Introduction

Chronic obstructive pulmonary disease (COPD) is a common chronic disease which causes significant morbidity and mortality. The best available data suggest that there are approximately 900,000 patients diagnosed with COPD in England and Wales1 and, allowing for known levels of under-diagnosis, the true number is likely to be around two million. Over 25,000 people died of COPD in the UK in 2007,2 representing around 5% of all deaths. Mortality from COPD in the UK is among the worst in Europe,3 and audit data from the UK indicate a mortality rate of 13.9% within 90 days of an admission for an exacerbation.4

There is a growing recognition internationally of the need for patients with advanced COPD to receive education about prognosis, end-of-life care and advance care planning (ACP),58 and such education is a neglected aspect of the management of non-malignant respiratory disease.912 The End of Life Care Strategy13 for England advocates that all patients with advanced, progressive, incurable illness should be given the opportunity to participate in ACP, and the recent Department of Health Strategy Document for COPD emphasises end-of-life care and structured hospital admission and specialist intervention.14 However, the reality is that few such patients participate in ACP.

Clinicians identify numerous barriers to this communication, such as patients being too old or too young, too well or too ill, lack of time and skills, and being unclear whose role it is.15,16 Patient barriers reported include the avoidance of difficult issues such as discussions about life and death17 and passive acceptance of COPD as part of normal ageing.18 A proposed way forward to aid discussions about ACP has been the identification of key transitions in end-of-life care.19,20 Such a transition in COPD may be admission to hospital for an exacerbation of COPD.18 Patients with COPD are subject to frequent exacerbations of the condition usually due to infection. Although some exacerbations can be managed in the community with adequate clinical and social support, a large number of cases are admitted to acute hospital beds. One in eight (130,000) admissions to hospital in the UK per year are related to COPD,6 with around 30% of patients admitted with COPD for the first time being re-admitted within 3 months.21

Admissions to hospital have been seen as missed opportunities for disease prevention.22 We wished to examine whether or not an admission to hospital for an exacerbation of COPD was an opportunity to discuss issues concerning resuscitation, ventilation, and ACP. We therefore designed a qualitative study to examine events leading to the admission, recall of discussions about ACP while in hospital, and the views of people with COPD and their carers about the context in which it is appropriate for healthcare professionals to discuss ACP.

Methods

The study population comprised patients with an admission for an exacerbation of COPD in the previous year to the local District General Hospital (Royal Devon and Exeter Hospital) during 2009. Cases were drawn from two general practices in Devon (a semi-rural practice of 16,500 and a city-based practice of 30,700 patients). Admissions were either identified from discharge data or from cases referred direct to the research team by their general practitioner (GP). Each potential case was reviewed by the GP with exclusion criteria including severe cognitive impairment, non-English speaking, and terminal illness from any cause at the GP's discretion. Patients were invited via a letter signed by their own GP with the offer of an interview to be conducted face-to-face or over the telephone. Patients were free to have a family member or friend participate in the interview. Ethical approval was obtained from the Exeter Research Ethics Committee. The interviews were conducted by an experienced non-clinical qualitative researcher. All interviews took place in the patients' homes.

An interview schedule (see Appendix 1 available online at www.thepcrj.org) was developed following a review of the research, practice, and policy literature and discussion with key informants. These informants included a respiratory physician, a patient with severe COPD and his wife, a professor of palliative care with a specific interest in COPD, a physiotherapist running a pulmonary rehabilitation course, and GP colleagues. The schedule was designed to explore events before, during, and after an admission. Patients were then asked if they were happy to discuss issues concerning death and dying and planning for the future. If in agreement, they were asked whether or not resuscitation (what to do if you stop breathing or your heart stopped), ventilation (helping you with your breathing either using a mask or having a tube down your windpipe), or ACP (planning for the future in case a similar situation occurs) had been discussed during the admission and on their views on the circumstances under which ACP would be discussed.

The audiotapes from each interview were transcribed verbatim and field notes regarding participant interactions were added where appropriate. Each transcript was read by two of the authors (DS and CS) and core themes were identified. These themes were further analysed by DS using a simple qualitative content analysis23 and constant comparison methods.24 This organised the data into a set of concepts and higher level themes and categories which were then presented to the research group for further refinement.

Results

In total, 38 patients were invited by letter and 18 agreed to be interviewed. Sixteen patients (12 men and four women) aged 58–90 years were interviewed; three had mild disease, six had moderate disease, six had severe disease and one had very severe disease. Two men with severe disease who had agreed to take part died before their interview could take place. One participant was a non-smoker, one a current smoker, and 14 were ex-smokers (Table 1).

Table 1 Patient details

Seven patients (four men and three women) aged 53–85 years declined by letter. Two had moderate disease, two had severe disease, and in three cases the severity of disease was not known. Four of these patients were ex-smokers and three current smokers. Two of these patients also died before the end of the study.

Four of the participants were interviewed in the presence of their spouse and full-time carer and one with their daughter who was their part-time carer. For eight it was their first admission with an exacerbation of COPD and the other eight had had at least one previous admission.

No reply to the invitation was received from 13 patients. No further attempt to contact them was made and their details were not sought.

Was ACP discussed in hospital?

No patients recalled discussion in hospital about issues of resuscitation, the possibility of being ventilated, and planning for future exacerbations. One patient had a directive kept at home asking not to be resuscitated. This was composed with the help of a community matron; however, the form did not accompany the patient to the hospital.

It's in the yellow book, there's a form in there for that purpose, no resuscitation … that's supposed to go with the person… but not very often do they (the hospital doctors) ask for it. [Patient 14]

Possible reasons for lack of discussion about ACP emerged in the following theme.

Hospital admission and discharge: chaotic and too ill to engage

The majority of patients (14 of 16) were admitted as emergencies by ambulance. The process of admission was recalled as being hurried with a minimum of discussion with the ambulance crew or attending GP.

I was seen by a doctor as far as I can remember and pushed into hospital. [Patient 8]

No chance to think, I was whipped in and that was it — “you're going”. [Patient 3]

We were whizzed straight into the Medical Emergency or surgical place … I couldn't tell you. [Patient 3]

The actual inpatient stay was seen as chaotic by most patients with a lack of recall about where care took place and by whom. This was compounded by patients being extremely ill, some experienced hallucinations, and many had no recall of the events of the first few days of the admission.

On reflection, patients felt that the admission had benefited them by providing reassurance and the relief of anxiety. Plans for discharge were another variable feature of the admission with some patients receiving follow-up instructions and others seemingly offered no specific follow-up (see Box 1).

Attitudes to advance care planning

All patients agreed to discuss ACP and related issues such as resuscitation, ventilation, and planning for future exacerbations. For many patients the emotional impact of discussing such concepts was too difficult and led to avoidance of decisions.

It's a subject I would prefer not to have to think about greatly. [Patient 10]

For some, ACP was clearly a new concept and on reflection was seen as important and discussions moved to a more problem-focused basis.

It's going to make me think… it's something I wouldn't have thought of. Interesting subject. [Patient 9]

It was also viewed as a difficult topic about which it might be hard to make concrete decisions. In terms of resuscitation, the view was frequently expressed that only if a successful outcome could be guaranteed would they consider the option. This position led to the idea of resuscitation being a medical decision to be made by clinicians.

Others found the issues concerning resuscitation, ventilation, and future episodes of illness difficult to discuss with the attitude that, if ‘it’ (cardiopulmonary arrest) occurs, then “let's hope it's quick”. Joint interviews revealed different agendas between patient and spouse at times, especially regarding the future course of the illness and managing further exacerbations (see Box 2).

Who to talk to — someone you know or someone who knows?

In answer to the question who should discuss ACP, there was a desire for someone they know and who knew them. Someone with a good knowledge of the condition was also desirable. This usually translated to that person being their GP. Some patients envisaged a discussion with their family being present to help the decision process (see Box 3).

A consultant or specialist nurse was seen as someone with the clinical knowledge but not necessarily the personal relationship. One person suggested a group setting might be a useful forum.

A discussion group with (consultant) and the respiratory staff there so that they could put things over and not hold anything back. [Patient 11, page 10]

Discussion

Main findings

From the descriptions of the hospital admission it is clear that most patients saw this as chaotic, bewildering, and lacking in continuity. In addition, being extremely ill made decision-making and recall of events very difficult. In the light of these findings, it was perhaps not surprising that none of the patients recalled issues about end-of-life care and their personal wishes being discussed with hospital staff during their admission.

The arrangement of suitable follow-up, either in primary or secondary care, varied from being haphazard to reasonably planned but, again, none of the subjects recalled discussions about end-of-life care resulting as a consequence of the admission. When asked directly by the interviewer, all participants were willing to talk about end-of-life issues such as resuscitation, ventilation, and what the future might hold. Responses to whether or not such things should be discussed fell into two broad categories. There were those patients who revealed an emotion-focused style of coping with the concept of death, where negative emotions were avoided by avoidance of the topic. Other participants, who adopted a more problem-focused style, felt such conversations should occur and were a welcome opportunity for them and their families to deal with difficult decisions.25 There was evidence that some patients — after an initial emotion-focused response — would welcome a more problem-focused style if supported by an empathetic health professional.

Most patients favoured their own GP as the person best placed to talk to them about end-of-life issues. Substantial value was placed on knowing the doctor, being known by them, and a relationship built on trust. The setting of such discussion varied, with most favouring the home or GP surgery in the period after admission. For some, a person with authority and a good knowledge base (consultant or specialist nurse) would also be a suitable person with whom to discuss such matters. Family involvement in such discussion was seen as beneficial by some participants.

The desire for resuscitation only to be attempted if a good outcome could be guaranteed, with a sense of giving responsibility over to the clinicians involved, was expressed a number of times.

Strengths and limitations of this study

This study was based on admissions to a single average-sized teaching hospital with well organised respiratory medicine services and outpatient care. It is likely that these findings would be reflected in other hospitals in the UK. None of the participants was from an ethnic minority, although the cohort did represent a varied demographic group. Patients were recruited from two practices that operated personal lists of patients and hence provided continuity of primary care. Patient experience may be different in practices with pooled lists where a patient may not have an identified usual GP. The number of participants and response rate was appropriate for a qualitative study of this nature and saturation of data was achieved. This was when analysis of new transcripts did not produce any new themes or categories and the content confirmed the findings from previous transcripts.

At the time of the study, patients were unlikely to have had contact with a community matron.

The purpose of this study was not to check the accuracy of recall but whether it was recalled at all, and therefore if it was an appropriate time for such discussions. The group considered triangulation by review of the hospital notes, but personal experience (DH) indicated that many of the discussions about ceilings of treatment and decisions regarding acute resuscitation are made soon after admission and hence, although it is likely that a resuscitation status would be documented, it would not be a reliable source of knowing what had been discussed.

Interpretation of findings in relation to previously published work

ACP has been shown to produce benefits to patients and carers,26 yet the implementation of ACP is difficult. The barriers identified by clinicians have been well documented15,17,27 and may be overcome with training and education.26,28 More recently, patient factors have been identified, with a Dutch study showing that patients with end-stage COPD considered their limitations as normal and did not realise there was anything that could improve their situation.29 Pinnock and colleagues found that patients with severe COPD accept COPD as a way of life rather than an illness, and that policy focus on identifying a time point for transition to palliative care has little resonance for people with COPD or their clinicians.18 With the difficulty of identifying a definite transition point to palliative care in the COPD illness trajectory, the concept of key milestones along a lifetime journey with COPD has developed.19,20 Thresholds at which ACP might be discussed have been described and include forced expiratory volume in 1s (FEV1) less than 30% predicted, oxygen dependence, and hospital admission for exacerbation of COPD.30

This study aimed to discover whether or not an admission to hospital for an exacerbation of COPD was an opportunity for initiation of discussion about ACP. It appears from the chaotic nature of the admission and the severity of the illness that this is not necessarily an appropriate time. However, the other aim of the study was to see if an admission was a potential trigger to starting such discussions after discharge, and this seemed to show promise. Ambivalent patient responses to end-of-life discussions have previously been observed in the UK and USA.17,31 Here, interviews with severely affected COPD patients after a hospital admission for exacerbation indicated that patients' information needs were diverse and sometimes ambiguous and contradictory. Many patients appeared to be ‘in denial’ regarding the life-threatening effect of their illness. A more positive outcome was observed in a community-based study of patients with stable COPD which revealed that discussions about artificial ventilation and cardiopulmonary resuscitation could be conducted without causing distress. The attitudes to resuscitation could not be predicted from parameters of respiratory disease severity or age.32

Previous work has indicated that the optimal context for end-of-life discussions was within a caring, trusting, long-term relationship which offered open and repeated negotiations for patients' preferences for information,33 with the family physician being well placed for providing such care.34

Our study group indicated a willingness to talk about ACP, and that having a relationship of trust with a clinician and the clinician having a good clinical knowledge of the subject could help move the patient from a position of emotional avoidance to that of being more problem-focused. Variations between patients suggest that interventions should be sensitive to the different ways in which people cope with emotional upset and that an incremental approach to such interventions should be adopted.35 Evidence from psychological reviews of health campaigns suggests that people respond to threatening situations with denial and/or avoidance, and may respond better if there is also a message of hope and that something can be done about the situation.36

Implications for future research, policy and practice

This research has highlighted the perceived chaotic nature of some emergency admissions for patients with serious respiratory illness. Although some of this may be due to the nature of the illness, it is possible that the organisation of the admission could be improved to facilitate patient care. Whether or not such an admission to hospital is a milestone that might open up opportunities for discussion about ACP needs to be explored. Further research is needed to establish what patients really want from ACP and the most appropriate time for discussions. Further investigation is needed to determine which healthcare professionals should be involved and the support required to facilitate this in daily practice.

Conclusions

Admission to hospital for an exacerbation of COPD was recalled as a chaotic experience by most patients in the study. After the admission the patients expressed an openness to discussions about ACP. However, the admission itself may not be the most appropriate time for this. The preferred person with whom to have the discussions was an informed and empathetic health professional. GPs often provide the relational continuity that patients value and are well placed to engage in such discussion.