Non-pharmacological intervention for gastro-oesophageal reflux disease in primary care

Intervention development stage (stage 1)

One hundred and seventy-nine of the original 669 patients responded, expressing an interest to take part in stage 1, or 2, or both stages of the study. Twenty-three patients were sampled, (12 women, 11 men, age range 31–86 years, representative of the demographic spread of the responder group) to participate in semi-structured interviews. Using a schedule developed from the published evidence, patients were interviewed in their homes by the lead author, about aspects of their reflux that caused most concern, the influence of lifestyle including diet, alcohol, smoking, activity, posture, and dinner to bed time, and the problems they experienced in managing their symptoms. It was confirmed that daily living patterns do appear to affect reflux symptoms, combining in complex and distinctive patterns for each GORD patient. Where patients are aware of behaviours that may exacerbate their symptoms, they are often unable to make changes because they do not know how to. This development stage is reported elsewhere.15

A literature review of patient education and self-management programmes in other chronic illnesses was then undertaken,16 and the evidence was combined with findings of the qualitative interviews, to develop an education intervention to enhance patient self-management of their symptoms, designed around aspects of chronic disease self-management that are known to be effective.11,17

At interview, participants identified that the most commonly occurring influences on their GORD symptoms were diet and stress. Aspects such as smoking, alcohol, exercise, and activity were introduced into the sessions, but because of constraints on time and funding this study focused attention on the self-identified issues. As the intervention was aimed at promoting behavioural change, the National Institute for Health and Clinical Excellence (NICE) principles for designing behavioural-change interventions were used as a framework.18

Group intervention stage (stage 2)

The intervention was designed as a 1.5-hour session per week, for 4 weeks (Table 1).

Of the original 179 responders, 53 had expressed interest in and were recruited to the intervention stage (stage 2) of the study. Eleven patients withdrew due to family commitments (n = 3), or holiday (n = 2), or did not attend (n = 6); 42 patients attended the groups (male = 19, female = 23, aged 31–86 years). Groups met on a weekday afternoon or evening. Group sizes varied (group 1, n = 26; group 2, n = 3; group 3, n = 8; group 4, n = 5), but membership per group remained consistent throughout the intervention. The attendance rate was 94%, with 6% of sessions missed due to hospital appointments or illness. No one who commenced withdrew from the intervention.

Diagnoses included reflux/reflux oesophagitis, GORD (n = 16), dyspepsia (n = 6), hiatus hernia (n = 10), Barrett's oesophagus (n = 4), and other (n = 6). The intervention was delivered in a group format at venues in the participants' neighbourhood or at their GP surgery, on 4 consecutive weeks in early 2009 by the lead author. The Brief Illness Perception Questionnaire (BIPQ),19 the GORD Impact Scale (GIS),20 and the Hospital Anxiety and Depression Scale (HADS)21 were completed during the first session, and again 3 months after the final session.

The Brief Illness Perception Questionnaire (BIPQ)

There was a significant change (P<0.001) in the overall BIPQ score following the intervention and in six out of eight domains (Table 3).

No significant change was seen in patients' perceptions of the impact of reflux on their lives before and after the intervention, or in how long they thought their illness would last. The median score in these domains prior to intervention suggests that patients perceived the illness to have only a moderate influence on their lives (median = 5). They also had a very realistic impression of the likely duration of their condition (median = 10, score of 10 = forever), a factor that is unlikely to change, irrespective of how well that condition is managed. The remaining domains indicate that following the intervention, patients felt more in control, believed that their treatment can help them, experienced fewer symptoms, were less concerned about their illness, and had a greater understanding of and were less affected emotionally by their reflux than before the intervention.

The GORD Impact Scale (GIS)

Patients indicated how often in the past week they had experienced the symptoms listed. There was a significant change (P = 0.008) in the overall GIS score following the intervention, and in six out of nine domains (Table 4).

No significant change was seen in patients' reports of pain in the chest or behind the breastbone, in symptoms preventing patients eating or drinking what they wished, or in the use of additional medication to control symptoms. The remaining six domains indicate that following the intervention, patients experienced significantly fewer of the established symptoms associated with reflux disease than they did previously.

The Hospital Anxiety and Depression Scale

There was no significant change in the overall anxiety or depression scores (Table 5). The median scores indicate that anxiety and depression among this group was within normal ranges for the general psychiatrically well population.

Number of monthly prescription repeats

GP database records confirmed that, for the 3 months immediately preceding and following the intervention, patients in the study requested a median of three repeat PPI prescriptions per 3-month period. No significant change in repeat prescription requests was seen.

Qualitative data

Three months after the intervention, patients were given the opportunity to make any free-text comments they wished about their involvement in the programme and experiences with their reflux since then. Content analysis revealed four main categories (Table 6).

Summary of main findings

The intervention improved patients' perception of their illness at 3 months, and improved their illness experience in many domains. The main improvements were in those domains testing patients' sense of control, perception, and understanding of their condition, and also in symptom severity. Despite a significant change in symptom scores, there was little change in patients' perceptions of how much GORD affected their life. This may be explained by attitudes towards their illness and a determination not to let it impact on their life, despite troubling symptoms. These findings suggest that GORD does affect people's lives and that the intervention may moderate this impact to some extent. The study may, however, have been underpowered to detect other significant differences.

The GIS scores indicate improvements in many of the symptoms associated with GORD. There was no significant difference in patients experiencing pain in the chest or breastbone, but this was an infrequently reported symptom before the intervention, leaving little room for improvement. There was also no improvement in symptoms that prevent patients from eating or drinking the foods that they like, which may reflect the fact that, despite reporting symptoms, patients did not let these interfere with what they wanted to consume anyway. However, qualitative comments from patients suggest a different picture — patients have clearly stated that they have changed eating and drinking habits (content and timing of meals) to successfully manage their symptoms. There was no change in the number of additional medications taken to control symptoms, although the patient group reported minimal use of additional over-the-counter remedies before the intervention. Of the remaining domains, those relating to sore throat, interrupted sleep, and activity show a trend towards significance, suggesting there may be some improvement. All other domains relate to classic symptoms associated with reflux, and there was improvement in all of these.

The GIS correlates well with quality-of-life scores such as QoLRad,21 with a higher score indicating a greater impact on quality of life. There was significant reduction in the median scores and the interquartile range on the GIS, suggesting a corresponding improvement in quality of life at 3 months.

The HADS did not detect any change in anxiety or depression. Since overall scores in both the anxiety and depression scale were within the normal range (0–7) for the population at baseline, there was little opportunity for improvement to be demonstrated.

No change was found in prescription use, although this was not the intention of the study.

Strengths and limitations of the study

These were self-selected patients who volunteered for the study, and most people approached did not opt in to the intervention. All intervention sessions were delivered by the same person, to ensure parity of experience across each patient group.

The study attempted to collect economic data but responders were found to be reluctant to divulge information, so that the majority of data were void due to incompleteness. This information could provide a useful insight into the ongoing financial impact of GORD on patients' lives. The study lacked a control group, so improvements detected might represent a placebo effect of intervening per se, rather than being attributable to the specific intervention, and the sample sizes were small.

A further large-scale randomised controlled trial would enable confirmation of the findings. The initial follow-up is short at 3 months; however, evidence suggests that if behavioural changes are going to be made, they will be made early and sustained.22 The authors plan to test this with further follow-up at 12 months.

Comparison with existing literature

The only previous trial of an education intervention for GORD was based in Norway,23 and found that increased knowledge about GORD had no impact on quality of life or use of health care. Improved quality of life was only seen in patients with education limited to primary school, suggesting that better educated patients are more knowledgeable about GORD and may manage their symptoms more effectively.24 Symptom severity or impact of knowledge on disease management were not measured, so it is not known whether that intervention improved the patients' perception and experience of their illness. Contrary to other patient education programme guidance, that study did not appear to include the key components that are known to support patient self-management of illness.17 The current study has adopted these proven approaches and enabled patients to make behavioural changes that lead to effective control of the lifestyle issues that influence their symptoms and lead to a corresponding improvement in quality of life.

Generic lifestyle advice for management of GORD does exist but the stage 1 findings15 suggest that patient symptom experiences are unique, and that patients may need help to change the behaviours that lead to their lifestyle habits. By incorporating behaviour-change strategies into the education intervention, this study has been able to demonstrate the potential of this approach in managing persistent GORD symptoms. Recent evidence suggests that PPIs may even promote the symptoms they are intended to control.25 If this is the case, maintaining patients on, or weaning them off PPIs without giving them a means of managing their symptoms will be very difficult. If, as suggested, the way forward with reflux management is to use more than one approach,9,26 an education intervention alongside a carefully managed PPI programme27 may have the potential to reduce the long-term use of PPIs and avoid the problem of rebound acid production as well as improving quality of life.

Implications for future research

These findings suggest that a GORD education programme can have an impact on patient symptom experience. Follow-up is planned at 12 months post-intervention, to determine whether changes seen at 3 months are sustained.

A randomised controlled trial with a larger cohort that tests delivery of the intervention in a range of formats is needed in order to develop a programme that is readily available in primary care, economically effective to deliver, and benefits patients.