Preventing Self-Harm: Mental Health Nursing

Published Date: 23 Mar 2015 Last Modified: 25 May 2017

Keywords: self harm treatment, self harm nursing, therapy self harm

The area of self-harm has been subject to extensive research since the 70s when professionals first noted the occurrence of the behaviour in psychiatric inpatients (Iwata et al, 1994). Since then the occurrence of self-harm has continued to increase. Each year approximately 220,000 people present to A&E after self-harming and it accounts for the most common reason for medical admission in females and the second most common cause for admission in males (Hawton et al, 2007). The prevention and minimisation of Self-harm within an acute inpatient setting will be the focus of this dissertation not only because of the high prevalence but because of an interest that has developed from my past clinical experiences.

The first chapter of the dissertation will review the relevant literature to explain the issues pertinent to self-harm. Research suggests that for the correct treatment and interventions to be proposed, knowledge of self-harm as an aspect of mental health nursing needs to be understood (Chapman et al, 2006; Klonsky and Glenn, 2008). In addition chapter one will explore the link between self-harm and a diagnosable mental disorder and the prevalence of the behaviour within an inpatient setting.

At the centre of mental health nursing is the one-to-one therapeutic relationship that nurses develop with their patients (Dooher, 2008). The concept of the therapeutic nurse-patient relationship evolved from the work of Hildegard Peplau in the 1950's (Forchuk and Reynolds, 1998). Therefore, Peplau's (1997) Theory of Interpersonal Relationships will provide the backdrop for the dissertation. Chapter two will critically discuss the importance of the therapeutic relationship when working with an individual who self-harms with the focus predominantly on the working phase of the nurse patient relationship. The working phase of the relationship consists of identification and exploitation, chapter two will focus on collaboratively working with the patient to identify what they would like to achieve from the nurse-patient relationship.

Chapter 3 will discuss the exploitation phase and the helping approaches used to minimise or prevent self-harm, currently there are only a few evidence based approaches devised specifically for self-harm (Nock, 2010). Harm reduction will be critically analysed given that self-harm is often the only coping strategy an individual has and removing this coping mechanism suddenly can lead to withdrawal from the therapeutic relationship (Pengelly et al, 2008). Furthermore, 'enhanced nursing observation' is the most commonly used nursing intervention to manage self-harm in an inpatient setting (Bowers et al, 2000). However, much criticism surrounds the use of nursing observations. Therefore, the literature will be reviewed and critically analysed to determine if it does reduce self-harm through therapeutic engagement.

Naturally all therapeutic relationships come to end. A therapeutic relationship can end for a number of reasons including, disengagement from the client, the nurse may be transferred to a different position or even death. Nevertheless the ending of the relationship is difficult (Sreevani, 2007). Chapter 4 will critically discuss the issues that may arise and the discharge planning process when transferring care to a community mental health team.

Chapter One

Introduction to self-harm as an aspect of mental health nursing

This chapter will explore the issues pertinent to self-harm. Self-harm will be defined and the behaviour will be explored to determine why and how individuals may self-harm.

A single definition of self-harm, as well as the vocabulary used to describe the behaviour, has been difficult for researchers and clinicians to identify because there is no generally accepted terminology and therefore various terms have been used in the literature (Ross & Heath, 2002). The term self-harm has often been used interchangeably with self-mutilation, deliberate self-harm, self-injurious behaviours, parasuicide and self-wounding. Weber (2002) identifies that the terms used often, such as self-injury, Para-suicide and self-destructive behaviour, reflect the theoretical standpoint of the clinician using them rather than the client who is harming. Furthermore the term used to describe the self-harm can imply a different meaning of the act, for example self-abuse suggests a psychodynamic understanding of the motivation for the behaviour whereas direct self-harm conveys the person has chosen to act in this way and that it is a behaviour rather than an illness. Onacki (2005) defines self-harm as a deliberate, repetitive, impulsive, and non-lethal harming of one's body. Whereas the National Institute for Clinical Excellence (NICE) guidelines for the short-term management of self-harm define self-harm as 'self poising or self-injury, irrespective of the apparent purpose of the act' (NICE, 2004, p7) The term 'deliberate' has been removed from their definition, in acknowledgement that some people may self-harm in a dissociative state and that intent varies from individual to individual and on each occasion that someone harms themselves. Their broad scoping definition aims to ensure people presenting with a wide range of self-harm will be offered access to a psychosocial assessment and appropriate support and follow-up.

Research has categorized self-harm and there is a general agreement based on the severity of the behaviour and the act itself. The most recent theoretical classification discussed in the literature is proposed by Favazza (1998). In this classification system, self-harming behaviour has been classified in to three observable categories, including major, stereotypic, and superficial/moderate based upon tissue destruction and the rate and pattern of the behaviour. According to Favazza (1998), major self-harm consists of rare acts in which a major amount of body tissue is destroyed for example a limb may be amputated in the act and stereotypic self-harm includes acts of moderate behaviours such as head banging, hitting, throat and eye gouging, and self-biting and are primarily rhythmic and repetitive. Superficial or moderate self-harm is the most common and varied type of self-harming behaviour and it comprises of acts of low lethality that occurs both sporadically and repetitively (Rao, et al 2008).

Whalen (2006) identifies superficial or moderate self-harm (herein referred to as self-harm) as the most common type of self-harm among the psychiatric inpatient population. Self-harm can take many forms including self-injury and self-poisoning. The most common form of self-injury is skin cutting with a variety of implements targeting areas of the body that can be covered by clothing and most likely the lower arm (Beer, 2010; Mannion, 2009). There is some evidence that cutting is more repetitive than other forms of self-injury (Lilley et al, 2008). Other forms include burning, scratching banging or hitting body parts and interfering with wound healing (Klonsky, 2007a). Self-poisoning is the intentional use of more than prescribed or recommended doses of any drug and includes poisoning by non-ingestible substances, overdoses of recreational drugs and severe alcohol intoxication where this seems to be intended as an act of self-harm. Furthermore, Substance misuse, physical risk-taking, sexual risk-taking and self-neglect are sometimes labelled as indirect self-harm (Royal College of Psychiatrists, 2010). In addition, when people who repeatedly harm themselves through cutting or taking overdoses are helped to overcome these behaviours, eating disorders or other self-damaging problems may emerge. With research suggesting that the prevalence of self-harm among patients with an eating disorder is 25% (Sansone et al, 2003). For the purposes of this dissertation the discussion will be predominantly focused on self-injury because research has identified this as the most common type of self-harm within an inpatient setting (e.g. James et al, 2012; Beer et al, 2010).

Self-harm has been identified as a key risk factor for suicide with approximately 50% of people who commit suicide having previously self-harmed (Cooper et al, 2005); this strong correlation may suggest why some researchers feel self-harm only exists when the intent is to kill oneself (Klonsky et al, 2003; Ross & Heath, 2002). However, Conaghan & Davidson (2002) are of the opposite view and suggest self-harm is a behavior with the outcome being not to kill oneself. This view is supported by Favazza (1998) who suggests that self-harm is a 'morbid' form of self-help that is opposing to suicide. Furthermore, Suyemoto's (1998) 'anti-suicide' model focuses on self-harm as an active coping mechanism to avoid suicide. Thus suggesting that suicide and self-harm are two very different phenomena and therefore in this dissertation self-harm will continue to be discussed without suicide intent.

Self-harm can occur because of a wide range of psychiatric, psychological and social problems. Meltzer et al (2001) found that individuals with current symptoms of a mental disorder are up to 20 times more likely to have self-harmed in the past. However, self-harm does not currently fulfil the criteria for an independent category of mental or behavioural disorder in the DSM-IV-TR (Diagnostic and Statistical Manual 4, Text Revision, American Psychiatric Association, 2000) or the ICD-10 (International Classification of Diseases, World Health Organisation, 1992). In both classificatory systems, the DSM-IV and the ICD-10, self-harm is only referred to as a symptom of Borderline Personality Disorder. Conversely, Klonsky et al (2003) discussed that self-harm is also commonly found in clients with a number of mental disorders. Haw et al (2001), for example, found that 92% of individuals receiving inpatient treatment for self-harm were also suffering from psychiatric disorders, with the most common being depression and anxiety disorders. The vast research that illustrates self-harm co-occurs with a variety of diagnoses, and not just Borderline Personality Disorder, has lead to a proposal for the DSM-V to include a new diagnosis specific to non-suicidal self-injury (Shaffer and Jacobson, 2009). Signifying that self-harm is an issue in itself irrespective of the co-occurring diagnoses.

Some of the other psychological or social problems stem from early childhood experiences and may include: sexual abuse, neglect, emotional and physical abuse, loss or separation and parental mental health problems (Skegg, 2005). Current psychological or social experiences, aside from a mental health diagnosis, rape, domestic violence and substance misuse have been found to increase the risk (Tuisku et al, 2009). The risk factors identified could have a significant impact on everyone, and even more so if someone has a mental health diagnoses, however not everyone predisposed to these experiences and/or with a mental health diagnosis self-harms. Therefore, these contributing factors alone do not control the behaviour. More recently, both academics and professionals have recognised that until there is an evidence base indicating why people self-harm, it is unlikely that the correct treatments and interventions will be proposed (Klonsky and Glenn, 2008). Fortunately, the functions of self-harm have been subject to increased research in recent years (Klonsky, 2007b).

Research has identified multiple functions of self-harm (Hanley et al, 2003). Converging evidence suggests self-harm occurs with a primary intent to alleviate negative emotions and to release tension (Klonsky, 2009). Utilising self-harm to reduce tension does seem to result in an immediate release and Crowe & Bunclarck (2000) have found biological evidence to suggest that a physiological stress reduction after an episode may last up to 24hrs. These findings emphasise the risk of repetition and further more why self-harm is a very complex behaviour to treat which will be discussed further in Chapter Three. Another prominent function is reported by Machoian (2001), who describes self-harm as a means to communicate the degree of pain that is being felt. It seems that those who self-harm feel that no one can offer emotional support or show they understand (Magnall, 2008). These beliefs suggest the importance of a therapeutic collaborative relationship because if self-harm is being used to communicate, the skills of a nurse should be adapted to overcome this. Although these functions are commonly identified to explain why someone may self-harm they are not inclusive thus emphasising the importance of a comprehensive assessment (both of these points will be discussed further in Chapter Two). Nevertheless self-harm is a maladaptive coping strategy (Linehan, 1993).

The demographics of individuals who self-harm, age, gender and ethnicity, may also give insight in to why people self-harm or at least help to identify who is more at risk. Self-harm can occur at any age but it is most prevalent in adolescents and younger adults. Martin et al (2010) found that self-harm typically begins in early adolescence around fourteen years of age and the disorder seems to peak between the ages of sixteen and twenty-five. Furthermore, studies which have specifically looked at self-harm within an adult acute in-patient setting have found an age range between 20 to 37 years old (e.g. Bowers et al, 2003; Chengappa et al, 1999). More recently, Beer et al (2010) identified that 69% of patients were under the age of 35 supporting that self-harm can indeed occur at any age but it is more predominantly seen in younger adults. In addition, whether an individual is male or female has been captured in research studies and until recently much of the research suggested that females are more likely to self-harm. For example Fox and Hawton (2004) indicated that females are four times more likely to self-harm whereas, Beer et al (2010) found there to be no specific gender prevalence. The variation in the findings suggests there is no specific relationship between gender and self-harm. However, gender is an important factor as it may determine the method used to self-harm. Females are most like to cut themselves and males are more likely to burn or hit themselves (Claes et al, 2007). Findings in relation to the ethnicity of people who self-harm are somewhat inconsistent; Bhardwarj (2001) discussed that Asian women are two to three times more likely to self-harm compared to other ethnic groups. Whereas Beer et al (2010) found the prevalence of self-harm to be disproportionately high in white ethnic groups. Conversely other research has identified no correlation between self-harm and gender (for e.g. Bowers et al, 2003). Thus, suggesting that ethnicity cannot always indicate if someone is likely to self-harm.

Additionally the prevalence of self-harm is much higher in in-patient settings, with up to 80% of a psychiatric inpatient sample self-harming (Nock and Prinstein, 2004) compared to a prevalence rate of 15-20% in the community (Heath et al, 2009). Of late, James et al (2012) reviewed the self-harm incident reports from adult psychiatric wards during 2009. The review established that 14,271 self-harm incidents were reported and self-injury in the form of cutting was the method used most. Acute services reported the highest amount of incidents (65%) with forensic services reporting 29%. However, acute services have a higher quantity of beds compared to forensics and therefore there are significantly more reports of self-harm from forensic settings (James et al, 2012). Nevertheless the high rate of self-harm occurring within psychiatric inpatient settings suggests that all nurses will experience someone who self-harms at some point during their career. Nijman et al (2005) estimated that during a year period 84% of mental health nurses will witness mild self-harm and 57% severe self-harm.

To summarise, self-harm is a maladaptive coping mechanism which predominantly starts in adolescents usually at age 14 and peaks at 16 through to the age of 25. However, within an adult acute inpatient setting the age range increases to 35 with self-injury in the form of cutting the lower harm being the most likely form of self-harm. Finally, the varied causes and functions that self-harm serves highlight the importance of in-depth therapeutic assessment which will be discussed in the next chapter.

Chapter Two

Critical Discussion of the therapeutic relationship and assessment tools used within self-harm.

Therapeutic relationships provide the foundations for mental health nursing practice with people who are experiencing threats to their physical and mental health (Reynolds, 2009, p.313). An effective therapeutic alliance is one of the key factors that help patients to develop alternative modes of coping with intolerable affects when habitual self-harm has become common (Walsh, 2007).

The concept of the therapeutic nurse-patient relationship evolved from the work of Hildegard Peplau in the 1950's (Forchuk and Reynolds, 1998). Peplau (1988) held the view that whilst relationships may contribute to dysfunctional behaviour, people can also heal within relationships. Peplau's interpersonal relationship theory of nursing focuses on the evolving therapeutic relationship between the nurse and client. Peplau (1997) identifies overlapping phases. Firstly, the orientation phase, in which parameters of the relationship are established and the initial trust develops. Secondly, the working phase, which includes problem identification and exploitation (making full use of the services of the nurse) and finally, the resolution phase, within which the nurse prepares the patient for termination of the therapeutic relationship (Forchuk, 1995). The current health care environment, including shortage of nurses, economic constraints and advances in pharmacology, has led some to question the relevance of Peplau's model in current nursing practice (Ziegler, 2005). For example, Hagerty and Patusky (2003) discuss that the long-term relationship development is no longer useful in most health care settings. Conversely, Gastman (1998) argues that the fundamental concepts found in the interpersonal relations theory are still useful.

Development of a therapeutic relationship has been identified as being particularly complicated where self-harm is involved. It is not unusual that self-injurers encounter professionals that respond ineptly to their behaviour. Many authors (e.g., Simeon and Hollander, 2001; Walsh, 2006) have discussed negative reactions, such as shock, disgust, fear and anxiety, that self-injurers experience when they first encounter professionals such as nurses. This may lead to individuals who self-harm being 'sceptical' of the continuity and longevity of the therapeutic relationship (Walsh and Rosen, 1988). Although these preconceptions exist both for the nurse and client, a therapeutic relationship is vital. The guidelines provided by the National Institute for Clinical Excellence (NICE, 2004) for the management of self-harm, suggest that a therapeutic relationship is often more important than the treatment. Furthermore although a positive therapeutic relationship is not the solution to self-injury, it provides a context in which problem solving and behaviour change can take place (Nafisi and Stanley, 2007). Therefore it is essential that the skills used in the orientation phase to develop trust are continued and expanded upon throughout the relationship. Shiner (2008) identifies it is important to maintain a non-judgemental, empathetic, open-minded and respectful attitude with the focus kept on the person and not on their self-harm behaviour. These underlying principles are congruent with the much earlier work of Peplau (1952) and Rogers (1957). The use of empathy has been identified as a key component when working with an individual who self-harms, the ability to communicate an understanding of what they are going through may alter the individuals' perception of adults and their previous experiences (Davies & Huws-Thomas, 2007).

For problem solving and behaviour changes to take place collaborative care is important to identify the needs individual to the patient. Collaborative working ensures that both the nurse and patient influence the decision making process (Ellis, 2009, p.145). Hence, Nurses should deal with the 'persons' description of their own immediate needs, this is something The Tidal Model recognises. The Tidal Model advocates an essentially curious and broad-minded stance towards people's problems. The model incorporates elements of alliance, whereby, the nurse is aware the patient already has the acquired skills and abilities to move forward and helps the patient to recognise and develop these through achievable therapeutic goals, which is essentially collaborative care (Barker and Buchanan-Barker, 2005). The philosophical underpinning of this model draws upon the earlier work of Peplau (1952) which further supports the relevance of Peplau's model in today's practice. A recent study by Cook et al (2005) illustrates the interpersonal transactions displayed when The Tidal Model is implemented, as being positive for recovery. This emphasizes the use of this model when mental health nursing is fundamentally aiding someone on their journey to recovery (Shepherd et al, 2008).

The working phase of Peplau's model is initially concerned with identification. During this phase of the therapeutic relationship the nurse in partnership with the client carries out an assessment to identify problems that require working on within the relationship (Peplau, 1988). Unfortunately, relatively little has been written on the formal assessment of self-harm. This absence is unfortunate because any effective treatment of self-injury must begin with a thorough and accurate assessment to identify needs and problems (Walsh, 2007).Nevertheless, a mental health assessment is a legal requirement under the Mental Health (Wales) Measure 2010 (MHM).

The MHM is primary legislation that was passed by the National Assembly for Wales in 2010. It recognises that the majority of people with mental health problems rarely need to be treated compulsorily under the Mental Health Act 2007 and voluntarily seek treatment (Welsh Assembly Government, WAG, 2010a). The Measure is different to the Mental Health Act as it does not provide for compulsory admission and treatment of people. The measure ensures that people are able to access appropriate mental health services and receive care that is co-ordinated by a named person. Part 2 of the Measure places duties on service providers, Local Health Boards and local authorities in Wales, to act in a co-ordinated manner to improve the effectiveness of services provided to an individual. The WAG (2010b) identified that clients often require help with aspects of their lives in addition to care and treatment, such as education and their physical health. This places demands on services that one discipline cannot meet alone, and it is therefore a requirement to have an integrated system of effective assessment, planning delivery and review, so that collaborative care can be provided to benefit the client. The framework for this integrated system is the Care Programme Approach (CPA).

Fundamentally, the CPA is a problem solving process designed to facilitate effective and efficient clinical management of a client from admission in to services through to discharge (Kingdom, 1994). In essence the CPA introduced no major changes in to psychiatric care and should have merely formalised good care. However many have found this not to be the case and the CPA has been subject to major criticism with it being unevenly implemented (Social Services Inspectorate, 1999) and clients have reported it to be 'invisible' and ineffectual (Webb et al, 2000; Rose, 2001). These findings may suggest that the use of the CPA is ineffective however the principles of the framework are now legitimate requirements under the Measure which should eradicate any criticism and negative preconception in the future. Furthermore Part 2 of the Measure, which applies to secondary services, does not prescribe a particular assessment process (WAG, 2012).Therefore, the CPA framework still has an existence if used in conjunction with the Measure's care and treatment planning documentation.

Assessment is the first stage of the CPA framework which needs to be personalised to suit the individual who is self-harming. A psychosocial assessment following self-harm, as outlined in the NICE (2004) guidance, is a necessary starting point for preventive interventions (Hawton et al, 1998). To carry out a psychosocial assessment means to recognise the importance of, and interrelationships between, psychological and social domains of the client's life. A recent study carried out by Bergen et al (2010) found a psychosocial assessment following self-harm was associated with a 53% decreased risk of a repeat self-harm episode in individuals without a mental health diagnose and a 26% decreased risk in those with a mental health diagnoses. These findings, although lower if an individual has a mental health diagnosis, support the importance of a psychosocial assessment. More importantly these findings identify that such assessment can reduce repetition of self-harm. However, the findings may be limited because some diagnostic factors, including previous or current mental health diagnoses, were not available. Nonetheless, the large sample may override this deficit and therefore the findings could be generalised to represent the wider population. Furthermore, previous studies (e.g. Hickey et al., 2001; Kapur et al., 2002) have also highlighted that psychosocial assessment appears to reduce repetition by 50%, which offers reliability to the more recent findings.

Although a psychosocial assessment has been proven to reduce the risk of repetition there are a high proportion of individuals who will continue to self-harm after an assessment (Hawton et al, 2003), with one percent of individuals going on to commit suicide in the 12 months following the assessment (Hawton and Fagg, 1988). Therefore, before formulating a care and treatment plan, a risk assessment to identify risk-factors for future self-harm is a nursing priority. Risk assessment is integral to the management of individuals with a mental disorder and a vital part of the CPA framework (Phull, 2012). Risk has been defined as the likelihood of an event happening with potentially harmful or beneficial outcomes for the self and or others (Morgan, 2000). In mental health nursing it is not uncommon for the 'event' to be referred to as behaviours resulting in suicide, self-harm, neglect and violence. Because of the high risk of death associated with these behaviours it may explain why until recently there has been little or no consideration of the positive potentials of risk taking. However, risks may need to be tolerated and managed, for longer-term positive gains (Morgan, 2004), in relation to self-harm this may be interpreted as allowing the patient to self-harm 'safely' initially which will be discussed further in the next chapter. The Best Practice in Managing Risk (Department of Health, DoH, 2007) identifies that positive risk taking aids recovery and by avoiding all possible risks it may be counterproductive creating more problems for the patient in the long term. Therefore positive risk management is essential and can be achieved by using a collaborative approach.

The ability to assess risk effectively is an essential skill for mental health staff working with patients who harm themselves. The assessment should include identification of the main clinical and demographic features known to be associated with the risk of further self-harm and or suicide, and identification of the key psychological characteristics associated with risk, in particular depression, hopelessness and continuing suicidal intent (NICE, 2004). The means of collecting this information includes using past clinical notes, information gathered from the psychosocial assessment, actuarial assessment tools and clinical judgement, as well as working closely with the patient to anticipate their future behaviour (Doyle 1999, Morgan 2004).

In clinical practice there are a number of risk assessment tools which provide practitioners with prompts to ensure that all aspects of the risk assessment are considered. Self-harm has been identified as a main predictor of suicide therefore risk assessment tools to help identify suicidal ideation would aid the nurse. The Becks Hopelessness Scale (Beck et al, 1974) has been recognised as a useful tool to predict repetition of self-harm and suicide (DoH, 2007). If repetition of self-harm is identified as being likely, the Functional Assessment of Self-Mutilation (Lloyd et al, 1997) assessment tool may be used to determine what type of self-harming behaviour may prevail. The incorporation of specific risk assessment tools has been described by some as creating the 'ideal' assessment (Brown et al, 2004). However, the usefulness of such tools has been questioned because such tools tend to focus on common behaviours and more often than not they have primarily been developed for research purposes (Walsh, 2007). Because of these limitations the use of risk assessment tools alone would not ensure a valid risk assessment and should be used to aid clinical judgement only.

Once a comprehensive psychosocial risk assessment has been completed, a care and treatment plan needs to be formulated and although this is a fundamental part of the CPA framework it is now a legal requirement under Part Two of the Mental Health (Wales) Measure 2010. Furthermore under the measure it is now a statutory requirement that a care coordinator is allocated to the patient as soon as possible. A care coordinator is responsible for developing, reviewing and revising the care and treatment plan and coordinating the care which is delivered by both themselves and others (WAG, 2010a). Qualified mental health professionals including nurses are eligible to be care coordinators. The new standardised care plan consists of eight domains, recognising that there are a number of aspects that collectively contribute to an individual's mental well being. These include accommodation, education and training, finance and money and medical and other forms of treatment, including psychological interventions (WAG, 2012). Although all domains are of importance, for the purpose of the next chapter, the interventions will be based on the aspect of 'medical and other forms of treatment, including psychological interventions'.

To summarise, therapeutic engagement consisting of a non-judgemental, open minded, empathetic approach is essential to develop a therapeutic relationship with a patient who self-harms. A therapeutic relationship provides the essential platform needed for collaborative care to take place which will encourage behaviour changes. Furthermore, research has identified that a psychosocial assessment reduces the risk of future self-harm, suggesting the assessment is a therapeutic intervention in its self. Additionally for a care and treatment plan to be effective a risk assessment is essential, not only to help minimise future risks such as suicide and neglect but to enable the patient to incorporate positive risks in to the plan to promote recovery. Finally with a care and treatment plan established and agreed the interventions incorporated can be initiated to minimise and reduce the self-harming behaviour which will be the focus of the next chapter.

Chapter 3

Critical discussion of the Helping Approaches used within a inpatient setting to prevent and or minimise self-harm

Currently, there are few evidence based interventions or prevention programmes specifically devised to reduce and minimise self-harm (Nock, 2010). Nevertheless any intervention or helping approach needs to be based on the best evidence available (Zauszniewski, 2012). Evidence based practice is the systematic use of current best evidence to make clinical decisions for patient care, which inevitably improves patient outcome (Sackett et al, 2000; Fineout-Ovedrholt, 2005). Therefore the interventions which have emerged from recent clinical research and those that are advocated in the NICE guidelines for the short and long term management of self-harm will be discussed. The focus will be predominantly on harm reduction and specialist nursing observations. Walsh (2012) has emphasised that any intervention must be relevant to the function that self-harm serves for the patient. Thus the rationale for the use of each helping approach will be evident in relation to the functions focused on in chapter one.

This stage of the therapeutic relationship is the latter part of the working phase known as exploitation (Peplau, 1997). During this phase the patient, after collaboratively working with the nurse to identify the felt need, starts to experience an improved sense of belonging as well as developing a capability for problem solving alone or with others. In addition, the patient begins to take advantage of the available interventions to meet their needs and begins to feel an element of control over their self-harming behaviour (Nathan, 2006).

The NICE guidelines (2011) suggest a number of interventions for the management of self-harm. Initially the guideline highlights that drug treatment should not be used as a specific intervention for the reduction on self-harm. This is because pharmacological treatments such as antidepressants have been found, in numerous studies, to increase the risk of self-harm in depressed patients already at risk of the behaviour (e.g. Donovan et al, 2000 and Martinez et al, 2005). Gunnell and Ashby (2004) found this heightened risk of self-harm is particularly prevalent in young adults. Therefore, pharmacological treatments should only be prescribed for associated mental health diagnoses and if depression is the co-occurring diagnosis then the increased risk of self-harm occurring should be considered. Psychological therapies such as cognitive behavioural therapy and dialectical behavioural therapy have been found to be useful treatment options (James et al, 2012). Nevertheless if self-harm has become habitual, repetitive and an imbedded maladaptive coping mechanism the self-harm may continue during the psychological therapy suggesting other nursing interventions need to be implemented in the mean time. The guideline advocates harm reduction if stopping self-harm is unrealistic in the short-term.

Although cessation of the behaviour remains the treatment goal for many professionals providing care to people who self-harm, this may not be realistic or possible in the short term for some individuals. As with other addictive or obsessive behaviours such as drug or alcohol addiction, it may be more realistic to expect attenuation rather than complete cessation of the self-harming behaviour (Shiner, 2008). The underpinning philosophy of self-harm reduction shares some similarities with harm reduction approaches found in drugs, alcohol and sexual health. In these areas it has become widely accepted that punitive or prohibitive approaches can actually increase harm by perpetuating stigma and driving self-harming behaviours, such as illicit drug taking, underground (Inciardi and Harrison, 2000; Marlatt, 1998). In a similar fashion, service users, clinicians and academics have sought to show that punitive or cessation approaches to self-harm can result in similar negative effects. The theory underlying this approach is that self-harm is a coping mechanism and if it is immediately halted with no replacement offered a more damaging activity may replace it (Pengelly et al, 2008).

Harm reduction describes a non-judgemental practical approach to self-harm, maximising safety and limiting harm. While not encouraging or condoning self-injury, harm reduction recognises the needs of individuals and promotes empowerment and self-management (Holley et al, 2012). This intervention is concerned with providing the individual with knowledge of how to self-harm 'safer' by introducing alternative coping strategies and less destructive means of self-harm. If it's not possible to provide alternative methods then advice should be given about how to keep 'safe', for example, a nurse should emphasise the risk of blood born viruses such as HIV if cutting implements are shared (Shiner, 2008). Furthermore, it should only be done in controlled environments such as in inpatient setting where clean sterile implements are provided with the patient being taught the safest way and where to cut (Pengelly et al, 2008). It's important to note that harm reduction techniques should not be advocated if self-poisoning is the means of self-harm because there are no safe limits in self poisoning (NICE, 2011). Harm reduction may also include suggesting alternatives to self-harming behaviours such as using ice or tight bands which are perceived as safer means to achieve the same result, namely temporary relief or a reduction in psychological distress (Edwards and Hewitt, 2011).

The adoption of harm reduction as an intervention may help the patient realise that their self-harming behaviour is problematic and is something that needs to change. Realisation that the behaviour is problematic is the first step that is needed to break the cycle of any addictive behaviour (Nixon and Heath, 2009). Changing behaviour, especially one that has been adapted as a maladaptive coping mechanism such as self-harm, is a complex process involving psychological, social and environmental issues. Prochaska and DiClemente (1983) identified stages in the process of behaviour change, these include, pre-contemplation, contemplation, commitment, action, maintenance, relapse and exit stage (appendix B). Nurses can utilise this model to help patients make changes by focusing on moving them one step further around the cycle.

The nursing skills used during harm reduction can be termed Motivational Interviewing. Motivational interviewing is a style of therapy that is particularly beneficial in the early stages of the cycle of change, predominantly contemplation (Banerjee et al, 2002). It consists of adopting an empathetic, person-centred and collaborative approach that incorporates skilful reflective listening and aims to raise individuals' awareness of the impact their behaviour has on their bio-psychosocial health and to increase motivation for positive change. In essence, Motivational Interviewing is a skill that encourages individuals to conclude for themselves that change is necessary. This skill has been shown to increase the likelihood of behaviour change, improve engagement in treatment, and enhance an individual's ability to maintain changes (Noonan, 1997). These findings have been extended to individuals not only with a need to change a behaviour but also positive effects have been found in individuals who have a co-occurring diagnosis (Kavenagh et al, 2004).

Pengelly et al (2008) reported that patients perceived 'safer' self-harm to represent a shift in nurses' attitudes towards the behaviour and if nurses no longer expect total cessation, but adopt a collaborative approach to reducing the frequency or severity of self-harm, the reduction of self-harm is much more likely in the long term. Furthermore, Duperouzel and Fish (2007) identified that all patients in their study felt they should be allowed to self-harm because it was their choice and 'right' to do so. The nursing staff involved in the study also thought that allowing a patient to self-harm in a controlled environment prevented patients finding other means to self-harm which may be more dangerous such as swallowing or inserting objects. Further research has also acknowledged that patients support a harm reduction approach (e.g. Holley et al, 2012; Hume and Platt, 2007). However, further research also suggests that the majority of MHP's who support this approach is limited (Edwards & Hewitt, 2011). This may be because nurses have a duty of care to their patients and therefore understand themselves as being under obligation to stop the self-harm behaviour (Nursing and Midwifery Council, 2008).

The controversial issues surrounding harm-reduction become more evident if the harming behaviour is done so with a nurse supervising. This initially stemmed from a 'supervised self- harm' regime implemented within St George's Hospital, Stafford, in 2006. If patients before admission harmed using knives or razors they were allowed to continue with these methods under the supervision of a trained nurse. This approach seems to challenge claims that self-harm is indisputably harmful, or at least seems to validate self-harm as an acceptable means of managing psychological distress. Nevertheless, Holley et al (2012) identified that four years after this regime was implemented there was a significant reduction in self-harm incidents reported. For example, in 2004 over a three-month period 59 incidents of self-harm were reported compared to just 7 incidents over the same three-month period in 2011. These findings suggest that harm reduction in the form of 'supervised self-harm' does actually reduce the behaviour, however it may be the case that less incidents are being reported if it has become something the ward condones. Furthermore these findings would be hard to generalise as there are many other variables that may have led to a reduction for example staff training specific to self-harm which may not be available in other settings. Nonetheless it is clear to see that harm reduction does work for some individuals (Birch et al, 2011).

Hewitt and Edwards (2011) considered the controversial issues surrounding 'supervised self-harm', they determined such stigma exists because its practice risks sending a message to both the self-injurer and other potential self-injurers that this is an effective way of controlling extreme distress. Supervised self-harm seems to sanction self-injury by providing a forum for it, so there are undoubtedly concerns about this strategy. Therefore, it would be quite natural to presume that 'supervised self-harm' would be seen as a more controversial intervention than other nursing interventions. However, this is not the case, both harm reduction without supervision and total cessation have both been subject to criticism and not seen in the patients' best interest (Gutridge, 2010).

To allow self-harm to occur within an inpatient setting without supervision highlights a number of problems, it may create a milieu that the behaviour has no dangers attached and superficial harming could become more extreme unintentionally if the correct guidance isn't given. In addition, Hewitt and Edwards (2011) proposed that total prevention can damage the therapeutic relationship because it conveys to the patient that they cannot be trusted. Thus, the likelihood of self-harm to occur secretly is increased; leading to feelings of guilt if this is the only established coping mechanism (Csipke et al, 2008).

Csipke et al (2008) carried out an internet based study with 946 anonymous participants. A large proportion of their sample (205 participants) felt guilt or shame towards their behaviour if they were expected to agree to total cessation. Nonetheless, if supervised harm reduction was implemented the guilt element was removed from the self-harm cycle, which proved to be an important component for treatment to be successful and self-harm only continued for a short-period of time. This study utilised an anonymous web based questionnaire to gather their findings which usually ensures a wide sample of the population responds openly and honestly (Wright, 2005). This may suggest the findings are reliable enough to be generalised. But on the other hand there is no way of validating the individuals' experiences and the questions may have been interpreted differently by different people. In addition, the findings were not unanimous suggesting that harm reduction is not an intervention for everyone and cessation may be the only way forward for some patients (James et al, 2012). Therefore, it appears, that harm reduction may be an appropriate helping approach if the patients' recovery goal is to control and reduce self-harm whilst addressing the underlying problems (Shiner, 2008). However, the research discussed which provides an evidence base for harm reduction is relatively new. Thus, in-depth guidance and local policies on harm reduction are not readily available across the board and the prevention of self-harm and ultimately suicide remains the primary goal of acute mental health care (Pengelly et al, 2008; Bowers et al 2005).

Research suggests that special observation is one of the most common nursing interventions utilised in inpatient care to prevent self-harm and suicide (Meiklejohn et al, 2003; Bowers et al, 2000). The Sainsbury Centre for Mental Health acute problems report (SCMH, 1998) found that more than half of patients were put under an increased level of observation during an inpatient stay with over one fifth spending at least one period under constant observation.. Furthermore, Bowers et al (2000) identified self-harm as the most common reason for a patient to be placed on an increased level of observation.

A variety of terms are used in the literature to refer to the procedure, some of the most common being 'therapeutic', 'special', 'close', or 'constant observation' (Ritter, 1989; Hardy and Mingella, 1997). Special observation will continued to be used because this appears to be the most common term utilised in the literature. The Standing Nursing and Midwifery Committee (SNMC,1999) defined observation as regarding the patient attentively, while minimising the extent to which they feel they are under scrutiny. Barre and Evans (2002) expanded on this definition and suggested there are two elements to special observation, with the primary purpose to maximise patient safety and to minimise risk, and the secondary purpose being an opportunity to provide therapeutic engagement. A nurse is allocated to care and manage for the patient for a specified time. The UK practice guidelines (SNMC, 1999) suggest three levels of special observation: intermittent, where the patient is checked at specific intervals; within eyesight, where the patient must be able to be seen at all times; and within arms length, where the patient must be in close proximity at all times. The latter two, within eyesight and arms length, are collectively termed constant special observation (Bowers et al, 2008). From here on in special observation will refer to constant special observation.

Although special observation has been identified as one of the most common nursing interventions for 'at risk' patients, there appears to be a number of variations in the policies governing how to implement such an intervention, which questions the evidence base. A national survey of observation policies among psychiatric inpatient wards found varied terminology and standards of record keeping, while 1 in 10 had no written policy (Bowers et al, 2000). Nelson et al (2001) identified that medical staff usually initiate the special observation, while other studies have found nursing staff are able to make the decision without consulting medical staff (for e.g. Duffy, 1995; Dennis, 1997). Furthermore, the range of staff involved in special observations also varies from qualified nurses to nursing assistants and when nursing staff are not available, bank and agency workers are commonly employed for special observations (Bowers et al, 2000). These common variations found in numerous studies pose as a concern when the purpose of a special observation is to intervene and offer support when the patient is most in need. Furthermore, the range of staff alone suggests that the patient is not getting the level of therapeutic intervention needed. These vast variations may provide reasons as to why research has identified that special observations are often inappropriately executed (e.g. Barre and Evans, 2002; Bowers et al, 2000).

Nelson et al (2001) regarded special observation as a fundamental skilled nursing intervention, however many patients do not perceive it as such. For example, evidence from Barker and Walker's (1999) study indicated that patients did not feel safe or supported under special observations. Similarly, Jones et al (2000) found that patients do not like the experience of being observed found it intrusive and more often than not the nurses did not talk to the patients. These findings suggest that specialist nursing observations as an intervention may be somewhat meaningless with no evidence of skilled nursing being applied. However, negative perceptions appear to become eradicated when engagement becomes the focus. In one study, 13 out of 20 patients identified positive feelings associated with constant observation. They viewed it as protecting them, and this was most emphatically felt when staff were perceived to be friendly and willing to help (Cardell and Pitula 1999).

Engagement during special observation is clearly an important factor for the intervention to be effective and have purpose for both the patient and nurse. Stevenson and Fletcher (2002) proposed that observation should be seen as a chance to have uninterrupted therapeutic engagement with a patient. Practice guidance has been published to help nurses provide 'safe and supportive observations' (SNMAC, 1999). Although published in 1999 this document continues to provide the back drop for current NICE guidelines (e.g. NICE, 2005). The practice guideline advocates that, ideally, the nurse and patient should know each other and the nurse should be familiar with the patient's history, social context, and significant events since admission. Furthermore acceptance and the willingness to listen must be displayed and incorporated with self-disclosure and the therapeutic use of silence (Butler, 2007). These elements enhance the probability of effective communication occurring which in-turn will provide an opportunity to improve the patients' mental state which might be compelling them to self-harm (Bouic, 2005).

Despite the purpose to prevent self-harm and suicide, if inappropriately executed, self-harm and suicide still occurs whilst special observations are in place. Meehan et al (2006) carried out a national clinical survey in England and Wales and found 3% of suicides occurred whilst under special observation. In opposition, when policies and the fundamentals of special observations are executed a reduction in self-harm has been recognised. For example Stewart et al (2009) found that self-harm was more likely to occur when not under special observation (61%) compared to 39 % of self-harming incidents occurring whilst special observations were in place. These findings were determined by analysing officially collected data over a period of 2 ½ years from 16 acute wards suggesting reliability. However self-harm is well documented as being a secretive act, signifying that the percentage of self-harm occurring when not under special observation may be higher giving more credibility to the intervention. Conversely, the findings displayed no change in the rate of self-harm occurring whilst under special observation (Stewart et al, 2009). Although no reasons were given this finding may suggest that the increased level of therapeutic engagement may not have been utilised to promote additional coping mechanisms through skilled nursing interventions.

Jones and Jackson (2004) proposed that nurses carrying out special observations should have the skills need to deliver brief psychosocial interventions to promote additional coping mechanisms. Problem solving therapy (PST) could be an appropriate approach given that 70% of all self-harm occurs because of an impulsive response to an interpersonal problem (Boncraft, 1977).PST is a brief talking therapy which helps the patient clearly define their problems which are provoking them to self-harm, the nurse can help the patient realise the link between their symptom (self-harm) and any current interpersonal problem. The nurse can assist the patient to brainstorm other alternative or solution to self-harm and implement and evaluate them in a structured way (Mynars-Wallice, 2002). Research has suggested that PST is a realistic and effective treatment for deliberate self-harm (e.g. Hawton et al, 1998, Townsend et al, 2001).

Recently, findings of a large study which included 136 acute psychiatric wards were published (Bowers and Simpson, 2007). This study, specifically, looked at whether special observation reduced self-harm, the findings identified that special observations had no impact on self-harming rates. However the use of intermittent observations decreased self-harming behaviours dramatically. Caution needs to be exercised when analysing the findings because they suggest special observations should be eradicated but constant observation remains a legitimate and necessary intervention in high risk situations (Appleby et al, 2006). The vast amount of inpatient wards included in this study suggests that the findings could be successfully generalised to provide a base for evidence based practice. Furthermore, to minimise any limitations both quantitative and qualitative methods were used to obtain the data. This combination ensures high reliability of data, understanding the contextual aspects of the research, flexibility and openness of the data collection, and a more holistic interpretation of the research problem (Williams, 2007).

The empirical findings from the above study may suggest why local health boards such as Abertawe Bro Morgannwg university health board have created local policies which advocate flexible observations (appendix A). Flexible observation allows both the nurse and the patient to have more control over the duration of observation. This is a stark contrast from special observation where the average mean time spent under observation has been found to be 72 hours whether the patient still requires the intervention or not (Bowers et al, 2008). Flexible observations are implemented and incorporated in to the patients care plan, they span the range of observation levels but they are reviewed more frequently for example every hour. The primary nurse works with the patient to identify the level of support and supervision that will manage the identified risk of self-harm at times of crisis (Bowers et al 2006). The patient should be encouraged to seek assistance when they feel the urge to self-harm so that the observation levels can be increased to minimise self-harm. The interventions and nursing skills needed when the patient requires extra engagement will depend on the individuals needs at the time.

Furthermore, of late, Mullen (2009) discussed that because of an 'observational culture' recovery-based values are becoming diminished. The main reason for this was because nurses felt they don't have time to provide psychosocial interventions. However if flexible observations are adopted with the primarily focus being engagement, psychosocial interventions will naturally follow and become an essential part of the observation level required. This in turn will reduce the need to self-harm whilst under observation and after the higher level of engagement has ended (Bowers and Simpson, 2007).

To summarise, both harm reduction and nursing observations can be utilised to minimise and or prevent self-harm. Harm reduction is clearly controversial but nevertheless it appears to have a place in minimising self-harm in the acute setting. Nursing observations can indeed reduce and prevent self-harm if the focus of therapeutic engagement is perceived as just as important as minimising risk. Furthermore, both approaches assist an individual on the journey to recovery to function as independently as possible and reduce the need for therapeutic engagement within an acute setting which will be the focus of the next chapter.

Chapter 4

Critical analysis of ending the therapeutic relationship.

Therapeutic relationships can end for a number of reasons including, death, disengagement and discharge. This chapter will focus on the transfer of care from the inpatient setting to a Community Mental Health Team (CMHT) whilst analysing why the termination phase has been identified as the most difficult but nevertheless the most important phase of the therapeutic nurse-patient relationship (Sreevani, 2007).

The termination is the final phase in the process of the therapeutic interpersonal relationship (Peplau, 1997). The therapeutic relationship enters this phase after the helping approaches utilised in the working phase have assisted the patient to achieve a level of health where they no longer require the support from the current setting or the nurse providing the care (Peplau, 1952; 1988). For example the patient has adopted a harm reduction approach to control and alleviate negative emotions and, if prevention was the goal, special nursing observations are being used less to prevent self-harm. It is a time for both the nurse and the patient to reflect together on what has been accomplished by the patient since the relationship began whilst identifying their needs for post discharge (Chesser, 2012). Therefore the patient is moving forward and a therapeutic relationship with another nurse in a different role would continue this journey further (O'Carroll et al, 2007).

Ending a therapeutic relationship can sometimes be challenging and closure issues can arise for both the nurse and patient (Shulman, 1999). During this phase negative emotions can arise for example anger and frustration and patients may attempt to prolong the relationship by re-focusing on resolved problems (Walsh, 2007). Thus leading to the nurse feeling frustrated if the resolved problems re-surface like new problems (Boyd, 2008). Nevertheless all therapeutic relationships must come to an end (Walsh 2007). Fortunately, the literature provides guidance to help nurses overcome these issues. For example Reynolds (2009) emphasises the importance of preparing the patient for termination at the beginning of the relationship, whilst focusing the inpatient stay on recovery and moving the patient forward to help them cope with everyday economic and social realities (Cleery et al, 2006). Furthermore, instead of readdressing the problems the nurse must assist the patient to recognise they already have the skills to overcome them and assist the patient to realise they may be experiencing feelings of anxiety about leaving the supportive relationship (Boyd, 2008).

In addition to addressing closure issues, the nurse and patient engage in planning for discharge and identify potential needs for transitional care (Peplau, 1997). Maramba (2004) defines discharge planning as a process which identifies and prepares for the patients needs after discharge from an inpatient facility. The nurses' role during the discharge planning process has been identified as critical because they are the primary care giver with the most knowledge about the patients post discharge needs (Nosbuch, 2011). However, Maramba (2004) acknowledged, although the nurse is integral to the process, the role appears to be ill-defined. Fortunately, several recent national policies have outlined the discharge planning process (e.g. DoH, 2010; NHS institute, 2009) which has given clarity to the nurses' role (Lees, 2010; Morris et al, 2012).

Morris et al (2012) identified continued assessment, patient advocacy and effective collaboration with the MDT including external agencies to be important aspects constituting to the role during the discharge planning process. Continued assessment is paramount, not only to assess the stability of the patients mental health, but to ensure that 'social' aspects that contribute to an individual's 'well-being' are established for discharge. The MHM code of practice (WAG, 2012) accentuates that all aspects of the care and treatment are of equal importance with some aspects taking longer to organise, for example if addressing needs and setting outcomes which may be met by the availability of appropriate accommodation upon discharge is a critical part of discharge planning, this should be undertaken as early as possible following an admission to hospital with referrals made to the relevant housing agencies.

Furthermore, the nurse in the role of patient advocate is essential because the treating physician is responsible for determining clinical stability for discharge and identifying post-hospitalisation needs and the nurse is in the best position to influence these decisions (Alghzawi, 2012). Additionally collaboration is essential for the continuation of care to be effective. The Department of Health (2001) recommend that all patients with a history of self-harm within the previous 3 months should be followed up within a week of their discharge which emphasises the importance of the continuity of care. For continuing care to be effective, integration of mental health and social services is essential. This collaboration has been a key aspect in an attempt to reform mental health services in the UK (DoH, 1999), with the aim to minimise patient and family distress and confusion arising from service discontinuity and transition of care (Sheppard, 1995). CMHT's are a key component of UK policy for integrated service delivery (DoH, 2002). Referrals should be made to the CMHT prior to discharge, in order to allocate a care co-ordinator to the patient. The allocated care co-ordinator prior to discharge should meet with the patient and contribute to their discharge plan (Rethink, 2010). The CPA framework is an important aspect in the continuity care it allows for all of the MDT to contribute giving the patient one unified assessment, care-plan, contingency and crisis plan which provides for a smoother transition (DoH, 1999; WAG, 2011).

The above aspects are all evident in the guidance published by the National Leadership and Innovation Agency for Healthcare (NLIAH, 2008). Passing the Baton (NLIAH, 2008) was published in response to statistics demonstrating that many transfers of care were delayed (NLIAH, 2007). Tulloch et al (2011) identified that patients spend a shorter period of time within inpatient settings compared to twenty years ago and as a result nurses must conduct assessments, create care and treatment plans and plan and execute discharge within a much shorter time frame (Siefert, 2011) With the latter, discharge planning, becoming somewhat neglected which in turn increases the length of stay (Muramba, 2004). A Delayed transfer of care (DTOC) can have a negative impact on the patient with significant implications for their independence (Welsh Audit Office, 2007). In essence, Passing the Baton (NLIAH, 2008) aims to reduce DTOC and promote effective transfers of care by promoting the six fundamental principles that facilitate effective transfers. These are communication, coordination, collaboration, consideration, creativity and integrity. All of which are of equal importance and need to be tailored to suit the individuals needs (NLIAH, 2008).

Fundamentally, the purpose of discharge planning is to provide the patient with all the resources necessary to function as independently as possible, in the least restrictive environment and to avoid re-hospitalisation (Boyd, 2008). However, Gunnell et al (2008) found that during the 12 months following discharge 6.5% of patients were re-admitted at least once because of self-harm, with one third of the admission occurring with the first four weeks after discharge. The strongest risk factor for self-harm after discharge was admission for this previously. These findings are in keeping with those for suicide in other previous UK studies (for e.g. Goldacre et al, 1993; Qin & Nardentatt, 2005).These studies highlighted that the risk of suicide after discharge from a psychiatric acute ward is around one hundred times greater than that for the general population. Research has suggested that poor treatment adherence and insufficient psycho-education contributes to an increased risk of self-harm and re-admission (Suziki et al, 2003; Thompson et al, 2003; Cuffel et al, 2002).

Psycho-education refers to the provision of information to the patient and their family in relation to their mental disorder and their self-harming behaviours (Wessely et al, 2008). Psycho-education grew out of the belief that people with