Factors associated with suicides in Wolverhampton: Relevance of local audits exploring preventability

Abstract

Suicide rates and associated risk factors are known to change over time. The periodic evaluation of suicides in a particular locality may identify specific issues that may help in prevention efforts. The profile of 146 consecutive cases with suicide, open and narrative verdicts between January 2004 and July 2011 in Wolverhampton was studied in order to explore associated factors. Specific information about methods, mental health issues and stressors was collected. There was considerable variation in suicides in different wards of the city, with higher occurrences in inner-city areas. Male suicides were three times more common, and they were significantly younger than in female suicides. Common methods were hanging (52.7%) and poisoning (21.9%) involving a wide range of drugs. Life events were reported in 52.1% suicides, most frequently relational problems (28.1%), followed by physical illness and bereavement. Mental-health factors were associated with 63.0% of suicides, predominantly depression (45.9%) and alcohol and drug abuse (6.8%). The majority of the suicide victims (63.0%) were not in contact with mental-health services, including: most men (68.8%), young suicides up to age 34 (78.6%), the elderly (69.2%) and Asians (70.0%). A considerable proportion of these suicides had mental-health issues, mainly depression (42.4%) and alcohol or drug abuse (6.5%). Significantly more of them (69.6% vs. 22.2%) had stressors compared with suicides known to mental-health services. The findings of the study highlighted risk factors that may help prioritising intervention initiatives. It appears that local suicide audits may complement national information on suicide in designing appropriate local suicide-prevention strategies.

Keywords

Suicide risk prevention method life event mental health

Introduction

Suicide prevention is a key national priority for all health and social services. Achieving reduction in suicide requires understanding the phenomenon of suicidality and continuously evaluating the issues surrounding it. It is known that rates, risk and protective factors for suicide may differ according to the sites, institutions, psychological profile and time periods.¹ It is evident based on the information available at national and regional levels in the UK that there are variations in suicide rates in different regions and different periods in the same region. This suggests a need for greater understanding about the local picture. Reviewing trends may provide the locally relevant issues that may be helpful for suicide prevention. Although local audits monitoring suicide data, trends, hot spots and so on have been considered as useful and been suggested by many, their utility has been debated.^2–4

Local information on suicides can be built upon the national data and can contribute by identifying any specific issues that may be useful to various organisations such as the primary care, public health, social care, voluntary organisations and secondary care services, including mental-health (MH) services and other agencies that contribute to suicide prevention. It addition, the recommendations can contribute to the local suicide prevention strategy. In this context, it was intended to study the profile of suicides through a local suicide audit to understand the factors that may be useful in suicide prevention.

Methodology

The audit was conducted in Wolverhampton, which is a multicultural city in the West Midlands region of the UK, with a population of about 250,000. Information about suicides was available from the coroner, and that was the primary source of information for this audit. The Suicide Audit Group of the MH services collected further information for those suicides known to MH. The data were analysed, and the findings were discussed periodically in multidisciplinary meetings and annually in suicide prevention symposia since 2005, as part of the learning process.

The sample presented here consisted of consecutive cases with suicide, open and narrative verdicts between January 2004 and July 2011 in Wolverhampton. Misadventure verdicts were not included. For the purposes of this audit and for the ease of description, the sample for this study is referred to as suicides.

The available information included age, sex, method of suicide and associated factors such as psychological or MH issues and stressful life events (LE). Type of MH or psychological issues and LE as observed by the coroner were noted. It was ascertained whether the individual was known to the local MH services.

Confidentiality was maintained regarding the data, which were anonymised for evaluation. Identifiable information was not used, and the information was codified. Results were communicated as aggregates only. The project was considered as an audit, and the confidentiality procedure was approved by the trust Caldicott guardian.

Results

Description of sample

The sample (n = 146) consisted of 37 (25.3%) females and 109 (74.7%) males, with a ratio of 1:2.95. A little over half (52.1%) of the sample received a suicide verdict, followed by open (43.8%) and narrative (4.1%) verdicts. There was no difference between the sexes (females: 51.4%, 43.2%, 5.4%; males: 52.3%, 44.0%, 3.7%, respectively). The sample had an age range of 16–88 years (M = 46.0 ± 17.6 years). The mean age of females was 51.1 ± 19.0 years and that of males was 44.3 ± 16.8 years (t = 2.06; p < .05). Sample characteristics are given in Table 1. Almost half (45.2%) of the suicides were observed in the 25–44 age group. A considerable proportion (27%) of female suicides occurred in the elderly (aged 65 and over), whereas most suicides in males occurred in the 35–44 age group (29.4%). For a sizeable proportion (39.0%) of the sample, the ethnicity was not known; from the available information, there were 46.6% Caucasians and 13.7% Asians in the sample.

Table 1.

Sample characteristics.

Variable	Female (n = 37)	Male (n = 109)	Total (n = 146)
Age group*
<25	13.5	8.3	9.6
25–34	8.1	22.9	19.2
35–44	16.2	29.4	26.0
45–54	10.8	14.7	13.7
55–64	24.3	10.1	13.7
≥65	27.0	14.7	17.8
Method
Hanging	43.2	56.0	52.7
Overdose	32.4	18.3	21.9
Jumping/multiple injuries	10.8	7.4	8.2
Carbon-monoxide poisoning	5.4	5.5	5.5
Other methods	8.1	12.8	11.6
Stressful life events	40.5	56.0	52.1
Mental-health factors	67.6	61.5	63.0
Types of mental-health factors
Depression	54.1	43.1	45.9
Schizophrenia and other psychoses	5.4	4.6	4.8
Drug and alcohol	5.2	7.4	6.8
Other specified disorders	13.3	4.6	6.8
Unspecified	8.1	11.0	10.3
Known to mental-health services*	54.0	31.2	37.0

Figures are in percentages.

p < .05.

Variation within the city areas

There are 20 wards in Wolverhampton, and considering suicides within the city, there were considerable variations in the proportion of suicide in the wards, which ranged from 1.5% to 11.2%. The average number of suicides in the wards was 6.7 ± 2.94. A small proportion of the sample (8.2%) had addresses outside Wolverhampton city areas. The average number of suicides in six inner-city wards was 8.2 (median = 9), whereas in other wards it was 6.2 (median = 5.5). There was a positive but weak correlation (r = 0.0512) with suicide numbers in the ward and the Mental Illness Needs Index 2000 (MINI) of the electoral wards.

There was a marked variation in suicides in different months, the lowest being in September (4.8%) and the highest in March (11.6%). It was observed that there was the lowest number of suicides (19.9%) in the autumn (September–November) and the highest (29.4%) in spring (March–May). However, there were variations according to sex, with most male suicides occurring in spring (32.1%) but most female suicides occurring in autumn (32.4%). Observation of suicidal deaths on different days of the week suggested that there were comparatively fewer deaths on Sundays (9.6%) compared with other days, which ranged from 13.0% to 17.1%.

Method of suicide

The most common method of suicide was by hanging (52.7%), which was more commonly used by males (56.0%) compared with females (43.2%). This was followed by overdose (21.9%), with more females (32.4%) compared with males (18.3%) using it. Other notable methods were falls from a height (6.2%) and carbon-monoxide poisoning (5.5%). Double the number of females (10.8%) used fall from a height compared with males (4.6%).

There were 32.4% female and 17.4% male suicides associated with medication overdose. A wide range of medications were used. These included amitriptyline, chlormethiazole, citalopram, clozapine, codeine, colchicine, coproxamol, dextropropoxyphene, digoxin, diltiazem, dothiapin, ibuprofen, lamotrigine, opiate, paracetamol, propranolol, quetiapine, valproate, zopiclone and zolpidem. In some of the suicides (3.4%), the type of prescription medications used were not clear, and in a few (2.7%) suicides, there was a combination of more than one medication or bleach. Besides bleach, another non-medicinal chemical substance used for suicide was aluminium phosphide, a common pesticide.

Methods in different age groups

The proportion of hanging in elderly suicides was 34.6%, which was considerably less than that in the younger population (56.4%). Around a quarter of the elderly (26.9%) used an overdose, which was comparable to that of suicides in those younger than 25 years of age (28.6%), but it was higher than other age groups. Methods such as drowning, carbon monoxide and asphyxia were observed more frequently in the elderly (7.7% each) compared with those in the younger age groups.

MH factors

MH or psychological factors were identified in 63.0% suicides (67.6% females and 61.5% males), with 11.6% (18.9% females and 9.2% males) having more than one type of factor. The most common of the MH factors was depression, which was noted in 45.9% of suicides (including different forms of depression and bipolar disorder), followed by alcohol and drug misuse (6.8%), schizophrenia and other psychoses (4.8%), anxiety disorders (4.1%) and other issues (2.7%). In 10.3% of suicides, although the psychological factors were identified and considered to be associated, they were not specified.

Stressful LE

Stressful LE or situations were reported in 52.1% of suicides, and 9.6% had more than one LE. Amongst sexes, LE were noted more in male (56.0%) than female (40.5%) suicides (not significant), and 11.9% of the former had more than one LE. More common LE in both the sexes were relationship problem (28.1%), physical health-related issues (9.6%), death of family or a friend (4.8%) and death of a spouse (2.7%). Legal issues (7.3%), financial difficulties (6.4%), divorce and separation (2.8%) and job-related problems (4.6%) were reported only in men in this sample.

Comparison of suicides known and not known to MH services

The majority of suicides (63.0%) were not known to MH services. Out of 37% known to MH services, 5.5% were referred to MH services, but their details were not available. However, they were considered as known to MH. Significantly more males (68.8%) than females (46.0%) were not known (p < .05) to MH services. The majority (78.6%) of younger suicides (up to the age of 34), and also the elderly suicides (69.2%) were not known to MH services. In contrast, in the 55–64 age group, more (65%) were known to MH services. The mean age of suicides known to MH was higher (48.6 ± 14.3) than those not known (44.5 ± 19.2), but this was not significant. Amongst those whose ethnicity was known in the sample, the majority of Asians (70.0%) and a little over half (52.9%) of Caucasians were not known to MH services.

There were more cases of hanging (58.7% vs. 42.6%) in suicides not known to MH, whereas overdose (16.3% vs. 31.5%) and fall from height (4.3% vs. 9.3%) were observed in almost double proportions in those who were known to MH. A considerable proportion (42.4%) of suicides not known to MH had MH issues which were mostly depression (31.5%) and alcohol and drug abuse (6.5%). In this group of suicides, significantly more suicides had LE (69.6% vs. 22.2%; p < .0001) compared with those known to MH.

Discussion

Suicide has been one of the top contributors to Years of Life Lost in Wolverhampton^5,6 and deserves detailed study. It is of interest to highlight that the suicide rate in Wolverhampton was higher than that in England in the initial half of the decade starting in 2001. This appears to have gradually reduced, and became similar to the national rate in the middle of the decade. Later in the decade and at the beginning of next decade, the rates, although similar, are in fact slightly lower than those for England.⁷

It is interesting to observe the changes in the local suicide profile compared with a previous study in Wolverhampton, which reviewed all post-mortem reports between 1976 and 1990.⁸ On various parameters, comparing the previous⁸ and current study, the findings were: male and female ratio (1.96:1 vs. 2.95:1); average age of male (47.2 vs. 44.3 years) and female (51.6 vs. 51.1 years) suicide; the common method of suicide – drug overdose (38.6% vs. 21.9%) and hanging (21.8% vs. 52.7%), respectively. These data show that there was a considerable increase in the proportion of male suicides and also a change in the most common method of suicide from overdose in the previous study to hanging in the current one. There was a history of psychiatric treatment in 64.5% of suicides in Scott’s study.⁸ However, in the index study, 37.0% were known to MH services, although 63.0% suicides had MH or psychological issues. It appears that a proportion people with MH or psychological issues in the community are not in contact with secondary MH services. However, it cannot be ascertained whether they have received support from other sources.

The results of this audit identified that men in the 35–44 age group and women aged 65 and over were more represented in the sample, thus highlighting the possibilities of higher risk in these groups. Elderly women and young men, especially those who are not in contact with appropriate services for their stress or MH issues, are particularly vulnerable. There were differences in proportions of suicides in different wards in Wolverhampton, with more suicides in the inner-city areas. This issue has been recognised, along with the inequality amongst different areas of the city in the terms of the MINI and deprivation index.⁶ The greater number of suicides in areas of generally higher social deprivation and unemployment highlights the social context of suicide.⁹ This suggests that focused remedial measures in these identified areas with higher suicide proportion may assist suicide preventive efforts.

Methods of suicide

It appears that while the proportion of drug overdoses is coming down, hanging is increasing as a method of suicide. This is similar to the national trend in the UK,¹⁰ although the proportion of hanging in Wolverhampton (52.7%) is higher than that in England (45.0%).¹¹ It has been suggested that the increase in hanging is due to restricted availability of other methods and a misconception that it is painless and quick.¹¹ Although hanging is relatively difficult to prevent in the community, measures should be taken to make it a never-event in an institutional setting such as inpatient wards or prisons. This might be achieved by a combination of strategies, for example by structural changes using building safety features, removing possible ligature points and materials, using collapsible curtain rails and so on.

Self-poisoning in Wolverhampton has decreased compared with past findings. This is similar to the national trend.^8,10 The range of medications used in the Wolverhampton sample was wide, and most of the overdoses used prescription drugs. This suggests that focusing on one or particularly few medications may not be effective. However, some attention can be given towards prescribing and dispensing practices. Medications which are less toxic in overdose may be chosen; and prescribing/dispensing smaller quantities at one time may help. However, that may not prevent people hoarding or having other over-the-counter medications.

While the proportions of suicides by drowning, carbon-monoxide poisoning, and firearms are lower compared with past observations in the city,⁸ falls from height (6.2%) are emerging as a new concern. It would be worthwhile tracing local hotspots to determine whether preventive measures around such sites might help.

Stressful LE

Stressful LE and situations are commonly associated with suicides.^12–18 In this study, around half (52.1%) of the suicides had identified stressors. Relational or interpersonal problems were most common in this sample. Interpersonal problems, both chronic and recent, have been reported in suicides.^13,14 The availability of support for relational problems and public education regarding this may aid in prevention efforts.

Physical health-related stress is commonly reported in suicides,^15,19,20 especially in elder suicides.^16,21 This was observed in the current sample. It is possible that in some cases, the impact of this stressor could be minimised with additional care and support.

The death of near and dear ones and spouses was observed in a considerable minority of suicides. This kind of stressor has been mentioned in other studies.^16,22 This observation highlights the risk of suicide in bereaved individuals. It might be worthwhile assessing and providing support to the vulnerable bereaved proactively.

Some of the LE were observed only in men in this sample, which included legal, financial, marital (divorce and separation) and job-related problems. All these have been reported in association with suicide.¹³ In the current sample, it appears to have a male predominance.

Although a wide variety of LE are implicated in suicides,^{12–16,23,24} some of them are encountered more often, as observed in this study. It may be argued that these stressors are ubiquitous and occur in spite of the availability of specific support systems. Even so, some targeted approach can be taken for some specific stressors, which may involve identifying individuals with these stressors and proactively supporting them. In addition, public education encouraging people undergoing these stressors to access particular support may help.

MH factors

Mental illnesses are independent risk factors for suicide, and it has been observed that recent LE may have a lesser causal role in those with severe mental illness.¹³ As observed in this study, stressors were significantly more associated with the suicide of those who were not known to MH services compared with those who were known. However, it was also observed that a considerable proportion did have MH or psychological issues who were not known to MH services. While stressors clearly play a role in suicides not known to have MH issues, it is plausible that LE and MH issues contribute additively to the suicide risk.

The majority of suicides in this sample were associated with MH factors. While it was obvious for those with mental illness and who were known to MH services, it was present in a sizeable proportion who were not known to MH services. The most common of the MH factors was depression. However, around one tenth of suicides were associated with MH factors of which the nature was not clear. It is known that MH issues, even depression, are often missed in the community.^25–27 This highlights the need for proactive identification, structured risk assessment in primary care and appropriate intervention.²⁸

Contact with MH services

In the index sample, a little over one third of suicides were in contact with MH services, which is comparable to another study of suicides by coproxamol poisoning.²⁹ According to the National Confidential Inquiry into Suicide and Homicide by People with Mental Illness (NCI), during 2002–2012 in England, 28% of general population suicides were in contact with MH services in the 12 months prior to their death.¹¹ Various factors may contribute to referral to MH services, for example higher psychiatric morbidity, improved identification in primary care, lack of other more appropriate services and so on. Some of the reasons of a high proportion being known to MH in Wolverhampton could be a higher MINI index in some areas and probably improved awareness amongst local general practitioners. It is a concern that most suicides involving men, young people up to the age of 34 years, the elderly and Asians were not known to MH services. Identification of stress and MH issues in these populations is an essential target for preventive actions. Access and utilisation of services by Asians, especially the elderly, is known to be low³⁰ and may need more focused attention.

Methods used by patients known to MH in England and Wolverhampton were comparable for hanging (41% vs. 42.6%),¹¹ although self-poisoning increased slightly (26% vs. 31.5%) and jumping/multiple injuries were slightly less (15% vs. 11.2%), respectively. As mentioned earlier, efforts to make hanging a never-event in hospitals, safer prescribing and managing local hotspots for jumping might be helpful.

Use of local audit on suicide

It is probable that the findings of this audit may be useful for all local stakeholders contributing to suicide prevention efforts as an information source with issues that are locally relevant and specific. In this process, it may contribute to identification of areas on which to focus in the local suicide prevention and MH promotion strategies. As most suicides currently are not known to MH services, detailed evaluations of these are needed in line of NCI, which may help to improve understanding and focus preventive efforts for this group of suicides.

It has been reported that information and training are helpful as suicide prevention activities.^31,32 Local suicide audit findings can provide information to be specifically covered in training programmes for a range of professionals. For example, the observations from suicide audits have been used to improve awareness amongst multidisciplinary professionals, including psychiatrists, psychologists, MH nurses, general practitioners, clinicians from primary care, public health and individuals from the voluntary sector in Wolverhampton annually since 2005. In these programmes, local issues are highlighted, training in risk assessment and management methods are provided and possible prevention strategies are discussed. The predominance of the stress of relationship issues and the large number of suicides with depression in the community have been specifically highlighted.

Limitations

Data for various relevant variables either were not available or were inadequate. This can be improved by arranging data collection from multiple sources, including the coroner’s office, GP practices, psychological and MH services. It may be beneficial to have a process established for local data collection, analysis and dissemination to all stakeholders so that preventive approaches are shared.

Conclusions

The results of this study have highlighted various factors associated with suicides, which may be the focus of local suicide prevention efforts. Identification of areas in the city with a higher number of suicides, trends of suicide methods, risk factors and profile of vulnerable individuals not in contact with appropriate services can help in preventive initiatives. This can suggest specific lines of enquiries and interventions, for example support for relationship stresses, bereavement, identification and appropriate support for depression in the community, informing people about local resources and encouraging them to seek help. In addition, the results may contribute towards formulating a local multi-agency suicide prevention strategy, identifying health promotion measures and as a resource for training programmes.

Footnotes

Acknowledgements

The author acknowledges the support from Suicide Audit Group, Mental Health Directorate of Wolverhampton City Primary Care Trust; HM Coroner, Wolverhampton; Departments of Postgraduate Psychiatry and Clinical Audits, Black Country Partnership NHS Foundation Trust; colleagues for support regarding data collection; and Karen Hirst for data entry and checking.

Declaration of conflicting interests

The author declares that there is no conflict of interest.

Funding

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

References

Soubrier

. Looking back and ahead. Suicidology and suicide prevention: do we have perspectives? World Psychiatry 2004; 3: 159–160.

Public Health England. Guidance for developing a local suicide prevention action plan, https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/359993/Guidance_for_developing_a_local_suicide_prevention_action_plan__2_.pdf (2014, accessed 16 February 2015).

Caley

Fowler

. Suicide prevention: is more demographic information the answer? J Public Health 2009; 31: 95–97.

Owens

Roberts

Taylor

. Utility of local suicide data for informing local and national suicide prevention strategies. Public Health 2014; 128: 424–429.

Wolverhampton Joint Strategic Needs Assessment, www.royalwolverhamptonhospitals.nhs.uk/pdf/WCPCT/Healthy_City/JSNA_Initial_document_2008.pdf (2008, accessed 21 March 2015).

Wolverhampton City Primary Care Trust. Wolverhampton Public Health Annual Report 2011. Wolverhampton City Primary Care Trust, www.google.co.uk/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&uact=8&ved=0CCEQFjAA&url=http%3A%2F%2Fwww.wolverhamptoninprofile.org.uk%2Fresource%2Fview%3FresourceId%3D35&ei=XI_kVMTZJ9HQ7AbFz4DoBQ&usg=AFQjCNG7kmwlk9Iw8u4EJttX8uuC_4LezA&sig2=zA5wuWnogPFQ7ad4dh6K9g&bvm=bv.85970519,d.ZGU (2011, accessed 18 February 2015).

Public Health Outcomes Framework, www.phoutcomes.info/public-health-outcomes-framework#gid/1000044/pat/6/ati/102/page/4/nn//par/E12000005/are/E08000031/iid/41001/age/1/sex/4 (accessed 20 August 2015).

Scott

. Suicide in Wolverhampton (1976–1990). Med Sci Law 1994; 34: 99–105.

Wolverhampton Joint Strategic Needs Assessment Update, http://wolverhampton.moderngov.co.uk/Data/Health%20Scrutiny%20Panel/201009231400/Agenda/06%20-%20Joint%20Strategic%20Needs%20Assessment.pdf (2010, accessed 18 February 2015).

10.

Office of National Statistics. Suicides in the United Kingdom, 2012 Registrations, www.ons.gov.uk/ons/dcp171778_351100.pdf (2014, accessed 16 February 2015).

11.

The National Confidential Inquiry into Suicide and Homicide by People with Mental Illness. Annual Report 2014: England, Northern Ireland, Scotland and Wales. July 2014. University of Manchester, www.hqip.org.uk/assets/NCAPOP-Library/NCAPOP-2014-15/NCISH-2014-annual-report.pdf (accessed 20 March 2015).

12.

Kar

. An analysis of life event burden associated with suicide attempt. Arch Indian Psychiatry 2012; 14: 43–51.

13.

Cooper

Appleby

Amos

. Life events preceding suicide by young people. Soc Psychiatry Psychiatr Epidemiol 2002; 37: 271–275.

14.

Appleby

Cooper

Amos

. Psychological autopsy study of suicides by people aged under 35. Br J Psychiatry 1999; 175: 168–174.

15.

Bastia

Kar

. A psychological autopsy study of suicidal hanging from Cuttack, India: focus on stressful life situations. Arch Suicide Res Off J Int Acad Suicide Res 2009; 13: 100–104.

16.

Heikkinen

Lönnqvist

. Recent life events in elderly suicide: a nationwide study in Finland. Int Psychogeriatr IPA 1995; 7: 287–300.

17.

Squires

Busuttil

. Suicide by fire in Scotland: 1980–1990. J Clin Forensic Med 1996; 3: 81–85.

18.

Voshaar

RCO

Kapur

Bickley

. Suicide in later life: a comparison between cases with early-onset and late-onset depression. J Affect Disord 2011; 132: 185–191.

19.

Qin

Hawton

Mortensen

. Combined effects of physical illness and comorbid psychiatric disorder on risk of suicide in a national population study. Br J Psychiatry 2014; 204: 430–435.

20.

Cooper

Kapur

Webb

. Suicide after deliberate self-harm: a 4-year cohort study. Am J Psychiatry 2005; 162: 297–303.

21.

Biermann

Sperling

Bleich

. Particularities of suicide in the elderly. A population-based study. Aging Clin Exp Res 2009; 21: 470–474.

22.

Heikkinen

Aro

Lönnqvist

. Recent life events, social support and suicide. Acta Psychiatr Scand Suppl 1994; 377: 65–72.

23.

Kar

. Profile of risk factors associated with suicide attempts: a study from Orissa, India. Indian J Psychiatry 2010; 52: 48–56.

24.

Kar

. Suicidality following a natural disaster. Am J Disaster Med 2010; 5: 361–368.

25.

Lemelin

Hotz

Swensen

. Depression in primary care. Why do we miss the diagnosis? Can Fam Physician 1994; 40: 104–108.

26.

Saver

Van-Nguyen

Keppel

. A qualitative study of depression in primary care: missed opportunities for diagnosis and education. J Am Board Fam Med 2007; 20: 28–35.

27.

Goldman

Nielsen

Champion

. Awareness, diagnosis, and treatment of depression. J Gen Intern Med 1999; 14: 569–580.

28.

Vail

Adams

Gilbert

. Investigating mental health risk assessment in primary care and the potential role of a structured decision support tool, GRiST. Ment Health Fam Med 2012; 9: 57–67.

29.

Hawton

Simkin

Gunnell

. A multicentre study of coproxamol poisoning suicides based on coroners’ records in England. Br J Clin Pharmacol 2005; 59: 207–212.

30.

Bhattacharyya

Kar

. Suicide in black and minority ethnic elderly in UK. J Geriatr Care Res 2014; 1: 14–16.

31.

Rutz

Wålinder

Von Knorring

. Prevention of depression and suicide by education and medication: impact on male suicidality. An update from the Gotland study. Int J Psychiatry Clin Pract 1997; 1: 39–46.

32.

Rihmer

Rutz

Pihlgren

. Depression and suicide on Gotland. An intensive study of all suicides before and after a depression-training programme for general practitioners. J Affect Disord 1995; 35: 147–152.