Abstracts of Papers/Posters presented at the one day scientific meeting in Edinburgh 25th June 2003

1. HOW DO GPs USE SYRINGE DRIVERS?

Dr. S. Barclay, Department of Public Health and Primary Care, University of Cambridge
Prof. C. Todd, School of Nursing, University of Manchester
Prof. I. Finlay, University of Wales School of Medicine, Cardiff
Dr. G. Grande, School of Nursing, University of Manchester
P. Wyatt, Velindre Hospital, Cardiff.

Research Aims:
Studies continue to reveal problems with symptom control at home at the end of life. This study aims:

1) To investigate the knowledge of syringe drivers (SDs) among GPs in Wales.

2) To develop a summative scale of GP SD knowledge.

3) To investigate the prediction of low scoring GPs on this scale.

Methods and sampling:
A questionnaire was mailed to a stratified random sample of 600 Wales GPs. To develop the summative scale, a panel of 15 Macmillan GPs and 10 Consultants in Palliative Medicine weighted both the importance of each question for GPs and the correctness of the responses received. Using a Thurstone methodology, importance and correctness were combined to weight respondents' replies, and summed to give a SD knowledge score for each GP.

Results:
Response rate was 68%. 56% of respondents had managed three or more patients with a SD in the last year. Over 75% were aware of the drugs commonly used in SDs. The distribution of scores on the knowledge summative scale was strongly negatively skewed, indicating a good level of knowledge among the majority of GPs. Two variables emerged from regression analysis as predictors of low-scoring GPs: smaller number of practice partners and medical school group. Knowledge was higher among GPs in Wales compared with GPs in East Anglia surveyed four years previously.

Conclusions:
Most GPs have a good understanding of SD use. Educational efforts are most needed among GPs in smaller practices and who are non-UK and London graduates.

2. EVALUATION OF PALLIATIVE CARE DAY THERAPY SERVICES: AN EXPLORATION OF THE PERSPECTIVES OF PATIENTS, INFORMAL CARERS, VOLUNTEER STAFF AND DAY THERAPY MANAGERS

J Low, R Perry and SM Wilkinson
Marie Curie Palliative Care Research and Development Unit, Royal Free & University College Medical School, London

Aim
To explore the perspectives of patients, informal carers, volunteer staff and day unit managers, involved in Marie Curie day therapy services and identify the important outcomes of day services for these different stakeholders.

Method
Focus groups were used to elicit stakeholders’ ideas, thoughts and perceptions about Day Services. These groups were held in 4/11 Marie Curie day units, which were purposively selected to represent the diversity of services offered by Marie Curie Cancer Care within the United Kingdom. The data obtained were transcribed verbatim and analysed using in-depth thematic analysis.

Results
The following themes were identified:
1) There was agreement between day unit managers and volunteers about the main aims of day services.
2) Carers and patients appreciated the access to specialist palliative care professionals and complementary therapies.
3) Patients benefited from peer support and social opportunities whilst the respite element was important to carers. These elements of the day services contributed to an increase in patients’ confidence and a perceived improvement in their Quality of Life, (QoL).
4) Patients and carers were anxious about being discharged and also wanted greater access to the day services.
5) There is evidence of some tension between managers and volunteers with regard to local policy initiatives.

Conclusion
Day services have a positive impact on patients and carers QoL. However access is limited to certain groups and there may be a need for in order for a more flexible configuration of services to meet the needs of a range of patients and carers.

3. REMOVING THE BOUNDARIES: SPECIALIST PALLIATIVE CARE FOR PATIENTS WITH HEART FAILURE

Gillian Horne, Doncaster and Bassetlaw Hospitals Trust, Doncaster
Sheila Payne, Trent Palliative Care Centre, The University of Sheffield, Sheffield

Introduction:
This abstract describes a study, which explored the views of staff in caring for patients with end-stage heart failure.

Research Aims:
To give health professionals opportunity to discuss reservations and identify educational needs in providing palliative care for patients with heart failure.

Sampling Frame:
5 Focus groups (59 health professionals).

Methodology:
A small prospective qualitative designed study using focus groups.

Data Analysis:
Data analysis used a grounded theory approach. Audiotapes were transcribed verbatim. Authors coded transcripts independently.

Results:
Palliative care was seen to be "reserved for cancer patients" and patients with heart failure were perceived to be "set aside". Generic staff lacked knowledge of palliative care and specialist palliative care staff lacked knowledge in the management of heart failure.

Development of services for patients with heart failure included: collaboration; different ways of working; shared care of patients; improved patient access to services; patient and family education and support; promotion of the palliative care approach in general settings.

Specialist palliative care staff felt their skills were transferable and extending current specialist palliative care services was preferable to setting up new services for patients with heart failure.

Conclusion:
Considering the needs of many in an ageing population with multiple diseases, begs the question of where to draw the boundaries of care within specialist palliative care. The answer to this question lies not in what diagnoses or diseases patients have, but what needs exist that can be met by and supported through advise from specialist palliative care services.

4. HOW DO PALLIATIVE CARE PHYSICIANS ASSESS AND MANAGE DEPRESSION?

Iain Lawrie, The Leicestershire Hospice, Leicester
Mari Lloyd-Williams, University of Liverpool Medical School, Liverpool
Fiona Taylor Leicester Medical School, Leicester

Introduction:
Depression is a significant symptom for approximately 25% of palliative care patients, but is frequently unrecognised and untreated. This study was carried out to determine how senior doctors working in palliative medicine assessed and managed depression and highlight any difficulties encountered when managing the depressed patient in palliative care.

Method:
A questionnaire containing both open and closed questions was sent to all palliative care units who had a designated medical person in the UK. This included questions on assessment and management of depression (pharmacological and non-pharmacological); difficulties encountered, their qualifications and current post.

Results:
The response rate was 66%; two-thirds (90) of respondents were consultants and a further 21 medical directors. The majority (73%) routinely assessed for depression with 27% using the HAD scale and 10% asking the patient "are you depressed"?. The most frequently prescribed medication was SSRI (80%). Less than 6% prescribed psychostimulants. Almost all respondents reported difficulties with the assessment and management of depression, which mainly focused on distinguishing symptoms of depression from sadness and whether it was appropriate to treat patients when life expectancy was short. Difficulties were identified in accessing psychiatric input by 47% of respondents - these were mainly around long delays between referral and appointments and lack of access to psychiatry.

Discussion:
Depression is identified as being a difficult symptom to manage by many senior palliative care physicians in the UK and suggestions will be made as to how some of theses issues can be addressed.

5. OPIOID INTOLERANCE: THE RELATIONSHIP BETWEEN WHITE BLOOD CELL COUNTAND POOR RESPONSE TO MORPHINE

JR Ross. D.Rutter, J Riley, E Rees, J. Hardy and K Welsh.
Dept. Palliative Medicine, Royal Marsden Hospital, London, and Dept Clinical Genomics, NHLI, Royal Brompton Hospital, London

Background:
Morphine is the opiate of choice for moderate to severe pain in cancer patients. However, 10-30% of patients require switching to alternative opioids because of inadequate analgesic effect or severe side effects. A recent retrospective study showed that switchers had significantly higher white blood cell counts (WCC) compared to controls. In order to investigate further the mechanisms relating to opioid intolerance, a prospective study was designed.

Methods:
We are currently recruiting cancer patients who have either been stable on morphine ≥1month (controls), or who are identified by the palliative care team to require switching to alternative opioids due to morphine intolerance (switchers). Demographic data, pain and side effect scores and biochemical and haematological data are being recorded and DNA stored for genetic analysis.

Results:
Interim data analysis of sixty four patients recruited to date (19 switchers and 45 controls), confirms the findings of the previous retrospective study showing elevation of WCC in switchers compared to controls (mean ± SD 8.4 ± 4.3 versus 13.2 ± 13.5; F test 4.59 p<0.003). This difference is independent of treatment with cytotoxics or steroids.

Conclusion:
Morphine receptors are known to be present on various cells of the immune system. We now propose to investigate the significance of WCC's in patients recruited to this study by measuring white blood cell subtypes and density of opiate receptors on these cell subtypes. We will also investigate associations with known genetic polymorphisms in the µ opioid receptor as these have been shown to influence both receptor number and function.

6. SYSTEMATIC REVIEW OF THE ROLE OF BISPHOSPHONATES IN METASTATIC DISEASE: SKELETAL MORBIDITY

Y Saunders, Royal Marsden Hospital, London
JR Ross, Royal Marsden Hospital, London
PM Edmonds, Kings College, London
S Patel, Institute of Child Health, London
K Broadley, Royal Marsden Hospital, London
SRD Johnston, Royal Marsden Hospital, London

Aim:
To systematically review the role of bisphosphonates in the reduction of skeletal morbidity in patients with bony metastatic disease.

Method:
We identified randomised controlled trials (RCTs) by searching electronic databases, scanning reference lists, and consultation with experts and pharmaceutical companies. Foreign papers were included. The inclusion criteria were: RCTs, patients with proven malignant disease and bony metastases, oral or intravenous bisphosphonate in the experimental arm, compared to another bisphosphonate, placebo or standard care, with one outcome of skeletal morbidity.

Results:
95 articles were identified; 47 citations relating to 30 studies fulfilled inclusion criteria.

Primary analysis: On meta-analysis bisphosphonates, compared with placebo, significantly reduced the odds ratio for vertebral, non-vertebral and combined fractures (#), radiotherapy (RT) and hypercalcaemia (­ Ca) but not orthopaedic surgery (OS) or spinal cord compression (SCC). Odds Ratio (95% confidence interval): Vertebral# 0.692 (0.570–0.840) p<0.0001; Non-vertebral# 0.653 (0.540-0.791) p<0.0001; Combined# 0.653 (0.547-0.780) p<0.0001; RT 0.674 (0.573-0.791) p<0.0001; SCC 0.714 (0.470-1.083) p=0.113; OS 0.698 (0.463-1.052) p=0.086 and ­ Ca 0.544 (0.364-0.814) p=0.003.

Sub-analysis over time: Different endpoints reached significance at different time points: RT and ­ Ca by 6 months, fractures between 6-12 months. OS reached significance at 12-24 months.

Time to first skeletal related event: Bisphosphonates significantly increase the time to first skeletal related event.

Survival: There was no survival benefit.

Conclusion:
Bisphosphonates significantly decrease all skeletal morbidity endpoints, except SCC. Bisphosphonates significantly increase the time to first SRE and should be started early when bone metastases develop and continued until no longer clinically relevant.

7. REFLEXOLOGY FOR SYMPTOM RELIEF IN PATIENTS WITH CANCER: A COCHRANE PAPAS SYSTEMATIC REVIEW

M Gambles, SM Wilkinson and D Fellowes
Marie Curie Palliative Care Research & Development Unit, Royal Free & University College Medical School, London

Abstract not available

8. INPATIENT PALLIATIVE MEDICINE – HOW GOOD IS THE EVIDENCE?

Paul W. Keeley, Beatson Oncology Centre

Background:
In 2002, Good & Stafford published a study attempting to show that inpatient palliative medicine is evidence based.

Aim:
To critically examine the papers cited in support of the claim that palliative medicine is evidence-based.

Method:
All the papers cited were re-examined by at least two reviewers. The papers were graded using a tool validated in the production of over 60 evidence-based national clinical guidelines. Differences of opinion about quality were resolved by discussion or independent arbitration.

Results:
Sample size Mean sample size was 160 (range 9-1404); 19/41(46%) had <50 subjects; 30/41 (73%) had <100.

Power Calculations:
16/25 (64%) of randomised controlled trials (RCTs), had no power calculations, making it difficult to draw conclusions about true differences between groups.

Quality rating and grade of evidence. These are tabulated below:

Of the RCTs 15/25 (60%) had a high risk of bias; of the other controlled studies 5/6 (83%) had a high risk of bias.

Applicability:
The studies were generally applicable to a population with advanced disease: 30/41 (73%) related to populations with advanced disease.

Conclusions:
This study demonstrates that studies cited in support of palliative medicine as an evidence-based specialty are of variable, and at times poor quality. This study seems to show that the claim that inpatient palliative medicine is evidence based is at best tenuous and at worst misleading.

9. "EVERYTHING HAS CHANGED" - EXPERIENCES OF YOUNG PEOPLE WITH RECURRENT OR METASTATIC CANCER

A.K Hodgson and O.B Eden
Young Oncology Unit, Christie Hospital, Manchester

Introduction:
There has been increasing interest in recent years in the palliative care needs of adolescents and young adults. However, few studies have looked at the experiences of young people with any kind of advanced disease and none has looked specifically at those of patients with recurrent or metastatic cancer.

Aims:
To describe and understand the experiences of a small cohort of young people with recurrent or metastatic cancer, looking particularly at how the illness has affected them, how they have coped with it, and their hopes and concerns for the future.

Method:
Face-to-face semi-structured interviews with consenting patients aged 16-24 years under the care of one oncology hospital. Interviews were audio-taped, transcribed and then analysed to elicit the main themes. Patients were re-interviewed 2 months later, to allow confirmation and further exploration of themes from the first interviews and to determine whether any issues had changed over time.

Results:
7 patients were interviewed, mean age 19 years. Most of the young people felt the illness had impacted on all areas of their lives. The main themes were:

adjustment to changed circumstances –

acceptance of the illness as part of life

appreciation of life and family

altruism towards others

abandonment by friends

anger towards certain people and situations

anxiety for the future

Conclusions:
This study gives detailed information about the experiences of young people with recurrent or metastatic cancer, and the problems they face. This information will help in the provision of appropriate and effective support to these patients.

10. WHAT SHOULD WE MEASURE WHEN ASSESSING THE QUALITY OF END-OF-LIFE CARE?

F. Aspinal, J.A.H. Addington-Hall, J. Chidgey, M. Dunckley and R. Hughes,
Department of Palliative Care and Policy, King's College London

Background:
Given increasing emphasis on user involvement, it is essential that quality assurance measures in end-of-life care address the issues that patients and their families, as well as professionals, consider important.

Aim:
To identify what issues dying patients, their families, and professionals think are important to measure at the end of life.

Methods:
Ten focus groups were conducted using an adapted nominal group technique. Groups comprised of health professionals (2), allied professionals (2), bereaved relatives (4) and patients (2). Participants identified, discussed, rated and ranked the issues raised. Issues were then thematically grouped (by two researchers) to enable cross group comparison. The most important themes were identified using the within-group ratings.

Results:
Identified issues fell into fourteen themes: preparation; dignity; communication; co-ordination and continuity; access to services; information; quality of life; relationships; carer support; symptom management; choice and control; questioning; safety and security; and maintaining a normal environment. Patients', bereaved relatives' and professionals' views varied about which issues were the most important: for example, patients rated 'preparation' as very important but other groups did not.

Conclusion:
There was some agreement between groups but bereaved relatives and health care professionals views of important issues at the end of life did not always reflect patients. When assessing the quality of services for terminally ill people and their families, it is important that a range of perspectives are considered and, in particular, that users' views are incorporated.

11. OLDER PEOPLE'S ATTITUDES TOWARDS DISCUSSING DECISIONS ABOUT CARDIOPULMONARY RESUSCITATION

Mr Gary Bellamy, Dr Jane Seymour, Dr Merryn Gott, Professor Sam Ahmedzai, Professor David Clark
Sheffield Palliative Care Studies Group, University of Sheffield, Trent Palliative Care Studies Group, Sheffield

Aim:
This paper examines older people's attitudes towards discussing resuscitation decisions with patients with palliative care needs. Understanding these is important in view of guidelines published in the UK in 2001.

Methods:
Interviews with 45 older people (29 women, 16 men) were conducted using a third party vignette. Recruitment was from 3 socio-demographically diverse general practices in Sheffield, UK. 2 respondents were &lt; 65 years; 16 were 65-74; 15 were 75-84; 12 were 85 years and over. 42 lived at home, 3 in residential/nursing accommodation. 9 described their health as excellent or very good; 12 as good and 24 as fair or poor.

Analysis:
Data derived from the interviews were audiotaped and transcribed. The interviews were analysed using a framework approach and aided by a software package: 'NUD*IST'. 36 themes were distilled into 7 core categories.

Findings:
32 respondents believed that doctors should discuss resuscitation decisions for the following reasons: 'preparation for uncertainty', 'the right of self determination', 'being a good doctor' and 'helping patients and their families'. 7 participants believed doctors should not engage patients in discussions since 'those who are suffering need the protection of family'. 4 participants were unsure. 2 participants stated that a discussion was too difficult to imagine. These patterns of response related to socio-demographic characteristics.

Conclusion:
These findings highlight important concerns. Discussions of a wider prognostic remit should be held with older people who are seriously ill and their families in which decisions about resuscitation may arise, but not in a precisely targeted way.

P01. HEALTH CARE PROFESSIONALS' EXPERIENCE OF DEPRESSION IN
PALLIATIVE CARE PATIENTS

R Perry, Marie Curie Centre Warren Pearl, Marie Curie Cancer
Care, Solihull, UK

Aim:
To explore Health Care Professionals' experiences and concepts of depression in patients admitted to a hospice.

Method:
Extended case study, collecting data from observation, documentary sources and 13 semi-structured interviews with health care professionals. The data was
analysed using the approach of Grounded Theory.

Results:
Four categories emerged from the data analysis:

1) assessment of depression;

2) management of depression;

3) professional development;

4) perception of depression in palliative care.

It was this last category, which emerged as the core one, weaving the data from the previous categories together. It examined the
premise that the participants' assessment and management of depression and their professional development was affected by their perception of depression.

Conclusions:
A greater understanding and recognition of depression and knowledge of appropriate management in patients with advanced disease may theoretically lead to improved quality of life for patients. There is a need to raise awareness about depression among the multi-disciplinary team who have day to day contact with the patients and education programmes should be planned which address the participants attitudes toward depression. It was evident that a multi-disciplinary team approach to care, which includes professionals specialising in the provision of psychological care, could enhance the care of depressed patients via a coordinated thorough assessment and the formulation of a structured management plan. In light of this research a multi-disciplinary steering group has been set up to formulate a working plan for the assessment and management of depression.

P02. TERMINAL AGITATION FOLLOWING WITHDRAWAL OF ANTI-DEPRESSANT THERAPY

Dr Priya. S. Iyer, Marie Curie Hospice. Newcastle-Upon-Tyne
Dr C. B. Regnard, St. Oswalds Hospice, Newcastle-Upon-Tyne

Aim of the Study:
To compare the incidence of terminal agitation in patients on selective serotonin re-uptake inhibitors with those on tricyclic antidepressants.

Materials & Methods:
We retrospectively looked at the case notes of patients who had died at a Hospice in the period January 2000-December 2002. We identified those who had been treated with antidepressants in the last 3 weeks of life. The doses of drugs used to control agitation and the level of agitation were obtained from patients' case notes.

Observations:
In the documentation reviewed to date that includes 118 cases, agitation was documented in 56% of TCA, 47% of control and in 76% of SSRI groups. The incidence of moderate to severe agitation was also higher in the SSRI group compared to the TCA group and controls. The dose of haloperidol used after stopping the antidepressant was higher in the SSRI group compared to TCA and control groups.

Conclusion:
Tapering the dose of antidepressants is recommended to avoid discontinuation symptoms such as insomnia and hyper-arousal, which may necessitate re-institution of the antidepressant drug. This becomes particularly relevant at the end of life when patients become unable to swallow, resulting in withdrawal symptoms causing distress to patients and their carers.

This pilot study shows increased incidence of terminal agitation in patients who have been on SSRIs as compared to TCAs and controls. In order to achieve statistical significance, the study is ongoing and full results will be presented. Larger prospective studies are needed to substantiate this finding.

P03. PREVALENCE OF VITAMIN C DEFICIENCY IN CANCER PATIENTS

Dr. Catriona Mayland, St James Hospital, Leeds
Dr. Mike Bennett, St Gemma's Hospice, Leeds

Aims:
To assess the prevalence of vitamin C deficiency within a group of hospice patients. To obtain a structured dietary history and assess relationship with plasma vitamin C.

Method:
Patients with advanced cancer were recruited from a large hospice. Data was collected on demographic details, physical functioning and smoking history. An estimate of the number of weekly dietary portions consumed equivalent to 40mg of vitamin C was obtained (recommended daily intake 1 portion per day). Plasma vitamin C was measured by a single blood sample. The study had local ethical approval.

Results:
Fifty patients were recruited (mean age 65.2 years, 28 female). Plasma vitamin C deficiency was found in 17 (34%). Dietary intake of vitamin C was correlated to plasma vitamin C (r = 0.479, p=0.001). Patients who consumed 7 or more dietary portions in the previous week (n=14) had twice the mean plasma vitamin C level than those who only consumed 6 or less (n=35 [39.91 vs 18.96 micromoles/l, p=0.025]). In the 17 patients who were vitamin C deficient, 15 (88%) had consumed less than 6 portions in the previous week. There was no correlation between plasma vitamin C, smoking history or physical functioning.

Conclusion:
Vitamin C deficiency is common in patients with advanced cancer and dietary intake strongly relates to plasma level. A daily portion of vitamin C appears to prevent biochemical deficiency. The clinical significance of vitamin C deficiency in this population has yet to be established.

P04. COST-EFFECTIVENESS OF BISPHOSPHONATES IN THE TREATMENT OF HYPERCALCAEMIA OF MALIGNANCY

D Wonderling, London School of Hygiene and Tropical Medicine, London
J Ross, Royal Marsden Hospital, London
Y Saunders, Royal Marsden Hospital, London
P Edmonds, King's College, London
K Broadley, Royal Marsden Hospital, London
C Normand, London School of Hygiene and Tropical Medicine, London

Introduction:
A recent systematic review reported that bisphosphonates normalise serum calcium in &gt;70% of patients with hypercalcaemia of malignancy, within 2-6 days. Time to relapse increases with higher doses of a given bisphosphonate, is doubled with pamidronate compared to clodronate or etidronate, and is greatest with zoledronate.

Aim:
To model the cost-effectiveness of different bisphosphonate regimens in the treatment of hypercalcaemia.

Methods:
RCT data on initial response rate and time to first relapse, for different drug regimes, were entered into a decision analytic model. We estimated the cumulative duration of normocalcaemia, life expectancy and health service cost, including the cost of hospital stay for treatment. The model was constructed such that a) after relapse, patients would have up to two further treatments with the same drug regimen; b) with each successive treatment both the baseline response rate and the time to relapse diminished by 1/3.

Results:
Bisphosphonates with the longest cumulative duration of normocalcaemia were most cost-effective. Zoledronate 4mg was the most costly but most cost-effective treatment (£69 per day of normocalcaemia). The estimates were, however, sensitive to the amount of time in hospital.

Conclusions:
Bisphosphonate therapy appears to be reasonably cost-effective in the treatment of hypercalcaemia. Newer more potent drug regimens, such as Zoledronate may be more costly than less potent regimens but appear to be more cost-effective. We would advocate that future RCTs collect data on cumulative length of stay and response to successive treatments to facilitate more accurate economic modelling.

P05. NALOXONE STAINING ON IMMUNE CELL SUBSETS CORRELATES WITH A LACK OF RESPONSE TO MORPHINE.

J Stebbing Chelsea and Westminster Hospital, London
J Riley Royal Marsden Hospital, London
JR Ross, Royal Marsden Hospital, London
D Rutter, Royal Marsden Hospital, London
D Henderson, Chelsea and Westminster Hospital, London
K Welsh, Royal Brompton Hospital, London

Purpose:
Intolerance to morphine is a significant problem in palliative care. The mechanisms underlying this process are poorly understood. In a recent study we observed that laboratory white cell counts (WCC) were higher in switchers compared to morphine treated controls. In an attempt to find out why, we investigated the phenotype and relative numbers of cells from both the innate and adaptive immune systems.

Patients and Methods:
We prospectively recruited 42 cancer patients: 30 who had been stable on morphine ≥1 month (controls), and 12 who required alternative opioids due to morphine intolerance (switchers). Blood was analysed by flow cytometry for levels of the following markers: CD3, CD4, CD8, CD16, CD19 and CD56. In addition we estimated opioid receptor numbers using FITC-conjugated naloxone staining.

Results:
Switchers, compared to controls, had higher total WCC (15.0 vs 8.9 x10⁹/l) and higher WBC receptor numbers, demonstrated by the mean fluorescence intensity of naloxone-FITC staining on lymphocytes (F test analysis 11.3, p=0.002), monocytes (F test 20.3, p=0.00006) and granulocytes (F test 16.2, p=0.0002). All patients were profoundly immunosuppressed compared to normal volunteers: total lymphocyte count (CD3) 661 vs 589 cells/mm³, B cell subsets (CD19) 105 vs 82 cells/mm³ in controls and switchers respectively. [normal ranges: CD3 800-2660, CD19 100-600 cells/mm³]

Conclusions:
FITC naloxone staining of white blood cells has a high positive predictive value for switching. Whilst the numbers are small, data are consistent with a recent report related to morphine addiction. Further studies are warranted to explore these results.

PO6. A RETROSPECTIVE STUDY TO IDENTIFY FACTORS PREDICTING MORPHINE INTOLERANCE IN CANCER PATIENTS

J Riley Royal Marsden Hospital, London
K Welsh, Royal Brompton Hospital, London
JR Ross, Royal Marsden Hospital, London
B Gwilliam, Royal Marsden Hospital, London
AU Wells, Royal Brompton Hospital, London
J Hardy, Royal Marsden Hospital, London

Introduction:
Up to 30% of patients do not receive the desired analgesic effect or suffer intolerable side effects from morphine and are often "switched" to alternative opioids. The aim of this study was to find out why.

Methods:
The study was conducted at a cancer centre. Data was collected retrospectively from 100 patients who tolerated morphine controls) and 77 patients who were identified by the palliative care team to be intolerant of morphine and therefore required switching to alternative opioids (switchers). We investigated whether currently logged data could fully explain the need to switch to an alternative opioid. Demographic details, cancer type and serum markers related to organ function were included in an analysis of biochemical and haematological parameters.

Results:
Patients >78 years old (p=0.03), with a high white blood cell or platelet count (p=0.002, p=0.003 respectively), or with severe organ impairment were more likely to switch to alternative opioids. The number of patients receiving cytotoxic therapy was not significantly different between groups. A stepwise logistic regression model based on age, white blood cell count, platelets, albumin and serum alkaline phosphatase accurately separated switchers and controls in only 68% of cases (equation pseudo R² 0.069).

Conclusion:
Biochemical parameters explain little of the variability in response to opioids. Haematological parameters, particularly the strong association of white blood cell count with switching warrant further investigation, as does the finding relating to age. We hypothesise that genetic variation may account for much of the unexplained variability in response to opioids.

P07. A RETROSPECTIVE SURVEY ON THE ISSUES RELATED TO METHADONE TITRATION OF HOSPICE INPATIENTS

SE Foster, H Hugel, J Skinner and JE Ellershaw
Marie Curie Centre, Liverpool, U.K.

Objectives:
Methadone use in patients with cancer-related pain resistant to other opioids is limited due to fears of toxicity, lack of a standardised conversion ratio, and time-consuming titration. This retrospective survey assessed the success rate and time for methadone titration and the morphine/methadone ratio.

Method:
We carried out a retrospective analysis of inpatients with uncontrolled cancer-related pain, who were converted from opioids to methadone, between January 2000 and January 2003. The Morley Makin titration regime was used (1). Successful titration was defined as achieving a stable methadone dose during inpatient stay. The morphine/methadone ratio was calculated using the 24hr-morphine equivalent of the opioid dose and comparing it to the stable total methadone dose.

Results:
The study included 31 patients. Nineteen were converted successfully. The median time period for conversion was six days (range 3-10). The median 24hr-morphine equivalent dose was 120mg (24-1000mg). The median total methadone dose was 30mg (range 1-120mg). The range for the morphine/methadone ratio was 1.2:1 to 40:1. Twelve patients failed titration due to lack of effect (n=7), death during titration (n=3), and drowsiness (n=2).

Conclusion:
In this sample, methadone titration was achieved in a short space of time in the majority of patients with opioid resistant cancer-related pain. Side effects were an issue in a minority. There is no reliable formula for dose equivalence, limiting titration to be carried out in a specialist, controlled setting.

Reference:
1. Morley JS, Makin MK, 'The use of methadone in cancer pain poorly responsive to other opioids' Pain Review 1998:5:51-58.

P08. MODIFIED-RELEASE OPIATES IMPROVE PAIN CONTROL AND HEALTH-RELATED QUALITY OF LIFE IN PATIENTS WITH CHRONIC REFRACTORY ANGINA

Dr Claire A Douglas, Dr Rodger Moore, Dr Austin Leach and Dr Michael R Chester
National Refractory Angina Centre, Cardiothoracic Centre-Liverpool, England

Introduction and Aim:
Patients with Chronic Refractory Angina (CRA) are essentially palliative. Despite maximum medication and revascularisation they have uncontrolled angina with significantly reduced Health Related Quality of Life scores. CRA is a treatment dilemma to cardiologists.
The first series of patients to achieve statistically significant improvements in Quality of Life and Pain scores following treatment with modified-release opiates are described.

Methods:
A prospective consecutive series of twelve new patients was surveyed at the opiate clinic of a regional refractory angina centre. Prior treatment with psychological therapies, TENS, temporary sympathectomies and mild opiates was unsuccessful. Questionnaires were completed on Health Related Quality of Life (Seattle Angina Questionnaire, Short Form-12, Euro-Qol) and Mood (Hospital Anxiety and Depression Questionnaire) before and three months following commencement of opiates.

Results:
Two patients stopped opiates due to side-effects. Results of the remaining ten patients were examined using the Students Paired t Test. The following scores improved with significance: Seattle Angina Questionnaire - exertional capacity (p&lt;0.03), angina frequency (p&lt;0.002), disease perception (p&lt;0.009); Short Form-12 Questionnaire - overall body pain (p&lt;0.0001), general health (p&lt;0.001), vitality (p&lt;0.05), social functioning (p&lt;0.048) and mental health (p&lt;0.036); Euro-Qol (p&lt;0.004) and Hospital Anxiety and Depression score for depression (p&lt;0.05).

Discussion:
One report on four patients suggests that modified-release opiates reduces hospital admissions in CRA. Otherwise, no further evidence exists for their use in this chronic palliative pain problem.

This prospective case series clearly indicates that modified-release opiates significantly improve quality of life and pain control for patients in whom conventional treatment has failed.

P09. VARIANT CJD AND OTHER NON-CANCER CONDITIONS: A SURVEY OF UK SPECIALIST PALLIATIVE CARE SERVICES

Dr Emma Jones, Guy's and St Thomas' Hospitals, London
Dr Teresa Beynon, Guy's and St Thomas' Hospitals, London
Professor Julia Addington-Hall, King's College London, London

Background:
Variant-CJD (vCJD) is a progressive neurodegenerative illness that emerged following the BSE epidemic in the UK. The Department of Health has recommended specialist palliative care for vCJD patients. Non-malignant diagnoses continue to represent a small proportion of patients receiving specialist palliative care, although they have been shown to have similar levels of need to cancer patients. There is little in the literature on the palliative care of vCJD patients.

Method:
A postal survey was sent to the clinical directors of all UK specialist palliative care services (n=652).

Findings:
An 81% response rate was achieved. 92% of responding services felt it was appropriate to accept patients with life-limiting neurological illness, 86% patients with CJD, 46% patients with young onset dementia and 35% patients with dementia. Respondents' concerns included:

1. Limited resources: inappropriate placements and fear of flooding of existing services

2. Safety and behavioural management of patients with dementia,

3. Limited knowledge and experience in caring for non-cancer patients

4. Identifying 'specialist palliative care need'

16% of services had cared for a CJD patient in the last 2 years, and 60% of these had had variant CJD. 2.5% had declined a patient with CJD.

Conclusion:
Most services are willing to take patients with vCJD, although less likely to take patients with other forms of dementia. Ageism may play a part in determining which patients are accepted. The main factors influencing specialist palliative care provision are an identifiable specialist palliative care need, and the desire to avoid inappropriate long-term placements.

P10. DOES THE TIMING OF OUR DRUG ROUNDS COMPROMISE SYMPTOM MANAGEMENT?

Dr John Chambers
Katharine House Hospice, Adderbury, Oxon

Our hospice performs drug rounds at 09:00hrs, 13:00hrs, 17.00hrs and 21.00hrs. The clinical effectiveness of syringe drivers is reviewed between 09:00hrs and 17:00 hrs, when suitable adjustments are made. This means that, for twelve hours each day, only "rescue medications" are available to maintain or restore symptom management. Does this compromise patient care?

The drug charts of 100 consecutive admissions were studied retrospectively. Indications for the use of rescue drugs for each patient were established from medical and nursing records. The time of administration of every rescue dose for every individual patient was noted. The Wilcoxon Signed-Rank Test for differences in medians was used with alpha=0.05 to test the null hypothesis that differences in the use of rescue medication between the "daytime" (09:01hrs and 21:00hrs) and "night-time" (21:01hrs and 09:00hrs) did not reach statistical significance.

More rescue doses were dispensed during the daytime for pain (386 vs. 253) and nausea and vomiting (63 vs. 45). The converse was true for respiratory rattle (30 vs. 43) and sedation (51 vs. 90). Figures for sedation were highly skewed by one patient who required 2 rescue doses during the daytime and 34 at night time but, regardless of this fact, none of the differences in the use of rescue medication between the daytime and night-time reached statistical significance.

This data suggests that symptom management is not compromised at night by performing the routine drug rounds within a twelve-hour daytime timeframe. The current timing of drug rounds will remain unchanged at this hospice.

P11. PAIN EXPERIENCED BY PATIENTS ATTENDING CANCER OUTPATIENT CLINICS: PREVALENCE, SEVERITY, AND IMPACT ON PSYCHOSOCIAL FUNCTION.

Dr Jean Potter, Kings College London
Professor Irene Higginson Kings College London
Dr Columba Quigley, Hammersmith Hospitals NHS Trust, London
Dr John Scadding, National Hospital for Neurology and Neurosurgery, London

Introduction:
Most epidemiological studies of cancer pain recruit patients from pain clinics or palliative care services. The natural history of cancer pain experienced by patients who attend general oncology outpatient clinics is less well known.

Aims:
To calculate the prevalence of pain and assess its severity, impact on psychosocial function and percent pain relief experienced in the previous 24 hours by cancer outpatients.

Methods:
Prospective study of patients attending four cancer clinics. Demographic and medical details were collected on all patients. Patients in pain completed the Brief Pain Inventory and were assessed by a palliative care doctor to identify cancer-related pain.

Results:
197 patients were approached; 94% took part. Mean age was 73 years (SD 11); 40% female; 78% caucasian.

37/185 (20%) had cancer-related pain. Age, sex, ethnic group and presence of pain were not affected by primary site of cancer. 31/37 (83%) had moderate to severe pain (>5 on 11 pt numerical rating scale). 43% experienced <50% pain relief in the previous 24 hours. 24% received Step 3 opioids. Pain was positively correlated with reduced mood (Pearson's r=0.67); enjoyment of life (Spearman's rho=0.57); relations with others (rho=0.51); work (r=0.43); sleep (rho=0.41) and general activity (r=0.36); (all p<0.03), but was not correlated with walking ability.

Conclusions:
Pain was a substantial problem that interfered with psychosocial function in one in five cancer outpatients. Furthermore, pain relief was often inadequate. Closer attention to pain assessment and treatment in oncology outpatients is of fundamental importance to patient-centred care throughout the disease trajectory.

P12. RETROSPECTIVE QUANTITATIVE ANALYSISOF THE AYRSHIRE HOSPICE PAIN ASSESMENT CHARTS

Jill Fyfe, University of Glasgow

Introduction:
National guidelines (SIGN 44) recommend the routine use of a pain assessment tool in the management of cancer pain. This study aims to improve understanding of the use of The Ayrshire Hospice Pain Chart: a uni-dimensional tool which uses a four category Likkert scale, using qualitative and quantitative techniques of data collection

Method:
Retrospective case note and drug chart review. Sample population cohort: 50 consecutively admitted Hospice in-patients. Primary objectives: review of pain classification prevalence and number of pains; pharmacological profile and documented pain assessments at days 1 and 7.Additional data: demographic, reason for admission, frequency of pain assessments, 24 hour opioid requirements, adjuvant prescriptions. Descriptive statistics were used to describe the sample characteristics and Wilcoxon Signed Ranks Tests and 2-sample T tests were used to test for statistical significance and comparison between groups

Results:
No statistical difference in pain at Day 1 and 7 was found using qualitative and quantitative analysis. No statistical difference in opioid doses or number or type of analgesic prescribed. Non significant differences were found in paracetamol prescribing in females.72% were prescribed the same number of analgesics.

Discussion:
A pain assessment tool is used routinely in The Ayrshire Hospice. Within the sample group documented pain assessments remained constant or worsened over the 7-day period Prescribing practice complied with Sign 44. Opioid doses remained stable. This would suggest that concomitant factors such as advanced, rapidly progressive disease and complex pain syndromes may contribute to the prevalence of pain in these patients.

P13. SYSTEMATIC ASSESSMENT OF UNCONSCIOUS PATIENTS AT END OF LIFE: A PILOT STUDY OF BENZODIAZEPINE PRESCRIBING AND USE OF THE AYRSHIRE HOSPICE COMFORT ASSESSMENT TOOL (CAT).

R Thorp, K Sherry and M Tadjali
The Ayrshire Hospice, Ayr, Scotland

Introduction:
The Ayrshire Hospice CAT has been developed to assess the comfort of unconscious patients. We describe a pilot study looking at documentation of comfort, level of unconsciousness and the use of benzodiazepines in the last week of life.

Aims:
To evaluate the level of comfort and unconsciousness of patients as measured by the CAT; to quantify and characterise nursing care administered to unconscious patients; to review daily doses of midazolam in the last week of life.

Method:
Retrospective review of 20 case notes. Data collected: demographic data; length of time on CAT until death; daily maximum, minimum and mean comfort scores and unconsciousness scores; number of nursing assessments; number of drug administrations; number of nursing care interventions; total daily dose of midazolam administered.

Results:
7 males (mean age 70), 13 females (mean age 56), age range: 24-96 years; mean daily dose of midazolam over last week: 10mgs (range 0-180mgs); mean daily dose midazolam day 7 (day of death): 37mgs, day 6: 35mgs; length of time on cat (ie unconscious) range: 0-8days (75% last 2 days); mean number of hours unconscious: 47.5; mean number of nursing observations: 8.7/24hrs; mean comfort scores (number of signs of distress from possible total of 7) day 7: 0.96, day 6: 0.68; mean unconsciousness scores (adapted gcs)day 7: 4.1, day 6: 4.3.

Conclusion:
This has been a useful exercise demonstrating that patients had a high level of comfort in the last 2 days of life and that nursing interventions remained intensive.

P14. PROFESSIONALS VIEWS ON COMPUTER-BASED CLINICAL OUTCOME MEASURES

R Hughes, A Sinha, J Addington-Hall, F Aspinal, M Dunckley and I J Higginson
Joint institutional affiliation:
Department of Palliative Care and Policy
Guy's, King's and St Thomas' School of Medicine
King's College London, London

Background:
Clinical outcome measures can help professionals address palliative care patients' needs, and aid in diagnosis, treatment and care management. Computerising this information has been proposed but little is known about the best ways to achieve this.

Aim:
To understand professionals' views on transforming one clinical outcome measure – the Palliative care Outcome Scale (POS) – for use on handheld computers.

Methods:
Eleven organisations providing palliative care were purposively sampled to implement the POS. These organisations comprised oncology wards (n=4), nursing homes (n=4), general medical ward (n=1), specialist hospital palliative care team (n=1) and community motor neurone disease team (n=1). Ten health professionals who led the implementation of the POS within these organisations took part in semi-structured interviews. These explored implementation processes and the potential for the POS to be integrated into computer systems. Interviews were audiotaped, transcribed verbatim and coded. These codes were subject to content analysis and interpreted.

Results:
Responses fell into three groups: 1. issues surrounding computerised outcome measures (attitudes, computer abilities, training and computer access); 2. perceived impacts (time, workloads and workplace efficiency); and 3. professionals' views on patients participation with computerised clinical outcome measures (patients' abilities and attitudes, illness severity and assisting patients).

Concluding Discussion:
These implications of the results are three-fold: 1. they reinforce existing research on clinical outcome measures and IT in health care; 2. identify special palliative care considerations; and 3. highlight the need for further research to understand the barriers and facilitators to implementing computerised clinical outcome measures in palliative care.

P15. THE DEVELOPMENT OF AN INSTRUMENT TO MEASURE THE INFORMATION NEEDS OF MEN WITH PROSTATE CANCER

Wolfram Jatsch, Nathan Hughes, Alison Pearce, Jeremy Dale and Chantel Meystre
University of Warwick, Coventry

Objectives:
The study aimed to develop and validate an instrument capable of ascertaining the information needs of men with prostate cancer, to investigate the importance attributed to specific areas of information, and to determine whether information needs were being met.

Design/Methods:
Several stages of consultation and modification were conducted to develop a questionnaire which was then piloted on prostate cancer patients (N=96). Respondents were asked to rate the importance of a range of topics of information, and to assess whether or not they had received sufficient information on these subjects.

Results/Findings:
Construct & content validity of the instrument were established. Internal consistency reliability was calculated using Cronbach's alpha and found to be satisfactory (0.91). A factor analysis revealed four discrete, individual factors, which together explained over 68% of the variance. These factors were titled 'Basics of prostate cancer care', 'Disease management', 'Physical well-being' & 'Self-help'.

While each topic was seen to be important by a majority of respondents, particular importance was attributed to information concerning 'Disease management'. However, within this factor, items concerning emergencies and worsening conditions revealed a very high level of discrepancy between the amount of information desired and the amount received.

Conclusion:
The basis of a tool capable of ascertaining prostate cancer patients' information needs has been developed which, with further piloting, would provide doctors with a means of ensuring that their patients receive levels of information adequate to their individual needs. Particular subject areas where information is seen to be lacking were identified for future consideration.

P16. BEREAVEMENT CARE IS AN ESSENTIAL PART OF PALLIATIVE CARE. WHAT SUPPORT SYSTEMS DO BEREAVED PEOPLE NEED?

Doreen Hall, Northumbria Healthcare NHS Trust Palliative Care Team, Wallsend, UK

A review of the literature identified inconsistencies between palliative care teams in providing support to bereaved individuals. Despite years of research, bereavement care remains a neglected area.

Aims:
1. To establish if there was a gap in support services for individuals bereaved through cancer

2. To explore the need for support networks

3. To examine which services will facilitate adaption to loss

4. To increase understanding of the needs of bereaved individuals

5. To identify gaps in current service provision

Method:
A qualitative approach was adopted using a purposive sample of 14 spouses bereaved through cancer. Data was gathered using semi-structured interviews, audiotaped, transcribed and analysed using Burnard's (1991) thematic content analyses.

Results:
Seven major themes emerged from the data collected:

1. The need for support- practical, emotional and informational

2. Timing of support

3. Who supported the bereaved individual?

4. The absence of professionals

5. What bereaved individuals want from the professionals

6. Illness associated with bereavement

7. The bereaved individual's choice- a one to one visiting service

Conclusion:
These results expand our knowledge of the gap in support services. They clearly give health care professionals an understanding of what bereaved individuals feel they need to assist in their grieving process. This needs urgent review if we are to continue to provide holistic care to families. At present bereaved individuals are neglected as a result of poor service provision. If those bereaved are to be effectively supported, recommendations for service delivery will have resource implications.

P17. THE EVALUATION OF USING AN ETHICS HISTORY WITH HOSPICE IN-PATIENTS

Dr Fliss Murtagh, Kings College Hospital, London
Dr Andrew Thorns, Pilgrims Hospices in East Kent, Margate

Definition:
Ethics History' refers to questions of patients about their wishes for illness information, family/carer information, and preferences for decision-making.

Objective:
To increase understanding of information giving and decision-making in the hospice population and to evaluate whether taking an 'Ethics History' improves patient satisfaction and staff confidence.

Method:
This study consists of interviews with in-patients and staff before and after introduction of an 'Ethics History', collecting quantitative and qualitative data. Contamination between control and intervention groups precluded use of an RCT design.

Setting:
Three hospices in Kent.

Outcome measures:
Patient satisfaction with illness information, family information-giving, and patient confidence in decision-making, all measured using simple yes/no questions and verified using qualitative data. Staff confidence in information-giving and decision-making, measured using linear analogue scores, and verified with qualitative data.

Results:
101 patient interviews and 165 doctor interviews were completed. The study shows information-giving and decision-making preferences in this population are similar to cancer populations, but with certain key differences. Satisfaction was high, but significant improvement occurred both with the way information was given and with family information-giving following Ethics History introduction. Doctor confidence also improved significantly. There was no evidence of additional distress. Extensive demographic data was collected to describe the population studied and clarify comparability of control and intervention groups.

Conclusion:
This study expands understanding of the particular preferences of hospice in-patients and how these differ from the well researched cancer populations. The 'Ethics History' also appears to be a valuable addition to the standard history for hospice in-patients.

P18. COMPARATIVE STUDY OF ONCOLOGY, PALLIATIVE MEDICINE AND GENERAL PRACTICE SPECIALIST REGISTRARS' COMPLIANCE WITH AND UNDERSTANDING OF ADVANCE DIRECTIVES

Dr Claire Stark Toller, Duchess of Kent House, Reading, England
Dr Marc Budge, Department of Gerontology, University of Canberra, Australia

Background:
Advance directives should enable patients to influence their treatment if they become incompetent. Little is known about their impact on decisions made by doctors training in different specialties, nor doctors' understanding of their legal status.

Method:
A questionnaire was administered at specialist registrar educational meetings. Respondents determined the treatment level to provide for a patient with Alzheimer's disease presenting with a myocardial infarction and no additional information (case1) or an advance directive requesting maximum therapy (case 2).

Results:
The response rate was 86.1% (31/36). In response to the advance directive, 28.6% (4/14) oncologists, 30.8% (4/13) GPs and 40% (2/5) of palliative medics increased treatment levels. Palliative medics (p<0.09) and GPs (p=0.095) gave more active treatment than oncologists who never gave intensive treatment.

Palliative medics strongly support the use of advance directives, contrasting GPs (p=0.019) and oncologists (p<0.019) who support their use. 28.1% (9/32) of respondents wrongly thought that a statute exists in British law regarding advance directives. 57.1% (8/14) oncologists would not comply with an advance directive written prior to a change in a patient's medical circumstances, while 61.5% (8/13) GPs responded ‘don't know'.

Medical school education was not important for palliative medics, but GPs were unsure of its impact (p=0.014).

Conclusions:
The advance directive influenced doctors across all specialties, palliative medics especially. Confusion exists about the legal status of advance directives, especially among GPs, and no one found medical school training influential. These findings have implications for the impact and usefulness of advance directives, and for medical education.

P19. PATIENT AND GENERAL PRACTITIONER PERSPECTIVES ON SPIRITUAL NEEDS

Elizabeth Grant and Scott Murray, University of Edinburgh

Introduction and aims:
Patients facing death have spiritual needs. Spiritual distress may correlate with poor health outcomes. We describe the range of spiritual needs patients experience, explore if and how these needs impinge on their well-being, elicit patient suggestions concerning who could meet these needs, and assess if patients' spiritual needs are recognised by their general practitioners (GPs).

Participants:
Twenty patients with advanced cancer or progressive non-malignant disease in the community and their GPs.

Methods:
Two in-depth interviews with each patient at three monthly intervals

Interviews with the GP after each patient interview.

Analysis:
Interviews were taped, transcribed and analysed according to the research questions and emerging themes, assisted by Nvivo, and overviewed by the multi-disciplinary steering group.

Results:
Patients described a deep sense of loss, and a sickening fear of the unknowness of death. They felt isolated, and no longer in control of their illness, nor of themselves.

They reported sleeplessness, anxiety, and inadequacy in expressing feelings. Those who had established cosmologies coped better than those whose religious commitment was nominal or whose worldview did not offer a coherent explanation of events. GPs, because they had experience of people dying, were often looked to for spiritual support. GP's recognised that those in spiritual distress were often less well, more anxious, and less reachable. Few GPs believed they had the time, understanding or training to help.

Conclusion:
Patients described significant existential needs, which aggravated psychological, social and physical distress. Providing spiritual care may improve overall health.

P20. NOT ANOTHER QUESTIONNAIRE! MAXIMISING RESPONSE RATE, PREDICTING NON-RESPONSE AND ASSESSING NON-RESPONSE BIAS IN A POSTAL QUESTIONNAIRE STUDY OF GPS.

Dr. S. Barclay, University of Cambridge
Prof. I. Finlay, University College of Wales School of Medicine, Cardiff
Prof. C. Todd, University of Manchester
Dr. G. Grande, University of Manchester
P. Wyatt, Velindre Hospital, Cardiff.

Research Aims:
Non-response is an important potential source of bias in survey research. This paper aims:

1) To investigate the effectiveness of follow-up procedures during a postal questionnaire study of GPs.

2) To investigate the use of publicly available data in assessing non-response bias.

3) To develop regression models predicting responder behaviour.

Methods and sampling.:
A postal questionnaire was mailed to a stratified random sample of 600 GPs in Wales concerning their training and knowledge in palliative care. A cumulative response rate graph permitted optimal timing of follow-up mailings.

Results:
A response rate of 67.6% was achieved. Using publicly available data concerning GPs, differences were found between responders and non-responders on several parameters and between sample and population on some parameters: some of these may bias the sample data.

Logistic regression analysis indicated medical school of qualification and membership of the Royal College of GPs to be the only significant predictors of response. Late responders had been qualified for longer.

Conclusions:
This study has several implications for postal questionnaire studies of GPs. The optimal timing of reminders may be judged from plotting the cumulative response rate: at least three reminders are worth sending. Few parameters predict GPs who are unlikely to respond; more of these may be included in the sample, or they may be targeted for special attention. Publicly available data may be readily used in the analysis of non-response bias and generalisability.

P21. THE POWYS MACMILLAN GP CLINICAL FACILITATOR PROJECT: RESULTS OF AN EVALUATION

Christine Ingleton, Philippa Hughes, Bill Noble and David Clark
University of Sheffield

Background:
A new model of a Macmillan General Practitioner Facilitator role has been developed to enhance palliative and cancer care in Powys, Wales. The scheme ran from April 1999-March 2002 and was the subject of an evaluation, commissioned by Macmillan.

Methods:
The evaluation was designed to be responsive to views of participants and used self-complete diaries, documentary analysis, in-depth interviews, monitoring of referral and prescribing patterns, postal surveys, both of local health care workers and of a sample of bereaved carers.

Findings:
During the lifetime of the project, 17 Powys GPs (from 94) participated in the scheme as Facilitators, a higher figure than anticipated. There was some evidence that the Facilitators had an influence on other GPs' opioid prescribing patterns. GPs also reported a doubling in the amount of time spent on palliative care work. Referrals to Macmillan Nurses also increased substantially during the scheme. Bereaved carers reported high levels of satisfaction with care given by GPs to relatives in the last year of life, though satisfaction levels were higher in relation to District Nurses and palliative care nurse specialists; reported satisfaction levels were unchanged during the scheme. Among those dying of cancer, Powys has a high proportion of deaths in Community Hospitals (48%), with a low proportion dying in hospice; however there were no perceived changes in place of death patterns during the period of the scheme.

Conclusions:
Facilitators appeared to be more successful in undertaking education work than in enhancing their own clinical skills.

P22. USING ETHNOGRAPHY TO RESEARCH PALLIATIVE CARE IN COMMUNITY HOSPITALS

Sheila Payne, University of Sheffield
Christine Kerr, University of Southampton
Sheila Hawker, University of Southampton
Frances Sheldon, University of Southampton
David Seamark, The Honiton Group Practice, Devon
Nikki Jarrett, University of Southampton
Helen Roberts, Southampton University NHS Trust
Helen Smith,University of Southampton
Paul Roderick, University of Southampton

Aims:
To evaluate the nature and quality of palliative care delivered to older people in community hospitals by eliciting the views of users and health professionals. This paper focuses on the methodological issues raised when using ethnography.

Background:
Despite the successful growth of hospices in the UK, there remains inequity of access to good end-of-life care, especially in general hospitals. Community hospitals are conveniently located and allow general practitioners (GPs) to provide continuity of care but little is known about the quality of end-of-life care provided.

Methods:
Building upon a survey of all UK community hospitals (n=478) and semi-structured interviews with 30 hospital managers, six in-depth organisational case studies are being undertaken. Overt non-participant observation of patient care, a questionnaire for health professionals plus semi-structured interviews with palliative patients and their carers are providing detailed, rich and varied ethnographic data.

Discussion:
Spending time in the field has enabled us to build trusting relationships with hospital staff and patients. Staff are encouraged to explain procedures and how practice has developed. However, problems have been encountered with ambiguous definitions of palliative care in this setting. Negotiating access is time consuming and ethical constraints have rendered access to respondents difficult.

Conclusions:
An ethnographic approach can supply in-depth data thereby facilitating greater understanding of practice. It is particularly useful when incorporated as a later stage of a larger multi-method project. However, it is a difficult research method to utilise in situations where transferring ideas across contexts can be problematical. Hospital bureaucracy and ethical restrictions must also be considered.

P23. COMPARING LOCATION OF DEATH AND DECISION MAKING OF PREFERENCE FOR LOCATION OF DEATH AMONG FIRST GENERATION BLACK CARIBBEAN AND NATIVE-BORN WHITE PATIENTS WITH ADVANCED DISEASE

Jonathan Koffman, Irene J. Higginson
Kings College London

Background:
Surveys show many patients with advanced disease prefer to die at home but no research has explored the black Caribbean perspective.

Study aim:
To compare actual place of death, preferred place of death and decision making among first generation and native-born white patients with advanced disease.

Design:
Retrospective comparative approach using an established questionnaire completed by relatives/friends of deceased patients.

Setting:
Inner-city health authority.

Data collection:
Demographic and clinical characteristics, actual/preferred location of death, and patient and family involvement this decision making.

Results:
106 informants representing black Caribbean and 110 native-born white patients contacted. Of these, 47% and 45% participated.12 (24%) black Caribbean and 9 (18%) white patients died at home. 34 (68%) black Caribbean and 27 (54%) white patients died in hospital (c ²= 8.56, P=0.04). 20 (43%) black Caribbean versus 12 (27%) white patients talked about location of death. Of these, 17 (85%) Caribbean versus 10 (83%) white patients wanted a home death. Although similar number of patients who died at home considered they had enough choice about their location of death this was not the same for those dying in other locations. Here, 18 (45%) white versus 6 (18%) Caribbean patients had a choice (c ²= 11.59, P=0.009). Problems with choice were also present for family members of deceased black Caribbean patients (c ²=9.94, P=0.007). Qualitative data illustrate findings from the survey.

Conclusions:
Although home deaths were realised by some black Caribbean patients extending patient and family involvement in decision making on place of death is required.

P24. USING ACTION RESEARCH TO RETRIEVE DYING FROM THE PERIPHERY OF NURSING HOME CARE: THE INTRODUCTION OF AN ‘INTEGRATED CARE PATHWAY FOR THE LAST DAYS OF LIFE'.

Jo Hockley, St Columba's Hospice, Edinburgh
Carol Logan, Lennox House Nursing Home, Edinburgh
Theresa Gibson, Lennox House Nursing Home, Edinburgh
N Hewitt, The Long House Surgery, Edinburgh
Sandra Huggins, Lennox House Nursing Home, Edinburgh
Laura Mingo, Lennox House Nursing Home, Edinburgh

With an increasing older population, there is growing attention on the care of older people dying in care homes. Statistics show that in some areas a fourfold increase in nursing home deaths has occurred over the last 10 years. Research into palliative care and older people dying in care homes is limited. There is a temptation to think that what has been developed for those dying from cancer can just be transferred to those in care homes dying from multiple pathology without any acknowledgement of the differing care environment.

The aim of the study was to explore the culture of dying in nursing homes and through an ‘action research' strategy bring about appropriate change. This presentation will look at ‘action research' as a relatively new strategy in healthcare research, and explore one aspect of the larger study inductively developed to bring about sustained change in an unstable environment, namely: the introduction of an ‘integrated care pathway (ICP) for the last days of life'. How the nursing home manager and wider multi-disciplinary team formulated the ICP group, the training necessary for its implementation and its evaluation will be described. Initial evaluation reveals a greater confidence to attend to the holistic needs of residents and their friends/families in the nursing home. It also confirms results from other aspects of the whole study that those dying from multiple pathology at the end-of-life may encounter a differing dying trajectory from those dying from cancer in mid-life.

P25. ‘REFLECTIVE DE-BRIEFING SESSIONS' FOLLOWING DEATHS IN NURSING HOMES: ENCOURAGING A CULTURE OF LEARNING USING ACTION RESEARCH

Jo Hockley, St Columba's Hospice, Edinburgh
Donna Gilmour, Care Commission, Lothians

With an increasing older population, there is growing attention on the care of older people dying in care homes. Statistics show that in some areas a fourfold increase in nursing home deaths has occurred over the last 10 years. Research into palliative care and older people dying in UK care homes is limited. What research has been carried out demonstrates not only that many nursing homes in the UK are isolated from palliative care education/training but also that traditional education may not demonstrably change practice.

The aim of this study was to explore the culture of dying in nursing homes and through an ‘action research' strategy collaboratively develop knowledge/care relating to palliative care issues. This presentation will look at ‘action research' as a relatively new strategy in healthcare research and explore one aspect of the larger study, namely: nursing home staff ‘reflective de-briefing sessions' following a death to encourage a culture of learning. The reflective process uses critical analysis of a resident's death where feelings, thoughts and actions are synthesised in order to encourage greater palliative care awareness. Extracts from the taped 'reflective de-briefing sessions' will be used as examples. Final evaluation of the ten sessions shows evidence not only of how increased knowledge and self-awareness changed practice, but also how the sessions provided increased communication and support. Reflection on practice encourages learning from within practice thus closing the theory/practice gap from an often-marginalised care environment.

P26. SURVEY OF PATIENT SURVIVAL WHEN TRANSFERRED FROM HOSPICE TO NURSING HOME

K Shorthose, St Peter's Hospice, Bristol, England

Introduction:
It has been shown that transfer of patients to nursing home (NH) can be a stressful event and that transfer of terminally ill patients from hospice may not meet the needs of the patients or carers.

There is increasing pressure within hospices to keep admissions brief. This can lead to difficult decisions being made regarding prognosis and appropriate funding.

Objectives:
To review the survival rate of patients transferred from hospice to NH and how this corresponded with funding.

Methods:
Retrospective survey of patients notes and computerised records at 2 inner city hospices with a total of 20 beds over a 28 month period.

Results:
There were 64 patients in total. Of these:

22 patients (34%) survived less than 2 weeks from transfer;

15 (23%) had been given a prognosis of less than 3 months for Health Authority funding;

5 out of 8 self-funded patients (63%) survived less than 2 weeks;

34 patients (53%) survived less than one month;

4 patients (25%) who were health Authority funded survived more than 3 months.

Conclusions:
A significant percentage of patients transferred to NH survive less than one month. These figures question the appropriateness of some transfers with our knowledge of the likely psychological effects and organisational burden on patients and carers.

We aim to further look at risk factors within this patient group, which may predict imminent death, and whether it would be realistic for these patients to remain at the hospice for terminal care.

P27. TEACHING TEAM-WORKING SKILLS

Dr. Mary Miller, Sir Michael Sobell House, Churchill Hospital, Oxford

Research aim:
To evaluate the module developed to teach registrars in this deanery

Introduction:
The ability to work within a team is an essential skill for doctors. Trainers found this skill difficult to teach and assess. Registrars found it difficult to extrapolate knowledge to practice and adapt knowledge for use in new or abstract situations. A module focusing on learning team-working skills was developed.

Research design:
Evaluation was undertaken by determining baseline knowledge and skills, learning during the module and learning at one year. A semi-structured questionnaire was used for the initial and final enquiries. Each registrar completed a project during the module.

Analysis of data: The questionnaires were studied by the author and themes distilled from the information provided.

Research findings:
66% returned the initial questionnaire and 50% the second. 16% completed their project.

Themes from the baseline questionnaire were:

• Skills we possess
• Skills we need to acquire
• Difficulties in applying knowledge and skills to our practice
• Our educational objectives

Evaluation of learning at one year revealed:

• Registrars were aware of their preferred style of working
• Had greater understanding of difficult team situations
• Continued to feel a lack of skills and strategies to cope with difficult situations.

Conclusion:
The module did not achieve the intended outcome of helping the registrars move from novice to expert in team working. This research has helped define the educational problems the trainers need to overcome to help registrars learn team-working skills.

P28. WHAT IS THE IMPACT OF USING THE LIVERPOOL CARE PATHWAY FOR THE DYING PATIENT ON JUNIOR MEDICAL STAFF?

Dr L. J. Chapman and Dr J. E. Ellershaw
Royal Liverpool University Hospital, Liverpool

Integrated care pathways (ICPs) are increasingly being used in health care. The Liverpool Care Pathway (LCP) is a multiprofessional document, with the purpose of optimising care of the dying and recording outcomes. There is a paucity of literature looking at the views of medical staff who use ICPs, with most studies concentrating on outcome measures.

Aim:
The aim of this study was to assess the impact of the LCP upon the Pre-registration House Officers (PRHOs) working in the Royal Liverpool University Hospital.

Method:
25 PRHOs completed the questionnaires as part of a teaching programme. They included nine items examining their views on the LCP, measured on a four point Likeart scale, from "strongly agree" to "strongly disagree". There was also space for additional comments, and a themed analysis of these was undertaken.

Results:
23 had experience of using the LCP. The Key findings from their responses included: that all the PRHOs agreed or strongly agreed that the LCP increased their ability to prescribe appropriately for dying patients, and 19 agreed or strongly agreed that it helped them to convert oral to subcutaneous medications. Views on team-working, ethical and communication issues were also sought.

Conclusion:
The overall conclusion was that the LCP empowered the junior medical staff in caring for dying patients and promoted multiprofessional team working. This questionnaire provides a model for assessing the attitudes of other health care professionals to the LCP, and will in time influence education programmes and the role of the hospital specialist palliative care team.

P29. PALLIATIVE CARE TEAMS IMPROVE THE SYMPTOMS OF CANCER PATIENTS WITHIN THE ACUTE HOSPITAL SETTING

Dr Barbara A Jack, Marie Curie Centre Liverpool
Dr V Hillier, University of Manchester
Professor A Williams, University of Wales, Swansea
Professor J Oldham, University of Manchester

Background:
Approximately 56% of cancer patients in England and Wales die in hospital. In response, hospital based palliative care teams, which aim to transfer the principles of hospice care to the acute hospital setting, have developed. Despite an increase in the number of hospital teams, there is little research that evaluates their effectiveness. The aim of this study was to assess the effect of the hospital palliative care team on symptom control.

Methodology:
A non-equivalent control group design, using a quota sample, investigated 100 cancer patients been admitted to a University Hospital for symptom control. 50 patients received hospital palliative care team intervention compared with 50 patients receiving traditional care. Data was collected using the PACA symptom assessment tool evaluating the symptoms of pain, anorexia, nausea, insomnia and constipation on 3 occasions (within 24 hours of admission/diagnosis or referral to the team, day 4 and day 7)

Results and Discussion:
Both groups demonstrated a statistically significant improvement in their symptoms (pain p<0.001; anorexia p<0.001; nausea p=.040; constipation p=.002) except for insomnia (p=.527). The intervention group had a greater improvement in all their symptoms, particularly for pain and anorexia for which there were no differences between the groups on the initial assessment. The results indicated the palliative care team had a positive impact on symptom control.

P30. HOSPITAL BASED PALLIATIVE CARE CLINICAL NURSE SPECIALIST IMPACT UPON DOCTORS AND NURSES

Dr Barbara A Jack, Marie Curie Centre Liverpool
Dr V Hillier, University of Manchester
Professor A Williams, University of Wales, Swansea
Professor J Oldham, University of Manchester

Background:
There has been an expansion in the number of palliative care teams based in the acute hospital setting. Although organisation of these teams varies both in structure and approach, clinical nurse specialists (CNS) are one of the key members. The last decade has seen an escalation in the United Kingdom of CNS, and following the NHS Cancer Plan, it is anticipated that the number of CNS in palliative and cancer care is likely to grow. Yet there is a paucity of research studies to demonstrate their effectiveness, and no studies that specifically focus on hospital based palliative care. This study aimed to explore the impact that the hospital based palliative care CNS had on doctors and nurses.

Methodology:
A qualitative evaluation using a non-probability sample design was selected for the study, encompassing 31 tape-recorded semi-structured interviews with senior nurses, consultants, junior doctors and nurses representing different grades and clinical areas. The data was analysed for emerging themes utilising a case and cross case analysis methodology.

Results and Discussion:
The results suggested the presence of the CNS is seen as beneficial to both medical and nursing staff, providing support and advice (symptom control, support after a bad death, supervision) and education (empowerment). Benefits that help non-specialist staff to attain the skills needed to provide optimal care for palliative care patients within the acute hospital setting.