Repeat non-fatal suicidal behaviour at Johannesburg Hospital

Objective. To describe the characteristics of non-fatal suicide behaviour (NFSB) in a group of patients and to determine factors, if any, that may be associated with repetition of this behaviour. Method. The study included all patients treated for NFSB at Johannesburg Hospital during the period August 2002 - October 2002. The information was gathered by means of a structured questionnaire designed to evaluate characteristics of the behaviour. Results. The study sample comprised 43 patients with NFSB (mean age 29.7 years, range of 16 - 75 years), of whom 26 (60%) were female. Sixty-three per cent of the patients overdosed with medication and 33% ingested household poisons. Events that precipitated the event included relationship problems (70%), illness (12%), financial difficulties (9%), and depressed mood (9%). In 65% of patients the behaviour was impulsive. Factors associated with non-fatal repetition included being in the 18 - 30-year age group (76%) (χ2 = 6.74, p < 0.05); being female (90%) (χ2 = 4.75, p < 0.05); having children (90%) (χ2 = 4.72, p < 0.05); a past psychiatric history (50%) (χ2 = 4.08, p < 0.05); and the current attempt deemed medically serious (50%) (χ2 = 6.67, p < 0.05). Conclusions. NFSB is a major problem in South Africa and the incidence is still increasing. Hospital-based interventions following admission are recommended to reduce repeat attempts in such patients. Significant factors associated with non-fatal repetition include among others, a history of a previous medically serious attempt and/or a known psychiatric illness.

About one-third of the general population has, at some point, experienced thoughts of self harm. 1 3 Deaths resulting from homicide and unintentional injury usually outnumber those resulting from fatal suicidal behaviour (FSB). 4Although it is estimated that worldwide about 25% of FSB is preceded by nonfatal suicidal behaviour (NFSB) in the preceding year, 5 importance of the latter is usually under-estimated.
There is considerable debate about the underlying causes and the manner in which biological and psychological factors interact in NFSB.Reported risk factors for this behaviour include a history of a psychiatric disorder, [6][7][8][9] advancing age, [10][11][12][13][14] living alone 15 or in a low-income area, 10 current mental illness, 11 somatic illness [16][17][18][19] and the abuse of alcohol and drugs. 15Other factors are exposure to various forms of stress 14,20 25 (adversity, discrimination, economic hardship, marital problems, and social disputes).There is also a significant association between having a friend or relative who committed FSB and the individual's attitude towards this behaviour. 26Young females have much higher rates of NFSB 23 but lower rates of FSB than males. 10,11,21 22 27 28mmon methods employed in NFSB include the ingestion of harmful substances (paraffin, pesticides or battery acid 22 25 ), overdosing with medicines, 23 24 use of sharp weapons, and attempted hanging. 26In many cases the behaviour can be categorised as demonstrative rather than genuine. 25The major intentions or reasons given for committing this behaviour are 'failing to solve problems' and 'mental illness'. 26previous episode of NFSB ranks as one of the major risk factors for future FSB, as does recent discharge from inpatient psychiatric care. 9,26 29About 1 in 6 patients repeat the behaviour over the next year, and 1 in 4 after 4 years.Nonfatal repetition is also associated with unemployment, increasing severity of suicidal ideation, previous psychiatric treatment and some personality disorders. 30,31uth Africa is a society in a state of transition and its citizens have been and continue to be severely traumatised. 32NFSB appears to be on the increase as people struggle to come to Results.The study sample comprised 43 patients with NFSB (mean age 29.7 years, range of 16 -75 years), of whom 26 (60%) were female.Sixty-three per cent of the patients overdosed with medication and 33% ingested household poisons.Events that precipitated the event included relationship problems (70%), illness (12%), financial difficulties (9%), and depressed mood (9%).In 65% of patients the behaviour was impulsive.Factors associated with non-fatal repetition included being in the 18 -30-year age group (76%) (χ 2 = 6.74, p < 0.05); being female (90%) (χ 2 = 4.75, p < 0.05); having children (90%) (χ 2 = 4.72, p < 0.05); a past psychiatric history (50%) (χ 2 = 4.08, p < 0.05); and the current attempt deemed medically serious (50%) (χ 2 = 6.67, p < 0.05).

Method
The study subjects included all patients treated for NFSB at the Johannesburg Hospital adult medical emergencies ward during the period August 2002 -October 2002.They were interviewed approximately 24 hours after admission or when medically stable.

Assessment
After obtaining informed consent psychiatry registrars interviewed the patients.All the interviewers received training before the study to improve inter-rater validity and reliability.Data were obtained using a structured questionnaire designed to evaluate subject characteristics (marital status, family history, employment status, highest level of education achieved, current accommodation, etc.) and factors associated with the behaviour (method employed, precipitating event, premeditation, leaving a suicide note, feelings about the outcome, etc.)The seriousness of the attempt was determined by the subject's physical status and by means of laboratory investigations.
All subjects gave written informed consent to participate in the study (in the case of subjects under 18 years of age, consent was obtained from the parents, with assent from the patient).
The study was approved by the Committee for Research on Human Subjects, University of the Witwatersrand.

Statistical analysis
Descriptive statistics of variables were computed as mean and frequencies (count and percentages).The two-sample t-test was used to compare the continuous characteristics (age) between the groups.The chi-square test and Fisher's exact test were used to determine the relation between categorical characteristics.
Pearson's correlation coefficient (r) was used to describe any correlations between variables.All analyses were done using the Statistical Package for Social Sciences 10.0 for Windows (SPSS Inc., Chicago, Ill.)A value of p < 0.05 was considered significant.
Seventy-two per cent of the attempts were deemed not medically serious and the majority of subjects were discharged from hospital for ambulatory care.Sixty-five per cent of the patients indicated that they had a desire to die at the time of attempt while 21% said they were trying to make a point.At the time of interview, 58% of the patients were glad that they had survived and the majority (77%) said they had no intention of repeating the behaviour.
There was a significant association between the seriousness of the behaviour and the behaviour being premeditated (p = 0.017), a past history of NFSB (p = 0.005), a family history of NFSB (p = 0.012), a family history of psychiatric illness (p = 0.032), and a history of sexual or physical abuse (p = 0.001).
Patients were significantly more likely to be admitted to hospital if they had a history of a previous NFSB (p = 0.009) or if they had a low level of education (p = 0.039).They were more likely to be treated as outpatients if they had a family history of psychiatric illness (p = 0.044) or if they had children of their own (p = 0.004).

Discussion
In contrast to previous reports of increasing age as a feature of patients with NFSB, [10][11][12][13][14] the patients in this study were much younger and predominantly in the 18 -30-year age group.

Higher incidences of this behaviour have recently been reported
3][14] Younger persons may be particularly at risk because of educational and socioeconomic demands, high unemployment rates and unmet expectations. 12ong the subjects in this study, relationship problems, financial difficulties, illness and depressed mood were reported as common precipitants of the behaviour.
South Africa is undergoing rapid urbanisation (95% of our subjects lived in the city) and the stress associated with this relocation may be a contributing factor. 30,31erdosing with common medicines (influenza preparations and analgesics) and the ingestion of household poisons were the most common NFSBs in our subjects.In most cities in the developing world similar medicines are often used for selfharm. 34 36In a case series from Cape Town the majority of patients stated that they used battery acid because it was readily available in car batteries used as a power source in their unelectrified houses, and also because they were aware of its destructive effects. 21,37,38Poisoning is a common form of deliberate self-harm and while suicidal intent is often far lower than in cases of self-immolation and hanging, the mortality rate is high owing to the toxicity of the agents used.Traditional medicines are a cause of accidental, but rarely intentional, poisoning. 39Outside the cities these methods are relatively uncommon and their prominence is displaced by pesticides, which are often fatal.
Although the majority of our patients indicated that they had a desire to die at the time of the attempt, they were glad to have  survived and said they had no intention of repeating the behaviour.1][42][43] Instead, the acts are used to express rage or hostility, or to gain revenge by causing distress to another person.In some cultures this may be seen as the only way to express one's anger with someone. 44,45People who do want to kill themselves often do not succeed; in contrast, others with little or no suicidal intent sometimes die as a result of their act. 46Many factors affect the outcome, including the degree to which the toxicity of the poison was understood, the speed with which the person comes to clinical attention, and the availability of effective medical treatment.

Conclusion
NFSB is a major problem in South Africa and the incidence is still increasing.In suggesting effective ways to prevent this problem, we must be realistic and aware of our limited resources.However, acknowledging the seriousness of the situation is a first step towards preventing this unnecessary behaviour.
A significant factor associated with non-fatal repetition includes a past psychiatric history and a medically serious attempt.Improved mental health care, particularly at community level, must be an important part of any strategy to reduce self-harm.
Approaches to primary prevention may include increasing peoples' coping skills -possibly offering coping-skill classes at school, and counselling in the community.
Moosa, MMed (Psych), FC Psych F Y Jeenah, MMed (Psych) M Vorster, PhD (Psych), MMed (Psych) Department of Neurosciences, Division of Psychiatry, University of the Witwatersrand, Johannesburg Objective.To describe the characteristics of non-fatal suicide behaviour (NFSB) in a group of patients and to determine factors, if any, that may be associated with repetition of this behaviour.Method.The study included all patients treated for NFSB at Johannesburg Hospital during the period August 2002 -October 2002.The information was gathered by means of a structured questionnaire designed to evaluate characteristics of the behaviour.
Africa and the incidence is still increasing.Hospital-based interventions following admission are recommended to reduce repeat attempts in such patients.Significant factors associated with non-fatal repetition include among others, a history of a previous medically serious attempt and/or a known psychiatric illness.

articles Volume 11
No. 3 December 2005 SAJP terms with the effects of former South African racial policies, and related sociocultural, socioeconomic and other pressures.It is necessary to ascertain the characteristics of NFSB across our various sociodemographic groups.In a previous report33 a group of patients with NFSB was compared with a control group without this behaviour.The aim of the present study was to describe, in this same group of patients, the characteristics of the NFSB and to determine factors, if any, that may be associated with repetition of this behaviour.

articles Volume 11
No. 3 December 2005 SAJP Hospital-based interventions after admission for NFSB are recommended in an attempt to reduce repetition.In South Africa a reduced number of inpatient beds has meant that medical staff are reluctant to admit patients judged to be at low physical risk; the latter are also often seen as difficult and unrewarding cases.Psychiatric services are increasingly reserved for those with serious mental illness, a term not taken to include most cases of NFSB.The current situation should not be allowed to continue, because NFSB represents a major social and clinical problem.At the least, large-scale intervention studies are required to inform practice and ensure that management of NFSB is less arbitrary in the future.

Funding was received from
the Department of Neurosciences, Division of Psychiatry, University of the Witwatersrand.articles Volume 11 No. 3 December 2005 SAJP

Table II . Factors associated with non-fatal repetition
*p < 0.05.

Table I . Characteristics of non-fatal suicide behaviour in 43 patients at Johannesburg Hospital
articles Volume 11 No. 3 December 2005 SAJP