Suicide-preventive training has shown to influence attitudes. This study aimed at investigating what impact other factors than knowledge might have on attitudes towards work with suicidal patients and suicide prevention. In 2007, 500 health-care staff working in a psychiatric clinic in Stockholm received a questionnaire with items concerning work with suicidal patients to which 358 (71.6%) responded. A set of attitude items were tested using structural equation modelling (LISREL). Three models were found to be satisfactory valid and reliable: Job clarity, Job confidence and Attitudes towards prevention. These were then used in regression analyses as dependent variables with predictors such as experience of work with suicidal patients, perceived sufficient training, age and gender. Perceived sufficient training was consistently the most important predictor for all three attitude concepts (p < 0.01, β = 0.559 for Job clarity; p < 0.01, β = 0.53 for Job confidence; p < 0.01, β = 0.191 for Attitudes towards prevention). Age was another significant predictor for Job clarity (p < 0.05, β = 0.134), as was experience of patient suicide for Job confidence (p < 0.05, β = 0.137). It is concluded that providing suicide preventive education is likely to improve attitudes towards the prevention of suicide, clarity and confidence regarding their role in the care for suicidal patients. These improvements may contribute to the prevention of suicide in health care settings.

Despite caregivers' impact on suicidal patients' compliance with treatment and suicide prevention, little is known about mental-health professionals' perceptions of work with suicidal patients. The roles of psychiatric staff's training and supervision in the care of suicidal patients were investigated by means of a postal questionnaire sent to a random sample of 1543 psychiatric staff members. The response rates were 71% for psychiatrists and 57% for nurses and assistant nurses. The responses of 53 psychiatrists, 164 nurses and 333 assistant nurses working with suicidal patients on a regular basis were compared and analysed using the Kruskal-Wallis test. Thirty-five per cent of the assistant nurses, 43% of the nurses and 74% of the psychiatrists who worked with suicidal patients on a regular basis perceived that they were sufficiently trained for this work, while 75% of the assistant nurses, 72% of the nurses and 34% of the psychiatrists received supervision in their work with suicidal patients. In spite of receiving supervision, nursing staff who perceived that they lacked training reported uncertainties in their work with suicidal patients to a larger extent than those who perceived that their training was sufficient. Uncertainties were significantly more prevalent among nursing staff than among psychiatrists. Basic and specific training in suicidology is needed and cannot be replaced solely by supervision, since psychiatric staff often have to deal with suicidal patients in emergency situations and must be able to rely on their own skills.

Reports on lifetime prevalence of suicidal behaviour may be biased in populations that are not reminded of these problems in everyday life. Data on past year prevalence of suicidal behaviour show clearly the similarity between the general population and the mental health-care staff.

A 200-hour academic, postgraduate training-of-trainers program in suicide prevention at Karolinska Institute for key mental healthcare staff, designed to enable them to enhance their co-workers' knowledge and job clarity, was evaluated in a panel study by means of questionnaires. Psychiatric staff working regularly with suicidal patients in clinics where key persons attended the course (n = 134, intervention group) were compared with staff working in clinics without participants (n = 166, control group). Perceptions of being sufficiently trained (p < .01) improved significantly among staff working in intervention clinics. Compared with the control group, the intervention group had a better understanding of essentials (p < .05); found instructions clearer (p < .01) and experienced fewer problems with superiors' differing views (p < .05) at follow up. Assistant nurses working in intervention clinics seem to have benefited most.