. It is presented as one table in the PDF version.

Australia (strategy implemented in 1995)*

	Average rate and (95% CI) for 5 years before	Average rate and (95% CI) for 5 years after	Trend [slope and (SE)] for 5 years before	Trend [slope and (SE)] for 5 years after	Difference between slopes [t-value and (p)]
Females	5.04 (4.64 – 05.45)	5.35 (4.97 – 05.73)	-0.21 (.11)	00.07 (.15	-7.09 (<.001)
Males	20.35 (19.48 – 21.22)	21.42 (20.10 – 22.75)	-0.16 (.34)	00.54 (.44)	-0.98 (ns)

*National Youth Suicide Prevention Strategy.

Finland (strategy implemented in 1992)

	Average rate and (95% CI) for 5 years before	Average rate and (95% CI) for 5 years after	Trend [slope and (SE)] for 5 years before	Trend [slope and (SE)] for 5 years after	Difference between slopes [t-value and (p)]
Females	11.80 (11.48 – 12.12)	11.36 (10.97 – 11.74)	0.09 (.12)	-0.07 (.15)	03.67 (.01)
Males	47.08 (45.31 – 48.85)	43.56 (40.77 – 46.35)	1.17 (.25)	-1.85 (.39)	11.46 (<.001)

Norway (strategy implemented in 1994)

	Average rate and (95% CI) for 5 years before	Average rate and (95% CI) for 5 years after	Trend [slope and (SE)] for 5 years before	Trend [slope and (SE)] for 5 years after	Difference between slopes [t-value and (p)]
Females	7.74 (7.08 – 08.41)	6.42 (6.01 –0 6.83)	-0.41 (.14)	-0.02 (.17)	-6.72 (<.001)
Males	22.44 (21.34 – 23.54)	18.12 (17.62 – 18.62)	-0.60 (.29)	-0.03 (.20)	-3.81 (.009)

Sweden (strategy implemented in 1995)

	Average rate and (95% CI) for 5 years before	Average rate and (95% CI) for 5 years after	Trend [slope and (SE)] for 5 years before	Trend [slope and (SE)] or 5 years after	Difference between slopes [t-value and (p)]
Females	9.71 (9.16 – 10.27)	8.24 (7.67 – 08.81)	-0.38 (.05)	-0.32 (.14)	-2.44 (.05)
Males	22.79 (21.58 – 24.00)	20.16 (19.48 – 20.84)	-0.76 (.23)	-0.34 (.19)	-3.81 (.009)

Top of page

Limitations

There are several limitations to the current analysis. Firstly, the wide variation in suicide rates in a five-year period results in a large degree of error in any statistical analysis. These fluctuations mean that substantial decreases in suicide mortality can fail to reach statistical significance. Secondly, determining the most appropriate point of 'implementation' of a strategy is an imprecise task. Particular activities may be well underway before the introduction of a formal strategy has been developed that ties these activities together. On the other hand, it may take several years before all the elements of a strategy are put into effect. Thirdly, only a small number of countries have implemented suicide prevention strategies. With the exception of Sweden, the introduction of the each country's strategy followed historically unprecedented levels of suicide mortality that may have been unsustainable in the long term. Hence, a contraction in rates may be misconstrued as evidence for an effective intervention. Additional analysis over a longer period and with a greater number of countries will be needed in the future to draw more convincing conclusions. Such an analysis will probably not be feasible for a further five to ten years as strategies in countries such as the England, Germany, New Zealand, and the United States come into effect.

While enthusiasm for anti-suicide activities is increasing throughout the world, this analysis suggests that declines in suicide seen in Australia and abroad cannot be validly attributed to the introduction of national suicide prevention strategies. Even if national plans are partly responsible for recent reductions, the component of those plans that have contributed to the reduction cannot currently be identified. There is little evidence to support the efficacy of any particular approaches or strategies. The effectiveness of prevention activities in reducing suicide mortality and morbidity has been the subject of recent literature reviews and commentaries (De Leo 2002a; De Leo 2002b; Gunnell & Frankel, 1994; Hawton et al., 1998). These documents indicate that scientifically sound evaluations of suicide prevention activities are scarce. Those evaluations that do exist generally lack sufficient sample size to detect meaningful effects and have inadequately defined outcome measures and control conditions.

Limiting access to means is an approach to suicide prevention that has the strongest evidence for efficacy in suicide prevention (Gunnell & Frankel, 1994) and is a core component of all national suicide prevention strategies. Potentially beneficial activities in this regard include detoxifying domestic (De Leo et al., 2002b; Kreitman, 1976) and car exhaust gases (Toseland, 1999; McClure, 2000), limiting the quantity of medications per pack (Hawton et al, 2001) and reducing prescriptions of lethal medications (Buckley et al., 1995; Ohberg et al., 1995), erecting barriers on bridges (Cantor & Hill, 1990), and limiting firearm ownership (Brent et al., 1991; Lambert & Silva, 1998). The rationale of restricting access to means is that by delaying death a person has an opportunity to reconsider their actions. However, the utility of these approaches have not been subjected to controlled studies and some have argued that restricted methods may be substituted by other more readily available methods (Amos et al., 2001; Lester, 1991; Rich et al., 1990). At any rate, restricting commonly used methods such as hanging is unrealistic beyond institutional settings such as prisons and hospitals (Cantor et al., 1996; Jordan et al., 1987).

Improved detection and treatment of depression has also been suggested as a potentially efficacious approach to suicide prevention, and is also a major component of all suicide prevention strategies. An educational programme for general practitioners coincided with a 60% reduction in suicide mortality on the island of Gotland in Sweden (Rutz et al., 1989; Rutz et al., 1992). However, suicide mortality rates were found to rise in the years following the education programme, suggesting other factors may have contributed to the initial reduction. Furthermore, this investigation may have been biased by the small size of the populations involved, and consequently by the extremely limited number of cases that defined as positive the outcome of the programme (two cases). In addition, the experience apparently had results only with female subjects (General Practitioners were educated for recognition and treatment of depression, and since female patients consult more frequently than males with their doctor, this may have contributed to explain results).

Promising approaches to prevention of suicide include problem solving therapy (Hawton et al., 1998), emergency access cards (Morgan et al, 1993), dialectical behaviour therapy (Linehan, 1993), neuroleptic medication (Montgomery et al, 1983), and telephone active outreaching in the aged (De Leo et al., 1995; De Leo et al., 2002c). However, these approaches need further examination through large-scale controlled trials to demonstrate their effectiveness. To date, most studies on the effectiveness of suicide prevention activities have used deliberate self-harm as an outcome measure. While deliberate self-harm is a significant risk factor for suicide, it is not analogous to completed suicide. Activities that have been shown to be effective in preventing self-harm must be evaluated for their potential to prevent suicide deaths before being widely adopted.Top of page

Figure 65. Suicide rates before and after the implementation of the National Youth Suicide Prevention Strategy in Australia

Top of page

Text version of Figure 65

Figures in this description are approximate as they have been read from the graph.

Figure 65 shows suicide rates in Australia before and after the implementation of the National Suicide Prevention Strategy in 1995:

• For females aged 15-24 years, the rate remained relatively stable, fluctuating between 3.5 per 100,000 and 7 per 100,000 for all years between 1980 and 1999 and remained constant at 5 per 100,000 from 1999 to 2001.
  
  
• For males aged 15-24 years, the rate increased gradually in a fluctuating manner from 17 per 100,000 in 1980 to 26 per 100,000 in 1987. The suicide rate remained relatively stable until 1995, when it began to increase, reaching 30 per 100,000 in 1997, before decreasing rapidly to 20 per 100,000 in 2000.
  
  
• For females of all ages, the rate remained relatively stable, fluctuating between 3.5 per 100,000 and 7 per 100,000 for all years between 1980 and 1999 and remained constant at 5 per 100,000 from 1999 to 2001.
  
  
• For males of all ages, the rate increased gradually from 16 per 100,000 in 1980 to 22 per 100,000 in 1987. The suicide rate remained relatively stable until 1995, when it began to increase, reaching 23 per 100,000 in 1997, before decreasing to 20 per 100,000 in 2000.
  
  

Figure 66. Suicide rates before and after the implementation of the National Suicide Prevention Strategy in Finland

Top of page

Text version of Figure 66

Figures in this description are approximate as they have been read from the graph.

Figure 66 shows suicide rates in Finland before and after the implementation of the National Suicide Prevention Strategy in 1992:

• For females aged 15-24 years, the rate remained relatively stable, fluctuating between 5 per 100,000 and 11 per 100,000 from 1980 to 1992. The rate dipped to 2 per 100,000 in 1993 before increasing to 8 per 100,000 in 1994 and remaining at approximately that level to 2000.
  
  
• For males aged 15-24 years, the rate increased gradually in a fluctuating manner from 39 per 100,000 in 1980 to 52 per 100,000 in 1990, before decreasing in a fluctuating manner to 30 per 100,000 in 2000.
  
  
• For females of all ages, the rate remained relatively stable around 10 per 100,000 for all years between 1980 and 2000.
  
  
• For males of all ages, the rate increased gradually from 42 per 100,000 in 1980 to 49 per 100,000 in 1990 and 1991. It then started decreasing to 34 per 100,000 in 2000.
  
  

Figure 67. Suicide rates before and after the implementation of National Suicide Prevention Strategy in Sweden

Top of page

Text version of Figure 67

Figures in this description are approximate as they have been read from the graph.

Figure 67 shows suicide rates in Sweden before and after the implementation of the National Suicide Prevention Strategy in 1995:

• For females aged 15-24 years, the rate remained relatively stable, fluctuating between 4 per 100,000 and 9 per 100,000 for all years between 1980 and 2001.
  
  
• For males aged 15-24 years, the rate increased in a fluctuating manner from 17 per 100,000 in 1980 to 20 per 100,000 in 1989, before decreasing in a fluctuating manner to 10 per 100,000 in 1992. The rate then increased to 13 per 100,000 in 1995, then decreased to 11 per 100,000 in 1998 before increasing again to 14 per 100,000 in 1999.
  
  
• For females of all ages, the rate remained relatively stable, fluctuating around 11 per 100,000 from 1980 to 1989, where it began to decrease gradually, reaching 8 per 100,000 in 1997 and remained stable to 1999.
  
  
• For males of all ages, the rate decreased gradually from 27 per 100,000 in 1980 to 19 per 100,000 in 1999.
  
  

Figure 68. Suicide rates before and after the implementation of National Suicide Prevention Strategy in Norway

Top of page

Text version of Figure 68

Figures in this description are approximate as they have been read from the graph.

Figure 68 shows suicide rates in Norway before and after the implementation of the National Suicide Prevention Strategy in 1994:

• For females aged 15-24 years, the rate increased gradually in a fluctuating manner from 4 per 100,000 in 1980 to 9 per 100,000 in 1999.
  
  
• For males aged 15-24 years, the rate increased gradually in a wildly fluctuating manner from 20 per 100,000 in 1980 to 28 per 100,000 in 1992. The rate then decreased to 21 per 100,000 in 1994, increased to 22.5 per 100,000 in 1995, and decreased to 20 per 100,000 in 1997 before increasing to 28 in 1999.
  
  
• For females of all ages, the rate increased very gradually from 7 per 100,000 in 1980 to 9 per 100,000 in 1988, before decreasing gradually to 6 per 100,000 in 1996. The suicide rate then increased to 7 per 100,000 in 1997 and remained stable to 1999.
  
  
• For males of all ages, the rate increased gradually from 18 per 100,000 in 1980 to 25 per 100,000 in 1988, before decreasing gradually to 18 per 100,000 in 1994 and increasing slightly to 19 per 100,000 by 1999.
  
  

Footnotes

² A statistically significant reduction in the slope of post-implementation suicide mortality trends as compared to that related to pre-implementation suicide mortality trends was adopted as an indicator of a probable positive impact of national suicide prevention strategies. It should be noted, however, that the correlational nature of the present study and analysis does not allow us to draw reliable conclusions regarding the presence of a causal relationship between national suicide plans and suicide mortality trends. Furthermore, the adopted indicator of impact was based on the assumption that the suicide mortality trends observed across five years before the implementation were stationary (i.e. would have remained the same across time if the national suicide prevention strategies had not been introduced).

prev page| TOC |next page