Suicide & Bipolar Disorder

Bipolar disorder (BD) is a common and severe major psychiatric illness with an estimated worldwide lifetime prevalence of between 1% and 5% (Goodwin & Jamison, 1990). It is associated with a high risk of premature mortality in general and a high risk of suicide in particular (Goodwin & Jamison, 1990). It has been estimated that as many as 10% of BD patients die by suicide, and as many as 40% report having made suicide attempts (Lopez, et al., 2001; Leverich, et al., 2003; Dalton, et al., 2003). Evidence also suggests that suicidal behaviour in BD patients has a higher lethality than in the general population (Tondo, Isacsson, & Baldessarini, 2003).

A number of risk factors for suicide attempts in BD have been identified in previous research, but less is known about risk for completed suicide or how such risk factors link together to precipitate episodes of suicidal behaviour (Hawton, et al., 2005). Risk factors for attempted suicide and self-harm in BD include: a family history of suicide (Lopez, et al., 2001), earlier onset of BD (Oquendo, et al., 2000; Lopez, et al., 2001; Dalton, et al., 2003), comorbid anxiety disorder, and substance misuse (Hawton, et al., 2008). For completed suicide only male gender, a history of suicide attempt and expressed hopelessness have been shown to be consistent risk factors (Leverich, et al., 2003; Tondo, et al., 2003; Simon, et al., 2007). Evidence of factors that reduce or protect against suicidal behaviour in BD, such as specific medical treatment, also remains limited and largely inconclusive (Baldessarini, & Jamison, 1999; Angst, et al., 2002). However there is consistent evidence for the efficacy of Lithium in reducing suicidal behaviours (Tondo, Hennen, & Baldessarini, 2001).

This study sought to fill the gaps in the current body of knowledge on risk factors for suicide in BD. Using existing datasets held by the National Confidential Inquiry and Manchester Self Harm Project which contain detailed information on a large number of BD patients, we hoped to build a more comprehensive picture of demographic, illness, and clinical management risk factors and associations between them. Factors found to be important would be investigated further in a series of in depth interviews, with BD patients who have made medically serious suicide attempts and the relatives of BD patients who have died by suicide, to explore how risk factors are linked together in a suicidal process.

Main research questions

• To understand the risk of suicide in bipolar disorder
• To understand the demographic, clinical, and management-related risk factors for suicidal behaviour in bipolar disorder
• To understand how these risk factors link together in episodes of self-harm or suicide

Study design

1. Analysis of existing databases, using controls where appropriate
2. Qualitative interview study of relatives of those who have died and bipolar disorder service users who have made serious attempts

Database review

Databases available for review include:

National Confidential Inquiry (NCI) into suicide: The main NCI suicide database contains detailed information on all people who die by suicide or who receive an open verdict at coroner’s inquest in England and Wales (1996 to present day) that were in contact with mental health services in the 12 months prior to death. Data is collected via questionnaire and covers; demographics, clinical history, details of the suicide act, details of care (in-patient and community services), details of final contact with services, events leading to suicide, and clinicians views on prevention. There are currently 1576 cases with a diagnosis of BD on the database (from a total sample of 15, 721).

Case-Control studies of suicide by in-patients and suicide within 3 months of discharge: Both case-control studies used a consecutive series of suicides extracted from the main NCI database, where the death had occurred either whilst in in-patient care or within three months post discharge from in-patient care. One control per case was randomly selected who satisfied the criteria of being alive and i) being in in-patient care on the date of the suicide of the corresponding case, or ii) of having been in in-patient care and being discharged on the same day as the corresponding case. The control questionnaire was an adapted version of the suicide questionnaire but with reference to index date rather than suicide date. There are 81 in-patient, and 43 post discharge cases/controls with a diagnosis of BD on these databases (total sample 960).

Psychological autopsy study: The psychological autopsy study used a consecutive sample of suicides extracted from the main NCI database. Data were extracted from coroners’ files, GPs medical records, and emergency department records. GPs also participated in semi structured interviews covering; clinical and social factors leading up to death, views on the mental health care the patient received and views on prevention. There are 21 cases with a diagnosis of BD on the database (from a total sample of 286).

Manchester self-harm project (MaSH): The MaSH project collects information on all patients presenting following a self harm episode to three Emergency Departments in the city of Manchester. The information collected includes details of the self-harm episode, socio-demographic characteristics, clinical characteristics, current mental state, clinical assessment of risk, and subsequent management. Few self-harm episodes result in an assessment by a mental health professional and only a minority result in formal psychiatric diagnoses, but these details are recorded and there are currently 102 episodes recorded with a diagnosis of BD (from a total sample of 32286).

Interview Study

To further explore factors associated with suicide and self-harm in BD the relatives of individuals with BD who died by suicide within 12 months of contact with services, and those with BD who have made a medically serious self-harm attempt would be interviewed. Topics to be covered in interviews would include a detailed history of psychiatric disorder, treatments received, experience of services, mental state in the weeks and days immediately prior to the episode, immediate antecedents, details of the episode, subsequent events, hospital management (for cases of self-harm), and views on what might have prevented the episode. Interviews would be recorded, transcribed and anonymised. The details of the interview schedule would be informed by existing literature and the results of the database review. Recruitment would take place in (Greater Manchester and Trent).