Suicide is a major cause of death worldwide and its prevention is an international priority.1, 2, 3 Social factors, such as unemployment and wider economic circumstances, are undoubtedly major determinants of suicidal behaviour, and not only at times of recession.4, 5, 6 However, psychological, biological, and clinical factors are also important.7, 8 In this context, what role do health services have in suicide prevention?
Adequate access to services and effective management of mental and substance-use disorders have been highlighted by WHO in the global attempt to reduce suicide rates and have been examined in recent studies.2, 9, 10 Research over many years suggests that most people who die by suicide could be suffering from a psychiatric disorder at the time of death, yet comparatively few are in contact with specialist services.11, 12 Previous research has identified the characteristics of people who die while under the care of services,13 and the elements of mental health service provision that could be associated with reduced rates of suicide, such as ready access to mental health professionals,14, 15 well developed community services,16 and specific policies for substance misuse.17 Our own previous research found that three service changes in particular (provision of 24 h crisis services, policies for people with drug and alcohol misuse, and a system of reviewing care after suicide deaths) were associated with lower suicide rates in England and Wales after their implementation.18 Other factors, such as absence of continuity of care and short hospital admission of less than a week, might increase suicide risk.19, 20
However, the evidence base is far from consistent—some studies have found no association between service provision and suicide,21 whereas others have found that particular service elements, such as levels of compulsory detention, were associated with higher suicide rates.22 Many studies have been purely ecological and have focused on service provision across large areas (eg, country or region) rather than at the level of the individual service provider.17 Few studies have examined the impact of service changes over time.
Generally, studies have considered few aspects of mental health service provision and have restricted themselves to delivery of care variables rather than considering the way services are organised.23 Internationally, safety has been highlighted as the first responsibility of health care.24, 25, 26, 27 Specific recommendations include staff being able to readily raise concerns about the quality of care, an emphasis on learning, and the importance of the strength of the organisation itself (as shown by factors such as staff turnover and patient complaints). Studies in general medical and residential care settings have suggested that poor staff satisfaction and staff turnover might be associated with higher mortality.28
Research in context
Evidence before this study
We searched MEDLINE and PsycINFO databases from inception to Nov 9, 2015, with a combination of keywords and subject headings (with the terms “suicide”, “suicide, attempted”, and “suicidal”, combined with “policy making”, “reduction”, “organisational policy“, “public policy”, and “health policy” and “mental health services”) to identify published studies in English and systematic reviews on suicide prevention. Studies reported that a number of factors (such as implementation of 24 h crisis services, the presence of community teams, policies for multidisciplinary review after suicide, and follow-up within 7 days of inpatient discharge) were associated with lower suicide rates. Mixed associations were found between suicide and implementation of national substance misuse policies, use and size of inpatient services, and staffing levels. Some aspects of service reorganisation (such as transfer of services or merging specialist teams, or inpatient admissions of less than 7 days) might have been associated with increased suicide risk. No studies examined the effect of service change in different organisational contexts.
Added value of this study
A number of service changes related to ward safety, community services, training on management of suicide risk, and the implementation of key policies and National Institute for Health and Care Excellence guidance were associated with a lower suicide rate after these policies had been introduced by mental health services. Implementation had a bigger impact in mental health services that had low rates of staff turnover but high rates of overall event reporting.
Implications of all the available evidence
Social, psychological, biological, and clinical factors are crucial determinants of suicide but system-wide change implemented across the patient care pathway could be a key strategy for improving patient safety. Just as important as the changes themselves, might be the organisational context in which these changes are introduced.
The interaction between service and organisational factors has not been examined. To our knowledge, no studies have examined the effect of implementation of service changes in different organisational contexts. For example, service changes might plausibly have less of an impact in health providers with an unhappy or constantly changing workforce.
In this study, we examined the relationship between service provision and suicide rates in all mental health services in England. Use of randomised controlled designs for this research would be extremely challenging. Therefore instead we used service developments as the basis of a natural experiment and opted for a descriptive design. The current study builds on earlier work18 by examining a wider range of service variables over a long period and for the first time considering the potential role of organisational variables. Although mental health care varies, many developed countries are dealing with similar issues, such as the move away from inpatient to community service provision, a focus on severe and enduring mental illness, ageing populations, and limited budgets.2, 29, 30 The findings of this study are therefore likely to have international relevance.
We had three main objectives. First, to examine the association between implementation of service changes and suicide. We hypothesised that service changes would be associated with improvements in patient safety as measured by a reduction in the rate of suicide. Second, to consider how wider organisational factors might affect suicide. We hypothesised that organisational factors such as staff and patient satisfaction and staff turnover would be associated with suicide rates. Third, to investigate whether the impact of service changes varied according to available measures of the organisational context in which they occurred. We hypothesised that changes would have less of an effect in mental health services that had markers of impaired organisational functioning.
