Mary Fox, research facilitator ^a

Grace Sweeney, research fellow ^b

Clifford Howells, medical director ^c

Jonathan Stead, research fellow ^b

a Mid Devon Research Group, Wyndam House Surgery, Silverton, Devon EX5 4HZ, ^b Exeter and North Devon Research and Development Support Unit, Noy Scott House, Haldon View Terrace, Exeter EX2 5EQ, ^c Prison Health Task Force, Wellington House, 133-135 Waterloo Rd, London SE1 8UG.

Correspondence to: Mary Fox M.Fox@exeter.ac.uk

Prison health care is undergoing a renaissance. Prior to the introduction of clinical governance a pilot scheme was set up to trial significant event audit, which is a key tool of clinical governance, in prison health care. This study sought to evaluate the pilot scheme. The design was qualitative, using non-participant observation of significant event audit meetings analysed by content analysis together with initial and follow-up semi-structured interviews analysed using grounded theory. The location of the study was three prison health care units, two in England and one in Wales. The

Participants were 78 members of the health care teams who were present in the 18 meetings observed; 14 of these were purposefully sampled for interview. The

results showed that a variety of staff attended and contributed to the meetings which produced many theoretical solutions to problems but fewer actions. The interviewees showed a dissimilarity in understanding and attitudes to SEA, ranging from the cynical to the enthusiastic. There was a pervasive belief that, whatever the merits of significant event audit, the prison system was too entrenched in its ways for significant event audit to effect major changes, but that relatively minor ‘in house’ reforms had been made. It was concluded significant event audit has the potential to aid the development of clinical governance in prison health care. A number of recommendations were made to help ensure its successful delivery.

Prison health care is undergoing a period of change.¹ Historically health care has been under the auspices of the Home Office, the service largely being delivered by professionally isolated GPs and health care officers with minimal clinical training.¹ The introduction of NHS trained nurses in the 1980s heralded the future with further steps currently being taken to clarify their role and to recognise and better equip the health care officers.²

The government is committed to "ensuring that prisoners have access to the same quality and range of health services as the general public receives from the NHS".¹ This challenging aim is being worked out through a formal partnership between the prison service and the NHS.³ As part of the reforms, prison health care is embracing the concept of clinical governance.^{3 4 5} Central to this ethos of continuous improvement is the process of reviewing practice and the events which occur within a particular health care organisation.

One means of doing this is through significant event audit (SEA). This occurs when, "Individual cases in which there has been a significant occurrence (not necessarily involving an undesirable outcome for the patient) are analysed in a systematic and detailed way to ascertain what can be learnt about the overall quality of care and to indicate changes that might lead to future improvements". ⁶ There is evidence to show that SEA can be an effective tool in primary care. In this setting it has been shown to contribute to team building, enhance communication and improve patient care; it is making a major contribution to the development of clinical governance. ^{7 8 9}

There has been no previous research on SEA in the prison health care unit setting. Therefore, by observing SEA meetings and evaluating prison health care staffs’ attitudes to and understanding of SEA, this study sought to find out:

SEA was piloted in three prison health care units, two in England and one in Wales. These prisons were selected as being broadly similar local prisons, housing a mixture of remanded and convicted men, where it might be expected to find comparable situations. The entire population of 89 members of staff of the health care units were invited to attend the SEA meetings. Of these 78 members were present over the 18 meetings observed. Those present were representative of all the staff groups involved in the health care units; they included 37 nurses, 15 health care officers, 9 doctors, 5 pharmacists, 3 administrators, 3 occupational therapists, 1 community psychiatric nurse, 1 member of the board of visitors, 1 chaplain and 1 governor. From those attending the initial SEA meetings, a purposeful sample was taken for interview from each prison, to reflect the maximum variation of professions involved so that their differing, possibly conflicting, perspectives might be represented in the data.⁹ The interviewees included four nurses, three health care officers, two doctors, two pharmacists, two occupational therapists and one community psychiatric nurse.

Six SEA meetings in each prison were observed, in a non-participatory way, to determine the nature of SEA in situ. A pro-forma documenting issues, process, solutions and actions was used to make detailed records. The data from these meetings underwent a content analysis to show the profiles of attendance, presenters and contents of the meetings.

Fourteen health care unit staff took part in semi-structured interviews after the first observation and again after the sixth. The interview guide was validated by prior use within primary care.⁸ The invitation to participate was by a letter which explained the study in some detail and gave potential participants assurance about the confidentiality of the conversations.

The participants were initially interviewed in the health care units in private locations as soon after the first SEA meetings as possible. The interviews were tape recorded and notes were taken. Participants were interviewed again after the sixth SEA observation, to see if their understanding and opinions of SEA had changed.

The reliability and validity of the data were ensured in a number of ways. Two methods of data collection, observation and interview, were used to provide method triangulation.¹⁰ Data were triangulated by collection occurring on three separate but comparable sites and by the use of multiple informants. ¹¹ Following preliminary analysis, the findings were presented back to each health care unit in the form of a negotiated feedback meeting where the participants were invited to comment on the findings.¹² The outcomes of these meetings were fed into the final analysis. Reliability was provided by meticulous collating of emerging themes and theories with the data.

The use of content analysis¹³ to analyse the observational data enabled a clear picture of the content and process of SEA in the health care units to be formed. Issues were coded according to type and solutions were categorised, according to Pringle’s earlier work.⁶

The verbatim transcripts from the interviews were subject to a qualitative, grounded theory analysis using Atlas.ti software (T Muhr, Berlin, Germany, 1994), conducted by MF. This involved a systematic, iterative process of coding and refining concepts, by constant comparison, until higher order categories emerge which collectively aid abstraction of a theory, rich and dense, that encompasses all the data.¹⁴ Co-authors were not involved in this process because it was believed that only the interviewer was close enough to this particular context to ensure correct interpretation.¹⁵

So what has been the impact of SEA and what is its future in prison health care? The results combine the data from all three prisons.

1. Observations

The observation of the 18 SEA meetings showed that 78 members of staff from variety of professions attended the meetings, with a mean attendance of 10.5 and a range of 6-16 (table 1). They discussed a range of issues (table 2); the outcomes were classified, 111 solutions were reached in the meetings but only 72 actions taken (table 3).

There is evidence from the observations of the meetings that SEA can and does bring about change, that it can be multidisciplinary and encompass a wide range of issues. It is noteworthy that along with clinical issues, items discussed most commonly concerned systems and communication issues, possibly highlighting the need for reform.

2. Interviews

The interview analysis revealed the story behind the figures presented in tables 1-3. They combine the results of the initial and follow-up interviews with the negotiated feedback sessions.

Opinions about SEA ranged from the enthusiastic to the cynical. Some reported the pleasure of seeing positive change take place and that SEA made them feel valued. However, most interviewees felt that SEA was a good idea that could not work in HCUs.

Interviewees thought that other staff did not understand what SEA was about and that in some prisons SEA had lost its format. It was perceived as simply as a safe place for a structured discussion, while this was valued, it lacked the dynamic of SEA. The perception remained that it was a management tool. Although solutions to problems were being found they were often not implemented and SEA had generally not made an impact on the day-to-day life of the staff. The ability to discuss matters with colleagues was valued although it was not generally felt that this had led to team-building, any team-spirit generated in the meetings was not being carried through in the units. As these misunderstandings may be partly the result of individuals missing the introductory sessions due to shift working, a more thorough introduction to SEA may be needed which accounts for this problem.

Where changes were acknowledged it was generally believed that these would not be sustained and could only have impact on health-care issues, so that any problems

that overlapped with the rest of the prison were insoluble by the team. This was because decision making was perceived to occur elsewhere in the prison ie by governors. Feedback of the outcomes of the meetings was generally poor and appeared to rely on minutes, which most people did not read.

Leadership was an issue, it was felt some training would be helpful and that having non-managers as leaders would make SEA more acceptable. The term ‘significant event’ was misleading to some who interpreted it as meaning major event eg. Suicide.

Another concern was the poor attendance at the meetings. Various ideas were put forward to improve this, from making attendance compulsory to providing a lunch or paying people to attend. In prisons where SEA was held on training afternoons this was not thought to be helpful as there were too many conflicting demands on staff time, it was strongly felt that SEA needed dedicated time. The people who attended the meetings tended to reflect the divisions within the staff, eg mainly ‘day care’ or managers, so SEA was to some extent seen by health care unit staff as a meeting for ‘us or them’.

The follow-up interviews revealed a small positive shift in attitudes. There was a group of staff in each prison who valued and understood the potential of SEA. However, for the majority it seems that it was just another meeting.

The overall impact of SEA has been to demonstrate that it has the potential to bring about change in prison health care. It has been shown to be an effective tool, however the challenge of whether it will be well used lie in the attitudes and understanding of health care staff. A comprehensive introduction to SEA is therefore vital. There was a feeling overall that a lot of change was going in prison health care and that SEA was part of an evolving system. SEA needs to be presented as a tool for managing change as well as monitoring events. It can enhance staff development and be a supportive, professional, multidisciplinary tool, which as part of clinical governance, can help to enable change in a positive way.⁷ If SEA is to succeed it requires acceptance throughout the prison hierarchy and to be introduced to the health care staff within the broader context of clinical governance.

From the above it can be seen that there are a number of avenues for developing the potential of SEA in prison health care. For SEA to be effective in health care it needs to be accepted as a vehicle for change throughout the prison, so that the many issues that overlap the rest of the system can be addressed. This requires a more comprehensive and intensive introduction process.

A more thorough introduction of SEA in prison health care units is needed to ensure that all the staff have a clear understanding of the purpose and power of SEA, including its’ place in the changing nature of prison health care with reference to clinical governance. This introduction should take into account the shift system, the degree of cultural change that is being suggested and the assumptions that SEA is a management tool and not about empowerment for those at the coal face.

The selection of the right people to chair the meetings is essential. SEA almost stands or falls on the quality of leadership. The chairperson needs to be committed to the process, accepted by the group and an effective listener and enabler of others. They should have an ability to stick to an agenda that is solution and action centred. Some felt that having management lead was a problem and that the meetings would be better supported if lead by the troops. The role of leader can be an isolating one. It is suggested therefore that a core team, representative of the staff groups is trained in each prison. This would be mutually supportive and allow the chairing to revolve. There may be times when internal support is not enough. It is therefore recommended that the leaders receive specific support, including small group facilitation. They will also need ongoing support about SEA: a help-line could be set up to give encouragement, expert advise and have a trouble shooting role.

There are further issues about the logistics of the meetings. It appears that they often conflict with other events. For SEA to be effective it should be given greater priority and dedicated time set apart for it, with a high expectation of attendance while acknowledging that normal business has to carry on, with clear links to continuing professional development.

It can be seen that careful management of the outcomes of SEA is required, so that all the team know of solutions generated and successes applauded. Simply relying on minutes to do this job is not realistic. Means need to be found to ensure that non-attenders hear of the outcomes of the meetings. It is important that solutions found are revisited at subsequent meetings to ensure their implementation and break the cycle of returning to the same issues.

The term ‘significant event’ has perhaps a different meaning in a prison to the NHS. Some have interpreted it to mean major events eg suicides and assaults. The term ‘audit’ is also problematical, with associations of accountability. It is suggested that the name be changed to Health Care Open Meeting.

The personal benefits of SEA need to be highlighted including: being heard, effecting change, increasing understanding of others, greater team-spirit and the breaking down of traditional barriers. SEA should be linked to personal learning plans for all staff.

It is self evident that the presence of a researcher will have had some influence on data collection. Some contributions may have been withheld or had a particular spin put on them. However, we believe that breadth of the study allowed for the overall picture obtained to be accurate.

This study opens up a number of avenues for future research, relating to the development of clinical governance in prison health care and the wider impact this may have on prison culture.

We would like to acknowledge and thank the members of staff in the health care units involved. Without their help and support this research could not have been carried out. Especial thanks are due to those who gave up their time to be interviewed.

Funding: this study was funded by H M Prison Service.

Conflict of interest: None.

Contributors: MF conducted the interviews and observations, and carried out the analysis and interpretation of the data. GS, CH and JS were involved in the conception and design of the study and critically revising the article. MF is the guarantor of the study.

1. Her Majesty’s Chief Inspector of Prisons. Patient or Prisoner? A new strategy

for health care in prisons. London: Home Office, 1996.

Nursing in Prisons. Summary and Key Recommendations. London: Department

of Health, 2000.

3. Joint Prison Service and National Health Service Executive Working Group. The

future organisation of prison health care. London: Department of Health, 1999.

4. Scally G, Donaldson L J. Clinical Governance and the drive for quality

improvement in the new NHS in England. BMJ 1998; 314:61-65.

5. Prison Health Policy Unit and Task Force. Prison Health Handbook. London:

Department of Health, 2000.

6. Pringle M, Bradley CP, Carmichael CM, Wallis H, Moore A. Significant Event

Auditing. Occasional Paper 70. Exeter. RCGP, 1995.

8. Westcott R, Sweeney G, Stead J. Significant Event Audit in practice; a

preliminary study. Family Practice, 2000; 17: 173-179.

9. Sweeney G, Westcott R, Stead J. The benefits of significant event audit in

primary care: a case study. Journal of Clinical Governance, 2000; 8: 128-134.

10. Mays N, Pope C. Qualitative research: rigour and qualitative research. BMJ

1995; 311:109-112.

11. Banister B, Burman E, Parker I, Taylor M, Tindall C. Qualitative methods in

psychology. Buckingham: Open University Press, 1994.

12 . McKeganey NP, Bloor MJ. On the retrieval of sociological descriptions:

respondent validation and the critical case for ethnomethodology. International

Journal of Sociology and Social Policy. 1981;1:58-59.

13. Henwood KL. Qualitative inquiry perspectives, methods and psychology. In:

Richardson JTE, ed. Handbook of qualitative research methods. Leicester: BPS

books,1996.

14. Glaser B, Strauss A. The discovery of grounded theory: strategies for qualitative

research. New York: Aldine, 1967.

15. Morse JM. ‘Perfectly healthy but dead’: the myth of inter-rater reliability.

Qualitative Health Research 1997; 7:445-447.