Long-term symptom severity in people with irritable bowel syndrome following dietetic treatment in primary care – a service evaluation.

Background: Evidence suggests dietary interventions can improve symptoms in people with 10 irritable bowel syndrome (IBS), but most data explores short-term (immediate) impact; data on 11 long-term (>6 months) impact are limited, especially from primary care settings. This study aimed 12 to investigate the long-term effect of dietetic-led interventions for IBS delivered in primary care. 13 Method: A service evaluation of a dietetic-led IBS clinic was completed, analysing data on 14 symptom severity, stool frequency and consistency, and healthcare input. Data were collected 15 before and immediately after dietary intervention as part of patients’ routine clinical appointments. 16 Long-term data was collected via a postal questionnaire at least 11 months later. 17 Results: 211 patients responded to the long-term follow-up questionnaire at 13 months (median; 18 interquartile range 12-16 months) post follow-up appointment. 84% had been advised to follow the 19 low FODMAP diet. All symptoms were reported significantly less frequently short-term, and all, 20 except heartburn and acid regurgitation, remained so long-term. The four most commonly reported 21 bowel symptoms reduced in frequency by 62% abdominal pain, 50% bloating, 48% increased wind, 22 and 49% urgency to open bowels (p<0.001). Percent patients reporting satisfactory relief of gut 23 symptoms was 10% baseline and 55% long-term follow up (p<0.001). Visits to the GP reduced 24 (96% vs 34% p<0.001), and to the gastroenterologist (37% to 12%; p=0.002) during the year prior 25 to long-term follow up compared to the year prior to dietary intervention. Conclusion: Patients with IBS, who received dietetic-led

including undiagnosed coeliac disease, non-coeliac gluten sensitivity and gastrointestinal food 65 allergy, can all present with similar symptom profiles to IBS, and are often misdiagnosed. 66 However, each of these conditions require different diets and variable levels of dietary stringency, 67 emphasing the need for specialist dietetic intervention ( 26-29 ). Coeliac disease must first be excluded 68 via appropriate tests ( 9 ), and exploring a patient's atopic history may provide an indicator for a 69 potential food allergy ( 29 ). Non-coeliac gluten sensitivity involves intestinal and extra-intestinal 70 symptoms that are triggered by gluten ingestion in the absence of coeliac disease and wheat allergy 71 ( 30 ) and may also include 'foggy mind', tiredness, headaches, fibromyalgia-like joint or muscle 72 pain, and leg or arm numbness ( 31 ). 73 The majority of available evidence demonstrates the benefits of diet, including the low FODMAP 74 diet, on IBS symptoms immediately following implementation, and up to 9 months afterwards ( 15-17,   , and support for this was delivered either by the 93 GP, or the general community dietetic clinics. Only those with intractable symptoms were referred 94 on to the specialist clinic. Those who attended the dietetic-led gastroenterology clinic were 95 assessed and counselled by a specialist gastroenterology dietitian. The patient attended at least two 96 dietetic appointments; an initial appointment for assessment and education on recommended dietary intervention; and a follow up appointment at least 4 weeks later when they had implemented the 98 advised dietary changes. Some patients attended subsequent appointments if further dietetic 99 intervention was recommended at their first follow-up. At the final follow-up appointment, patients 100 were educated on how to complete relevant food challenges, and advice was provided on food 101 reintroductions and long-term self-management.

102
All patients seen in the clinic between May 2013 and April 2017 were included. Data was collected 103 at three time points: prior to their initial appointment (baseline), prior to their final follow-up 104 appointment (short-term follow-up) (both of which were part of the routine clinical care), and 105 approximately 11 months later (long-term follow-up) via postal questionnaire.

106
Ethical approval via the UK Health Research Authority was not required because it was deemed an 107 evaluation of the dietetic service. Local approval was given by [blinded for peer review] to carry 108 out the data collection. Patients were seen in the dietetic-led gastroenterology clinic by one of three specialist 112 gastroenterology dietitians. As per routine clinical practice, a medical, social and diet history was 113 completed along with an assessment of gut and non-gut related symptoms, followed by a discussion 114 regarding previous treatments and dietary habits. Following careful consideration of these 115 parameters, and in consultation with the patient, at the initial appointment one of the following  The following data were collected from the clinical notes for baseline and short-term follow-up and 129 from the questionnaire for the long-term follow-up:  At baseline and long-term follow-up, patients were asked to recall, in the previous 12 months, how 145 many times they had visited their GP or gastroenterologist for their IBS symptoms and whether they 146 had any investigations for gut symptoms. Patients were also asked to confirm whether or not they 147 were currently taking any prescribed medication for their gut symptoms.  reasons than IBS, therefore 211 (44%) patients were analysed at long-term follow-up. The mean age 179 was 53.6 years (sd=15) and 182 (86%) were female. The median duration from baseline to short-180 term follow-up appointment was 9 weeks, (interquartile range 9-13 weeks), and the median duration 181 from short-term to long-term follow-up was 13 months (interquartile range 12-16 months). Five 182 patients (2%) were sent postal questionnaires before the planned 11 months, due to an 183 administrative error. In 38 patients (18%) there was a >6 month delay in sending out questionnaires 184 after their final appointment due to other work priorities at the time.

185
As shown in Table 1, the majority of patients (84%) were advised to follow the low FODMAP diet, 186 either in isolation or combined with an additional dietary intervention. This is similar to the 187 proportion of the original cohort (n=547) who were advised to follow the low FODMAP diet (81%). At baseline, only 23% of patients reported a normal stool consistency (BSFS 3 or 4) ( Table 2). The 216 most common stool types were mixed and loose stools (BSFS 5-7), and the least common was 217 constipation (type 1-2). At short-term follow-up the proportion of patients reporting normal stool 218 consistency significantly increased to 49% (p<0.001). At long-term follow-up this reduced to 45% 219 but remained significant when compared to baseline (p<0.001).

220
At baseline, 74% of patients reported a normal stool frequency (between once every 3 days and 3 221 times a day) ( Table 2). This significantly increased to 89% (p<0.001) at short-term follow-up and 222 82% (p=0.005) at long-term follow-up.  provided information on the number of times they had seen a gastroenterologist in the previous 12 229 months at baseline and 125 at long-term follow-up (Table 2). Similar to GP visits, the proportion 230 visiting a gastroenterologist at least once reduced from 37% to 12% (p=0.002).

231
Whether investigations for gut symptoms occurred was reported by 130 patients at baseline and 232 long-term follow-up. Endoscopic investigation was the most common type, followed by ultrasound 233 ( Table 2). At baseline, 49% patients reported having at least one investigation in the previous 12 234 months and 18% reported multiple investigations. At long-term follow-up this reduced to 17% and 235 5% respectively (p<0.001).

236
Results for medication usage showed a similar pattern; 57% reported using prescribed medication 237 for their gut symptoms at baseline, and this reduced to 49% at long-term follow-up, however this 238 was not significant based on our defined criteria (p=0.034) ( Table 2).  this area may to lead to the most significant improvements in the quality of life of IBS patients. 283 We also showed that reported healthcare usage significantly reduced in IBS patients in the period symptoms for many years, and due to lack of effective treatment options, they had repeatedly 300 visited their GP and had repeated referrals to secondary care over many years. Therefore, there is 301 the potential to reduce healthcare usage in all age groups by offering effective dietetic treatments.  However, further studies including randomised control trials (RCTs), are required to assess the 326 mechanisms for the other diets used in clinical practice, along with the long-term implications and 327 safety, before the diets can be included in formal guidelines. 328 It may be that the improvements seen in patients following alternative diets (not low FODMAP) 329 was due to the diets being effective treatments for alternative diagnoses, rather than an effective 330 treatment for IBS. IBS is difficult to diagnose due to the vague symptoms, thus this diagnosis may 331 not always be accurate. Undiagnosed coeliac disease, non-coeliac gluten sensitivity and 332 gastrointestinal food allergy all present with similar symptom profiles to IBS. Further research is 333 needed to explore this area, however, this real-life service evaluation supports the view that a 'one-334 size fits all' approach to dietary treatment of patients who present with IBS is not appropriate.

335
Dietitians, especially those with expertise in gastroenterology, can play an essential role in the 336 appropriate assessment and effective delivery of the dietary treatment options for IBS patients.

337
The main limitation of this study is that as it was an observational service evaluation, it is not using medication for their gut symptoms at baseline and long-term follow-up, we can not exclude 344 medication as playing a role in improvements seen. Due to this, further RCTs are needed that 345 explore benefits of dietary treatments on IBS management, which also take into account these other 346 factors. Other limitations to this study includethe increased risk of non-response bias associated 347 with a postal questionnaire design, as those patients who decided not to respond to the questionnaire 348 at long-term follow-up may differ from those who did. Questionnaire designs also increase the risk 349 of recall bias, which can lead to a deviation from true results. In this study, we included patients 350 who were referred for IBS, but we did not apply strict ROME IV criteria for inclusion. This was