Barrett's oesophagus: A qualitative study of patient burden, care delivery experience and follow‐up needs

Abstract Background Barrett's oesophagus (BO), a precursor to oesophageal adenocarcinoma, requires long‐term endoscopic surveillance. The rising incidence of this chronic disease has implications for service provision and patient burden. Few studies have explored BO patients’ personal burden, care delivery experience and participation in health‐care delivery decisions. Objective To identify and explore factors impacting BO patients’ health‐related quality of life, follow‐up needs and views on new models of follow‐up care. Design An exploratory qualitative approach was adopted using semi‐structured, in‐depth, one‐to‐one interviews, audio‐recorded and transcribed verbatim. Patients undergoing BO surveillance, at a single NHS hospital, were recruited using purposive sampling with the aim of achieving maximum variation. Data were analysed using framework analysis approach, supported by NVivo Pro 11. Results Data saturation occurred after 20 participant interviews. Ten subthemes and three main themes emerged from the analysis: (a) burden of disease—symptom control, worry of oesophageal cancer and surveillance endoscopy; (b) follow‐up experiences—follow‐up care, at this NHS hospital, was found to be inconsistent and often inadequate to meet patients’ needs, in particular a lack of disease‐specific information; and (c) follow‐up needs—participants sought enhanced communication, organization and structure of care. They highly valued face‐to‐face interaction with a specialist, and the concept of direct secondary care access in‐between endoscopies was reassuring to participants. Conclusions This qualitative research provides an in‐depth account of the patients’ perspective of BO, the effectiveness of follow‐up care and patient opinion on new follow‐up systems.


| INTRODUC TI ON
In contrast to many other cancers in the Western world, the incidence of oesophageal adenocarcinoma (OAC) has increased over the last three decades [1][2][3] with no significant change in survival over the last 10 years. 4 In an attempt to address this imbalance, Barrett's oesophagus (BO) has been identified as a key opportunity to intervene and prevent OAC. With clearer referral guidelines 5 and national public health campaigns (Public Health England "be clear on cancer"), 6 the diagnosis of this precursor for OAC will continue to increase. 7 Without reliable individual risk stratification, the majority of patients with BO undergo long-term endoscopic surveillance, which has implications for future health-care provision and lifelong patient burden. Few studies, predominantly quantitative in design, have demonstrated significant reductions in BO patients' health-related quality-of-life (HRQOL) scores. However, many of these are now outdated, lack generalizability and have used measurement tools not specific to BO. 8 Only in recent years have international guidelines, the British Society of Gastroenterology (BSG) and American College of Gastroenterology, advised consultation and counselling of newly diagnosed patients prior to surveillance enrolment. 5,9 Historically, BO patients are likely to have received inconsistent care from poorly informed or even disengaged physicians. 10,11 The effects of historic follow-up and current care pathways on patients remain unknown.
Traditionally, the providers of new health-care developments have controlled their design and implementation. This archaic "doctor knows best" attitude to health-care delivery and research has begun to change in the NHS over recent years with a keener focus on patient-centred, effective and safe clinical care. [12][13][14][15][16] One area where user involvement appears to have its greatest influence is when drawing upon patients' experiences, particularly in chronic disease settings. Previous engagement with patients to identify and address their follow-up needs has dramatically changed the landscape of care in some chronic diseases. Most notably, within gastroenterology, have been the developments in inflammatory bowel disease (IBD) care. In 1991 17 Probert et al conducted a questionnaire survey regarding disease counselling preferences in 59 patients with IBD. This landmark paper identified a significant number (60%) who required further information regarding their condition. They also found that many patients would be happy with a trained nurse consultation and identified a need for more rapid access to services. Since then, the role of the specialist IBD nurse has evolved and has been proven to reduce admissions, emergency attendances and outpatient appointments leading to large cost savings. 18 These improvements likely reflect enhancements in professional-patient relationships, patient disease-specific knowledge, self-care and medication compliance.
These endpoints, however, are somewhat harder to measure. More recent research in IBD follow-up care found that patients desire more active involvement in their care and are keen to explore more novel follow-up alternatives, for example virtual clinics. 19 Although the disease profiles, patient demographics and treatments may differ dramatically between chronic diseases, there are valuable commonalities to draw from these patient involvement strategies and service improvements. In particular, these include the processes used to involve patients and seek alternative or enhanced ways to educate, follow up and communicate with patients.

| Aims
• To identify and explore factors impacting BO patients' HRQOL.
• To identify and explore the follow-up needs of BO patients.
• To explore patients' perceptions and attitudes to new models of follow-up care.

| ME THODS
This exploratory qualitative research forms part of a concurrent mixed-methods study, using both qualitative and quantitative data collection tools, to explore the impact of BO on patients' HRQOL, 20 their experiences of follow-up care and attitudes towards service developments in line with the preliminary research needed when developing complex interventions. 21 This qualitative approach attempts to understand the social phenomena in natural circumstances, with an emphasis on exploring meanings and views of participants. 22 The study design incorporates the consolidated criteria for reporting qualitative research guidelines 23 (see Appendix S1 for further details).

| Ethical considerations
Prior ethical approval for this study was obtained from the Health Research Authority Yorkshire and Humber ethics committee (REC reference number 16/YH/0035).

| Participants and setting
Individuals with BO, enrolled in surveillance at a single general NHS hospital, were targeted because they were readily accessible within the constraints of the study team geography. Participants were purposively 24 recruited with the aim of achieving maximum variation in terms of disease duration, age and gender even though this is a malepredominant disease. Recruitment continued until a point where data  28 Participants were recruited face to face at their surveillance endoscopy, via telephone or postal invite. There was no prior contact between researchers and participants before recruitment.

| Data collection
Semi-structured, in-depth, one-to-one interviews were undertaken by JB (average time of 40 minutes, range 21-76 minutes).
The status of the interviewer (postgraduate research doctor) was made aware to all participants. An interview topic guide was developed from a prior literature review 8 and expert opinion (please see Appendix S1). Interviews focused on the impact of surveillance, physical and psychological symptoms, experiences of follow-up care, follow-up needs and new models of follow-up care.
New models of care included a dedicated BO service and patientinitiated consultation by means of telephone or virtual clinic. All interviews were conducted in a private seminar room to provide a non-clinical atmosphere. Interviews were audio-recorded, transcribed verbatim and anonymized prior to analysis. Participant's demographics and disease-specific information were also collected from their medical notes and endoscopy reports. Field notes were taken at the time of each interview. These written recordings captured important verbal and non-verbal information which can be overlooked once the content is transcribed. This is an important step to keep the context of the interview.

| DATA ANALYS IS
A thematic analysis was conducted on all data, using a framework approach 29 supported by NVivo Pro 11 (QSR International (UK) Limited, Cheshire UK). The key steps are outlined in Figure 1. This widely used approach 30 allows rigorous analysis without losing transparency or site of the initial raw data. Initial emerging themes were identified from the first four interviews. These themes, alongside topics raised from the interview F I G U R E 1 Framework analysis Identifying Initial Themes • Transcripts and audio recordings of the first 4 interviews were analysed to identify a long list of initial themes and concepts.

Conceptual Framework
• The emerging recurrent themes and topics from the interview guide formed the Thematic Framework. • This Framework allowed further classification and organization of the remaining raw data (verbatim transcripts) Labelling (Indexing) • The fine detail of every transcript paragraph or sentence was coded (labelled, tagged, indexed) to a theme from the framework. • During this process some themes from the initial framework were refined to achieve a more accurate fit. guide, formed the conceptual framework (Table 1). This framework was then applied manually to the raw data in a process called indexing. Field notes were linked to the content with clear associations between themes recorded for later use in descriptive analysis. The fully indexed raw data were then displayed in thematic charts allowing greater focus and distillation of the detail in each subtheme (see Appendix S1). Each column of the thematic chart was then subjected to descriptive analysis and further interpretation of the data to recognize patterns and explanations.

| Rigour
The following steps were taken to ensure rigour. Firstly, none of the participants had prior clinical contact with the researchers. The topic guide was reviewed by all researchers to ensure appropriateness of the content. Field notes were taken during each interview to ensure grounding of the content during analysis. Finally, two initial verbatim transcripts were analysed by two different researchers (JB and MH, a qualitative research specialist with a clinical background in nursing) to confirm the data were within the remit of the study and the initial emerging themes identified were consistent and fit the data captured. Preliminary findings were discussed between JB, MH and YA who agreed upon the relevance of the data and credibility of the analysis. Consensus on themes was reached through discussion.

| RE SULTS
Data saturation, the point where no new information emerged from the data, 31 occurred after 20 participant interviews, the demographics of which are displayed in Table 2. In total, this process generated three overarching themes and 10 subthemes ( Figure 2). Considering the aims of the study, the results will be discussed under the three main themes: (a) burden of disease, (b) follow-up experiences and (c) follow-up needs. Information describing each theme is given and supplemented with original verbatim quotes (

| Worry and anxiety of oesophageal cancer
Some participants are able to put thoughts regarding cancer "to the back of their mind" or approach cancer risk pragmatically with a "what will be will be" attitude.

| Inadequate follow-up at diagnosis
Participants' experiences of secondary care follow-up at the time of their diagnosis were inconsistent and in the majority of cases inadequate for their needs. The majority of patients received a brief interaction post-endoscopy either from the endoscopist or from the nurse at discharge. In some instances, BO was not discussed at all. In these cases, participants received notification via a copy of their endoscopy report or subsequent letter. In one case, the participant was unaware of the diagnosis until they were asked to attend for their next surveillance endoscopy. Such inconsistencies and inadequacies could be predicted considering the BSG has only recommended outpatient clinic follow-up since their latest guideline publication in October 2013.
However, those who did receive clinic follow-up also reported mixed experiences with some feeling the clinic was too time-pressured, with a lack of emphasis on Barrett's and left with unanswered questions.  where to find trusted sources online. This finding was present in both younger and older participants. Some patients expressed concerns that improved disease-specific knowledge may heighten anxieties regarding oesophageal cancer and were least likely to seek additional information preferring to adopt an "ignorance is bliss" approach. In comparison, overestimators of cancer risk were linked to heightened anxieties and worries of cancer, whereas those who correctly viewed their risk as low, generally, appeared to have less worry.

| Greater disease specific knowledge
The major unmet need identified was disease-specific knowledge,  (Table 4). Although those who received copies of their endoscopy report did not find them very informative, they did find the associated diagrams and pictures of their oesophagus both useful and interesting.

| Value of seeing a specialist
When asked about improving delivery of care and reflecting on their past experiences, it was clear that patients highly value face-to-face interaction with a specialist. This probably reflects past inadequacies

(re dedicated clinic) "" I think that's what is really needed to be quite honest, from my point of view.
There's just not enough information out there, concrete information. I think it gives more confidence to the patient, rather than just saying "look at this information leaflet and follow that to the best you can."" (66yr, male) P (re dedicated list) "That would be good because, obviously, the man with the camera is just doing one after another probably different procedures, like I said he's no specialist in Barrett's. I mean when you're going in, they had to ask me what I am coming in for. I think it will be a lot better. Obviously, if they're more trained in Barrett's they know what they're looking for." (58yr, female) R (re nurse care provider) "that would be good as long as they specialise in that particular area. Because for example when you ask your GP, sometimes he won't want to commit or wrongly advice you, and sometimes he'll probably just look on Google.

| Perceptions of new models of follow-up care
• Dedicated Barrett's oesophagus clinic and endoscopy Patients were asked about their views on the implementation of a dedicated Barrett's service. This service, run by a health-care professional (gastroenterologist or nurse specialist) with a specialist interest, would encompass both surveillance endoscopy and an outpatient clinic. All participants responded positively to this concept. In particular, they liked the face-to-face contact with a specialist and thought it could potentially solve the continuity of care issues currently faced.
When asked specifically about the provider of this care, the majority of patients would be happy to see either a specialist doctor or nurse.

| D ISCUSS I ON
This study aimed to explore BO from the patients' viewpoint, in particular the impact on health-related quality of life, experiences and effectiveness of follow-up care and opinion on new follow-up systems. To our knowledge, this represents the most in-depth account of BO patients' perspective of disease impact in a UK NHS setting.
The most striking finding relates to patients' experiences of follow-up care. Historic and current follow-up for BO appears inconsistent and often inadequate to meet patients' needs and expectations.
This has led to poorly informed patients with limited or inaccurate disease-specific knowledge. Some potential impacts, identified in this study, include reduced confidence and ability to self-manage symptoms and heightened cancer-specific worry. Very few studies exist which assess levels of patient education in BO. Concerningly, TA B L E 4 Disease-specific knowledge; patient uncertainties in 2008, Murphy and colleagues reported <50% of patients with concurrent OAC and BO diagnoses were aware of their BO diagnosis despite an average of more than seven previous endoscopies. 32 Improved patient knowledge in IBD appears to have positive and detrimental effects with greater knowledge associated with adaptive coping strategies but also higher anxiety levels. 33,34 This reflects those who report an "ignorance is bliss" attitude too improving disease-specific knowledge in this study. Even participants with longer-term diagnoses voiced unmet needs and questions regarding their condition. This finding questions the current BSG guidance which only recommends new patients attend an outpatient clinic. 5 The role of a Barrett's clinic may be much broader than this, by giving all patients the option of attending clinic after their surveillance endoscopy would capture patients seeking more information and guidance about their condition. In some cases, it may also provide a platform for addressing poor symptom control or an opportunity to discuss the appropriateness of discontinuing surveillance. The latter may be vital when one considers the number of patients who may have been enrolled in surveillance inappropriately 35,36  professional. This two-way discussion should cover both the professional (for BSG clinic agenda, see Appendix S1) and patient agendas ( Participants also liked the concept of a "safety net" in the form of a patient-initiated consultation service. This probably reflects the potential impact of uncontrollable symptom flares, length of time inbetween endoscopies and doubts over primary care ability to deal with their concerns in a timely manner. Patients had a strong preference to a telephone-based system rather than an "impersonal" vir-

| Strengths and limitations
The study utilized a number of steps to ensure rigour in its design; however, there are some limitations. Firstly, participants in this study were recruited through a single district general hospital population.
Therefore, one must be cautious when generalizing these findings, especially those relating to organization and structure of care which may differ significantly elsewhere. However, most UK NHS hospitals provide BO care in a similar ad hoc fashion and experience the same issues organizing and budgeting follow-up care provisions. Secondly, the study did not take a longitudinal approach to identifying BO impact over the life course. However, participants were recruited until the researchers were happy that thematic saturation was achieved with good variation of age, disease duration and gender. Variation in socio-economic status and health literacy was not formally sought, and this may be an area for future research to clarify. Thirdly, the data captured may have been influenced by the status of the interviewer. 52 Fourthly, only two interviews were coded by two separate researchers which may introduce bias; however, there was a strong correlation between findings and all authors reviewed and agreed upon the final themes and credibility of the analysis. Finally, all participants were "white British" and English-speaking, so one must be cautious when translating these findings to more diverse ethnic populations. A greater number of male than female participants could be viewed as a limitation; however, this is a disease predominantly affecting men with a male/female sex ratio of 1.96/1 reported in a meta-analysis. 53

| CONCLUSIONS
This qualitative research provides an in-depth account of the patient perspective of BO in an NHS setting. Key potential impacts on patients include symptom control, worry of oesophageal cancer and burden of surveillance endoscopy. These factors must be considered when implementing future care pathways, designing clinical trials or developing a BO-specific patient-reported outcome measure. Follow-up care, at this NHS hospital, was found to be inconsistent and often inadequate to meet patients' needs.
Patients require greater disease-specific information, enhanced communication, organization and structure of care. To improve patient experiences, we recommend the design, implementation and prospective assessment of a complex care intervention, which encompasses dedicated BO surveillance, outpatient clinic and telephone direct access line.

ACK N OWLED G M ENTS
This study forms part of a larger research project which has received unrestricted external funding support from Covidien (Medtronic).
Covidien have not influenced the design of this study, writing of the manuscript or decision to submit for publication.

CO N FLI C T O F I NTE R E S T
Nothing to disclose.

AUTH O R CO NTR I B UTI O N S
All authors significantly contributed to this work. The concept and