patient dangerousness and the views of nurses on low dependency wards

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of the copyright holders. Permission will usually be given for reproduction of graphs and tables if appropriately referenced. This publication is intended for use by researchers for backup or support of their own data and may not be utilised or presented in the original format.

A copy of this report is downloadable in Portable Document format

Click here to obtain a copy (or right mouse click and save target as)

You will need the appropriate version of the Adobe Acrobat Reader program to view the file. The program is freely available if you do not already have it.

The following study examines patient dangerousness and the views of nurses on low dependency wards at Ashworth Hospital, one of the three Special Hospitals in England and Wales which cater for patients who are a danger to themselves or others, and may or may not have criminal propensities. The study had four aims each related to the other and following a logical process; firstly, to identify the nurses' views on the assessment of patient dangerousness; secondly, to discover how nurses formulate their views on future patient dangerousness; thirdly, to discover the contribution of other disciplines when formulating these views; and lastly, to determine if such views so formulated are tested out.

The study was prompted by the dearth of research on nurses and the assessment of future patient dangerousness. The writer hypothesises that dangerousness is central to the assessment process for nurses working in forensic psychiatry and that nurses have a central role to play in the multi-disciplinary assessment of future patient dangerousness.

The sample was chosen purposively (N=51) from qualified nurses on the four low dependency wards at Ashworth Hospital. A survey of the nurses was conducted to collect qualitative data using a questionnaire developed specifically for the study. This consisted of mainly checklist and Likert scale questions and the responses were quantified for analytic purposes. A response rate of 60.78% (N=31) was achieved.

Indicators from the results show that the nurses felt nurses could accurately assess future patient dangerousness and that they should be actively involved in the assessment of it. Value was placed on areas of assessment and members of the multidisciplinary team which influence their views on future patient dangerousness, and a number of areas were identified through which they test these views out.

Indicators from the study suggest that there is a need for some form of assessment instrument for nurses to use to formalise these views and a number of areas for future research are suggested.

Professor Val Reed, RGN, RMNH, RNT, BEd (Hons), MA, PhD, Research Consultant, for academic supervision during the study and assistance with the research proposal.

Dr Tom Mason, RMN, RGN, RMNH, BSc (Hons), PhD. Lecturer in Forensic Nursing/Research, Ashworth Hospital, for assistance in developing the questionnaire.

Mr Bernard Moran, RMN, BSc (Hons) Nursing Studies, Cert in Adult Beh Psych, ENB 770, Advanced Nurse Practitioner, Ashworth Hospital, for assistance in developing the questionnaire and clinical supervision during the study.

Future patient dangerousness and how nurses formulate their views on it was chosen as a research topic as it has been a particular area of interest to the writer for many years and relevant to the forensic nursing field in which he is employed. Dangerousness can be defined as the potential to cause serious physical and psychological harm to others (Hamilton, 1982). It also includes fear inducing, impulsive and destructive behaviours (Hepworth,1982; Gunn,1982 ).

Nurses formulate their views on patients through the nursing process. Assessment is the fundamental part of this process, defined by Barker (1985) as 'something to do with estimating the character of something or someone'. Barker went on to say the process is concerned with 'estimating our patients in terms of who or what they are'. For nurses working in the forensic field the level of dangerousness is central to this assessment process. Dangerousness and its treatment is seen by Owen (1991) as the legitimate concern of psychiatry and psychiatry legitimates that treatment.

Forensic psychiatry, and especially the special hospitals are environments constantly under the scrutiny of the media and the general public. According to Marra et al. (1987) 'an increased burden has been placed upon mental health providers to protect the public, by identifying dangerous persons and taking the proper professional action'.

Although there have been many studies examining various aspects of patient dangerousness, very few include nurses in their analysis. However, Sepejak. et al. (1984) reported in their study that, of the five disciplines studied, nurses scored low in predictive accuracy of future patient dangerousness. Therefore, it is difficult to develop any direct theory relating to nurses and the assessment of patient dangerousness. Most of the studies appear to adhere to one of two opposite but related theories in discussing their results on patient dangerousness. Monahan (1988) summarises this well in a review of literature on risk assessment, where it was found that, for every study which found predictive accuracy, there was one that found it was no better than chance.

It is nevertheless necessary to predict future patient dangerousness. Every one in the clinical team has a part to play in this; and nurses especially are in a good position to assess their patients' level of dangerousness, since they spend twenty-four hours a day with the patients, getting to know them, building rapport, developing trust and rehabilitating the patients through community experiences.

From the writer's own experience it is becoming more frequent for Mental Health Review Tribunals to request reports from nurses. This must involve some future risk assessment, especially if the patient is on a rehabilitation programme. Additionally, nurses' levels of accountability are increasing as the profession develops. They are accountable to managers, the patients, their relatives, the media, the general public, the clinical team and, not least, to themselves.

However, when asking nurses how they arrive at their views about how dangerous a patient is, many a time one hears "it's a gut feeling". The writer feels that the process is more scientific than this; and that every view formulated must have some factual basis. Views are not innate; they are developed as a result of the information an individual has gathered over a period of time. It is thus important to establish how nurses working in the forensic setting actually arrive at their views on patient dangerousness.

An extensive review of the literature on dangerousness and risk assessment highlighted a number of areas perceived as being relevant to the study undertaken. Although there is a lack of studies involving nurses in their sampling, there are studies of related disciplines which were of use in the questionnaire design.

Firstly, there are studies which include psychiatrists in their sample. Hostility, agitation, previous assaults, suspiciousness (Werner. et al. 1984), history of violence (Menzies. et al. 1981), forensic history (Menzies. et al. 1981; Quinsey and Maguire.1986), social history (Jackson.1986), impulsive behaviour (Segal. et al.1988), recent incidents, substance abuse, and current offence and surrounding circumstances (Quinsey and Maguire.1986) are all shown to be indicators of future patient dangerousness.

Secondly, there are studies which include psychologists in their sample. Violent fantasies, ruminations around the index offence, current mental health, mental health history, sexual behaviours (Marra. et al.1987), uncooperative behaviour, hostility, suspiciousness, impulsive behaviour (Werner and Meloy.1992), history of violence (Cooper and Werner.1990), forensic history (Cooper and Werner.1990; Quinsey and Maguire.1986; Marra.et al.1987), current offence and surrounding circumstances (Cooper and Werner.1990; Quinsey and Maguire.1986), substance abuse (Quinsey and Maguire.1986; Marra.et al.1987; Werner and Meloy. 1992) and recent incidents (Quinsey and Maguire.1986) are all shown to be indicators of future patient dangerousness.

Hepworth (1982), a social worker at Broadmoor Hospital, highlighted current mental health, current behaviours and attitudes, forensic history and social history as areas which should be taken into account when assessing future patient dangerousness. Feinstein and Plutchik (1990) suggested that current thoughts, recent behaviours, past history of violence, anti-social destructive behaviours and the ability to cooperate should all be included in any assessment of patient dangerousness.

Probably the most useful paper found was that by Pollock and Webster (1990), which introduced a model for assessing patient dangerousness. The main areas highlighted as predictors were established patterns of violence, social history, anti-social value systems, underlying hostility, sadistic orientation, surrounding circumstances of the offence, drug and alcohol abuse, mental health history, suspicion, irritability, potential for change, current attitudes, motivation, socially acceptable values and goals, and self attitudes.

A non-experimental approach was used in the study. This approach is more widely used within the social sciences and highly appropriate for use in health care research (Reid,1993). This approach was chosen for the following reasons: firstly, there is no manipulation of variables; secondly, the approach is valuable for exploring social dimensions of health as well as attitudes and interpersonal relationships; thirdly, the approach is very powerful in studies intended to be descriptive; and fourthly, it is appropriate in the investigation of qualitative issues where theory generation is an aim.

The research was divided into three logical phases over a period of seventy weeks.

Weeks 1-30 saw Phase One of the research. During this time the writer prepared and submitted the proposal to the hospital Research and Ethics Committees. Parallel with this was establishment of contact with the ward managers and members of the sample, giving a brief résumé of the research and seeking permission to carry out the research on the wards.

Throughout this phase the literature review was conducted, paying particular attention to theory, methodology, instruments and models in relation to dangerousness. The writer designed the questionnaire and submitted this for validation. The pilot study was included in this phase and the reliability of the instrument was tested. Modifications of the questionnaire were made towards the end of this phase.

Weeks 31-60 saw phase two of the research and included the clinical start date. Sampling of all the low dependency wards in the Mental Illness Units at Ashworth; distribution and subsequent collection of the questionnaire; and analysis of the data collected were all included in this phase. Throughout this phase the literature review continued.

Weeks 61-70 saw phase three of the research. This time was devoted to completion of the research report and its submission to the writer's examining body. At completion of this phase there will be feedback of the results to the participants in the study.

The sample was obtained using a purposeful approach from the four low dependency wards of the Mental Illness units at Ashworth Hospital (N=51).

This form of sampling was used as opposed to a more random one as the writer wanted to look at views of nurses on a particular dependency level of ward. Low dependency wards are areas within a special hospital where more trust is placed in the patients, who are generally less problematic and may be nearing the end of their stay in the hospital. The writer's own ward is low dependency, a pre-discharge ward in one of the mental illness units. Therefore, the level of future patient dangerousness could be seen as extremely significant in these areas of the hospital. The criterion for inclusion in the sample was that the participant should be a qualified nurse involved in the assessment, identification, planning and evaluation of individual care given to patients who are considered to be at a lower dependency level.

Data were collected by means of a postal survey. A copy of the questionnaire developed for the study was sent to each qualified primary nurse on the four wards.

By using a postal survey, data were secured at minimal time and expense; but more importantly, the respondents were given a sense of privacy. The writer, who works on the wards and knows many of the sample group, is aware how at present ward based staff are very apprehensive and suspicious of anyone asking them questions. It was, therefore, felt that there would be a greater chance of obtaining data by use of an anonymous postal questionnaire, where any threat may be viewed as minimal.

It had to be accepted that there could be problems with a low response rate from the use of a postal questionnaire. To increase the chances of obtaining a higher response rate particular attention was paid to layout and length of the questionnaire and the time it would take to complete it. A covering letter was sent out with each questionnaire and a follow-up letter was also sent out three weeks later.

There were other problems which had to be taken into consideration with using a postal survey: responses of those who may respond may differ significantly from those who don't; and the data may be superficial. To compensate for this the questionnaire was constructed as a checklist, mainly answered by use of a Likert type scale. There was ample opportunity for respondents to address any of their own issues at the end of certain questions. The writer felt that although this would place some control over the research, it was a necessary step to establish a base for every respondent and make analysis of the data simpler.

When measuring attitudes a researcher needs to work from the basic assumption that the questions will have to have the same meaning for all and that the responses will have to be quantified in some way. The writer tried to tackle this problem by submission of the questionnaire to a small panel of experienced researchers and subject specialists. There was also a pilot study carried out on a small sample representative of, but not included, in the main study sample. The questionnaire consisted of checklist and rating scale type questions. It was hoped that this would minimise any misunderstanding of the content of the questions during the study.

The two Likert scales developed for use on the questionnaire sought to differentiate how strongly the respondents felt about or valued the questions being asked. This was done by having strongly disagree/agree on the scale as well as disagree/agree, and very unimportant/important as well as unimportant/important.

According to Polit and Hungler (1993) "an instrument can be said to be reliable if its measures accurately reflect the true measures of the attribute under investigation". As the research instrument was collecting information that was qualitative in nature the responses that were given on the questionnaire could not be treated as stable. Opinions and views may change over a period of time following the influences of new information received. Therefore, although the data received were quantified for ease of analysis, reliability tests for use on quantitative data were not used. Lincoln and Guba (1985) suggest another method for use under these circumstances and this was the one chosen by the writer. Reliability tests were conducted at the pilot study stage, when following analysis the respondents were requested to give qualitative validation of their replies.

Polit and Hungler (1993) define validity as "the degree to which an instrument measures what it is supposed to be measuring". There are a number of processes involved in establishing the validity of an instrument but it has to be accepted that this is an extremely difficult thing to achieve with a new instrument that is designed specifically for use in a small study. The writer endeavoured to establish face and content validity by submitting the questionnaire to a small panel of subject experts/experienced researchers. This panel was asked constructively to criticise the instrument; and their suggestions were included before submitting the proposal to the Hospital Research Committee. This committee consists of many experienced researchers/subject experts who again gave constructive criticism of the instrument; and their suggestions were also included in the final instrument design.

The analysis of any data collected allows a researcher to explain the findings in a logical and understandable manner. Traditionally, approaches to data analysis have been either qualitative or quantitative. However, more researchers are now using a combined approach. Filstead (1979) offers a sound argument for using a combined method of data analysis:

As the writer's study was descriptive in its approach and collected qualitative information from the respondents in order to discover, describe and compare themes and most of the data was able to be quantified and analysed by a computer statistical package, this combined method of data analysis was the preferred and most logical one to use.

The preferred method for analysing data once they were quantified was by means of descriptive statistics. The use of descriptive statistics is seen by Rowntree (1981) as for summarising or describing a sample, rather than being concerned with generalising from the sample to make assumptions to a wider population. This approach was selected for a number of reasons. Firstly, it allowed the data to be organised into a comprehensible form enabling the writer to gain an overall feel for the data; Secondly, the data were able to be presented in graphical and tabular form, therefore enabling important features to be summarised; and thirdly, as the data was inherently qualitative, it allowed development into qualitative descriptions of general attitudes, themes and trends to unfold, thus enabling the writer to generate theory from the results. However, generalisations from the results could only be made to the sample targeted by the study; and inferences could not be made about a sample beyond the data collected (Oldham, 1993).

Therefore, data analysis includes a mixture of both quantitative and qualitative analysis and included the following:

The writer obtained permission to undertake the study from each respective ward manager. Permission was also obtained from the Hospital Ethics Committee.

Each nurse in the sample was informed, in writing, exactly why the study was being undertaken and assured of complete confidentiality. No name was required on the questionnaire. The completed questionnaires were kept in a locked cupboard whilst at the writer's place of work and were only read by the writer and his supervisor. All the data that were gathered were coded and entered onto a computer for purposes of statistical analysis and protected by a password known only to the writer.

Of the fifty-one questionnaires sent out to the sample a total of 60.78% (N=31) were returned completed. Background information was collected in order fully to describe respondents by nursing position, gender, length of time on present ward, dependency level of previous ward, and length of time in forensic psychiatry.

Figure 1 below shows the totals for each nursing position held by the respondents and is representative of the total positions held on the sampled wards. To maintain confidentiality the senior enrolled nurse was included with the enrolled nurses for purposes of description and analysis of the data.

From the total respondents 9.7% (n=3) were female and 90.3% (n=28) were male. It was unknown to the writer how the sample were distributed by gender prior to the study but this result is somewhat representative in a hospital such as Ashworth, as the majority of the nursing staff employed there are male.

Of the respondents 45.2% (n=14) had been on their current ward for over three years whilst the dependency level of their previous ward ranged from high 29% (n=9) to low to medium 6.5% (n=2); with one respondent answering night services area. Tables 1 and 2 below show the total frequencies for all of the responses for both of these questions.

Time on present ward

n (%)

Over 5 years

Between 3 - 5 years

Between 1 - 3 years

Up to 1 year

5 (16.1)

9 (29.0)

12 (38.7)

5 (16.1)

Total

31 (100.0)

Dependency of previous ward

n (%)

High

Medium to high

Medium

Low to medium

Low

Night services area

9 (29.0)

6 (19.4)

2 (6.5)

7 (22.6)

1 (3.2)

Total

31 (100.0)

The most encouraging result obtained from the background information was the length of time the respondents had been working in forensic psychiatry, although it was unknown if all this was whilst employed at Ashworth. The results are shown in Figure 2 below. The higher proportion of the respondents (67.7%) have worked within forensic psychiatry for longer than ten years and a further 29% for longer than five years. Only 3.2% (n=1) had worked for less than five years. Therefore, the writer proceeded on the assumption that the information obtained through the questionnaire was from nurses with many years of experience working in forensic psychiatry who have established their views on future patient dangerousness based on this lengthy experience.

Initially the questionnaire asked if a research-based instrument was used for assessing future patient dangerousness. All respondents answered no to this question. Following on from this a statement was made by the writer that "nurses can accurately assess future patient dangerousness". Respondents were asked to comment on this statement using a Likert scale. Responses were mixed but were skewed towards agreeing with the statement. The results are shown in Figure 3 below. Of the respondents, 71% (n=22) either agreed or strongly agreed that nurses could accurately assess future patient dangerousness; whilst 29% (n=9) either disagreed or were undecided that they could. A two-cell contingency test showed a significant result beyond a 0.02 level (Chi-square=5.45, d.f=1, two-tailed).

Figure 3: If a research-based instrument was used for assessing future patient dangerousness.

Next a statement was made by the writer that "nurses should be actively involved in assessing future patient dangerousness". Respondents were again asked to give their views on this statement by use of a Likert scale. Responses to this statement were extremely skewed to one end of the scale with 96.8% (n=30) either agreeing or strongly agreeing that nurses should be involved and only 3.2%(n=1) strongly disagreeing. The results are shown in Figure 4 below. A two-cell contingency test showed a significant result beyond a 0.001 level (Chi-square=27.3, d.f=1, two-tailed).

Figure 4: Nurses should be actively involved in assessing future patient dangerousness

The respondents were asked to identify from nineteen given factors, each of which could form part of a nursing assessment in forensic psychiatry, how important each was in influencing their views on future patient dangerousness. Each factor had been drawn from the literature reviewed as important in determining future violence or dangerousness. The results are shown in Table 3.

Table 3: Level of importance respondents placed on selected factors in influencing their views on future patient dangerousness

Factor	Literature source	Unimportant n (%)	Undecided n (%)	Important n (%)	Very important n (%)
Fear inducing behaviour	Hepworth 1982)	2 (6.5)	2 (6.5)	13 (41.9)	14 (45.2)
Forensic history	Quinsey & Maguire (1986)	0 (0)	0 (0)	16 (51.6)	15 (48.4)
Index offence and surrounding circumstances	Pollock & Webster (1990)	0 (0)	2 (6.5)	14 (45.2)	15 (48.4)
Impulsive behaviours	Hepworth (1982)	0 (0)	0 (0)	17 (54.8)	14 (45.2)
Mental health history	Pollock & Webster (1990)	0 (0)	1 (3.2)	15 (48.4)	5 (48.4)
Current mental health	Marra.et al. (1987)	0 (0)	0 (0)	13 (41.9)	18 (58.1)
Social history	Pollock & Webster (1990)	1 (3.2)	5 (16.1)	22 (71)	3 (9.7)
Sexual behaviours	Marra.et al. (1987)	0 (0)	3 (9.7)	18 (58.1)	10 (32.3)
Fantasies/expressed thoughts	Marra.et al. (1987)	0 (0)	6.5)	15 (48.4)	14 (45.2)
Ruminations around offence	Marra.et al. (1987)	1 (3.2)	6 (19.4)	15 (48.4)	9 (29)
Alcohol\substance abuse	Pollock & Webster (1990)	0 (0)	1 (3.2)	18 (58.1)	12 (38.7)
Specific and repeated patterns of behaviour	Pollock & Webster (1990)	0 (0)	0 (0)	14 (45.2)	17 (54.8)
All incidents in psychiatric history	Feinstein & Plutchik (1990)	3 (9.7)	3 (9.7)	20 (64.5)	5 (16.1)
Recent incidents	Quinsey & Maguire (1986)	0 (0)	3 (9.7)	17 (54.8)	11 (35.5)
Current behaviour	Hepworth(1982)	0 (0)	1 (3.2)	15 (48.4)	15 (48.4)
Current attitudes	Pollock & Webster (1990)	0 (0)	3 (9.7)	15 (48.4)	13 (41.9)
Underlying hostility	Pollock & Webster (1990)	0 (0)	1 (3.2)	13 (41.9)	17 (54.8)
Beliefs about rules and regulations	Pollock & Webster (1990)	1 (3.2)	5 (16.1)	17 (54.8)	8 (25.8)
Motivation to change	Pollock & Webster (1990)	0 (0)	0 (0)	16 (51.6)	15 (48.4)

The first noticeable point is that none of the respondents viewed any of the areas as being very unimportant, although some felt that social history, ruminations around offence, and beliefs about rules and regulations (3.2%), fear inducing behaviour (6.5%), and all incidents in psychiatric history (9.7%) were unimportant. All areas were viewed as important or very important by some of the respondents - very important ranges from current mental health (n=18) to social history (n=3), and important ranges from social history (n=22) to fear inducing behaviour, current mental health and underlying hostility (n=13). All respondents viewed forensic history, impulsive behaviours, current mental health, specific and repeated patterns of behaviour, and motivation to change as either important or very important. Some respondents were undecided on index offence and surrounding circumstances, mental health history, sexual behaviours, fantasies/expressed thoughts, alcohol/substance abuse, recent incidents, current behaviour, current attitudes, and underlying hostility but none viewed them as unimportant.

A number of further areas were identified by the respondents as important in influencing their views on future patient dangerousness. Those that were identified as very important were egocentricity, insight into their actions, support family/others, insight into illness, recidivism, and insight regarding medication. Those that were identified as important were age group, history of employment, behaviour towards females/children, level of patients intelligence, what effects prescribed drugs have had, do any original symptoms remain, family background, and how the patient responds to stressful situations.

Respondents were asked how important they viewed discussion with various disciplines working within the hospital. The responses are outlined in Table 4. All respondents viewed their own nursing team and medical staff as either important or very important. A higher number of the respondents viewed all the disciplines as very important or important, range n=31 to n=20 except staff education department where more viewed this as either undecided, unimportant or very unimportant (n=17). A considerable number of the respondents were undecided in their views on staff education department (n=10), social workers (n=8), and psychology staff (n=5).

Table 4: The importance of other disciplines in influencing respondents' views on future patient dangerousness

Discipline	Very unimportant n (%)	Unimportant n (%)	Undecided n (%)	Important n (%)	Very important n (%)
Own nursing team	0 (0)	0 (0)	0 (0)	2 (6.5)	29 (93.5)
Other nursing teams	0 (0)	0(0)	3(9.7)	18 (58.1)	10 (32.3)
Medical staff	0 (0)	0 (0)	0 (0)	8 (25.8)	23 (74.2)
Psychology staff	0 (0)	1 (3.2)	5 (16.1)	8 (25.8)	17 (54.8)
Social worker	0 (0)	1 (3.2)	8 (25.8)	10 (32.3)	12 (38.7)
Patients education department	1 (3.2)	5 (16.1)	5 (16.1)	16 (51.6)	4 (12.9)
Occupations staff	0 (0)	2 (6.5)	5 (16.1)	17 (54.8)	7 (22.6)
Staff education department	3 (9.7)	4 (12.9)	10 (32.3)	9 (29)	5 (16.1)

The respondents also identified significant others as important in discussing their views with. Those identified as very important were domestic, family, family/friends, individual patients and direct family or guardians; and those identified as important were staff from admission source, patients' self portrayal and patients' visitors.

The respondents were asked to identify how they tested their views on future patient dangerousness. The responses are shown in Table 5 below. All respondents reported that they tested their views out with medical staff, with a two-cell contingency test showing a significance level beyond 0.001 (Chi-square=31, d.f=1, two-tailed). Some respondents reported that they tested their views out with all the given areas/disciplines. All except during clinical supervision, by using research and with social workers, showing a significance level beyond 0.001 from a two-cell contingency test; at handover (Chi-square =17.06, d.f=1, two-tailed); with associate nurses (Chi-square =17.06, d.f=1, two-tailed); with other nurses (Chi-square =18.61, d.f=1, two-tailed); with psychology staff (Chi-square =11.65, d.f=1, two-tailed). Not all respondents reported that they would use the individual care plan to test out their views (yes n=26, no n=5), a two-cell contingency test showed a significance level beyond 0.001 (Chi-square =14.23, d.f=1, two-tailed).

Table 5: How respondents test their formulated views on future patient dangerousness.

Test information out	Yes n (%)	No n (%)	Total n (%)
At handover	27 (87.1)	4 (12.9)	31(100)
During clinical supervision	17 (54.8)	14(45.2)	31(100)
With associate nurses	27 (87.1)	4(12.9)	31(100)
With other nurses	28 (90.3)	3 (9.7)	31(100)
With medical staff	31 (100)	0 (0)	31(100)
With psychology staff	25 (80.6)	6 (19.4)	31(100)
With social workers	21 (67.7)	10(32.3)	31(100)
Using care plan	26 (83.9)	5 (16.1)	31(100)
Using research	10 (32.3)	21(67.7)	31(100

The other identified ways in which the respondents would test their views involved the patients' mental state, PCTM (Patient care team meeting), rehabilitation excursions and the way the patient's personality was viewed in general.

Much of the debate surrounding future patient dangerousness seems to revolve around whether or not we can accurately assess it; and that for every study which finds predictive accuracy, there is one that finds it no better than chance (Monahan, 1988). The study results suggest that the respondents quite significantly agreed that "nurses could accurately assess future patient dangerousness", contrary to this Sepjak. et al. (1984) found that nurses scored low in their predictive accuracy of future patient dangerousness. Why did the nurses feel so strongly that they could predict the future dangerousness of their patients? One explanation for these views could be that the nurses, having spent many years working in forensic psychiatry, have seen many patients returned to Ashworth, having re-offended, when they were of the opinion developed from their own assessment process that they were not ready for conditions of lesser security or discharge.

Although the study did not intend to show how accurately nurses can assess patient dangerousness, it did try to establish their views; the processes they use to determine these; and the way they evaluate them. As far as the writer is aware this is the first study to try to establish this, with most of the related research involving either psychologists, psychiatrists or social workers in its sampling.

The writer previously put forward the theory that perceived dangerousness has a central role to play in the assessment process for nurses working in the forensic setting. The writer believes that nurses have a central role to play in the overall multi-disciplinary assessment of future patient dangerousness. The results of the study show that the nurses believed that "nurses should be actively involved in assessing future patient dangerousness". Surprisingly, none of them were in fact using any research-based instrument for assessing future dangerousness; which suggests either that they don't take the assessment of future patient dangerousness seriously; or, more positively, that they are not aware of any instruments which they could use for this purpose.

If the nurses felt that they could accurately assess future patient dangerousness, and should be actively involved in the assessment process, how exactly were they doing this? The writer previously put forward the theory that nurses formulate their views on a factual basis and not from what is often called "a gut feeling". Are the nurses using this "gut feeling" or are they using some other processes? The writer put forward nineteen factors to the nurses that had been taken from related research and tried to establish how important each was to the nurses in influencing their views on future patient dangerousness. Each of these areas had been shown by the related research to be a high indicator of future patient dangerousness.

Probably the first area put forward (fear inducing behaviour) is the closest one that can be linked to "gut feeling". This is valued highly by many of the nurses. To the writer this is the starting point; for if an experienced nurse working with a patient feels fearful at times in his/her presence then it is the nurses' duty to try to establish where this fear is coming from. Further assessment using the nursing process is the next logical step forward.

The starting point for this quite often is through the history of the patient to establish if this is 'normal' behaviour for the patient. Previous research highlighted in the literature review suggests that forensic history, social history, mental health history, alcohol/substance abuse, all incidents in psychiatric history, and index offence and the surrounding circumstances are all areas of the patient's past history which have been shown to be indicators of future patient dangerousness.

A high value was placed on all these areas by the nurses through how important they felt each was in influencing their views on future patient dangerousness. Generally, forensic history, index offence and the surrounding circumstances, and mental health history were valued more highly.

This part of the nursing assessment could be viewed as a basis from which to interpret the patients' more recent or current behaviours. A number of factors were put forward to the nurses which could be included in this group. These were impulsive behaviours, sexual behaviours, specific and repeated patterns of behaviour, recent incidents and overall current behaviour. Although sexual behaviours could also be a significant part of the past history of the patient. All areas were valued as important by the nurses in influencing their views on future patient dangerousness, with all the areas valued highly by many of them.

Interlinked with all the previous twelve factors are the factors which could be viewed as attitudinal and part of the current health of the patient. These include current mental health, fantasies/expressed thoughts, ruminations around the offence, underlying hostility, beliefs about rules and regulations, overall current attitude, and motivation to change. On the whole the nurses valued all these factors highly with current mental health and underlying hostility rating highest.

To the writer the fact that the nurses are using the factors suggested to influence their views indicates that, although a "gut feeling" may create a doubt in their mind, they then go on to clarify this further by use of some of the factors previously identified, be it through an un-researched assessment format or their own thought processes. However, the results suggest that this is not where the process ends for the nurses.

It appears from the results that the nurses, when formulating their views on future patient dangerousness, place great value in discussing these views with members of the multi-disciplinary team. They value discussion with their own nursing team highest; with the medical staff and social workers rated highly. Psychologists were not rated quite so highly but this could well be because there is a certain lack of psychologists at Ashworth at present. In the writer's experience, with so many patients in need of their time they are not regular visitors to the mental illness wards.

When it comes to testing these views out once they have been formulated the nurses report that they use a number of the disciplines and methods available to them. Again other nurses rated highly with results suggesting that this may take place at the handover period with associate nurses. All the nurses reported that they would test their views out with the medical staff, which was what the writer expected them to report. However, surprisingly, not all of the nurses would address their formulated views on future patient dangerousness through the individual care plan. This result was somewhat troubling to the writer as surely all views a nurse formulates on a patient, be it in relation to future patient dangerousness or any other factor, should be tested out using the patients' care plan and consequently evaluated at a later date. After all, future patient dangerousness is central to the nursing assessment of patients detained in a high security hospital and deemed a danger to themselves or others. Clinical supervision was also utilised by many of the nurses to test their formulated views out; which could suggest that they hold some value in being supervised in their practice in such a difficult and challenging field of nursing.

The results of the study need to be treated with some caution and cannot be generalised to all nurses who work within forensic psychiatry. The sample was chosen purposively from a predetermined target population and the views can, therefore, not be treated as representative of nurses working at Ashworth as a whole. The sample was small and drawn from one level of patient dependency, namely low, where the patients are generally more settled and co-operative with their treatment plans. Different results may have been obtained from nurses working in higher dependency units in the hospital where the dangerousness of an individual may be more obvious.

However, the background information showed that the respondents were from all the grades of qualified nursing staff working on the low dependency wards at Ashworth and that a large proportion of them had been working in this dependency level for over twelve months. It may be assumed that this is time enough to gain some sort of experience working with this particular patient group. The more encouraging information obtained was that all but one of the respondents had worked within forensic psychiatry for over five years, with substantially more of these employed for over ten years. Therefore, it could be assumed that the information obtained from the study could be viewed as coming from nurses experienced in the field of forensic psychiatry.

To conclude, it appears that nurses who are currently working within the low dependency wards on the mental illness wards at Ashworth hospital feel that they have a major role to play in multi-disciplinary assessment of future patient dangerousness. They are utilising many of the areas that other professions are using in their assessment of patient dangerousness, the professions who may be called as expert witnesses to assist in process of assessing the dangerousness of an individual. Although they are not utilising any research based instrument to formulate their views, they do utilise all the available resources and the experience of other disciplines to assist them in the process.

A number of issues have emerged from the study which have implications for future research.

Barker, P J. (1985). Patient Assessment in Psychiatric Nursing, pp 2-18, 360. Croom Helm, London.

Cooper, R P. and Werner, P D. (1990). Predicting Violence in Newly Admitted Inmates: A Lens Model Analysis of Staff Decision Making. Criminal Justice and Behaviour, 17 (4), pp 431 - 447.

Feinstein, R. and Plutchik, R. (1990). Violence and Suicide Risk Assessment in the Psychiatric Emergency Room. Comprehensive Psychiatry, 31 (4), pp 337 - 343.

Filstead, W J. (1979). Qualitative methods - a needed perspective in evaluation research. In: Cook, T D and Rechardt, C S. (eds). Qualitative and Quantitative Methods in Evaluation Research. Sage, London pp 33-48.

Gunn, J. (1982). Defining the Terms. In: Hamilton, J R. and Freeman, H. (Eds). Dangerousness: Psychiatric Assessment and Management. Gaskell.

Hamilton, J R. (1982). A Quick Look at the Problems. In: Hamilton, J R. and Freeman,H.(Eds). Dangerousness: Psychiatric Assessment and Management, pp 2. Gaskell.

Hepworth, D. (1982). The Influence of the Concept of 'Danger' on the Assessment 'Danger to Self and Others'. Med.Sci.Law,22(4),pp 245-254.

Jackson,M W.(1986). Psychiatric Decision-Making for the Courts: Judges, Psychiatrists, Lay People? International Journal of Law and Psychiatry, 9,pp507-520.

Marra, H A., Konzelman. and Giles, P G. (1987). A Clinical Strategy to the Assessment of Dangerousness. International Journal of Offender Therapy and Comparative Criminology, 31 (3), pp 291-299.

Menzies,R J., Webster,C D. and Butler,B T.(1981). Perceptions of Dangerousness Among Forensic Psychiatrists. Comprehensive Psychiatry, 22(4),pp387-396.

Monahan, J. (1988). Risk Assessment of Violence Among the Mentally Disordered: Generating Useful Knowledge. International Journal of Law and Psychiatry, 11, pp 249 - 257.

Oldham, J. (1993). Statistical Tests (Part 1): descriptive statistics. Nursing Standard, 7 (43), pp 30-35.

Owen, D. (1991). Foucault, psychiatry and the spectre of dangerousness. Journal of Forensic Psychiatry, 2 (3), pp 238-241.

Polit, D F. and Hungler, B P. (1993). Essentials of Nursing Research: Methods, Appraisal, and Utilization. 3rd ed. J B Lippincott Company, Philadelphia pp 244 - 255.

Pollock,N. and Webster,C.(1990). The clinical assessment of dangerousness. In: Bluglass,R. and Bowden,P.(Eds). Principles and Practice of Forensic Psychiatry. Churchill Livingstone Section VII.3 pp489-497.

Quinsey,V L.and Maguire,A. (1986). Maximum Security Psychiatric Patients: Acturial and Clinical Prediction of Dangerousness. Journal of Interpersonal Violence, 1(2),pp143-171.

Reid, N. (1993). Health Care Research by Degrees. Blackwell Scientific Publications, Oxford pp 32 - 35.

Rowntree, D. (1981). Statistics without Tears: A Primer for Non-mathematicians. Penguin Books, London pp 19,21.

Segal,S P., Watson,M A., Goldfinger,S M. and Averbuck,D S. (1988). Cival Commitment in the Psychiatric Emergency Room II. Mental Disorder Indicators and Three Dangerousness Criteria. Archives of General Psychiatry, 45(8),pp753-758.

Sepejak, D S., Webster, D C. and Menzies, R J. (1984). The Clinical Prediction of Dangerousness: Getting Beyond the Basic Questions. In: Muller, D J., Blackman, D E. and Chapman, A J. (Eds). Psychology and Law, pp 113-123. John Wiley & Sons Ltd.

Werner,P D., Rose,T L., Yesavage,J A. and Seeman,K.(1984). Psychiatrists' Judgments of Dangerousness in Patients on an Acute Care Unit. American Journal of Psychiatry, 141(2),pp263-266.

Werner, P D. and Meloy, J R. (1992). Decision making about dangerousness in releasing patients from long-term psychiatric hospitalization. Journal of Psychiatry and Law,20(1), pp 35-47.

Lord Allen .(1982). Dangerousness and Public Policy. In: Hamilton, J R. and Freeman, H. (Eds). (1982). Dangerousness: Psychiatric Assessment and Management. Gaskell. Chapter 7, pp 42 - 45.

Beck,J C., White,K A. and Gage,B. (1991). Emergency Psychiatric Assessment of Violence. American Journal of Psychiatry, 148(11),pp1562-1565.

Binder,R L. and McNeil,D E.(1988). Effects of Diagnosis and Context on Dangerousness. American Journal of Psychiatry, 145(6),pp728-732.

Bottoms, A E. (1982). Selected Issues in the Dangerousness Debate. In: Hamilton, J R. and Freeman, H. (Eds). (1982). Dangerousness: Psychiatric Assessment and Management. Gaskell. Chapter 5, pp 29 - 37.

Cirincione, C., Steadman, H J., Robbins, P C. and Monahan, J.(1992). Schizophrenia as a Contingent Risk Factor for Criminal Violence. International Journal of Law and Psychiatry, 15, pp 347 - 358.

Gottlieb, P. and Gabrielsen, G. (1990). The Future of Homicide Offenders: Results from a homicide project in Copenhagen. International Journal of Law and Psychiatry, 13, pp 191 - 205.

Halleck, N H. and Petrilla, J. (1988). Risk Management in Forensic Services. International Journal of Law and Psychiatry,11, pp 347 - 358.

Harding, T. and Montandon, C. (1982). Does Dangerousness Travel Well? A cross-national perspective on medico-legal applications. In: Hamilton, J R. and Freeman, H. (Eds). (1982). Dangerousness: Psychiatric Assessment and Management. Gaskell. Chapter 8, pp 46 - 52.

Hicks, C M. (1990). Research and Statistics: A practical introduction for nurses. Prentice Hall, London, p 73.

Hoge,S K., Sachs,G., Appelbaum,P S., Greer,A. and Gordon,C. (1988). Limitations on Psychiatrists' Discentionary Cival Commitment Authority by the Stone and Dangerousness Criteria. Archives of General Psychiatry, 45, pp764-769.

Jackson,M A.(1989). The Clinical Assessment and Prediction of Violent Behaviour: Toward a Scientific Analysis. Criminal Justice and Behaviour,16(1),pp114-131.

Krakowski, M., Volavka, J. and Brizer, D. (1986). Psychopathology and Violence: A Review of Literature. Comprehensive Psychiatry, 27 (2), pp 131 - 148.

Koenraadt, F. (1992). The Individualizing Function of Forensic Multidisciplinary Assessment in a Dutch Residential Setting: The Pieter Baan Centre Experience. International Journal of Law and Psychiatry,15, pp 195 - 203.

Lidz,C W.,Mulvey,E P.,Apperson,L J.,Evanczuk,K.andShea,S. (1992). Sources of Disagreement Among Clinicians' Assessments of Dangerousness in a Psychiatric Emergency Room. International Journal of Law and Psychiatry, 15(3),pp237-250.

Lipsedge, M. (1994). Dangerous stereotypes. Journal of Forensic Psychiatry,5 (1), pp 14 - 19.

Lovell, A M. and Scheper-Hughes, N.(1986). Deinstitutionalization and Psychiatric Expertise: Reflections on Dangerousness, Deviancy, and madness. International Journal of Law and Psychiatry, 9,pp 361 - 381.

MacCulloch,M.(1982. The Health Department's Management of Special Hospital Patients. In: Hamilton, J R. and Freeman, H. (Eds). (1982). Dangerousness: Psychiatric Assessment and Management. Gaskell. Chapter 17, pp 101 - 105.

MacCulloch, M.,Bailey J.,Jones, C. and Hunter, C.(1994). Nineteen male serious reoffenders who were discharged from a special hospital: III. illustrated administrative issues. Journal of Forensic Psychiatry, 5 (1),pp 63-81.

Mason, S A. (1993). Employing quantitative and qualitative methods in one study. British Journal of Nursing, 2 (17), pp 869 - 872.

McNiel,D E.and Binder,R L.(1987). Predictive Validity of Judgements of Dangerousness in Emergency Civil Commitment. American Journal of Psychiatry, 144(2),pp197-200.

McNiel,D E., Binder,R L. and Greenfield,T K. (1988). Predictors of Violence in Civally Committed Acute Psychiatric Patients. American Journal of Psychiatry, 145(8),pp965-970.

Mills, M J. (1988). Cival Commitment: The Relationship Between Perceived Dangerousness and Mental Illness. Archives General Psychiatry, 5, pp 770 - 772.

Monahan,J. (1982). The Prediction of Violent Behaviour: Toward a Second Generation of Theory and Policy. American Journal of Psychiatry, 141(1), pp 10 - 15.

Monahan,J. (1993). Limiting Therapist Exposure to Tarasoff Liability. American Psychologist, 48 (3), pp 242 - 250.

Ortet-Fabregat, G., Perez, J. and Lewis, R. (1993). Measuring Attitudes Towards Prisoners: A Psychometric Assessment. Criminal Justice and Behaviour, 20 (2), pp 190 - 198.

Pollock, N L.(1990). Accounting for Predictions of Dangerousness. International Journal of Law and Psychiatry,13, pp 207 - 215.

Prins,H.(1990). Dangerousness:a review. In:Bluglass,R.and Bowden, P.(Eds). Principles and Practice of Forensic Psychiatry. Churchill Livingstone Section VII.4 pp499-505.

Quinsey, V L. and Cyr, M. (1981). Perceived Dangerousness and Treatability of Offenders. The Effects of Internal Versus External Attributions of Crime Causality. Journal of Interpersonal Violence, 1(4),pp458-471.

Rofman,E S., Askinazi,C. and Fant,E.(1980). The Prediction of Dangerous Behaviour in Emergency Civil Commitment. American Journal of Psychiatry, 137(9),pp1061-1064.

Rossi, A., Jacobs, M., Monteleone, M., Olsen, R., Surber, R W., Winkler, E L. and Wommack, A. (1986). Characteristics of Psychiatric Patients Who Engage in Assaultive or Other Fear-Inducing Behaviours. The Journal of Nervous and Mental Disease, 174, (3), pp 154 - 160.

Segal,S P., Watson,M A., Goldfinger,S M. and Averbuck,D S. (1988). Cival Commitment in the Psychiatric Emergency Room III. Disposition as a Function of Mental Disorder and Dangerousness Indicators. Archives of General Psychiatry, 45(8),pp759-763.

Segal,S P., Watson,M A., Goldfinger,S M. and Averbuck,D S. (1988). Cival Commitment in the Psychiatric Emergency Room I. The Assessment of Dangerousness by Emergency Room Clinicians. Archives of General Psychiatry, 45(8),pp748-752.

Segal, S P., Watson, M A. and Nelson, S. (1987). Consistency in the application of cival commitment standards in psychiatric emergency rooms. The Journal of Psychiatry and Law, pp 125 -148.

Siegel, S and Castellan, N J, Jr. (1988). Nonparametric statistics for the Behavioural Sciences. 2nd ed. McGraw-Hall Book Co, Singapore. pp 33-36.

Strachan,J G.(1982). Psychiatric Assessment of the Dangerous Offender in the Netherlands. Med.Sci.Law, 22(1),pp16-20.

Tidmarsh, D. (1982). Implications from Research Studies. In: Hamilton, J R. and Freeman, H. (Eds). (1982). Dangerousness: Psychiatric Assessment and Management. Gaskell. Chapter 3, pp 12 - 20.