We aimed to determine the number and characteristics of psychiatric patients receiving electroconvulsive therapy (ECT) who had subsequently died by suicide. Data were collected on an 8-year (1999–2006) sample of suicide cases in England who had been in recent contact with mental health services. Of 9752 suicides, 71 (1%) were being treated with ECT at the time of death. Although the number of patients who received ECT had fallen substantially over time, the rate of suicide in these individuals showed no clear decrease and averaged 9 deaths per year, or a rate of 10.8 per 10,000 patients treated. These suicide cases were typically older, with high rates of affective disorder and previous self-harm. They were more likely to be an in-patient at the time of death than other suicide cases. Nearly half of the community cases who had received ECT had died within 3 months of discharge. Our results demonstrated that the fall in the use of ECT has not affected suicide rates in patients receiving this treatment. Services appear to acknowledge the high risk of suicide in those receiving ECT. Improvements in care of these severely ill patients may include careful discharge planning and improved observation of in-patients in receipt of ECT.
Electroconvulsive therapy (ECT) is considered an effective short-term treatment for patients with severe depressive illness (UK ECT Group, 2003). It is usually an end-stage therapy, recommended for those who have not responded to other therapies and whose severity of illness may be life-threatening (National Institute for Health and Clinical Excellence, 2003). However, guidelines of the American Psychiatric Association task force state that ECT should not be reserved for use as a “last resort” but promotes its use as a first-line treatment when there is a need for a prompt and definitive response, when patients have responded well to previous ECT administrations, and when the treatment is preferred by the patient (Task Force on Electroconvulsive Therapy, 2001).
In England, there has been a reduction in its use over the past few decades. In 1999, 2835 patients received ECT over a 3-month period, equating to an annual estimated total of 11,340 patients (Department of Health (DoH), 1999). By 2002, this annual figure had dropped to an estimated 9088 patients (DoH, 2003). These figures may be underestimates, however, as administration of ECT may not always be documented and there has been no formal national monitoring of ECT activity in the UK (DoH, 2003). A more recent study in 2006 confirmed a further decline in ECT administrations overall, although reported an increase in the proportion of patients receiving ECT who had been detained under the Mental Health Act (Bickerton et al., 2009). Possible reasons for the decline in its use include the availability of newer antidepressants, improved care in the community and earlier recognition of mental illness (Eranti and McLoughlin, 2003). In the US, the use of ECT has also declined since the 1970s but there are still an estimated 100,000 ECT procedures conducted each year (Hermann et al., 1995).
There is good evidence that ECT is more effective than antidepressants in reducing depression and suicidal intent among patients with severe depression (UK ECT Group, 2003 and Kellner et al., 2005). The incidence of suicide attempt has also been found to be lower in patients treated with ECT compared to antidepressants (Brådvik and Berglund, 2006). The impact of ECT on completed suicide, however, is less clear, particularly as the proportion of those who die by suicide who have received ECT is very low. In Finland, for example, Isometsä and colleagues found that, over a 12-month period, only 2 of the 1397 suicide cases had received ECT within 3 months of death, representing 0.14% of all suicides (Isometsä et al., 1996).
The majority of studies have found no relationship between ECT and completed suicide (Milstein et al., 1986, Black et al., 1989 and Sharma, 1999). For example, of 30 patients who died within 14 days of receiving ECT between 1993 and 1998, Shiwach et al. reported 8 cases (27%) had died by suicide, but that there was no association with the treatment — the rate of suicide being high in those with severe depression (Shiwach et al., 2001). However, Munk-Olsen and colleagues reported an increased risk of suicide (relative risk = 4.82, 95% CI 2.12–10.95) in patients who had received ECT in the week before death compared to non-ECT patients (Munk-Olsen et al., 2007). The authors acknowledged this finding was based on a small sample (6 suicide cases) and may have been explained by selection bias in that patients treated with ECT were already at high risk of suicide. In addition, this finding may be related to early treatment response mechanisms that have been suggested with drug therapies. For example, antidepressants are known to improve depressive symptoms but they may also possibly increase suicidal thoughts and behavior in the early stages of treatment (Jick et al., 2004 and Hall, 2006). Explanations include a delay in any therapeutic effects until later in the treatment phase, or that antidepressants are prescribed at a time when depressive symptoms are acute and therefore the risk of suicidal behavior is already increased (Jick et al., 2004). Alternatively, it may be that for some patients receiving pharmacological treatment a rise in energy levels precedes an improvement of depressive symptoms and they therefore may have the motivation to carry out suicidal acts (Bostwick, 2006). Suicides following ECT may also be an indication of treatment failure.
There have been no large detailed studies describing the characteristics of patients treated with ECT who have died by suicide. Our aims were firstly to determine the number of suicide cases by people under mental health care who were receiving a course of ECT treatment at the time of death in England over an 8-year period. Secondly, to compare the social, behavioral and clinical features of these suicide cases with those who had not received ECT. The study was carried out as part of the National Confidential Inquiry into Suicide and Homicide by People with Mental Illness (Appleby et al., 2001).
