Retrospective controlled study of inpatient ECT: does it prevent suicide?

Verinder Sharma
Journal of Affective Disorders
56 (1999) 183-187


Background: This study examined the use of ECT among inpatients who committed suicide at a provincial psychiatric hospital. Methods: A total of 45 psychiatric in-patients who committed suicide at a provincial psychiatric hospital were compared with a gender, age and admission diagnosis matched group of 45 hospitalized patients to examine the use of electroconvulsive therapy during the last 3 months of hospitalization. Results: No difference in the utilization of ECT was found in the two groups. Limitations: Retrospective design and small sample size. Conclusions: We failed to demonstrate that ECT had prevented suicide in hospitalized patients. Future prospective studies with large sample size are needed to further examine this question.

1. Introduction

Electroconvulsive therapy (ECT) is recommended as an initial treatment for suicidal depression (American Psychiatric Association, 1993). It is more effective than alternate treatments for severe depression and its rapid onset of action reportedly results in a reduction in the immediate risk of suicide (Goodwin and Jamison, 1990a). ECT is also an important treatment option in some bipolar patients but there is conflicting information whether it leads to acceleration of cycle frequency (Kukopulos et al., 1980; Winokur and Kadrmas, 1989). Studies have shown a lower incidence of suicide among depressed patients treated with ECT as compared with other treatment modalities (Avery and Winokur, 1976, 1978). ECT may also lower mortality and suicide risks in schizoaffective patients (Tsuang et al., 1979), however, the results of another study showed the treatment modality had little relationship with subsequent mortality including suicide (Black et al., 1989).

The utilization rate of ECT in patients who subsequently commit suicide is quite low. Only two of the 1397 suicide victims in Finland over a 12-month period were treated with ECT during the final 3 months. This figure is particularly low considering the prevalence of severe major depression at the time of suicide. A diagnostic study of a random sample of 229 suicides in this study showed a presence of major depression in 31% of the cases and 39% had been hospitalized at some point during the final three months. Given the extremely small number of patients committing suicide within 3 months of receiving ECT, the authors of this Finnish study concluded that ECT has a preventative effect on suicide (Isometsa et al., 1996). However, there was no control group and the base rate of ECT use in Finland during the study period was not specified. Barraclough et al. (1974) also reported that ECT had been neglected in some patients who eventually committed suicide, even though it had previously been successful.

In this study we present results of a case control study examining the use of ECT in a group of 45 in-patients who committed suicide at a psychiatric hospital during the period of 1969-1996 inclusive.

2. Material and methods

In a previous study we examined the risk factors for suicide among 44 in-patients at a provincial psychiatric facility who committed suicide between January 1969 and December 1995 by conspiring with an age and sex matched control group of 45 in-patients (Sharma et al., 1998). Sixty-five patients committed suicide during the study period but 21 patients who killed themselves 3 days or more after leaving the hospital on an authorized or unauthorized leave were excluded. This exclusion criteria was applied to enable us to carefully examine the course of the illness proximal to suicide. The study period for the current study was extended to December 1996. The patient admitted next after the suicide victim matched for gender, age (5 years) and admission diagnosis (DSM-IV) served as the control. The hospital charts which included reports from previous hospitalizations were studied carefully and data extracted. The suicide and the control groups were compared using one-way analysis of variance (ANOVA) and chi-square on various sociodemographic and clinical variables such as age, sex, marital status, education, employment, living arrangements, number of previous hospitalizations, number of comorbid diagnoses, age at onset of psychiatric illness and length of index admission. The suicide and control groups were compared for the use of ECT during the three months prior to suicide or discharge including the number of patients receiving this treatment modality, the number of ECTs, the type of ECT and the time period between ECT and suicide or discharge.

3. Results

There were 31 males and 14 females in the suicide group. Twenty-one patients were single, 13 were married and 11 were divorced or separated. The mean age at the time of suicide was 38.94.6 years. The majority of the suicide victims suffered from mood disorders. Major depression was present in 19 (45.2%) and 12 (26.7%) patients were given the diagnosis of bipolar disorder. Schizophrenia was present in six (13.3%) patients and schizoaffective disorder in four (9%) patients. Four (8.9%) patients received a diagnosis other than the ones mentioned above. No significant differences were found between the two groups on sociodemographic: variables as shown in Table 1.

3.1. Electroconvulsive therapy

Fight patients in the suicide group and four in the control group received ECT during the index admission. One patient from the suicide group and two from the control group were excluded having received ECT mom than 3 months prior to the time of suicide/discharge. No patient among the 21 suicide victims who were excluded received ECT. Table 2 shows clinical data for the suicide victims and control subjects who received ECT during the 3 months before suicide or discharge respectively. Three patients committed suicide following completion of a course of ECT, two suicided during a course of ECT, one was receiving maintenance ECT, and in one patient ECT was discontinued due to lack of efficacy. Three patients had bilateral ECT, three had unilateral ECT and one had an unknown type of treatment. These patients had failed to respond to adequate trials of psychotropic drugs prior to receiving ECT Using one-way ANOVA, there was a trend for a significant difference for time between ECT and suicide/discharge (F = 4.2, P = 0.08) and number of ECTs during index admission (F = 4.0, P=0.09).

4. Discussion

Nearly 16% of suicide victims received ECT during the final three months. This figure is considerably higher than the one reported by Isometsa et al. (1996) who examined all completed suicides in Finland within a 12-month period. The number of patients hospitalized at the nine of suicide was not specified in this study. The higher rate of utilization of ECT in our study was likely due to two reasons. First, the majority of the patients suffered from a mow disorder especially major depression which is the commonest indication for ECT. Second, the psychiatric disorder was severe enough to require hospitalization. In another Canadian study of suicide amongst psychiatric in-patients, 13 out of 37 patients received ECT during the index admission (Roy and Draper, 1995). Eleven of these patients suffered from schizophrenia and had been in hospital for extended periods of time. However, there was no statistically significant difference between the number of suicide victims and controls receiving ECT.

All of the suicide patients (3) who completed ECT were considered non or partial responders. Only one control subject was considered to have shown a favourable response. It can be argued that the patients who committed suicide suffered a relapse after showing initial improvement but this was not the case. It is of note that dim of the partial/nonresponders to ECT had previously benefitted from this treatment Diminished response to subsequent trials of ECT has been noted in patients with histories of multiple episodes (Goodwin, and Jamison, 1990b).

Two patients committed suicide during the course of bilateral ECT which is generally considered more effective than unipolar ECT. Both of time patients were on lithium and one on the combination of lithium and carbamazepine prior to initiation of ECT. The rather abrupt discontinuation of mood stabilizers in these patients which was done to minimize the risk of confusion may have been a factor in exacerbating the risk for suicide. Both were highly apprehensive about ECT and refused to have further treatments due to memory impairment following the first treatment Suicide during the early course of ECT in a small number of patients was also reported by Barraclough et al. (1974). Suicidal patients who fail to respond to a trial of ECT or those who decline to have further treatments because of concerns regarding memory impairment should be viewed as high risk candidates. Patients may view ECT as the last resort and may be at a particularly high risk for suicide in case of a lack of favourable response to this treatment modality. The risk and benefit of discontinuation of mood stabilizers should be carefully weighed in patients receiving ECT.

In summary, findings of this study did not show a difference in the utilization, rate of ECT among hospitalized patients who committed suicide and a group of matched controls during the last three months of hospitalization. The majority of patients who committed suicide following completion of the course of ECT did so within a month and those who declined to have further treatment killed themselves within a week of the final ECT. This is the only controlled study to my knowledge to examine the utilization rate of ECT among inpatients who commit suicide. However, there are some methodological problems in this study including the retrospective design, small sample size and relatively low exposure to ECT yielding limited statistical power. Due to rather small number of patients in the two groups, it was not possible to compare the technical data on ECT.


Dr E. Persad is thanked for his help with the chart reviews and Karen Kueneman for her assistance in data collection and analysis.


American Psychiatric Association, 1993. Practice Guideline for Major Depressive Disorder in Adults Am J. Psychiatry 150(S), 1-23.

Avery, D., Winokur, G., 1976. Mortality in depressed patients treated with electroconvulsive therapy and antidepressants Arch. Gen. Psychiatry 33, 1023-1029.

Avery, D., Winokur, G., 1978. Suicide, attempted suicide and relapse rates in depression: occurrence after ECT and antidepressant therapy. Arch. Gen. Psychiatry 35, 749-753.

Barraclough, B., Bunch, J., Nelson, B., Sainsbury, P., 1974. A hundred cases of suicide: clinical aspects. Br. J. Psychiatry 125, 355-373.

Black, D.W., Winokur, G., Mohandoss, E., WooIson, R.F., Nasralab, A., 1989. Does treatment influence mortality in depressives? A followup of 1076 patients with major affective disorders. Am. Clin. Psychiatry 1, 165-173.

Goodwin F.K., Jamison, K.R., 1990a. Manic-Depressive Illness. Oxford Press, New York, p. 778.

Goodwin, F.K., Jamison, K.R., 1990b. Manic-Depressive Illness. Oxford Press, New York. pp. 660-661.

Isometsa, E.T., Henrikisson, M.M., Heikkinen, ME, Lonnqvist, J.K, 1996. Completed suicide and recent electroconvulsive therapy in Finland. Convuls. Ther. 12 (3), 152-155

Kukopulos A, Reginaldi D, Laddomada P, Floris G, Serra G, Tondo L., 1980.
Related Articles Course of the manic-depressive cycle and changes caused by treatment. Pharmakopsychiatr Neuropsychopharmakol.. 13:156-67.

Roy, A., Draper, R., 1995. Suicide among psychiatric hospital in-patients. Psychol. Med. 25, 199-202,

Sharma, V., Persad, E., Kueneman, IC, 1998. A closer look at inpatient suicide. J. Affect. Disord. 47, 123-129.

Tsuang, M.T., Dempsey, G.M., Fleming, J.A., 1979. Can ECT prevent premature death and suicide in 'schizoaffective' patients? J. Affect. Disord. 1, 167-171.

Winokur G., Kadrmas, A., 1989. A polyepisodic course in bipolar illness: possible clinical relationships. Comp. Psychiatry 30 (2), 121-127.