Whats Borderline Personality Disorder

Published Date: 02 Nov 2017

1. INTRODUCTION.

The aim of this dissertation is to investigate whether education on borderline personality disorder (BPD) is helpful in improving the care and management of people with a diagnosis of BPD by causing an attitudinal change in staff.

What is Borderline Personality Disorder?

The ICD-10 (WHO 2010) and DSM-IV-TR (APA 2000) collectively describe BPD as a mental disorder whose symptoms include "frantic efforts to avoid abandonment, disturbance of self-image, impulsivity without considering consequences, transient paranoid ideation, an impairment of psychosocial functioning with a tendency towards self-destructive behaviour, including self-harm and suicide and mood instability with aggressive emotional outbursts" (APA 2000, WHO 2010).

BPD has both an environmental and biological basis (Bokian 2002, National Institute for Health and Clinical Excellence (NICE) 2009). Up to 66% of BPD patients have a psychosocial history of childhood abuse, with sexual abuse mostly implicated (Bokian 2002, Linehan 1993, National Collaborating Centre for Mental Health (NCCMH) 2009).

One of the core problems within BPD is emotional dysregulation. Oversensitivity to emotional stimuli manifests within the fronto-limbic system of the brain via the neurotransmitter serotonin (Ruocco et al 2012). This inhibits the ability to regulate destructive urges, aggression and affect (NCCMH 2009, Royal College of Psychiatrists (RCP) 2009).

A disturbance in early attachment formation within an environment where the experiences of the child (e.g. sexual abuse) are dismissed or invalidated, coupled with a predisposition toward emotional dysregulation creates a cycle of intense affect and inadequate coping skills (Linehan 1993) that present in BPD.

General population prevalence rates of BPD 0.7% (Coid et al 2006). However BPD is disproportionately represented within mental health and criminal justice systems (McVey & Murphy 2010). In mental health, BPD is the most prevalent category of personality disorder, with 20% of the inpatient population (NCCMH 2009, RCP 2009), whilst 25% of female prisoners are diagnosed with BPD (Fazel & Danesh 2002, Sansone & Sansone 2009).

Working with BPD; what are the challenges?

BPD generates a large amount of distress for both the patient, and those nursing them. Psychological distress is experienced with this disorder, which is usually managed by the use of dysfunctional coping mechanisms such as drug addiction, risky sexual encounters and self-harm (McVey & Murphy 2010).

In the area of staff attitudes towards patients with BPD, the focus of care is predominantly managed by nursing staff, however it is true to state that all mental health disciplines find working with BPD challenging (Bland et al 2003, Bowers 2002, Cleary et al 2002, James & Cowman 2007, McVey & Murphy 2010).

Patients with BPD have frequently been excluded from services, possibly due to a lack of confidence and skills within the healthcare workforce (NICE 2009), where lack of and inadequate training is often cited as an issue for staff managing this difficult patient group (McVey & Murphy 2010). Current literature states that nursing staff do not feel the level of training is adequate in preparing them for the role of managing and treating BPD (Cleary et al 2002, Deans & Meocevic 2006, James & Cowman 2007, Krawitz 2004). Healthcare policy requires mental health services to become more responsive to the needs of patients with BPD (National Institute for Mental Health in England/Department of Health (NIMHE/DH) 2003) in an attempt to end marginalisation within services.

Due to the entrenched nature of BPD, working with patients can be both slow and labour intensive. A twelve month period of therapeutic relationship building may be needed in order for the patient to feel safe enough to engage in any meaningful work (Murphy & McVey 2010). Therefore, a lack of post-qualifying education fails to provide the opportunity for staff to develop the skills, knowledge and attitude for working with BPD patients.

2. RESEARCH METHOD.

2.1 Development of the research question.

This dissertation originates from my previous experience of working with borderline personality disorder (BPD) women in a female secure unit as a nursing assistant. I noted that some staff displayed negative attitudes towards patients with BPD. These attitudes lead to people with BPD being stigmatised as attention seekers, manipulative, disrespectful and unappreciative, which led me to consider whether education surrounding BPD would help alleviate some of the negative attitudes displayed, thus improving the standards of care provided.

People with BPD give rise to the notion of the ‘unpopular patient’ (Pilgrim 2001, Wright et al 2007) who can evoke powerful negative emotional and attitudinal responses from staff (McVey & Murphy 2010). These negative responses lead to challenges in the nursing of BPD patients and denial of services on the grounds of non-treatability and lack of adequate care (James & Coleman 2007).

2.2 Rationale for a literature review.

This study is based around the hypothesis that education on BPD can have a positive effect on the attitudes of nurses working with this difficult client group. The rationale of conducting a literature review as my methodology is justified in that it provides a comprehensive critical in-depth analysis of pre-existing research literature (Walliman & Appleton 2009, Aveyard 2010), it negates the lengthy process of gaining ethics committee approval (Aveyard 2010), time constraints would not have allowed me to formulate and administer a primary research project thoroughly and after reading around the subject it became apparent that I would be able to manage to answer my question by summing up already published material (Walliman & Appleton 2009).

2.3 The literature review.

In order to establish what is already understood about staff attitudes towards BPD, I undertook an initial opinion based exploration of the relevant literature. This showed that there are still underlying issues of negativity around BPD which posed my initial question of ‘Would staff attitudes improve with more education surrounding personality disorders?’

As such I conducted a full literature review to provide an overall view of how education can provide a caring environment for people with BPD. The literature review will be managed by the use of following steps:

Identify a research topic relevant to the programme being studied.

Identify what is already known about the subject.

Identify a research question that is neither too wide nor too narrow.

Complete a systematic search of the literature.

Review the strength and weaknesses of the research methods.

Synthesise the findings and sum up the literature.

Suggest implications for practice and possible future research

Aveyard (2010).

2.4 Search Strategy.

From the initial literature review the proposed research question asked ‘How effective is education aimed at improving staff attitudes towards self-harming behaviour in patients with a diagnosis of borderline personality disorder?’, as the emphasis within mental health nursing is to build and maintain effective positive therapeutic relationships (DH 2006, Barker & Barker 2009), challenge inequality and promote recovery (DH 2004).

Thesaurus and ‘Mesh’ term searches were incorporated to identify all of the relevant search terms available. Searches were carried out using truncation (*) to allow for different word endings. Synonym searches using the Boolean logic operator ‘OR’, thus ensured any alternate terminology and spellings were included helping to broaden my search and increases the number of results. Combinations of the key terms using ‘AND’ were then applied to provide focus to the search (Table 1).

Table 1: Initial search terms.

Search Term

Combined with

"Education OR Training"

"Staff"

OR

"Nurs*"

OR

"Healthcare work*"

Or

"Mental health Nurs*"

"Self Harm" OR "Self Injury" OR "Self Mutilation"

"Personality Disorder" OR "pd" OR "Borderline Personality Disorder"

"Attitudes" OR "Perceptions" Or "Views"

PsycInfo, CINAHL, BNI, ASSIA, Cochrane Library and the Pubmed data bases were systematically searched. PsycInfo, CINAHL, BNI were used due to familiarity and containing primary, healthcare related research, advice from colleagues directed me towards ASSIA and Pubmed and Cochrane Library was used to identify any recent reviews covering my research question, thus providing originality to my own dissertation. To ensure the systematic search was not affected by spelling mistakes or misplaced truncation the terms in Table 1 were copied and pasted into each database. Further articles were found both by hand searching relevant journals and the scrutinising of references. This snowball sampling technique is considered effective in achieving a comprehensive literature search (Aveyard 2010, Walliman & Appleton 2009).

The keywords listed in Table 1 were used to identify trials and reviews relevant to the study aim. However, very little literature was found that was relevant to staff attitudes, self-harm and BPD. I concluded that the key terms were narrowing the question too much. Using both self-harm and BPD as keywords over-narrowed the search strategy and produced too few articles.

Aveyard (2010) addresses this by explaining that research addressing one factor only is more realistic at undergraduate level study. Therefore a second database search using the same keywords but excluding "Self Harm" OR "Self Injury" OR "Self Mutilation" was performed as BPD and self-harm are, in many cases, inexorably linked as a way of emotional regulation (Kleindienst et al 2008). Therefore I settled on the more manageable question of Is targeted clinical education helpful in improving staff attitudes towards patients with a diagnosis of borderline personality disorder?

In light of my revised question my aims and objectives for this study concentrate on three areas. I will examine how BPD patients are seen as unpopular and challenging and the attitudes that are portrayed towards them, discuss educational input’s effect on attitudinal change and the availability and efficacy of educational programmes.

The revised keyword search (Table 2) was systematically applied using identical databases as the initial search (Table 3). NHS Evidence and Google scholar were additionally searched simply as a means of accessing full text articles that were not available within other databases.

Table 2: Revised search terms.

Search Term

Combined with

"Education OR Training"

"Staff"

OR

"Nurs*"

OR

"Healthcare work*"

Or

"Mental health Nurs*"

"Personality Disorder" OR "pd" OR "Borderline Personality Disorder

"Attitudes" OR "Perceptions" Or "Views"

Table 3: Electronic Databases.

PsycINFO

CINAHL (Cumulative Index to Nursing and Allied Health Index)

BNI (British Nursing Index)

ASSIA (Applied Social Sciences Index and Abstracts)

COCHRANE LIBRARY

PUBMED

NHS Evidence (accessed as a means of full text access not available in other databases)

Google Scholar (accessed as a means of full text access not available in other databases)

The systematic searching of the databases produced the following results (Table 4).

Table 4: Database Results.

Search keywords and synonyms

Results

PsycINFO

Results CINAHL

Results

BNI

Results

ASSIA

Results

COCHRANE

Results

UK PUBMED CENTRAL

Results

NHS

Evidence

#1

Education OR Training

583,122

369,437

31,281

92,262

41,569

996,036

1,633,922

#2

Personality Disorder OR pd OR Borderline Personality Disorder

44,508

5932

387

560

17,314

5,080,743

61,044

#3

Attitudes OR Perceptions OR Views

747,215

110,171

21,803

3310

26,387

672,551

761,621

#4

Staff OR Nurs* OR Healthcare Worker Or Mental Health Nurs*

139,109

573,663

197,278

70,406

26,350

909,553

430,790

#1 and #2

2978

612

43

361

860

86,668

2721

#2 and #3

4308

221

70

451

658

73,628

1968

#1 and #2 and #3

405

46

12

40

99

10,018

207

#2 and #3 and #4

214

71

70

47

56

8272

94

#1 and #2 and #3 and #4

51

23

12

11

24

3216

24

2.5 Inclusion and exclusion criteria.

Inclusion and exclusion criteria were applied (Tables 5 & 6) to help distinguish between relevant and irrelevant information (Glasper & Rees 2012). This stopped me becoming sidetracked with articles not directly related to my question (Aveyard 2010), therefore keeping the review focused and articles down to a manageable number. Amongst prisoners within the custodial system (n = 88,179) (Bermen 2012), 14% to 25% of males and 20% to 26% of females meet the criteria for BPD (Fazel 2002, NICE 2007, Sansone & Sansone 2009). Mental health nurses provide services within prisons and work closely with the prison officers, as such I decided to include prison officers and mental health in-reach teams in order to examine the attitudes of all staff working with BPD.

Table 5: Inclusion criteria.

Inclusion Criteria

Rationale

Articles published between 2002 and present only.

To provide more up-to-date research as opposed to historical, an inclusion a period of ten years was decided upon.

English language publications only.

Due to being an English speaker only I felt it pertinent to only allow English language only articles. This is noted as a limitation of the study later.

Primary research directly related to the chosen topic.

Primary research papers are undertaken according to an accepted scientific method (Aveyard 2010).

Use of European, Australian and American Articles.

Due to the paucity of literature from the United Kingdom, I decided to include other countries in an endeavour to search as comprehensively as possible.

Key terms in title and abstract.

Searching for Keywords within both title and abstract provides more comprehensive search results (Aveyard 2010, Walliman & Appleton 2009).Table 6: Exclusion criteria.

Exclusion Criteria

Rationale

Articles published pre 2002.

To provide the most relevant, up to date information any research that is older than 10 years was not considered for this dissertation.

Unpublished (Grey) literature.

Searching for unpublished (Grey) literature is beyond the scope of an undergraduate researcher.

Full text documents available as purchase only.

Due to monetary constraints, it was not possible to include any non free articles. However this is stated as a limitation to the research process later.

Primary research whose focus was self-harm.

Although there is a distinct link between BPD and self-harm, it is not exclusive. Self-harm will still be discussed however, the primary focus will remain on the effects of education on attitudinal changes.

Secondary research (added after discussion with dissertation supervisor).

Primary research should be the focus of a dissertation. A secondary source is one step removed from the ideas referred to within the literature and should be avoided (Aveyard 2010).

From the original forty-six papers identified, twenty-one did not meet the inclusion criteria. Of the remaining twenty-five papers, thirteen were identified as either secondary research papers (9) or primary research that was not directly related to my question (4) and as such were excluded from the list. This left a total of twelve papers that would be critiqued, coded and themed.

2.6 Critical appraisal of evidence.

Critical appraisal is an important step in the review process (Katrak 2004, Walliman & Appleton 2009) and helps evaluate the quality of the evidence. Critical appraisal is defined as:

A structured process of examining a piece of research in order to determine its strengths and limitations, and therefore the weight it should have in your literature review.

Aveyard 2010:93.

As such I followed Aveyard’s (2010) guidelines for summarising the findings from the articles (Table 7) using the Critical Appraisal Skills Programme (CASP) (CASP 2010a,b,c) for qualitative and quantitative papers as they provide "a consistent approach to the critique of research" (Aveyard 2010:99).

Research questionnaires can be either qualitative or quantitative dependent on the questions asked (Fenton 2009, McLeod 2008). Therefore Greenhalgh’s "ten questions to ask about a paper describing a questionnaire study" (Greenhalgh 2010:178) was employed to critique the questionnaire papers.

Table 7: Summary of Information: Papers selected for full text retrieval (after Aveyard 2010: 128).

Author / Date

Aim of Study / Paper

Type of Study / Information

Main Findings / Conclusion

Strengths / Limitations

Cleary M, Siegfried N &

Walter G (2002)

To obtain baseline data to provide direction for developing and planning education.

A 23 item questionnaire and statistical analysis.

N=516 from various occupational levels and settings. Measurements of knowledge, attitudes, experience and willingness to attend training.

PD patients are challenging. 76% of nursing staff support the need for continued staff education and support. 76% of respondents would find further training helpful with 95% stating that they would readily participate in further training.

Standardised approach to data collection using an established collection programme covering different settings and professions.

Noted missing data and poor response rate of 44% (n=229). Would have been more effective if combined with another research method such as semi-structured interview to provide more in-depth answers. No indication of pilot of questionnaire.

Krawitz R (2004)

To assess the effect of training workshop on clinical attitudes towards working with BPD.

A survey questionnaire administered pre and post workshop training. N=910 from public mental health and substance abuse services attending workshop training over an 18 month period.

Workshop training was effective in achieving positive attitudinal change in clinicians working with patients with BPD. This showed the achievability of attitudinal change through training towards positivity of working with BPD patients.

Large sample size and good response rate of 62% (n= 418) and the exclusion of participants who attended other training on BPD to avoid confounding variables to attitudinal change (n=241). Good longitudinal follow up of 18 months (6 monthly follow up).

Survey questions not tested for validity and reliability. Sample biased towards only those who wanted to attend the workshops. No indication of pilot of questionnaire prior to use.

Author / Date

Aim of Study / Paper

Type of Study / Information

Main Findings / Conclusion

Strengths / Limitations

Bowers L, Carr-Walker P, Paton J, Nijman H, Callahan P, Allan T & Alexander J (2005)

To identify what events and experiences influence the attitudes of prison staff working on a dangerous and severe personality disorder (DSPD) unit.

Semi-Structured interviews using open questions.

N=66 from mixed gender prison officers working across an assessment and treatment unit.

Positive factors on staff attitudes included a greater understanding of the PD, improvement in prisoner behaviour, education, and staff support. DSPD units need to adopt timely treatment programmes, clear philosophy and regime, staff training and clinical supervision.

A longitudinal study with semi-structured interview at baseline, eight and sixteen months. Transcribed tapes and recognised software used for analysis (NUD*IST 6).

Fair participant number (n=66) produced a total of 96 interviews for analysis.

Semi-structured interview with open ended questions allowed for a more in depth understanding of attitudinal change and precipitating factors.

Questions in the semi-structured interviews reflected on the previous eight month period, allowed for recall bias. This bias was not investigated and shows some weakness in the research validity. Respondents may have created explanations for attitudinal change based on other external factors.

Bowers L & Allan T (2006)

To assess the components of attitude to personality disorder and factor analysis of feelings generated within nursing staff by PD.

Questionnaire comprising of 37 items set on a 6 point Likert scale.

N=904, over three sets, from various occupational levels and settings.

The largest influences on staff attitudes towards BPD are organisational and suggest that training, supervision and staff support structures promote positive attitudinal changes better than simply employing positive staff.

Good psychometric study comprising of the measurement of knowledge, abilities, staff attitudes & personality and educational input.

Good participant numbers (N=904) using recognised diagnostic tools for analysis.

Overrepresentation of non-registered staff who were non responders which may skew results from this staff group as uninformed response to the behavioural aspects of PD may be liable to overly judgemental and negative attitudes.

Author / Date

Aim of Study / Paper

Type of Study / Information

Main Findings / Conclusion

Strengths / Limitations

James P D & Cowman S (2007)

Aims to describe nurses’ knowledge, skills and attitudes towards BPD to provide guidance toward development of services and education.

A survey questionnaire consisting of five sections of demographics, frequency of contact, understanding of diagnosis and treatment, confidence and role understanding.

N=157 of mental health nurses.

BPD is regularly encountered by mental health nurses and is considered difficult to manage. Services need to develop in line with best practice. Nurses need to be provided with appropriate education in meeting the needs of this client group.

Use of previously established and recognised questionnaire. A small pilot study was used to assess content and validity (n=15). Inclusion of recognised rating scale (Likert) and multiple choice allowed for more depth answers.

Small study participant group (n=157) with a 41.4% response rate (n=65). All respondents worked in the same geographical location, as such caution needs to be used when attempting to generalise results.

Single use methodology used to collect data. A more in-depth analysis could have been achieved by used of multiple collection methods.

Commons-Treloar A J & Lewis A J (2008a)

To assess the attitudes of mental health and medical emergency medicine clinicians towards patient with BPD.

A purpose designed and established questionnaire and assessment tool to quantify attitudinal levels and demographic information between two differing groups of staff.

Significant differences were found between mental health and emergency medical staff when looking at the attitudes, with emergency medical staff having poorer attitudes. However both groups showed a moderate improvement in attitude when provided with training on the disorder

N=140 participants across three health services covering both general and mental health care. Purpose designed demographic questionnaire coupled with a standardised and established questionnaire on attitudes towards self harm (ADSHQ) adapted (with permission) to be only used with patients with a diagnosis of BPD. Good response rate mean of 90.8% (n=127) (community 89.4% & inpatient 92.3%).

Non-representation of unqualified staff within the study disregards a large portion of the workforce.

Author / Date

Aim of Study / Paper

Type of Study / Information

Main Findings / Conclusion

Strengths / Limitations

Commons-Treloar A J & Lewis A J (2008b)

To assess the impact of targeted clinical education on BPD on the attitudes of staff towards BPD diagnosis.

A RCT using purpose designed and established questionnaires and assessment tool to quantify attitudinal levels and demographic information in staff dealing with patients who self harm and have a diagnosis of BPD.

Statistically significant improvement in staff attitudes were found within both staff groups following the targeted educational program. This shows a necessity to provide regular access to the targeted clinical education of staff working with patients with a diagnosis of BPD.

Both pre and post education scores rigorously analysed to obtain data using a recognised analytical tool (ANCOVA) to compare pre and post educational attitudes within the ADSHQ adapted (with permission) to be only used with patients with a diagnosis of BPD.

Non-representation of unqualified staff within the study disregards a large portion of the workforce.

Woollaston K & Hixenbaugh P (2008)

To explore nurses relationships with BPD patients from their own perspective.

Semi-structured interviews and thematic analysis of results of staff attitudes and beliefs surrounding the diagnosis and how to best provide care.

The data contained within this article appears to suggest that negative attitudes are borne out of staffs feelings of helplessness. The need to provide nurses with appropriate training in managing this groups is paramount.

Good clear methodology and explanation. Clear explanation of the thematic analysis process providing clarity and validity to the study.

Small participation number (n= 6) and researcher had previously worked with 66% of participants (n=4) which could have affected their inclination to express themselves freely and candidly.

Author / Date

Aim of Study / Paper

Type of Study / Information

Main Findings / Conclusion

Strengths / Limitations

Purves D & Sands N (2008)

To investigate staff attitudes of psychiatric staff towards BPD with a view to looking at how education and supervision can produce attitudinal change.

35 item Questionnaire to assess global attitudes towards personality disorders using close ended and fixed response questions.

Overall negative attitudes were reported within the questionnaires. A positive relationship between education and attitudinal response is recorded within the literature. Negative attitudes directly relate to inequality of care.

Recognised and validated questionnaire used (APDQ) using Likert scaling. Information analysis using recognised and validated software programme (SPSS).

Good participation size (n= 110) with a return rate of 54% (n=61).

No indication of pilot of questionnaire prior to use. Based on a single geographical study and no data from non responders recorded.

Commons-Treloar A J (2009)

To examine the impact of two educational programs on clinical attitudes towards BPD.

Both a purpose designed and established questionnaire and assessment tool with a six month follow-up.

Significant improvements in attitude were noted immediately after both educational programmes. However, this improvement was not maintained in the group who attended cognitive-behavioural program where as the psychoanalytical; group participant maintained a significantly improved attitudinal change towards the diagnosis of BPD.

Recognised and validated questionnaire used (ADSHQ) using Likert scaling.

Use of both qualified and non qualified staff.

Good use of follow-up questionnaire to validate continued attitudinal changes.

Although a validated questionnaire, the ADSHQ only allows for the respondent to reply positively or negatively. Having no neutral response applies some restrictions to the results.

Participation size (n=140) with an 18% dropout rate (n=118) is relatively small for this type of study and would need to be replicated on a larger scale to improve its validity.

Author / Date

Aim of Study / Paper

Type of Study / Information

Main Findings / Conclusion

Strengths / Limitations

Maltman L &Hamilton L (2011)

A preliminary evaluation of the effects of workshop training in changing attitudes towards personality disorder in prison staff.

A recognised and validated 37 item, self-report questionnaire using Likert scale administered pre-training and after a two month interval.

Workshop training showed little improvement in attitudinal change. However, significant differences were noted on the security vs vulnerability scale with participants reporting feeling less fearful, anxious and manipulated by PD offenders.

The ADPQ questionnaire is known to provide a robust structure, good psychometric properties and validation.

Missing data was evaluated to ensure the evaluation sample represented the entire trainee group using a t-test.

Low sample size of this research (n= 67) Entrenched negative attitudes went untested within the studay and post training data was difficult to collect, indicating some potential self-selection bias within the group.

Brief interviews of participants would have helped to determine prior understanding of PD to help complete the ADPQ as it asks for previous emotional reactions to PD.

McGrath B & Dowling M (2012)

A qualitative study of psychiatric nurses attitudes towards working with BPD patients.

In-depth interviews incorporating thematic analysis of open ended questions.

Participants perceived BPD patients in a negative light, providing minimal care and distancing themselves from patients. Understanding the origins of BPD, produces a higher level of empathy. Education and workshops were discussed as a way of improving this understanding.

Multiple data collection method used (semi-structured interview & staff-patient interaction response scale (SPIRS)). Allowing for more in-depth participant narratives. Verbatim quotes from the interviews allowed insight into the main issues that have been identified in the analysis.

Only one mental health unit used within the study limiting validity for global attitudes. Self-selecting sample produced only those interested in working with BPD. Those who did not participate may have differing views.2.7 Strengths and limitations of the literature papers.

Methods utilised in gathering information varied and were not consistent in their measuring of data. A lack of information on bias appeared to be an issue in most papers. By discussing any bias in a study allows the reader to make a "judgement on the worth of the evidence" being presented (Glasper & Rees 2012:74). Low response rates for questionnaires were discussed by many researchers as a limitation to their studies, as were attrition rates in follow up studies. Attrition and response rates need consideration as these may distort the results (Hill & Spittlehouse 2003) thus reducing the validity of the studies.

The majority of papers did not discuss the implementation of a pilot questionnaire. This may show a weakness in the research (Walliman & Appleton 2009); however piloting may have been undertaken but not reported. The importance of pre-testing questionnaires on a small group of participants prior to use within a main study is vital as this allows testing of the questions and addresses any issues that may arise (Walliman & Appleton 2009)

Some studies examined the experiences of prison officers who work with BPD patients where others looked at mental health nurses and general nurses. This brought up individual perceptions from each group, however the strength here lays in the fact that these staff populations, in most cases, work closely in collaboration with each other.

2.8 Synthesis of article findings.

After the appraisal of the papers (Appendix 1), they were compared using thematic analysis to assign codes to the findings. Themes were then developed using a theme matrix (Crookes & Davies 2006) (Appendix 2).

Coding of papers helped to identify interesting features of the information for my research (Winch et al 2008).

A code ...is most often a word or short phrase that symbolically assigns a summative, salient, essence-capturing, and/or evocative attribute for a portion of language-based or visual data.

Saldana 2013:3

Coding identifies short segments of data that appear interesting, and refer to data that can be assessed regarding the phenomenon being researched (Tuckett, 2005). I organised my data into meaningful data codes such as ‘challenging patients, staff hostility and training increases positive attitudes’ (appendix A2.1). Davies (2006) explains that the key rules for coding are to code for as many patterns as possible, code data inclusively by keeping some of the surrounding data, ensure that context is not lost and data extract may be coded many times as necessary.

When the data had been coded I began to re-focus the data. By considering how different codes corresponded and combined, I was able to produce my first level of themes (e.g. overburdened, hopelessness and futility being grouped under staff feelings). This produced my initial thematic map showing the different codes and possible themes they represent (Saldana 2013) (Appendix A2.2).

Further reviewing of themes allowed me to incorporate separate themes, for example positive attitudes, negative attitudes and staff feelings into an overarching theme of staff attitudes. This produced a final thematic map of my data (appendix A2.3).

I then produced a detailed analysis of my themes by considering how each theme related to each other. I was then able to define my themes and the names I would give them in my final chapters (appendix A2.4).