Trends in suicide case fatality in Italy, 1983–2007

Abstract

The proportion of suicide attempts ending up in the death of the attempter was used in past studies as an index of suicide lethality, or case fatality. This study aimed at investigating whether case fatality of suicide has decreased in Italy over the latest 25 years using available data, as an alternative hypothesis to the proposed general decrease in suicidal behavior resulting from better identification and treatment of people with mental disorders. The official data on completed and attempted suicides by males and females in Italy, from 1983 to 2007, were analyzed with joinpoint regression analysis, to identify the points (i.e., “joinpoints”) where linear trends changed significantly in direction or magnitude. It should be noted that only the most severe suicide attempts are recorded in Italian official statistics. Suicide rates decreased in both sexes, particularly from 1990 onward. Attempted suicide rates increased progressively in males, while in females they reached their peak in 1996–1998 and then decreased. In both sexes suicide case fatalities significantly decreased from 1990 onward. Improved survival after a suicide act is probably the main reason behind this favorable trend. The spreading of emergency services may prevent suicide.

Introduction

Suicide is the lethal outcome of behaviors intentionally initiated by the deceased to cause his/her own death. According to a widely agreed operational definition, suicide is any “death from injury, poisoning, or suffocation where there is evidence (either explicit or implicit) that the injury was self-inflicted and that the decedent intended to kill himself/herself” (O'Carroll et al., 1996); while suicide attempt refers to “a self inflicted, potentially injurious behavior with a nonfatal outcome for which there is evidence (either explicit or implicit) of intent to die; a suicide attempt may result in no injuries, injuries, or death (Silverman et al., 2007). Within this framework, a suicide act has been defined as any self-inflicted injury, whether fatal or nonfatal (O'Carroll et al., 1996). The proportion of suicide acts ending up in suicide is used as an index of lethality, or case fatality.

This measure is an approximation of the real lethality of suicide acts, since official records register the most severe suicide attempts only: many suicide acts do not require hospital care, or go undetected because they are hidden by the physician, who agrees to the desire of the patient or his/her family not to record the act out of fear of stigmatization (Hatcher et al., 2009). Some studies found that up to 30% of suicide attempts did not receive medical treatment, or were treated by general practitioners without referral to a higher level of care (Card, 1974, Van Caasteren et al., 1993).

Past studies found that men and elderly people had the largest case fatality proportion (Elnour and Harrison, 2008, Jansen et al., 2009), principally because they tend to choose more lethal methods, with firearms, drowning, hanging and jumping from heights as the most lethal (Miller et al., 2004, Reisch et al., 2008, Chen et al., 2009).

A concomitant severe somatic illness may decrease the chances of surviving a suicide attempt: this might be one reason for higher suicide case fatality in the elderly (Juurlink et al., 2004). High case fatality in the elderly can also be explained by more elderly people living alone, who therefore have fewer chances to be rescued and hospitalized after their suicide attempt (O'Connell et al., 2004). Surveillance and prompt rescue interventions are critical to survival after suicide attempt: for example, hanging is a method with a very high case fatality (70%), but most of those who are taken to hospital alive, survive (Gunnell et al., 2005).

Recent studies pointed to a consistent decrease in suicide rates in the general population in the latest 20 years, attributed principally to better identification and treatment of mental disorders (Isacsson, 2000, Rihmer, 2001). In Italy, too, lower suicide rates were reported in both males and females (Rocchi et al., 2007, Vichi et al., 2010). Nevertheless this decrease observed worldwide could be merely the result of better survival after a suicide attempt because of improved medical protocols of emergency care, and the spreading of emergency services across countries, even to places that lacked such services before (e.g., rural areas: see Hirsch, 2006).

So far no study has investigated the issue. One study found a decline in the number of deaths by self-poisoning between 1970 and 2000 in Denmark, and the decline was related to restrictions on the availability of carbon monoxide, barbiturates and dextropropoxyphen, which are considerably more lethal than other self-poisoning methods (Nordentoft et al., 2006). In Germany, too, a decline in the number of suicides by self-poisoning was observed between 1991 and 2002, while the number of suicides achieved with firearms or cutting instruments remained stable, suggesting that case fatality with the highest lethality is less influenced by improvement in emergency protocols of care (Baumert et al., 2008). In Australia, suicide case fatality remained stable for firearm cases from 1993 to 2003, but a decline was reported in suicide case fatality by hanging/suffocation, self-poisoning, cutting instruments and crashing a motor vehicle (Elnour and Harrison, 2008). In Japan, from 1980 to 2000 the ratio of attempted suicide to completed suicide rates changed little in males but increased in females twofold to fivefold, indicating higher survival after a suicide attempt (Yamamura et al., 2006).

In Italy there is no code to record a suicide attempt at the time of hospitalization. However, police statistics include a section on suicide attempts, since in Italy both the instigation to suicide and assistance in suicide are crimes: therefore all self-injuries that come to the knowledge of the police or the judicial authority are investigated to ascertain possible third-party involvement in the mechanics of the event. Past studies found that attempted suicides are three to ten times more numerous than completed suicides (Spicer and Miller, 2000, Shenassa et al., 2003, Chen et al., 2009). However, Italian official police statistics report more completed than attempted suicide cases, because only the most severe suicide attempts undergo police investigation, i.e. those that produced social alarm or became publicly known because of the dynamics of the attempt (e.g., crashing a motor vehicle, jumping from the roof of a public building). Less severe suicide attempts that do not jeopardize the life of the attempter are unlikely to be signaled to the police. Therefore the suicide attempts recorded in the Italian police statistics resemble more failed suicide than deliberate self-harm, as is instead recorded in the hospital statistics of other countries. Past studies found that completed suicides and medically serious suicide attempts are overlapping populations (Beautrais, 2001). Data from Italian police statistics can thus offer a clue on suicide case fatality from the perspective of real, severe suicidal intent.

This study was intended to investigate whether case fatality of suicide has decreased in Italy over the latest 25 years based on available data, as an alternative hypothesis to the proposed general decrease in suicidal behavior – hence fewer suicide acts – resulting from better identification and treatment of people with mental disorders.