Search

Introduction Thiopurines (TP) have an established role in the maintenance of steroid-free remission in IBD. Their role in reducing the need for surgery in younger populations has been recently highlighed. 1 TPs are associated with an increased risk of lymphoma in the elderly but their role in reducing the risk of surgery in patients with elderly-onset IBD (EO-IBD) is not established. We therefore aimed to evaluate the impact of TPs on the risk of surgery in EO-IBD, using a nationally representative research database. Methods This was a retrospective descriptive study using data from the Clinical Practice Research Datalink (CPRD). Incident cases of IBD were isolated using previously validated Read codes. Patients were included if they had codes for ulcerative colitis (UC) or Crohn’s disease (CD), and their date of diagnosis was on/after their 60 th birthday. Information extracted included sex, age at diagnosis, smoking status, and prescription data for IBD medications. Finally, we obtained data on our main outcomes: colectomy in UC and first intestinal surgery in CD. A multivariate Cox proportional hazard model was generated to deduce the risk of surgery according to TP use. The model was adjusted for potential confounders including sex, smoking, 5 ASA use and early steroid use (corticosteroid within 3 months of diagnosis). In a subgroup of TP users, a second proportional hazard model was used to define the impact of duration of TP treatment on the risk of surgery. Results 2758 and 1349 were identified with EO-UC and EO-CD respectively, with a mean follow up of 4.9 years. Mean age at diagnosis was 71.8 years. Table 1 demonstrates the summary demographic and prescription data for both groups. TP use was associated with an increased risk of colectomy in EO-UC (HR 3.5, CI 2.28–5.35, p Conclusion TP use for longer than 12 months is associated with a reduced risk of colectomy in EO-UC but in contrast does not appear appear to impact on the risk of first intestinal surgery in EO-CD. This is an important observation since the risk of TP-associated lymphoma is substantially increased in this age group. Further work is needed to evaluate the role of TPs in EO-IBD. Reference 1 Chhaya V, et al . Impact of early thiopurines in 2770 children and young people diagnosed with IBD: a national population-based study. Aliment Pharmacol Ther . 2015 Oct; 42 (8). Disclosure of Interest C. Alexakis: None Declared, S. Saxena Grant/research support from: SS is funded by a National Institute for Health Research Career Development Fellowship (NIHR CDF-2011-04-048). The views expressed are not necessarily those of the NHS, NIHR, or Department of Health. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript, V. Chhaya: None Declared, E. Cecil Grant/research support from: EC is funded by a National Institute for Health Research Career Development Fellowship (NIHR CDF-2011-04-048). The views expressed are not necessarily those of the NHS, NIHR, or Department of Health. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript, R. Pollok: None Declared

Introduction Chronic Hepatitis B (HBV), a major cause of liver disease, is underdiagnosed in the UK with pockets of high prevalence in areas with large immigrant populations. Identifying individuals through a screening programme dedicated to high risk individuals would be beneficial if that target population could be defined. Our aims were to interrogate a Primary Care database (SystmOne) establishing the best way of identifying individuals suspected of having had a HBV test to exclude them from the target population. Methods We used four search terms: age (>18), ethnic code, birth place, language spoken. Read codes (RC) were applied to identify individuals previously tested/diagnosed with HBV. RC were identified by using ‘HBV’ as a broad search term. Results We identified 4256 individuals aged ≥18. 718 (18%) were excluded because of lack of demographic data; 3210 (75%) qualified as the target population. 81 RC pertaining to HBV were found and generated 413 ‘hits’, only 224 ‘hits’ (54%) satisfied our criteria. 206 individuals (6.4%) had HBV serology recorded, 9/3210 (0.28%) were coded for chronic HBV, a further 2 HBsAg positive patients (0.9%) had no RC. After reviewing individual notes electronically we established the following 8 RC to be the ‘most useful’: XaIq7, XaFuS, X306n, X306i, 43B4, XaMBL, XaG1R, XaPEy which would have identified the latter. Conclusion We have identified the 8 most useful RC to help with a potential screening programme. Using all available HBV RC would be too time consuming. This study highlights that very few high risk immigrants are currently tested for HBV. Disclosure of Interest None Declared.

Introduction Using real-world healthcare, routinely collected information we investigated the prevalence of irritable bowel syndrome (IBS) and associated consumption of healthcare resources in the Scottish National Health Service. Methods A consultation records database covering 56 general practitioner (GP) practices in Scotland (approximately 255,000 people; 5.8% of the population) was used to identify consultations primarily due to IBS with GP or practice nurses between April 2009 and March 2011 (Read codes: IBS, constipation, diarrhoea). Read codes suggesting other causes of diarrhoea or constipation and patients with inflammatory bowel disease, diverticular disease, coeliac disease or bowel cancer were excluded. A national primary care prescription database was used to identify use of antispasmodics, laxatives or antidiarrhoeals. Referrals to outpatient clinics and acute hospital admissions between January and December 2011 in antispasmodic-treated patients were also analysed using national datasets. Results Based on consultation records, an estimated 341,180 adults (≥18 years) in Scotland suffer from IBS, representing an estimated prevalence of 7.7% (females: 9.8%; males: 5.5%). Patients consulted 1.05 times during 2010/11 for IBS symptoms and consulted more frequently overall than the general population (9.3 vs. 4.7 times per annum). During 2011, 142,738 adults received ≥1 prescription for antispasmodics, most commonly mebeverine (40.1%), hysocine butylbromide (35.7%) or peppermint (18.0%), giving an estimated prevalence of antispasmodic-treated IBS of 3.4% (females: 4.7%; males: 2.0%). One-third of these patients were also prescribed laxatives (24.5%), the antidiarrhoeal loperamide (6.7%) or both (2.5%). Of the antispasmodic-treated patients, 11,645 (9.0%) visited a gastroenterology outpatient clinic in 2011 (11.7% of all gastroenterology clinic attendances) and 1,869 (1.3%) were acutely admitted to hospital due to IBS or symptoms likely to be associated with IBS, most frequently constipation (80.3%). The average length of hospital stay was 2.1 days (2.4 days for admissions due to constipation). A further 1,141 people with no record of antispasmodic prescriptions were admitted with a primary diagnosis of IBS. Conclusion Based on consultation records, the estimated prevalence of IBS in Scotland is 7.7%, and 3.4% based on prescription records; these estimates exclude patients who do not consult a GP or who use laxatives alone or over-the-counter medicines. The prevalence estimates and levels of treatment in this study were consistent with other published information, and show that patients with IBS consume significant healthcare resources. Study commissioned by Almirall UK. Disclosure of Interest S. McTaggart Grant/research support from: This study was undertaken as a commercially contracted piece of work for Almirall UK, G. Wyper Grant/research support from: This study was undertaken as a commercially contracted piece of work for Almirall UK, L. Harkins Grant/research support from: This study was undertaken as a commercially contracted piece of work for Almirall UK, I. Bishop Grant/research support from: This study was undertaken as a commercially contracted piece of work for Almirall UK, M. Bennie Grant/research support from: This study was undertaken as a commercially contracted piece of work for Almirall UK, M. Rance Employee of: Almirall.

Introduction Irritable bowel syndrome (IBS) is often a diagnosis of exclusion, with poor diagnosis coding in primary care. This makes identification of eligible research participants challenging. We present the methodology development of a multi-centre, observational, retrospective research study ongoing in primary care, designed to overcome the challenges of IBS patient identification. Methods Study feasibility was conducted by pH Associates (research consultancy; study coordinators) for Almirall UK Ltd (Sponsor) using medical opinion, clinical coding searches and NIHR Clinical Research Network expertise. FARSITE, a software tool for identification of research participants in primary care developed by the Greater Manchester Comprehensive Local Research Network and North West e-Health, was used to screen anonymised primary care records for potential eligible patients. Search criteria: patients aged 18–60; combination READ code symptoms indicative of IBS and prescription of IBS drugs 01/01/2009–31/12/2011. GP practices with eligible patients were invited to participate, with GPs reviewing clinical records of the FARSITE-generated list of patients to apply full eligibility criteria for final patient selection. Inclusion criteria: medical diagnosis of IBS or meeting ROME III criteria; provision of consent. Exclusion Criteria: diagnosis excluding IBS; IBS symptoms secondary to other condition; IBS medications for non-GI symptoms. The study is ongoing in 8 GP practices in Salford and Greater Manchester (Ethical approval 13/LO/0692). Results FARSITE feasibility search using READ code for IBS identified 50 (0.02%) patients. Combining READ codes with symptom and prescriptions criteria selected 4714 (1.9%) From these, 3 GP practices each screened 10 random patient records for eligibility and 12/30 (40%) were found eligible. Eligibility READ codes were revised following feasibility. Following study approvals, FARSITE identified 1089 potential eligible patients at the 8 participating practices, of which 297 (27.3%) were eligible and approached for consent for participation. Main reasons for non-eligibility were symptom characteristics not meeting ROME III criteria or not confirmed as IBS by medical opinion. Conclusion Identification of patients with IBS using READ code is sub-optimal in primary care. A combination search of READ codes with symptom and prescription data via FARSITE has enabled potential participants to be identified with a reasonable screening failure rate. FARSITE is a valuable research tool aiding study feasibility by reducing the need for manual patient identification. Disclosure of Interest I. Caldwell: None Declared, J. Collins: None Declared, M. Rance Employee of: Almirall UK Limited, R. Dew Conflict with: Commissioned by Almirall UK to provide research design, conduct analysis and scientific editorial services.

Introduction Previous studies have assessed the relationship between body mass index (BMI) and healthcare use in primary care settings but have been limited to individuals with pre-defined diseases or those with a recorded diagnosis of malnutrition. 1–3 This study using data from The Health Improvement Network (THIN) database (UK) aimed to examine GP-related healthcare use and hospital admissions according to two BMI categories (underweight and normal weight) across all individuals registered to GP practice. Methods Two cohorts of individuals registered with GPs were randomly selected from the THIN database (Cohort 1 BMI ≤19.9 kg/m 2 n2632, Cohort 2 BMI 20.0 to ≤24.9 kg/m 2 n2632). Cohorts were matched for age (52 y (22.4)), gender (81% female) and GP practice. Individuals were included if they had at least two BMIs recorded from May 2010 to May 2011 and were registered for at least 12 months. Individuals were excluded if they; were living in institutions, had conditions affecting weight for example, oedema, had a BMI 2 , or receiving palliative/end of life care. GP-related healthcare use (practice visits, home visits, out of hours visits, telephone consultations) and hospital admissions over 1 year were retrieved using read codes and locate flags; and a simple cost analysis was undertaken using unit costs. 4 Results A significantly greater proportion of individuals with a BMI ≤19.9 kg/m 2 had at least one; home, out of hours visit or telephone consultation with the GP practice (Abstract OC-038 table 1), and had significantly more practice visits than individuals with a BMI 20.0 to ≤24.9 kg/m 2 (8.5 vs 8.3 visits respectively p=0.022 Mann–Whitney U). Similar results were seen for hospital admissions, with a significantly greater proportion of individuals with a BMI ≤19.9 kg/m 2 being admitted to hospital (Abstract OC-038 table 1). Compared to individuals with a BMI 20.0 to ≤24.9 kg/m 2 , mean GP contact costs per person per year were increased by £29.54 and admissions by £34.62 per year. 5 Conclusion This study using data from THIN shows that individuals registered with their GP, with a low BMI (≤19.9 kg/m 2 ) use significantly more healthcare resources than those with a normal BMI (20.0 to ≤24.9 kg/m 2 ). The contribution of disease types and severity, social and nutrition-related factors needs to be further evaluated. Competing interests K Ashman Employee of: Nutricia, A Cawood Employee of: Nutricia, R Stratton Employee of: Nutricia. References 1. Martyn CN , et al. Clin Nutr 1998; 17 :119–23. 2. Edington J , et al. Proc Nutr Soc 1999; 58 :655–61. 3. Guest JF , et al. Clin Nutr 2011; 30 :422–9. 4. Curtis L . Unit cost of health and social care. 2010.