pure.southwales.ac.uk€¦ · web viewreport. spiritual support in end stage heart failure: a...

REPORT

SPIRITUAL SUPPORT IN END STAGE HEART FAILURE: A RANDOMISED CONTROLLED FEASIBILITY STUDY

Linda Ross, Reader1

Jackie Austin, Consultant Nurse2

Paul Jarvis, Statistican1

Sara Pickett, Health Economist3

1School of Care Sciences, Faculty of Life Sciences and Education, Universityof South Wales

2 Heart Failure and Cardiac Rehabilitation Services, Nevill Hall Hospital, Aneurin Bevan University Health Board (ABUHB)

3Swansea Centre for Health Economics, Swansea University

(Correspondence: [email protected])

Funded by the Nevill Hall Thrombosis & General Research Fund

Copyright © University of South Wales, 2017

1

All rights reserved. No part of this publication may be reproduced, stored in a retrieval

system, or transmitted in any form or by any means without written permission from the

publishers.

Linda Ross, Jackie Austin, Paul Jarvis, Sara Pickett have asserted their right to be identified as

the authors of this work in accordance with the Copyright, Designs and Patents Act 1988.

ISBN: 978-1-909838-28-4

2

CONTENTS

PageEXECUTIVE SUMMARY 5

1. BACKGROUND 7

2. AIMS AND OBJECTIVES 9

3. METHODS 103.1 Process 10

3.2 Control group: standard care only 10

3.3 Intervention group: standard care plus spiritual support 11

3.4 The volunteers 11

3.5 The study measures 11

3.6 Analysis and data cleaning 13

3.7 Ethical considerations 14

4. RESULTS 154.1 Description of the sample 15

4.2 Checking if randomisation was successful 16

4.3 Information needed to inform the design of a future RCT 174.3.1 Sample size 17

4.3.2 Time and other practicalities 204.3.2.1 Volunteer uptake and drop-out rates 204.3.2.2 Time 204.3.2.3 Nurse training 20

4.3.3 The measures 204.3.3.1 Effect size with the WHO-SRPB Tool 204.3.3.2 How the measures relate to one another 214.3.3.3 Suitability of the measures for use with this 22

patient group

4.4 Effect of the intervention on study outcomes and trends 23for further exploration

3

4.4.1 Effect of the intervention on:4.4.1.1 spiritual wellbeing 234.4.1.2 depression 254.4.1.3 anxiety 264.4.1.4 health related quality of life 26

4.4.2 Patient satisfaction with the intervention 29

4.4.3 Cost analysis4.4.3.1 Intervention costs 294.4.3.2 NHS resource use 314.4.3.3 Health related quality of life outcomes 344.4.3.4 Exploratory economic analyses 344.4.3.5 Exploratory sensitivity analyses 35

5. DISCUSSION 35

6. CONCLUSION AND RECOMMENDATIONS 38

7. REFERENCES 40

8. Appendix: Dissemination 45

4

EXECUTIVE SUMMARYThe primary aim of this prospective randomised controlled feasibility study was to make recommendations to inform the design of a future RCT to investigate the effect of spiritual support on specified outcomes in patients with end stage heart failure.

47 patients with end stage heart failure (NHYA IV) were randomised to control (standard care, n=25) or intervention (standard care plus spiritual support, n=22) groups. Spiritual support consisted of a 1 hour discussion facilitated by trained volunteers using the ‘Spiritual Enquiry Tool’ at 2 monthly intervals over 6 months (i.e. 4 times in total).

Participants completed measures of spiritual wellbeing (WHO-SRPB Field Test Instrument), depression/anxiety (HAD), and health related quality of life (EQ-5D-3L) at 0, 2, 4, and 6 months. Purpose designed questionnaires captured NHS resource use and potential confounding factors (change in circumstances, life events, symptoms, medication) at 2, 4 and 6 months, in addition to demographic information at baseline and intervention patients’ satisfaction with the spiritual support service at month 6.

Secondary aims were to investigate the effect and cost effectiveness of spiritual support on the above mentioned outcomes if the sample size was sufficient, or to explore trends worthy of further investigation if the sample size was insufficient.

Primary findings133 eligible patients were invited to take part and 104 accepted an information pack, giving an uptake of 35%, The resulting sample size was small (n=47), with 38 complete data sets [18 controls and 20 intervention] As a result of poor uptake, attrition and missing data (for example associated with fatigue, co-morbidity and death) the ability to detect significant changes in study outcomes with the intervention in this feasibility study was compromised; a finding consistent with the National Institute of Health Research (Evans et al.; 2013).

Recruitment and data collection took longer than expected because of patient and staff related factors; 18 months to recruit 47 patients and 2 years to collect the data rather than the estimated 9 months to recruit 65 patients and 15 months to collect the data.

Considerable effort and time was needed by the research team to recruit patients and to keep them in the study, and by the volunteers to deliver the intervention. Of the 8 trained volunteers, 2 dropped out because of ill health and lack of transport. Inclusion of an administrator and a dedicated research nurse is recommended.

The measures were, on the whole, suitable for this patient group, but another QOL measure, such as the Kansas City KCCQ-12, may be additionally useful to capture other aspects of life important to patients, such as ‘enjoyment of life’ which are not included in the EQ-5D-3L. The WHO-SRPB Field Test Instrument was quite long and complex so some patients needed help filling it in. It would be worth testing a shorter simpler measure of spiritual wellbeing in another study.

5

With the WHO-SRPB Field Test Instrument, domain spirituality scores ranged from 4.0 to 18.8 (average 12/13), with the biggest change in any individual of 7 points. Differences in scores between intervention and control groups ranged from 0.3 to 1.4.

Spiritual wellbeing was negatively correlated with anxiety (Rho ranging from -.306 to -.385, p<0.1) and depression (Rho ranging from -.342 to -.648, p<0.05); a finding which supports the existing literature, suggesting that spiritual wellbeing mitigates against anxiety and depression. Spiritual wellbeing may be associated with better QOL, but this was not supported by the limited data from this study.

Secondary findingsNo significant effects were identified for the intervention (spiritual support) on spiritual wellbeing, QOL, anxiety or depression.

The following trends require further investigation:• Spiritual support may have a positive effect on QOL and anxiety; both were improved

in the intervention group at 0-2 months (QOL increased +4 and anxiety -1.2) and were worse in control group at 0-2 months (QOL -8, anxiety +0.8). Alternatively there may be a group allocation effect so we would recommend that study measures are completed prior to start of the intervention. The intervention group’s QOL declined less than the control group’s overall.

• Withdrawal of spiritual support from the intervention group (at the end of the study) may have a negative effect. Depression scores increased in intervention patients (+0.9) and decreased in control patients (-0.6) at the end of the study (months 4-6). We would recommend that measures continue to be completed after the intervention has stopped.

• Spiritual support may lower NHS resource use; a lower resource cost of £204 per intervention patient was noted over the study period.

• Spiritual support may be cost effective if rolled out to more patients as part of routine care, rather than being offered on a cost per head basis within a trial. All service delivery and programme costs were £875 per patient, over 68% of these costs were attributed to training., If the intervention were offered to twice the number of patients over a longer period of time, the data suggests that the provision of spiritual support may meet the NICE ‘willingness to pay’ threshold of £20-30,000 per QALY gained.

• Spiritual support was valued by those receiving it.

Nurses were initially uncomfortable with/lacked confidence in having end of life conversations with patients, indicating the need for training.

6

1. BACKGROUND There is growing evidence demonstrating association between the spiritual part of a person’s life and their mental and physical health (e.g. Koenig et al 2012). Research indicates that spiritual wellbeing mitigates against depression and anxiety (e.g. McClain, et al 2003; Koenig, 2009) and is associated with enhanced coping and quality of life (e.g. Walsh, et al 2002; Tarakeshwar, et al 2006; Balboni, et al 2007).

There is also evidence that the psychological and spiritual domains are distinct but related. Whereas the psychological domain is concerned with affect, cognition, self-esteem and body image, the spiritual domain is concerned with people’s spiritual/ religious/ personal beliefs, and their sense of connection, meaning, wholeness and integration (O’Connell and Skevington, 2010). Examination of the range of definitions of spirituality across disciplines involving diverse groups reveals common elements: hope and strength; trust; meaning and purpose; forgiveness; love and relationships; belief and faith; peoples’ values, morality, creativity and self-expression (McSherry and Ross 2010; RCN, 2011).

It is recognized that the spiritual may come into particular focus during times of crisis such as during illness, uncertainty and loss, whether from reduced functional ability, death or threats of these (e.g. NHS Education Scotland [NES], 2009). Spiritual care is, therefore, an important part of health care and features within healthcare policy and guidance internationally, (e.g. WHO, 2002; WHO, 2012) at European level (Equality Act 2010; European Commission, 2010) and within the United Kingdom (UK) nationally (e.g. Department of Health [DH], 2009; DH, 2015) and locally (e.g. NES, 2009; Scottish Government, 2009; Welsh Government, 2010; Welsh Government, 2015). In the NHS in the UK, spiritual care is provided by healthcare chaplains as well as by other professions such as nurses (QAAHE, 2001; NMC, 2010; RCN, 2011; RCN, 2012). Spiritual care has been defined as:

‘that care which recognizes and responds to the needs of the human spirit when faced with trauma, ill health or sadness and can include the need for meaning, for self-worth, to express oneself, for faith support, perhaps for rites or prayer or sacrament, or simply for a sensitive listener’ (NES, 2009; p6)

and there is evidence that it is valued by patients (Murray et al, 2004; Ross, 2006; Cockell and McSherry, 2012).

Based upon this evidence, guidelines now include spiritual care as an essential component of end of life care (NICE, 2004; DH, 2015; NICE, 2015). The main focus of these guidelines, however, has been on palliative care of cancer patients rather than on those suffering from chronic conditions such as advanced heart failure.

The Welsh Government acknowledges the importance of developing a new pathway for patients with heart failure; one with the potential to gain the greatest population benefit (Welsh Government, 2017). Advanced heart failure is a complex syndrome and dying from

7

the condition can be a lengthy and difficult process, often involving severe exhaustion, prolonged uncertainty for both patients and carers, and additional psychosocial and spiritual challenges (Basta, 2004; Lehman, 2006; Park, 2008; NICOR, 2012). The notion of ‘total pain’ (Saunders, 1996), has been applied to the experience of chronic heart failure and includes spiritual pain in which there is a lack of inner peace and personal integrity (Goebel, et al 2009).

In 2004 the Department of Health (DH) called for further research on spiritual care of people with end stage heart failure (DH, 2004). Few studies had been conducted in this field at that time. In a prospective qualitative study conducted in Scotland on total needs, Murray et al. (2004) found that spiritual needs were important to the sample (n=20); many were seeking for meaning and purpose in the final stage of life and carers struggled to cope with the suffering of their loved ones. The opportunity to talk about spiritual issues when offered was welcomed, however professionals sometimes inadvertently caused spiritual distress by undermining people’s sense of self-worth and identity. Unlike those suffering from terminal cancer, people with heart failure experienced fluctuations in spiritual distress until their death (Murray et al 2007, 2010). The authors acknowledged that their data related to spiritual needs was limited and, like the DH (2004), recommended a more in-depth exploration of this part of life.

A study was therefore conducted (funded by the Nevill Hall Thrombosis & General Research Fund) focused specifically on the spiritual issues and care needs of end-stage heart failure patients/carers’ in Aneurin Bevan University Health Board (ABUHB), South Wales in 2008/9 (Ross and Austin 2013). By involving patients/carers directly the study responded to local directives at that time (Welsh Government, 2005; Gwent Healthcare NHS Trust, 2006). The 16 patients interviewed on 4 occasions over a year were struggling with a range of spiritual/existential concerns alongside the physical and emotional challenges of their illness. Isolation and loneliness was a significant issue; some patients saw virtually no-one from one interview to the next. Many were struggling alone with questions relating to existence, death/life after death. Other concerns related to: maintaining independence/dignity/control; anxiety & depression; hope of a brighter future, self-worth and value; relationships, love and forgiveness. As patients’ condition deteriorated the emphasis shifted from ‘fighting’ the illness to making the most of time left. These concerns were not being fully addressed by heart failure service provision.

A focus group was held with key stakeholders (palliative care [n=4], heart failure [n=4], chaplaincy [n=1], parish nursing [n=1], patients from previous study [n=3]) in 2010 to consider how this gap in service provision might be addressed (funded by Strategic Insight Programme for HEFCW Collaborative Third Mission Fund) and highlighted that a spiritual support home visiting service would be valued, especially by patients living alone. A randomised controlled trial was needed to test if offering a spiritual support service in addition to standard care improved spiritual wellbeing and quality of life and reduced depression and anxiety in this patient group as the literature review indicated may be so. However, a feasibility study was needed to inform the design of such a study.

8

2. AIMS Primary aimThe primary aim of this study is to gather necessary information to inform the design of a follow-on randomised controlled trial (RCT) to investigate the effects of offering spiritual support to end stage heart failure patients on a range of outcomes including spiritual wellbeing, anxiety, depression and quality of life. Information is needed about:

1. Likely sample size: uptake and drop-out ratesA 24% uptake rate was obtained in a previous study (Ross and Austin 2013). However, clinic staff responsible for recruiting had little briefing about the research and participants responded to the university rather than the hospital. The uptake rate may be improved with proper briefing of staff and if the proposed service is offered as part of the current hospital care package for this patient group.

Indication of the likely patient drop-out rate is also needed to ensure sufficient analysable data in accordance with power calculations for the intended study measures.

2. Time and other practicalities of conducting the intended study.It is important to have an idea of the time required to recruit patients, deliver the intervention, train and support volunteers as well as identifying any other issues that might impact on the conduct of a follow-on study. Volunteer up-take and drop-out rates will also be noted.

3. The intended measures- Effect size with the WHOQOL-SRPB Test Instrument. There is little published research on the WHO SRPB tool (WHO, 2002) which measures spiritual wellbeing. Published studies have not reported outputs in a consistent manner, reporting values for different facets within the tool but not necessarily calculating overall values in the way recommended by the WHOQOL-SRPB User’s Manual (WHO, 2002). It is not possible therefore to estimate an effect size which can inform a power calculation for a future study. Results from Hsu et al. (2009) and Boero et al. (2005) provide a guide. Although these studies were performed in different contexts, examining different interventions from those we are proposing, differences between groups of 0.3 or higher were often observed on individual facets of the SRPB scale. We anticipate effect sizes of similar magnitude in this study; this feasibility study will establish more precise figures to inform a subsequent study.

-How the measures relate to each other. For example there is indication from the literature that spiritual wellbeing is associated with higher quality of life and mitigates against depression/anxiety.

-Suitability of the measures for this patient group.

9

Secondary aimsThe secondary aims are to:

4. Investigate the effects of offering spiritual support on outcomes including spiritual wellbeing, anxiety, depression and quality of life if the sample size is sufficient, or to explore trends worthy of further investigation if the sample size is insufficient.

5. Determine patient satisfaction with the intervention.

6. Conduct a cost analysis of the intervention. Costs of training volunteers and of delivering the intervention will be assessed and an exploratory investigation of between group differences in use of NHS and other relevant resources will be undertaken. The intent is not to conduct a formal cost effectiveness analyses. Instead, it is planned that the cost of the intervention (training and delivery) plus/minus any group differences in NHS resource use will be described in relation to health related quality of life outcomes. This will allow for costing of the service, and a tentative exploration of the relationship between costs and effects.

3. METHODS3.1 ProcessEthical (NHS and University) and Health Board approvals were obtained in March 2014. All English speaking end stage heart failure patients (NYHA IIIb/IV) in Aneurin Bevan University Health Board (including Caerphilly, Newport, Monmouth, Torfaen and Blaena Gwent localities) with mental capacity who were able to provide consent were eligible to take part in the study. Patients were recruited by heart failure specialist nurses during routine follow-up appointments at clinic or at home. Patients were told about the study and were offered an information pack containing a participant information sheet and reply slip which they completed and returned to the Principal Investigator (PI) at the University of South Wales (USW) in a pre-paid envelope if they wished to take part. One hundred and four packs were given out and 47 patients returned slips and were prospectively and randomly allocated to receive standard care only (control group n=25) or standard care plus spiritual support (intervention group n=22) using a random numbers list generated by Microsoft Excel (uptake rate of 35%).

3.2 Control group: standard care onlyStandard care for patients with end-stage heart failure consisted of regular review by a heart failure specialist nurse at out-patient clinic or home depending on the patient’s clinical condition. The focus of care was based on symptom management achieved through support and drug titration (Austin et al. 2005). Consent and baseline data were obtained by the Principal Investigator (PI) (LR) at first appointment. Packs of questionnaires were left with participants to complete at months 2, 4 and 6. They received a reminder phone call the week before data collection was due and posted their completed forms to the PI using pre-paid envelopes.

10

3.3 Intervention group: standard care plus spiritual supportThe intervention group received standard care plus spiritual support provided by a trained volunteer at months 0, 2, 4 and 6. Written consent was obtained prior to first data collection by a volunteer who was a retired doctor or nurse. Volunteers collected questionnaire data from patients at each visit prior to providing the spiritual support. Spiritual support consisted of approximately an hour’s discussion using a purpose designed Spiritual Enquiry Tool used in a previous study (Ross and Austin, 2013), based upon the work of Murray et al. (2004) and Ross (1997). The tool included 4 elements: establishing comfort; exploration of the impact of the illness; exploration of meaning, value, purpose and faith; discussion of the future (the tool is available from the authors on request).

3.4 The volunteersFifteen volunteers were recruited by two retired GPs with a pastoral care background who were members of Partnerships in International Medical Education (PRIME, http://www.prime-international.org/home.htm), a charity committed to integrating rigorous science and compassionate care for the whole person through educating healthcare professionals across the world. Volunteers went through an application process and undertook a training course totalling 20 hours (Table 1). Eight volunteers expressed an interest in helping with the study. The Ethics Committee required that volunteers undertake a further half day CPR training course. Volunteers also obtained Disclosure and Barring Service (DBS) checks and Honorary Contracts with ABHUB. This new voluntary spiritual support home visiting service was then registered as a charity with Gwent Association of Voluntary Organisations (GAVO, http://www.gavowales.org.uk/).

Table 1: Volunteer Training Programme (20 hours)

These sessions are provided by retired GPs HM and BB, a psychosocial counsellor (JH) and a lead chaplain (MM). Training includes presentation of the concepts listed below, role-playing scenarios, as well as some case discussions. Personal reflective exercises are also used. Attention also focuses on inappropriate behaviours e.g. proselytising and sharing of personal beliefs.

Part 1 1. Introduction (2hrs)2. Basic listening skills (2hrs)3. Building on listening skills (2hrs)4. Creating understanding – being there and the power of presence (2hrs)5. Crisis and grief support – understanding grieving and loss (2hrs)Part 2Training in use of Spiritual Enquiry Tool (Ross, 1997; Murray et al. 2004) and familiarisation with the study measures involving role play.1. Introduction to spiritual enquiry (2hrs)2. Setting the scene and establishing comfort and exploring impact of the illness(2hrs) 3. Exploring meaning, value, purpose, faith and focus on the future (2hrs)4. Rounding up and putting it all together (2hrs)5. Protection of Vulnerable Adults and Lone Working Policy training (2hrs)

3.5 The study measuresThe study measures and their frequency of collection are outlined in Table 2. As far as possible, standardised measures were used and were selected to balance the need to

11

answer the study questions with keeping patient burden to a minimum. The WHO-SRPB Field Test Instrument (WHO, 2002) was chosen to measure spiritual wellbeing because of its extensive development across over 40 countries and its focus on the broad perspective of spirituality, not just religion. It also provided the opportunity to gather data on the how this patient group might score on the measure in the absence of any published data. The Hospital Anxiety and Depression Scale (HAD, Zigmond and Snaith, 1983) was chosen as it is a well validated and reliable tool and is short. Purpose designed questionnaires were used to gather information about demographics (Demographic Questionnaire), patient satisfaction with the spiritual support service (Satisfaction with Service Questionnaire) and potential confounding factors (Confounding Factors Questionnaire).

Table 2: The Measures

Item being measured Measure & details Frequency of completion

Demographic information Purpose designed Demographic Questionnaire

Baseline by volunteer/PI.

Spiritual wellbeing WHO-SRPB Field Test Instrument (WHO, 2002). 32 items rated 1 ‘not at all’ to 5 ‘An extreme amount’. 8 facets measuring connection, meaning, awe, wholeness, strength, peace, hope, faith. Total score.

0, 2, 4, 6 months by participant.

Anxiety and depression Hospital Anxiety and Depression Scale (Zigmond and Snaith, 1983). 14 items scored 0-3; 7 items relating to anxiety and 7 items relating to depression.

0, 2, 4, 6 monthsby participant

Health related quality of life EQ-5D-3L (The EuroQol Group, 1990). 5 dimensions each scored 1 to 3. Total utility weighted score.

0, 2, 4, 6 monthsby participant

NHS resource use Questionnaire

5 short questions on number of NHS contacts since baseline. Purpose designed.

2, 4, 6 monthsby participant

Degree to which the spiritual support service was valued

Satisfaction with Service Questionnaire. Purpose designed 5 point Likert Scale. (Completed in private, placed in a sealed envelope and given to the volunteer).

At final visit by intervention participants

Potential confounding factors Purpose designed confounding factors Questionnaire noting changes in the following since last visit:-major life events-circumstances-symptoms-medication

2, 4, 6 months by participant

In order to undertake the cost analysis aspect of the study intervention costs were recorded prospectively. These included cost of training, intervention delivery, and debriefing sessions. Additionally, NHS resource use data were captured via a bespoke questionnaire using standard methods and sources, including PSSRU unit costs (Curtis and Burns, 2015; Curtis and Burns, 2016), the British National Formulary (Joint Formulary Committee, 2017) and NHS reference costs (Department of Health, 2016). This questionnaire captured patient

12

contact with primary and secondary care, and antidepressant medication costs over the previous two months. Patient recall has been shown to be a valid method for collecting NHS resource use data (Thorn et al, 2013).

The primary outcome used to inform the exploratory economic analysis was a preference based measure calculated from the EQ-5D-3L (The EuroQol Group, 1990). The EQ-5D-3L health related quality of life instrument is preferred by the National Institute for Health and Care Excellence (NICE) for the economic evaluation of NHS interventions. The tool captures patient reported outcomes in two ways. Firstly, patients are asked to report the assessment of their health state, focused on five dimensions of health: mobility, self-care, usual activities, pain/discomfort and anxiety/depression. This generates a five figure profile, which can be converted to an individual utility score that represents the patients current health related quality of life (overall score). Secondly, patients are required to self-rate the value of their health state on a visual analogue scale (VAS), which ranges from 0 (worst imaginable health) to 100 (best imaginable health). EQ-5D-3L scores were converted to health utilities using a tariff provided by the EuroQol group derived from UK social preference surveys (van Reenen and Oppe, 2015).

3.6 Analysis and data cleaningForty seven patients were enrolled into the study. Missing data was a feature of the data collected, as is often the case, particularly in very ill patient groups as noted by the National Institute of Health Research (Evans et al, 2013). In order to minimise imputation bias as far as possible within the study constraints, the data set was cleaned to remove patients with baseline data only (i.e. where no follow-up data were available). Nine patients were removed in this way, mainly because of illness or death. This resulted in 20 patients remaining in the intervention group and 18 in the control for analysis purposes (38 in total). A further 7 patients dropped out ; 4 after 2 months and 3 after 4 months because of death (6) or feeling better (1).

Measures were scored and entered into SPSS version 22 for Windows. Written responses to open questions were categorized, assigned codes and entered into the database.

Descriptive statistics and Repeated Measures ANOVA were used to explore within and between group differences in the main study outcomes.

For the cost analysis, within-group means were imputed for four missing time-point costs (a total of four costs were imputed for three patients – one time point cost was missing for two patients and two time point costs for the third). It was acknowledged that this simple approach to imputation, like any other, introduces the potential for related bias. However, given that economic analyses use means, and that the standard error is not an influential factor in deterministic economic analyses, it was considered a suitable method for imputation. Linear imputation was used in two instances for EQ-5D generated QALYs (at two time points for one patient). This choice of imputation method was guided by written reports within the patient questionnaires, which indicated that the patient’s health related quality of life continued to improve with time. Resulting utilities were used to conduct area

13

under the curve analyses to produce outcomes in the form of quality-adjusted life years (QALYs).

Descriptive statistical analyses were undertaken in SPSS to present a descriptive profile of costs and outcomes. Additionally, a basic MS Excel model was developed to tentatively explore the relationship between costs and effects from an economic perspective. However, it needs to be re-emphasised that this by no means represented a formal cost effectiveness analyses. Instead, these analyses served to explore and estimate differences in effect size between costs and outcomes between the two groups, and the relationship between cost and effects, with the view to inform a future study design.

3.7 Ethical considerationsEthical approval was obtained from the University of South Wales Faculty Ethics Sub-Committee in January 2014 and NHS ethics approval was granted by NISCHR RES in March 2014 (REC 14/WA/0075. IRAS ID 145664). Health Board research scrutiny and risk review approvals were also obtained in March 2014 (Ref RD/1239/14). A number of ethical concerns were identified as outlined below.

• That participants may be identified. Consent forms were filed in patients' study records held at the University by the PI. Contact details were kept in locked filing cabinets at the University or at the Spiritual Support Co-ordinator’s office. Patients were identified by number and not name on all study questionnaires and data files.

• Burden of time incurred in taking part in the study. Participants were informed of the likely time envisaged in taking part in the study before they agreed to do so. The number and length of questionnaires was kept to a minimum.

• Becoming upset. Exploration of spiritual concerns has potential to elicit deeply emotional responses from patients. Volunteers were trained in conducting the discussions in a sensitive and sympathetic manner. Should participants become upset the counselling and chaplaincy services at Nevill Hall Hospital were available. These services were also available to volunteers.

• Disclosure of sensitive information. It was agreed that should a patient disclose sensitive information to a volunteer, e.g. mention of suicidal thoughts, the volunteer would inform the patient they would have to inform the psychosocial counsellor for the Heart Failure Team. When patient scores on the HAD Scale indicated extreme anxiety or depression (score of >10), this information would be passed on to the psychosocial counsellor for the Heart Failure Team.

• Voluntary participation without prejudice. Participation was entirely voluntary. Care for patients who declined to take part or dropped out was not affected.

• Obtaining consent. Consultants were asked for permission to approach all suitable patients identified from the clinic records to invite them to take part and GPs were informed if any of their patients agreed to take part. Patients returning a slip to indicate their willingness to take part were contacted by the PI to answer any

14

questions about the study prior to arrangements being made to collect written consent. At this point patients were told which group they had been allocated to (each new slip received was given the next slot on the random numbers list). Intervention patients were told that their contact details would be passed to the co-ordinator for the Spiritual Support Service who would allocate a volunteer to them (next available volunteer from the list to ensure equal distribution of workload) and arrange for a retired nurse or doctor to obtain written consent. The PI arranged to visit control patients to obtain written consent.

• Supplying and sustaining a group of volunteers able to provide a sensitive spiritual support service without proselytising or trying to impose their own personal beliefs on vulnerable patients. Sensitive volunteers were needed who behaved professionally and did not proselytise. Volunteers were recruited by two retired GPs (members of the GMC) and a palliative care nurse (member of the NMC) bound by their regulatory body codes of practice. Volunteers were already known to the GPs/nurse and were deemed to be suitable. A rule was set that, should a patient ask them about their personal beliefs, they would decline to answer the question and would instead steer the conversation back to the patient. Volunteers undertook a rigorous 20 hour training programme based upon Age UK charity and local chaplaincy training guidelines (Table 1). They were screened in the same way as other volunteers working within the Health Board having DBS checks, references supplied, occupational health clearance and Health Board honorary contracts in place.

• Ensuring that patients’ DNR preferences were respected. In order to ensure that patients’ DNR preferences were respected, volunteers needed to know if a DNR was in place for the patients they were visiting. This information was obtained from specialist nurses and included in participants’ research records prior to first visit. Should a patient with a DNR in place collapse during a visit, volunteers would call an ambulance and keep the patient comfortable until its arrival. CPR would not commence. Should a patient without a DNR in place collapse during a visit volunteers would call an ambulance and would commence CPR until arrival of the ambulance. Health Board training in and certification of competence in CPR was obtained by all volunteers.

4. RESULTS4.1 Description of the sampleForty seven participants were enrolled from the 5 localities in ABUHB (Caerphilly, Newport, Monmouth, Torfaen and Blaenau Gwent), however 9 withdrew shortly after the first data collection point meaning that data were collected across multiple time points for 38 patients (18 control, 20 intervention), 31 of whom participated for the entire duration of the study (4 time points). Four patients participated for only 3 time points and 3 participated for 2 time points. Of the 38, 22 were male and 16 were female, they were aged between 56-90 (mean 80) years, 17 were married, 17 were widowed and 4 were divorced. Twenty three claimed to have a religious affiliation (6 Church of England/Anglican, 6 Church in Wales, 3 Baptist, 2 Christian, 2 Presbyterian, 1 Congregational, 1 Roman Catholic, 1 Free Church, 1 non-conformist) and 15 were practising. Twenty lived with someone, such as a spouse/partner

15

or son/daughter, whilst 18 lived alone. Most (27) required some form of care from a combination of family and health and social care staff; only 11 had no help.

4.2. Checking if randomisation was successful.In order to check for bias and be assured that randomization had worked appropriately it was important to check that there were no significant differences between the control and intervention groups at the start of the study.

Tables 3 and 4 show that there were no significant differences between groups at baseline in demographic characteristics (gender, religious affiliation, practice of religious affiliation, whether living alone or with someone) (Table 3) or in the study measures namely; depression and anxiety (HAD), spiritual well-being (WHO-SRPB) and health related quality of life (EQ-5D-3L)(Table 4).

There were no significant differences between the 2 groups in reported changes in life events, circumstances, symptoms (deterioration) or medication in the first 2 months of the study (Table 3).

It can be concluded that randomisation has worked and there is no significant bias.

Table 3: Checking for differences between groups at baseline or in first 2 months (Chi square test)

Control N (%) Intervention N (%) chi sq p

GenderMale 13 (52) 13 (59.1)

0.238 0.626Female 12 (48) 9 (40.9)

Religious Affiliation

Yes 14(56) 12(54.5)0.01 0.92

No 11 (44) 10 (45.5)

Practicing Religion

Yes 10 (4) 6 (27.3)0.903 0.637No 5 (20) 6 (27.3)

N/A 10 (40) 10 (45.5)

Living with

Alone 10 (4) 11 (50)

0.586 0.746Spouse/ Partner 13 (52) 9 (40.9)

Son/Daughter 2 (8) 2 (9.1)

Life eventsYes 5 (27.8) 3 (15.8)

0.784 0.376No 13(72.2) 16 (84.2)

CircumstancesYes 0 (0) 1 (5.3)

0.974 0.324No 18 (100) 18 (94.7)

Deterioration in Symptoms

Yes 3 (16.7) 0 (0)3.446 0.063

No 15 (83.3) 19 (100)

MedicationYes 6 (33.3) 5 (26.3)

0.218 0.641No 12 (66.7) 14 (73.7)

16

Table 4: Checking for differences between groups in study outcomes at baseline (Mann Whitney U test)

Control Median

Intervention Median U p

HAD Depression Month 0 8 6.5 233 0.368HAD Anxiety Month 0 6 6.5 226.5 0.3

SWB Month 0 13.9 13.3 250.5 0.601EQ5D VAS Month 0 50 60 258.5 0.928

EQ5D overall Month 0 .49 .62 257 0.700

4.3. Information needed to inform the design of a future RCT

4.3.1 Likely sample size: uptake and drop-out rates Recruitment and data collection took much longer than expected. It was estimated that 65 patients (from a total of 270) would be recruited over 9 months at a rate of 8 per month (Table 5) with data collection taking 15 months. In reality it took 18 months to recruit 47 patients (from a total of 133, Table 5) and 2 years to collect the data. The uptake rate was 35% (47 of 133).

Table 5 Actual recruitment compared with expected (October 2014-March 2016, 18 months)

MonthNo each month

(expected)

Running total (expected)

No. told about study

No. accepting pack

No. enrolling

Running total

(actual)1 8 8 25 22 13 132 8 16 16 16 11 243 8 24 1 1 1 254 8 32 8 8 3 285 8 40 5 4 2 306 8 48 20 12 5 357 8 56 13 8 1 368 8 64 8 5 1 37

9 8 72 (stop at 65) 4 3 0 37

1014

(recruitment drive)

11 3 40

11 1 0 0 4012 1 0 0 4013 2 2 0 4014 3 2 2 4215 2 2 2 4416 4 4 1 4517 4 4 2 4718 1 0 0 47

TOTAL 65 133 104 47 47

17

Of the 47 patients enrolled, 9 dropped out before or just after collection of baseline data (3 were too ill, 4 died, 1 not interested, 1 found it too difficult to face) meaning that 38 took part in the study (18 control, 20 intervention). Complete data sets were obtained from 31 patients. Three patients dropped out after 2 months and 3 further patients dropped out after 4 months (5 died and one felt better). Data collection should have taken 6 months per patient. Whilst the majority of patients (23) completed the study within that time frame, 6 patients took between 9 and 19 months to complete the study. This was mainly because of co-morbidity and frequent hospital admissions. For 2 patients this was because of volunteer illness. The effort needed from the research team to obtain the data was considerable. For example one patient required 28 calls to friends/family and 7 calls to health care professionals (HCPs) over 14 months to collect 2 sets of data whilst another required 14 calls to friends and family, 2 calls to HCPs, a visit and discussion at the advisory group meeting over 12 months to collect 3 data sets.

Reasons for disparity between the estimated and actual figures and measures employed to enhance recruitment

It was estimated that around 270 patients would be available to recruit from, but in reality this was 133. The reasons for this disparity are complex and are summarised in Table 6, together with actions to enhance recruitment.

The estimate of 270 was based on a trawl of clinic records in 2012 and recruitment figures from our previous study in 2010. Data collection for this study did not start until 2014 because of the time taken to secure funding and to obtain ethical approvals. In hindsight a more up-to-date estimate, which would have taken account of the shifting landscape described in Table 6, should have been undertaken and is recommended for future studies.

Table 6: Reasons for slow recruitment and action taken

18

Reasons for slower than expected recruitment1. Change in pattern of the illness with patients staying well for longer (with new treatments such as device therapy) meaning that it took longer for them to deteriorate to NYHA IIIb/V (inclusion criterion for study).2. Co-morbidity and morbidity in this patient group. Several patients agreed to take part but then did not give consent due to feeling too ill or not fully understanding what was involved in taking part. Others were very frail with hearing difficulties or difficulty writing, making it difficult for them to hold a conversation or fill in the questionnaires. So although initially willing, the burden of participating proved too great. Some patients died after providing baseline data.3. Holiday periods, annual leave, and study leave of staff running clinics.4. Two of the 5 Nurses (recruiting from clinics) were on long term sick leave.5. Business of clinics.6. Heavy staff workloads.

Action taken to enhance recruitment (date action started)1. Regular recruitment updates sent to nurses throughout the study.2. To ease staff workloads the approach to control patients was altered. Instead of control patients being recruited by specialist nurses during routine appointments it was decided that the PI would visit all control patients to obtain consent and arrange for data collection. This had cost implications because of the additional travel and time required but was manageable (from October 2014).3. Nurses trawled their caseloads to ensure that all eligible patients in the Health Board had been invited to take part. Nurses on phased return to work after sickness absence were supported in undertaking this task by other team members (April-November 2015).4. Nurses went through clinic lists in advance to identify potential recruits so that none were missed during busy clinics (from November 2015).5. The PI undertook an analysis of the number of patients recruited from each locality. One locality had only recruited 2 patients and there were none from the local vicinity of the hospital. The clinic lists were scrutinised in detail with a view to identifying suitable patients in these areas (October 2015).6. The PI followed up every patient listed on the weekly spreadsheet to ensure they had been invited (from January 2015 onwards).7. The study administrator phoned every patient who was potentially interested in taking part (from clinic lists emailed weekly), to ask if they had reached a decision, thus ensuring everyone was followed up (from January 2015 onwards).8. The chaplain passed on the names of potential patients to the clinical team following his ward rounds as a ‘double check’ (from October 2015 onwards).

Other actions considered but not implemented.1. To ease the burden on nurses of recruiting during busy clinics the PI offered to attend clinics to speak to patients and give out packs. In reality it was decided this would not make the best use of time due to unpredictability of potential recruits (at various points throughout the study). 2. Consideration was given to increasing recruitment by targeting patients fluctuating between NYHA III and IIIb status. This was considered inappropriate because fluctuations are almost exclusively due to fluid overload which can be easily treated resulting in rapid return to status III (October 2015).

4.3.2 Time and other practicalities

19

4.3.2.1 Volunteer uptake and drop-out ratesEight volunteers undertook the training and two dropped out early on in the study because of ill health and lack of transport.

4.3.2.2 Time to deliver the intervention First visits to obtain written consent lasted between 30 and 140 minutes with the average being 55 minutes. Subsequent visits to deliver the intervention lasted around an hour every two months. Additional time was taken phoning to arrange or rearrange visits and in travelling to and from patients’ homes.

4.3.2.3 Nurse training Nurses indicated they felt unprepared for engaging patients in conversations about spiritual and existential concerns and felt the need for more training.

4.3.3. The measures4.3.3.1 Effect size with the WHO-SRPB Tool Mean domain scoresTable 7 shows the mean domain spirituality scores for the control and intervention groups at all 4 time points. The scores remain fairly constant except for a non-significant drop of 2 points (1.7 after correction for confounding factors) in the control group in the first 2 months. The difference in scores between control and intervention groups ranged from 0.29 (month 4) to 1.4 (month 0). Subscale scores are not presented in this study but are available on request from the authors.

Table 7: Mean Domain Spirituality scores for the WHO-SRPB

Control or Experimental Mean Std. Deviation N

Domain Spirituality Month 0

Control 13.852 1.832 15

Experimental 12.433 3.673 16

Total 13.120 2.972 31


Control 11.832 3.275 15


Total 12.338 3.093 31


Control 12.711 3.015 15


Total 12.562 3.190 31


Control 12.126 3.252 15


Total 12.321 3.056 31

Descriptive statistics

20

Table 8 shows that SWB domain scores ranged from 4.89 –18.88 with means around 12-13 for each month and a SD of around 3.

Table 8: Descriptive statistics for WHO-SRPB

Domain Spirituality

Domain Spirituality

month 2

Domain Spirituality

month 4

Domain spirituality month 6

NValid 47 38 35 32

Missing 0 9 12 15Mean 13.032 12.495 12.689 12.367

Std. Deviation 2.857 3.091 3.067 3.0167Minimum 6.00 5.11 4.89 6.22Maximum 18.78 17.11 18.88 18.00

Change in scoresTable 9 shows that the biggest change in score within any individual was 7 points. A potential deterioration of around 6-7 points was observed within individuals compared with baseline scores and a potential improvement of 4-5 points was observed over the same time period.

Table 9: Changes over time in comparison with baseline for WHO-SRPB

Domain Spirituality Change months 0 to 2

Domain Spirituality Change months 0 to 4 or

2-4

Domain Spirituality Change months 0 to 6 or

4-6

NValid 38 35 32

Missing 9 12 15Mean -0.850 -0.604 -0.752

Std. Deviation 2.69444 2.054 2.544Minimum -6.33 -6.00 -7.11Maximum 4.89 3.11 4.78

4.3.3.2 How the measures relate to one anotherThere is indication from the literature that spiritual wellbeing is associated with higher quality of life and mitigates against depression/anxiety. The analysis of our data in these respects is shown in Table 10.

21

Table 10: Correlation (spearman) of Spiritual wellbeing (WHO-SRPB Spirituality domain score) with anxiety (HAD), depression (HAD) and health related quality of life (EQ5D VAS and overall scores)

Month 0 (p) Month 2 (p) Month 4 (p) Month 6 (p)Depression -0.455 (0.001) -0.38 (0.019) -0.342 (0.045) -0.648 (<0.001)

Anxiety -0.332 (0.022) -0.306 (0.062) -0.364 (0.031) -0.385 (0.03)EQ5D VAS -0.224 (0.135) -0.096 (0.573) -0.115 (0.516) 0.252 (0.164)

EQ5D Overall 0.54 (0.717) 0.358 (0.027) 0.199 (0.252) 0.060 (0.744)

Our results show that there are strong significant negative correlations between spiritual wellbeing and anxiety and depression at nearly all time points (all p<0.05, except Anxiety at Month 2, p<0.1), a finding which supports the existing literature indicating that spiritual wellbeing mitigates against anxiety and depression.

Our results correlating spiritual wellbeing and health related QOL are less conclusive in supporting the literature which has suggested a positive association between spiritual wellbeing and quality of life. We found non-significant positive correlations of varying strength between spiritual wellbeing and health related QOL overall scores and mainly negative correlations with QOL VAS scores.

4.3.3.3 Suitability of the measures for use with this patient group.On the whole, the measures worked well with this frail, very ill patient group. We were concerned about patient burden so kept the number of questionnaires to be completed to a minimum.

The WHO SRPB tool is quite long and some patients required help with filling it in. Some also seemed to struggle to substitute their ‘own personal belief system’ in place of ‘spiritual being’ ‘faith’ or ‘spiritual strength’ as instructed by some of the questions. It seemed that some would sometimes score these questions as zero indicating that the question was not relevant to them. It may be worth trialling a shorter less complex spiritual wellbeing measure in future studies.

As discussed above, the addition of a quality of life measure more suited to the profile of this client group may also be worth considering. It would, however, be important to include a measure like the EQ5D if an economic analysis is to be undertaken.

Given the loneliness in this patient group (Ross and Austin, 2013), the addition of a measure of loneliness may be worth considering in a future study.

On a practical note, questionnaires were printed double sided. Often patients only completed one side and had to be contacted and asked to complete the missing page. We would, therefore, recommend printing questionnaires single-sided using large print for this patient group, many of whom struggled to see and write.

22

4.4 The effect of the intervention on study outcomes and trends for further exploration

4.4.1 Effect of the intervention on spiritual wellbeing, anxiety, depression, and health related quality of life. To test for changes in spiritual wellbeing, anxiety, depression and quality of life over time and between groups, a Mixed ANOVA test was performed using changes in life events, circumstances, deterioration in symptoms and medication during the first 2 months as covariates.

4.4.1.1 Spiritual wellbeing (WHO-SRPB): changes over time (within subject) and between groups

The results of the Mixed ANOVA test are shown in Table 11. Sphericity is not assumed, so the Greenhouse-Gesser correction is applied, resulting in F=.632, p=.552, indicating that there was no significant change in spiritual wellbeing over time.

Table 11 Changes in spiritual wellbeing (WHO-SRPB) over time and between groups

SRPB Control (n=15) Experimental (n=16) Total (n=31)Mean SD Mean SD Mean SD

Month 0 13.852 1.832 12.433 3.673 13.120 2.972Month 2 11.832 3.275 12.812 2.938 12.334 3.093Month 4 12.711 3.015 12.423 3.439 12.562 3.190Month 6 12.126 3.252 12.505 2.954 12.321 3.055

There was no significant interaction effect between time and group (F=2.18, p=0.117, Table 12), but the control group showed a marked drop of 1.7 points on the WHO-SRPB between baseline and month 2 after correcting for changes in life events, circumstances, deterioration in symptoms and medication during the first 2 months (Table 12 and Figure 1). Covariates were based on the first two months because that was when a marked change was noted in the control group prior to correcting for the 4 confounding variables.

Table 12 and Figure 1. Mean spiritual wellbeing scores after correcting for changes in life events, circumstances, deterioration in symptoms and medication during the first 2 months

SRPB Control ExperimentalMean 95% CI Mean 95% CI

Month 0 13.976 12.24-15.71 12.316 10.64-13.99Month 2 12.243 10.55-13.93 12.426 10.79-14.06Month 4 12.920 11.15-14.68 12.227 10.52-13.93Month 6 12.696 11.06-14.33 11.970 10.39-13.54

23

Figure 1

Patients who reported a deterioration in their symptoms between baseline and month 2 (n=3) showed a clinically significant drop of 4.8 points in their WHO-SRPB score although this was not statistically significant, the sub-sample being very small. Similar effects were not observed at 4 and 6 months (Table 13).

Table 13. Comparison of spiritual wellbeing scores and deterioration in symptoms between baseline and months 2, 4 and 6

Deterioration in symptoms Month 2 (n=3)

No deterioration in symptoms Month 2 (n=34)

Mean SD Mean SDMonth 0 14.037 1.779 13.291 2.940Month 2 9.259 1.091 12.779 3.099







24

4.4.4.2 Depression (HAD): changes over time (within subject) and between groups

Table 14 shows that there was no significant change in depression scores over time (F=0.545, p=0.653) or between groups (F=0.373, p=0.547).

Table 14: Changes in depression (HAD) over time and between groups

HADDepression

Control (n-=15) Experimental (n=16) Total (n=31)Mean SD Mean SD Mean SD


There was no significant interaction effect between time and group (F=0.743, p=0.530), using changes in life events, circumstances, deterioration in symptoms and medication during the first 2 months as covariates (Table 15 and Figure 2). Although not significant, experimental patients had higher depression scores at month 6 compared with control patients (1.5 points higher) and they also showed a marked increase in depression (+0.9) between month 4-6 compared with control patients who showed a marked decrease (-0.6).

Table 15 and Figure 2. Mean depression scores after correcting for changes in life events, circumstances, deterioration in symptoms and medication during the first 2 months

HAD DepressionControl Experimental

Mean 95% CI Mean 95% CIMonth 0 6.959 5.38-8.53 7.101 5.57-8.62Month 2 7.011 5.16-8.85 7.677 5.89-9.45Month 4 7.069 5.61-8.52 7.060 5.65-8.46Month 6 6.454 4.74-8.16 7.949 6.29-9.60

Figure 2

25

4.4.1.3 Anxiety (HAD): changes over time (within subject) and between groups

Table 16 shows that there was no significant change in anxiety scores over time (F=0.153, p=0.927), between groups (F=1.791, p=0.193).

Table 16: Changes in anxiety (HAD) over time and between groups

HADAnxiety

Control (n=15) Experimental (n=16) Total (n=31)Mean SD Mean SD Mean SD


Table 17 shows that there was no interaction effect between time and group (F=0.959, p=0.416), using changes in life events, circumstances, deterioration in symptoms and medication during the first 2 months as covariates. Although it was not significant, the control group showed a marked increase in anxiety scores from baseline to month 2 (+0.8) and the experimental group showed a marked decrease in anxiety scores from baseline to month 2 (-1.2).

Table 17. Mean anxiety scores after correcting for changes in life events, circumstances, deterioration in symptoms and medication during the first 2 months

HAD Anxiety Control ExperimentalMean 95% CI Mean 95% CI


4.4.1.4 Health related quality of life (EQ-5D-3L VAS and overall score): changes over time (within subject) and between groups

As described earlier (see ‘The study measures’), the EQ-5D-3L has two scores, an overall score (combining the first 5 questions) and a VAS score (perception of health state today). Analyses of both of these scores are reported.

EQ-5D-3L VASTable 18 shows that there was no significant change in EQ-5D-3L VAS scores over time (F=0.583, p=0.628) or between groups (F=0.0001, p=0.991).

26

Table 18: Changes in EQ-5D-3L VAS over time and between groups

EQ-5D-3L VAS

Control (n-15) Experimental (n=16) Total (n=31)Mean SD Mean SD Mean SD


There was no interaction effect between time and group (F=1.342, p=0.267) using changes in life events, circumstances, deterioration in symptoms and medication during the first 2 months as covariates (Table 19, Figure 3). However, from baseline to month 2 the intervention group’s scores increased by 4 points and the control group’s scores decreased by 9 points after correcting for covariates.

Table 19 and Figure 3. Mean EQ-5D-3L VAS scores after correcting for changes in life events, circumstances, deterioration in symptoms and medication during the first 2 months

EQ-5D-3L VAS Control ExperimentalMean 95% CI Mean 95% CI


Figure 3

27

EQ-5D-3L overall Score:Table 20 shows that there was no significant change in EQ-5D-3L overall scores over time (F=0.851, p=0.470) or between groups (F=0.434, p=0.485).

Table 20: Changes in EQ-5D-3L overall score over time and between groups

EQ-5D-3L Score

Control (n-15) Experimental (n=16) Total (n=31)Mean SD Mean SD Mean SD


There was no interaction effect between time and group (F=0.659, p=0.424), using changes in life events, circumstances, deterioration in symptoms and medication during the first 2 months as covariates (Table 21 and Figure 4), however, there was a non-significant drop of 0.85 (after correcting for covariates) among the control group between months 4 and 6 which cannot be immediately explained.

Table 21 and Figure 4. Mean EQ-5D-3L Overall scores after correcting for changes in life events, circumstances, deterioration in symptoms and medication during the first 2 months

EQ-5D-3L Score Control ExperimentalMean 95% CI Mean 95% CI


Figure 4

28

4.4.2 Patient satisfaction with the intervention Of the 20 patients who received the intervention (spiritual support), 18 completed the ‘Satisfaction with Service Questionnaire’ which was returned to the research centre in a sealed envelope to minimise social desirability bias, and 2 died before completion. 14 respondents valued the service ‘an extreme amount’ (5) or ‘very much’ (9), three ‘moderately’ and 1 ‘a little’. They particularly valued the opportunity to talk to someone (13) and to share their faith (2). Fourteen felt that the service should be offered to future heart failure patients (10 ‘completely agree’, 4 ‘agree’) and 4 were ambivalent (‘neither agree nor disagree’). Other comments participants made are listed in Table 22.

Table 22. Other comments made by respondents

Comment No. of timesPositive feeling/feeling good

Great rapport with volunteer

Volunteer was caring

Unable to talk to family so good to talk to someone

Volunteer was a good listener

Volunteer was non-judgemental

Hope funding found to provide service

Will miss visits

Service is good company for elderly who are alone

May benefit some people

3

2

2

2

1

1

1

1

1

1

29

4.4.3 Cost analysis4.4.3.1 Intervention costsThe training costs associated with the intervention over the study period are detailed in Table 23. These include costs of trainers, venue, training materials and volunteer attendance. Research costs were isolated and not included in the costing. While the GP trainers, the training venue and volunteers did not incur a financial cost, from an economic perspective they all have associated opportunity costs. The cost of a Band 3 community-based counsellor was therefore used as a proxy for the volunteers, since this would likely represent the health professional banding required, should this service be delivered in the future by the NHS (or equivalent). Market prices were used for the other two costs (see Table 24 for details and sources).

The cost of training was calculated at £12,027.40, or £601.37 per patient. However, caution must be exercised when considering this cost, given that this is an upfront semi-variable cost. Refresher training is not an anticipated feature of this service (i.e. once volunteers are trained they will not require retraining). If the study had continued for a longer period and had recruited more patients, this would have resulted in a lower cost per patient. A future study will allow for further exploration of this service cost, and for a more accurate representation of training cost per patient.

Table 23: Training costs

PersonnelResource

usageUnit cost Cost

Unit cost source

Comments/Assumptions

GP Trainers 60 hours £85.50 £5,130Study team and PSSRU

2016

14% oncosts added to £300 half day fee (0.5

days = 4 hours) – as per PSSRU guidance

Psychosocial counsellor

3 hours £32 £96 PSSRUCosted as a Band 5

counsellorHeart Start

Band 7 nurse2 £75 £150

PSSRU 2016

Costed as a Band 7 nurse specialist

Training materials

- - £270 Study team -

Overheads (venue) for GP Trainer

sessions

60 hours £6.63 £397.80Various: online

Average of a random selection of church hall

costs in Wales

Volunteer attendees

224 £25 £5,600PSSRU 2016

Costed as a Band 3 community based healthcare worker

(which includes counsellors)

Volunteer out-of-pocket

685 miles £0.56 £383.60 PSSRU 2016

NHS mileage unit costs used

30

travel costsTotal training cost £12,027.40

Total training cost per patient (n=20) £601.37

Using the same proxy for volunteer time, and assuming an average of 12 minutes travel time for each of the 75 visits (in line with PSSRU assumptions for GPs), the total service delivery costs were calculated as £2,125.20 for the study period. This resulted in a mean service delivery cost of £106.26 per patient. Again, this cost is influenced by study design. If patients had received more than four visits, then the costs would have increased at a cost of £27.60 per visit. Likewise, if a future service model delivered fewer visits per patient, this cost would be lower. A future study could possibly involve exploration of the impact of varying the number of visits on cost-effectiveness.

Debriefing sessions were delivered on eight occasions during the study period, by a psychosocial counsellor and a chaplain. If this service were to be rolled-out on a volunteer-delivered basis, these sessions would continue to be provided. However, if the NHS were to deliver the service instead, these sessions would no longer be necessary (i.e. an NHS delivered service would not incur these costs). Only half of the volunteers attended all of the sessions. The total economic cost for these sessions was calculated at £3,411.92, and £170.60 per patient. Again, caution should be exercised when considering this cost. If the number of patients visited were higher, this would reduce the cost per patient.

All of the service delivery and programme costs are detailed in Table 18, together with unit cost sources and any assumptions applied. These reveal a cost of £875.47 per patient. As can be seen, the majority of this cost can be attributed to training. As aforementioned, this cost has the potential to be considerably lower per patient, if the trained volunteers were to deliver to a larger number of patients over a longer time period – which is a distinct possibility within a non-research context.

Table 24: Service delivery and programme costs

Service component

Total costMean cost per

patientSources/Assumptions

Training £12,027.40 £601.37 (see Table 23)

Intervention delivery

£2,125.20 £106.26

£25 per hour Band 3 community based healthcare worker (excluding travel) -

PSSRU 2016. Travel costs added - assumed 12 mins per patient travel time - in line with

GP and district nurse assumptions.

Debriefing £3,411.92 £170.60Psychosocial counsellor costed as a Band 5

counsellorTotal cost of

service£17,509.32 £875.47

4.4.3.2 NHS resource use

31

Table 25 details the unit costs used to calculate NHS resource use costs, together with their source and any assumptions or proxies applied.

The resource use data collected during the duration of this study suggest lower mean costs in the intervention group (as detailed in Table 26). However, non-parametric Mann-Whitney U tests revealed that the differences between groups were not statistically significant. When all time points are compared p=0.718. Again, caution must be exercised when reviewing these results. Typically a maximum cut-off of p=0.05 is used as the basis of accepting the robustness of differences in two groups of data; a p value greater than 0.05 suggests that differences are not significant enough to accept with confidence. However, these statistics can be influenced by sample size; the larger the sample the more robust and reliable the outcomes of the tests. Given the small sample included in this study, these results should be considered informative rather than definitive; since they can assist in informing power calculations for the intended future study, which will formally assess cost-effectiveness.

Table 25: Unit costs used in the analyses

Cost type Unit cost Source Notes/AssumptionsHome visits

Heart failure nurse £38.38 PSSRU 2015

£75 per hour nurse specialist (community) in PSSRU 2015 - not inflated as other similar costs unchanged between 2015 and 2016. Average patient contact time for a practice nurse in PSSRU 2015 is 15.5 minutes – this

duration was assumed for the Heart Failure Nurse. Travel costs added - assumed 12 mins

per patient travel time - in line with GP estimates.

Own or other GP £91.26 PSSRU 2016 and 2015

£3.90 per minute patient contact PSSRU 2016. 11.4min home visit as per 2015 PSSRU

(no info in 2016 version). PSSRU 2015 estimate 12 minutes travel time per home

visit for GPs.

District nurse £31.27 PSSRU 2015

£67 per hour Community/District Nurse in PSSRU 2015. Average patient contact time

for a practice nurse applied, in addition to the above noted travel time assumptions.

Member of the frailty team £292.44 Jones et al,

1999

This can be a visit from a doctor, nurse social worker or occupational therapist. This team

when used in the acute stage is likened to Hospital at Home - as it aims to prevent

admissions to hospital. A proxy of hospital at home has therefore been used, given that

there are no published unit costs for Frailty Teams. Mean daily cost inflated to 2016

value = £292.44, using CPI inflation calculatorOther £45.60 PSSRU 2016

and 2015Calculated average of community-based

health care staff = £38/hour. 1 Hour consultation. Travel costs added (same

32

assumptions as above).

Other: paramedic £181.00NHS Ref

Costs 2015-2016

Cost of ambulance: See and treat or refer

GP surgery visits

Own or another GP £36.00 PSSRU 2016 N/A

Practice nurse £11.11 PSSRU 2016 £43 per hour PSSRU 2016 - assumed 15.5 minute consultation as per PSSRU 2015.

Other £7.41 PSSRU 2016 £24/per hour Band 3 Nursing assistant. Assumed 15.5 minute consultation.

Hospital visitsOutpatients

department (OPD) £134.00 PSSRU 2016 - Weighted average of all outpatient attendances

A&E department £137.64

NHS Reference

Costs 2015-2016

Weighted average of all emergency medicine attendances (excluding visits by patients dead

on arrival)

Medical admissions unit £733.31

NHS Reference

Costs 2015-2016

Costed as day case

Other £134.00 PSSRU 2016 Costed as OPD - as no information provided to indicate what other is

Non-elective Inpatient - SHORT STAY £615.83 NHS

Reference Costs 2015-

2016

These are end-stage heart failure patients. The assumption has been made that

admissions are emergencies.

Non-elective Inpatient - LONG STAY £3,054.14

Non- elective excess bed days £298.41

Antidepressant medication

Sertraline £1.75 BNF 2017

In the absence of medication details, this medication has been assumed – given that it

well suited to patients with heart failure. Also assumed dose of 50mg per day.

Table 26: Resource use (excluding intervention costs)

GroupMean Primary

Care costs(SD)

Mean SecondaryCare costs

(SD)

Mean antidepressant

medication costs (SD)

Mean totalcosts(SD)

Month 2Control Group (n=18) £431.78

(£1,034)£1,446.27(£2,068)

£1.17(£1.58)

£1,879.21(£2,446)

Intervention Group (n=20) £206.95(£302)

£427.55(£962)

£0.53(£1.28)

£635.02(£963)


(£187)£448.06(£1,123)

£0.93(£1.53)

£603.92(£1,162)

33


£1,394(£2,289)

£0.37(£1.10)

£1,711.38(£2,683.53)


(£393)£1,294.13(£2,215)

£1.40(£1.77)

£1,497.32(£2,255)


£1,134.92(£2,341)

£0.39(£1.13)

£1,369.52(£2,369)

All time pointsControl Group £746.26

(£1,504)£2,947.88

(£3,966.58)£3.21

(£4.50)£3,697.35

(£4,231.47)Intervention Group £718.90

(£860)£2,773.28(£3,421)

£1.23(£3.27)

£3,493.41(£4,231)

Incremental mean costs -£27.36 -£174.60 -£1.98 -£203.94

4.4.3.3 Health related quality of life outcomesHealth related quality of life data reveals a higher mean EQ-5D-5L utility score for the intervention group at baseline, and that utility scores declined over time in both groups (Table 27). However, the data also seems to suggest that health related quality of life declined less in relative terms in the intervention group over the study period. Again, Mann-Whitney U test p-values reveal no statistically significant differences between groups at each time point. However, as noted above, these values should be treated with caution given the small sample size.

Table 27: Patient health related quality of life outcomes – all time points

Group n Mean SD p-valueBaseline

Control 18 0.435 0.4011

Intervention 20 0.487 0.3142 months

Control 18 0.417 0.3580.654

Intervention 20 0.487 0.3224 months

Control 17 0.389 0.317 0.594Intervention 19 0.460 0.298

6 monthsControl 15 0.322 0.340 0.202

Intervention 18 0.470 0.351

4.4.3.4 Exploratory economic analyses The estimated incremental cost per QALY for any intervention can be compared with the willingness to pay (WTP) threshold of £20,000 - £30,000 per QALY gained currently used by NICE, to explore the potential of whether or not it is likely to be ‘cost effective’ and offer value for money (NICE, 2013).

34

The results of exploratory deterministic cost effectiveness analyses are record in Table 28. It must again be emphasised that this is not a formal assessment, as this is merely a feasibility study. These tentative results suggest that the intervention would not meet the aforementioned WTP threshold. Probabilistic sensitivity analyses, using 1000 iterations, similarly suggest that the probabilities of being cost effective at WTP threshold of £20,000 and £30,000 are 32% and 47% respectively. However, these results are representative of the research study, rather than of the service should it be rolled out.

Table 28: Exploratory cost effectiveness analyses

Control Intervention DifferenceCost per patient £3,697.35 £4,373.02 £676QALYs 0.095 0.115 0.020ICER (£/QALY gained) £33,381.25NB*ICER does not compute using the figures in the table due to rounding.

4.4.3.5 Exploratory sensitivity analysesSensitivity analyses can be useful in exploring the sensitivity of the results to changes in key parameters. Notably, one of the key drivers for this analysis was training costs, and these are significantly influenced by the small sample size of this study. If training costs are reduced from £601.37 to £300.69 to crudely estimate the cost effectiveness of the service had the intervention been delivered to twice as many patients, then the ICER becomes £18,526.24 per QALY gained, and falls within the NICE threshold. Similarly, if the debriefing costs are deducted to get a feel for the cost effectiveness should the NHS deliver this service in the future, the ICER again reduces to a level within the threshold range to £24,952.94 per QALY gained. These analyses are simplistic and exploratory only. However, they are insightful in terms of identifying which factors are likely to drive cost effectiveness.

5. DISCUSSIONThe study provided useful information for planning of future studies with this patient group in a number of ways.

Fewer patients were recruited than anticipated and recruitment and data collection took much longer than expected for a complex range of reasons related to changes in the illness trajectory of this group (patients staying well for longer before reaching end-stage by which time other co-morbidities were taking their toll), high co-morbidity and mortality in this group, as well as staffing issues coupled with increased workloads. We had estimated that 65 patients would be recruited over 9 months with data collection taking 15 months. In reality it took 18 months to recruit 47 patients (of which 38 took part) and 2 years to collect the data. The uptake rate of 35% is better than in our previous study (24%) possibly because of better staff engagement and commitment with the intervention being offered within existing hospital provision. The problem of poor uptake and high drop-out is recognised by NIHR as a major challenge in conducting research with very ill patients but is to be expected if the right patients have been recruited (Evans et al, 2013).

35

The follow-up and input from staff to maximise recruitment and to gather the study data was considerable and included: exhausting every means possible to identify recruits (this took up a huge amount of nurses’ time whilst carrying heavy workloads); many phone calls to remind respondents to complete questionnaires; re-sending lost questionnaires; following up incomplete questionnaires, re-arranging visits because of illness or patients just not feeling well enough on the day of the appointment (reflects the fluctuating nature of heart failure). This burden was shared by the clinical and research teams but required the additional appointment of an administrator. We would, therefore, recommend the inclusion of an administrator and a research nurse in a future study.

The commitment of volunteers was considerable, not only in delivering the intervention (about an hour every 2 months), but in meeting all of the other requirements of their role. They underwent over 20 hours training in interviewing, lone working and becoming familiar with the subject area. They had to have DBS checks, health board research passports and undertake additional CPR training. They also undertook a lot of travelling to carry out the visits as patients were located right across the health board (some journeys were around 30 miles round trip). Having a car proved necessary and was partly the reason for two volunteers dropping out. It became necessary to provide debriefing opportunities for volunteers on a quarterly basis where challenges they were facing could be discussed e.g. how to end a visit when a patient wanted it to continue or when the study was nearing an end, how to deal with presence of a third party during an interview, maintaining personal and professional boundaries. The need for this debriefing was not anticipated but was important for volunteer support and also for ensuring parity in delivery of the intervention and should be built into future study planning. Despite the high commitment required of the volunteers they found the experience positive and valuable and expressed regret at the study ending.

The study measures were, on the whole, suitable for this patient group, but we would recommend the addition of another quality of life measure, such as the Kansas City Cardiomyopathy Questionnaire (KCCQ-12), which captures additional aspects of life important to this particular patient group, such as ‘enjoyment of life’ and ‘lifestyle’. It may also be worth trialling a shorter less complex spiritual wellbeing measure and including a measure of loneliness as discussed above. We would also advise single sided printing of questionnaires in large print (font size 14 minimum).

We have provided much needed data on variation in scores on the WHO RFPB tool (WHO, 2002) with an intervention in end stage heart failure. We reported domain spirituality scores averaging 12-13 and ranging from 4.9-18.9, with the biggest change within any individual of 7 points. Differences in scores between control and intervention groups ranged from 0.3 – 1.4 which is in keeping with those anticipated using figures from Boreo et al. (2005) and Hsu et al. (2009). This data should be useful to future researchers looking to use this measure with this patient group in calculating what effect size they can expect and the sample size needed to produce significant results.

Even with such a small sample, our results reinforce findings in the literature (e.g. McClain et al, 2003; Koenig, 2009) that spiritual wellbeing is negatively correlated with depression and anxiety and seems to mitigate against it. The positive association between spiritual

36

wellbeing and quality of life that is suggested in the literature (e.g. by Walsh, et al. 2002; Tarakeshwar, et al. 2006; Balboni, et al .2007) was less conclusive in our analysis. This may be because of our small sample size but it may also be because different research studies have used different measures of both of these complex phenomena (spiritual wellbeing and quality of life) making comparability difficult.

No significant effect of spiritual support on spiritual wellbeing (WHO SRPB), anxiety/ depression (HAD) or health related quality of life (EQ5D VAS and overall score) was found in this feasibility study. The study was limited by its small sample size (n=47) so it is possible that effects may exist but would need a bigger sample to become evident. Some interesting trends were, however, identified which are worthy of further exploration.

First, the control group showed a marked worsening in SWB (-1.7 points), anxiety (+0.8 points) and Health Related QOL (-8.0 points on VAS) between months 0 and 2, whilst the experimental group showed a marked improvement in QOL (+4.0 points on VAS) and anxiety (-1.2 points) at this time. This raises the question of whether there is a group allocation effect (effect of knowing which group you are in). For example, the worsening of scores in study measures in the control group in the first 2 months of the study may have been due to disappointment at not having visits. Some patients had verbally expressed their disappointment at being allocated to the control group and some had dropped out because of it. The control group were visited only at baseline to obtain consent and be given the questionnaires. It is interesting that the drop in scores occurred between that visit and the next data collection which was completed alone. Similarly, anticipation of visits may have caused the intervention group to feel less anxious and to rate their QOL more highly. In order to test for a possible group allocation effect we would recommend that study measures are completed before patients are allocated to groups.

Second, depression scores increased in experimental patients (+0.9) and decreased in control patients (-0.6) at the end of the study (months 4-6) raising the question of whether anticipation of withdrawal of the intervention had a negative impact on intervention patients’ mood. Some patients said how much they had enjoyed the visits and did not wish them to end (additional comments Table 16). In order to test for any effect of withdrawal of the intervention we would recommend that measures continue to be completed after the intervention has stopped.

Third, our previous study had identified loneliness was a major issue for this patient group (Ross and Austin, 2013) and was a similar theme running through intervention patients’ comments on the Satisfaction with Service Questionnaire, so it may be worth considering inclusion of a measure of loneliness in a subsequent study. This is particularly relevant given the recognition by NICE of loneliness as ‘a serious public health issue’ affecting elderly people and calling for volunteers to provide home visiting and befriending programmes, similar to the service we trialled in this study (NICE, 2016).

Fourth, although the sub-sample was very small (n=3), SWB dropped by 4.8 points in patients reporting deterioration of symptoms between months 0-2 raising the question of whether there could be a link between SWB and symptoms. We would recommend, therefore, inclusion of a symptom severity measure at all time points, including baseline, as

37

well as before group allocation and after withdrawal of the intervention to enable further investigation of this possible link in a future study.

Patients valued the spiritual support service. Although not statistically significant, there was indication that these patients incurred lower costs in NHS resource use and showed less decline in quality of life compared with control patients, two trends which are worth investigating further, especially as a recent state of the art review found palliative care reduced healthcare costs and improved quality of life (Mcllvennan and Allen, 2016). Although the spiritual support service was not cost effective when delivered to such a small group of patients as part of this research study using the NICE cost effectiveness criteria, it may prove to be cost effective (by reducing per patient training and debriefing costs) if offered to more patients over a longer period of time, as would be the case if it became part of service delivery. This possibility is also worthy of further exploration. Despite the limitations noted above, these cost analyses have provided valuable guidance for the design of a future economic evaluation, particularly in terms of identifying observed trends and guiding future data collection. It will be particularly important in a future study to fully explore training costs, given their influence on costs and the relationship between costs and effects. Related to this, it will be necessary to explore staff turn-over, and how training costs are most likely to be distributed over time. It may also be beneficial to explore alternative proxies for volunteer time, given that there is no one agreed approach to this amongst health economists.

The study was undoubtedly limited by its small sample size. The co-morbidity and morbidity in this vulnerable patient group means the potential sample size is likely to be small, even when the entire population is targeted. The resulting sample size may, therefore, introduce an element of bias in that the views and experiences of the intended population may not be fully captured. The NIHR (Evans, et al 2014) recognises this limitation and takes the view that a study involving patient groups such as this with limited attrition due to death or worsening illness could be criticised for recruiting the wrong patients. Every qualitative study has an element of researcher bias. This was minimised by providing volunteers with the same training and with a protocol to follow during patient visits. Role play and quarterly debriefing sessions provided opportunity for variation between individual responses to be standardised as far as possible.

6 CONCLUSION AND RECOMMENDATIONSReturning to the overall aim of this feasibility study, we have provided valuable information to inform the accurate planning, costing and design of a future NHS randomised study to investigate the effects of offering spiritual support to end stage heart failure patients on a range of outcomes including spiritual wellbeing, anxiety, depression and quality of life. We have also identified interesting trends worthy of further investigation. Table 29 summarises our recommendations.

Table 29: Summary of recommendations for future study planning and trends for further exploration

1. Repetition of the study with a larger sample to identify the effect of spiritual support on

38

spiritual wellbeing, anxiety, depression and quality of life.2. Inclusion of additional/other measures.

Inclusion of a measure of loneliness inclusion of an additional QOL measure which captures other aspects important to this

patient group e.g. Kansas City Cardiomyopathy Questionnaire (KCCQ-12) Inclusion of a measure of symptom severity at all time points including before allocation

to groups and at baseline Trial the use of a shorter less complex measure of spiritual wellbeing.

3. Practicalities Accurately estimate likely number of recruits using up-to-date figures and conduct this

exercise again if there is a delay in the study starting. Allow adequate time for data collection Include administrative and research nurse support Include time and staff resources for training and debriefing of volunteers Where possible target retired health care professionals as volunteers Print questionnaires single sided with minimum 14 font type. Provide training for nurses in having end of life conversations

4. Exploration of trends namely: If there is a group allocation effect. We would recommend taking study measures before

patients are allocated to groups to determine if knowledge of allocation to control group has a negative effect and if allocation to intervention group has a positive effect.

If there is a lasting negative effect (in this study, increase in depression of almost 1 point) on patients in receipt of the intervention. We recommend taking study measures after the intervention has stopped to determine if knowledge of withdrawal of the intervention (visits) has a lasting negative effect on the experimental group.

If spiritual wellbeing and symptoms are linked. We would recommend inclusion of a symptom severity measure at all time points including baseline and after cessation of the intervention.

If spiritual well-being and quality of life are linked, possibly using measures included in previous studies to enable comparison

If patients receiving spiritual support incur lower mean costs in nhs resource use compared with controls

If health related quality of life declines less in patients receiving spiritual support compared with controls

If the intervention is cost effective when offered within routine care including full exploration of volunteer training costs and staff turn-over.

39

7. REFERENCESAustin, J., Williams, W.R., Ross, L., Moseley, L., Hutchinson, S. (2005) Randomised controlled trial of cardiac rehabilitation in elderly patients with heart failure. European Journal of Heart Failure, 7, pp. 411-417.

Balboni, T.A.1., Vanderwerker, L.C., Block, S.D., Paulk, M.E., Lathan, C.S., Peteet, J.R., Prigerson, H.G. (2007). Religiousness and spiritual support among advanced cancer patients and associations with end of life treatment preferences and quality of life. Journal of Clinical Oncology, 26: pp. 555-560.

Basta L (2004) End of life medical treatment of older cardiac patients. American Journal of Gerontological Cardiology, 13; 313-15

Boero ME, Caviglia ML, Monteverdi R, Braida V, Fabello M, Zorzella LM (2005) Spirituality of health workers: a descriptive study. International Journal of Nursing Studies, 42, 915-921.

Cockell N and McSherry W (2012) Spiritual care in nursing: an overview of published international research. Journal of Nursing Management, 20, 958-969

Curtis L., Burns, A. (2015) PSSRU: Unit Costs of Health and Social Care 2016. Accessed on 13/4/2017 at: http://www.pssru.ac.uk/project-pages/unit-costs/2015/

Curtis L., Burns, A. (2016) PSSRU: Unit Costs of Health and Social Care 2016. Accessed on 13/4/2017 at: http://www.pssru.ac.uk/project-pages/unit-costs/2016/

Department of Health (2004) Coronary Heart Disease Collaborative. Supportive and palliative care for advanced heart failure. Leicester: NHS Modernisation Agency

Department of Health (2009) Religion or belief. A practical guide for the NHS. London: Crown

Department of Health (2015) The NHS Constitution for England. London: Crown

Department of Health (2015) One chance to get it right. London: Crown

Department of Health (2016) NHS reference costs 2015 to 2016. National schedule of reference costs: the main schedule. Accessed on 9/3/2017 at: https://www.gov.uk/government/publications/nhs-reference-costs-2015-to-2016

Equality Act 2010. [online] Available at: http://www.legislation.gov.uk/ukpga/2010/15/pdfs/ukpga_20100015_en.pdf [Accessed 6 June 2016].

European Commission (2010) EUROPE 2020. A European strategy for smart, sustainable and inclusive growth. Brussels: European Commission

40

http://www.legislation.gov.uk/ukpga/2010/15/pdfs/ukpga_20100015_en.pdf

https://www.gov.uk/government/publications/nhs-reference-costs-2015-to-2016

http://www.pssru.ac.uk/project-pages/unit-costs/2016/

http://www.pssru.ac.uk/project-pages/unit-costs/2015/

Evans, C; Stone, KA; Manthorpe J; Higginson IJ (2013) MRC guidance on developing and evaluating complex interventions: application to research on palliative and end of life care. London : National Institute for Health Research School of Social Care Research

Goebel, JR; Doering, LV; Lorenz, KA; et al (2009) Caring for special populations: total pain theory in advanced heart failure: applications to research and practice. Nursing Forum, 44, 175-185.

Gwent Healthcare NHS Trust (2006) Clinical Futures Strategy. www.wales.nhs.uk/sites3

Hsu P, Krageloh CU, Sheperd D, Billington R (2009) Religion/spirituality and quality of life of international tertiary students in New Zealand: an exploratory study. Mental Health, Religion & Culture, 12, 4, 385-399

Lehman, R. (2006). ‘Heart failure and palliative care a team approach’, in Johnson, M. and Lehman, R. (eds,) The need for palliative care in heart failure. Oxford and Seattle: Radcliffe PublishingJoint Formulary Committee (2017) British National Formulary, March 2017. London: BMJ Group and Pharmaceutical Press. Accessed 9/3/2017 at: http://www.medicinescomplete.com

Kansas City Cardiomyopathy Questionnaire (KCCQ-12) http://cvoutcomes.org/pages/3214 accessed 7 June 2017

Koenig, HG. (2009) Research on religion, spirituality and mental health: a review. Canadian Journal of Psychiatry, 54, 283-291.

Koenig, H; King, D; Carson, V (2012) Handbook of religion and health. New York, Oxford University Press

McClain, C.S; Rosenfeld, B. and Breltbart, W. (2003). Effect of spiritual well-being on end-of-life despair in terminally ill cancer patients. Lancet; 361:1603-1607.

McIlvennan, C and Allen, L (2016) Palliative care in patients with heart failure. British Medical Journal, 352:i1010 doi: 10.1136/bmj.j1010

McSherry, W. and Ross, L. (2012) ‘Nursing’. In: Cobb, M; Rumbold, B; and Puchalski, C (eds). Oxford Textbook of Spirituality in Healthcare. Oxford; Oxford University Press.

Murray, S.A.; Kendall, M.; Boyd, K. et al (2004) Exploring the spiritual needs of people dying of lung cancer or heart failure: a prospective qualitative interview study of patients and their carers. Palliative Medicine, 18, 39-45

Murray, S; Kendall, M; Grant, E et al (2007) Patters of social, psychological and spiritual decline toward the end of life in lung cancer and heart failure. Journal of Pain and Symptom Management, 34, 4, 393-402

41

http://cvoutcomes.org/pages/3214

http://www.medicinescomplete.com/

http://www.wales.nhs.uk/sites3

Murray, S; Kendall, M; Boyd, K et al (2010) Archetypal trajectories of social, psychological and spiritual wellbeing and distress in family care givers of patients with lung cancer: secondary analysis of serial qualitative interviews. British Medical Journal, 304: c2581 doi:10.1136/bmj.c2581

National Institute for Health and Care Excellence (2004) Guidance on Cancer Service. Improving supportive and palliative care for adults with cancer. London, NICE

National Institute for Health and Care Excellence (2013) Guide to the Methods of Technology Appraisal 2013. Accessed 9/4/2017 at: https://www.nice.org.uk/process/pmg9/resources/guide-to-the-methods-of-technology-appraisal-2013-pdf-2007975843781

National Institute for Health and Care Excellence (2015) Care of dying adults in the last days of life. https://www.nice.org.uk/guidance/ng31 accessed 7 June 2017

National Institute for Health and Care Excellence (2016) Guidance on loneliness in older people https://www.nice.org.uk/news/article/nice-calls-on-third-sector-and-public-bodies-to-work-together-to-prevent-loneliness-in-older-people accessed 6 June 2017

NHS Education for Scotland (2009) Spiritual care matters. An Introductory resource for all NHS Scotland Staff. Edinburgh; NHS Education for Scotland

National Institute for Cardiovascular Outcomes Research (2012). National Heart Failure Audit April 2010-March 2011. London; University College London

Nursing & Midwifery Council (2010). Standards for pre-registration nursing education. London: NMC

O’Connell, K and Skevington, M (2010) Spiritual, religious and personal beliefs are important and distinctive to assessing quality of life in health: A comparison of theoretical models. British Journal of Health Psychology, 15, 729-748.

Park, C.L. (2008) Estimated longevity and changes in spirituality in the context of advanced congestive heart failure. Palliative and Supportive Care, 1, 3-11.

Quality Assurance Agency for Higher Education (2001) Benchmark statement: health care programmes. Gloucester, QAAHE

Ross, L. A. (1997) Elderly patients’ perceptions of their spiritual needs and care: a pilot study. Journal of Advanced Nursing, 26, 710-715.

Ross, L. A (2006) Spiritual care in nursing: an overview of the research to date. Journal of Clinical Nursing, 15, 7, 852-862

42

https://www.nice.org.uk/news/article/nice-calls-on-third-sector-and-public-bodies-to-work-together-to-prevent-loneliness-in-older-people

https://www.nice.org.uk/news/article/nice-calls-on-third-sector-and-public-bodies-to-work-together-to-prevent-loneliness-in-older-people

https://www.nice.org.uk/guidance/ng31

https://www.nice.org.uk/process/pmg9/resources/guide-to-the-methods-of-technology-appraisal-2013-pdf-2007975843781

https://www.nice.org.uk/process/pmg9/resources/guide-to-the-methods-of-technology-appraisal-2013-pdf-2007975843781

Ross, L and Austin, J (2013). Spiritual needs and spiritual support preferences of people with end stage heart failure and their carers: Implications for nurse managers. Journal of Nursing Management, 23, 1, 87-95

Royal College of Nursing (2011) Spirituality in nursing care: a pocket guide. www.rcn.org.ukAccessed 27/7/17

Royal College of Nursing (2012) Spirituality in nursing care on-line resource, http://www.rcn.org.uk/development/practice/spirituality/about_spirituality_in_nursing_care. Accessed 27/7/17

Saunders, C. (1996) Into the valley of the shadow of death: a personal therapeutic journey. British Medical Journal, 313, 1599-1601.

Scottish Government (2009). Spiritual care and chaplaincy. Edinburgh, The Stationery Office

The EuroQol Group (1990). EuroQol-a new facility for the measurement of health-related quality of life. Health Policy 16, 3, 199-208.

Tarakeshwar, N; Vanderwerker, L.C; Paulk, E et al. (2006) Religious coping is associated with the quality of life of patients with advanced cancer. Journal of Palliative Medicine, 9, 646-657.

Thorn, J.C.; Coast, J; Cohen, D et al. (2013) Resource-use measurement based on patient recall: issues and challenges for economic evaluation. Applied Health Economics and Health Policy, 11, 3, 155-61.

van Reenan, M and Oppe, M (2015) EQ-5D-3L User Guide: Basic information on how to use the EQ-5D-3L instrument. Version 5.1. Rotterdam, EuroQol Research Foundation: 1-25. Accessed on 08/06/2017 at: https://euroqol.org/wp-content/uploads/2016/09/EQ-5D-3L_UserGuide_2015.pdf

Wales Audit Office (2008) The management of chronic conditions by NHS Wales. Cardiff: Auditor General for Wales Walsh, K; King, M; Jones, L (2002). Spiritual beliefs may affect outcome of bereavement: prospective study. British Medical Journal, 324, 1551-1554.

Welsh Government (2005) Designed for Life. A world class health service for Wales. http://new.wales.gov.uk/topics/health/health_social_care_strategies

Welsh Government (2010) Standards for spiritual care services in the NHS in Wales 2010. Cardiff: Crown

Welsh Government (2015) Health and Care Standards. Cardiff: Crown

43

https://euroqol.org/wp-content/uploads/2016/09/EQ-5D-3L_UserGuide_2015.pdf

https://euroqol.org/wp-content/uploads/2016/09/EQ-5D-3L_UserGuide_2015.pdf

http://www.rcn.org.uk/

Welsh Government (2017) Heart Conditions Delivery Plan. Accessed 18/7/17 http://gov.wales/topics/health/nhswales/plans/heart_plan/?lang=en

World Health Organisation (2002) WHOQOL-SRPB Field Test Instrument. Geneva: WHO

World Health Organisation. Cancer. Palliative care, www.who.int/cancer/palliative/en/ accessed 7/6/17.

Zigmond, A.S and Snaith, R.P (1983). The hospital Anxiety and Depression Scale. Acta Psychiatrica Scandinavica, 67, 6, 361–370

44

8. APPENDIX: Dissemination

Copies of this report are available in the following libraries:

•The British Library•The Bodleian Library, Oxford•The University Library, Cambridge•The National Library of Scotland, Edinburgh•The Library of Trinity College, Dublin•The National Library of Wales, Aberystwyth.•The University of South Wales

Dissemination as of July 2017Other papers are currently being prepared for publication and for presentation at other conferences.

Conference papers.

Ross L & Austin J (2016) Spiritual support in end stage heart failure: research in progress. 4th International Conference of the British Association for the study of Spirituality. Can spirituality transform our world? New frontiers in understanding and exploring contemporary spiritualties. 23-26 May, Chancellors Hotel & Conference centre, Manchester.

Austin J (2016) Keynote address at the Annual British Association of Cardiac Rehabilitation Conference 2016. ‘Game changing – using research in clinical practice to improve patient care’, Cardiff, October

Ross L & Austin A (2015) Spiritual support in end stage heart failure. British Association for Cardiovascular Prevention and Rehabilitation Annual Conference 2015, The Palace Hotel Manchester, 1-2 October 2015 (Invited) (Paper rated highest of conference).

Austin J (2015) Keynote address at Aneurin Bevan University Health Board Research & Development Conference 2015. ‘Engaging with patients, clinicians and academics to improve care’. September.

Posters*This poster is shown overleaf.

*Linda Ross, Jackie Austin, Paul Jarvis, Sara Pickett (2017). Spiritual support in end stage heart failure (ESHF): a randomised controlled feasibility study. Aneurin Bevan University Health Board Research & Development Conference 2017. ‘Research – a core activity’. Christchurch Centre, Newport. 20 June 2017.

45

*Linda Ross, Jackie Austin, Paul Jarvis, Sara Pickett (2017). Spiritual support in end stage heart failure (ESHF): a randomised controlled feasibility study. Erasmus+ EPICC Project Multiplier event. 19-20 April 2017. Staffordshire University.

*Linda Ross, Jackie Austin, Paul Jarvis, Sara Pickett (2017). Spiritual support in end stage heart failure (ESHF): a randomised controlled feasibility study. Faculty of Life Sciences & Education Research Conference 2017. 12 April. Conference Centre, University of South Wales, Pontypridd.

Ross, LA, Austin, J, Jarvis P, Alam F, Pickett S (2016) Spiritual needs and spiritual support in end stage heart failure: 2 studies. Sharing Information to Drive Innovation - All Wales Review of Cardiac Services Day. 16th November. SWALEC Stadium, Cardiff with VC Link to Venue Cymru, Llandudno

Ross, LA, Austin, Jarvis P, Alam F, Pickett S (2016) Spiritual needs and spiritual support in end stage heart failure: 2 studies. 23 March. FLSE Spring Research Conference, Conference Centre, USW.

46

pure.southwales.ac.uk€¦ · web viewreport. spiritual support in end stage heart failure: a...

Documents