PTU-052 Clinical factors predicting outcome for double balloon enteroscopy: experience from a tertiary centre

Introduction

Double balloon enteroscopy (DBE) has revolutionised our ability to manage small bowel pathology, however there is paucity in the literature on factors that influence outcome. Our aim was to evaluate factors that contributed to improving diagnostic yield and subsequent management.

Method

We performed a retrospective review of patients who underwent DBE from July 2006–October 2014. Data was collected on demographics, indications, procedural details, diagnosis and subsequent management. Tolerability scores were assessed using a validated scoring system (1-no discomfort 2- slight discomfort 3-extremely uncomfortable 4–unbearable).

Results

A total of 399 DBE procedures (mean age 56; 50% male; anterograde n = 256, retrograde n = 143) were performed over 124 months. The commonest indication was obscure gastrointestinal bleeding (OGB, n = 201) and suspected Crohn’s disease (n = 122). Capsule endoscopy (CE) was done prior to DBE in 91% of patients. One patient had a small bowel lesion noted on DBE, which was missed on CE. Demand for DBE has risen gradually in tandem with the diagnostic yield (p < 0.0001, r = 0.9). The diagnostic and therapeutic yield across all indications was 49% and 24% respectively. On univariate analysis, cardiovascular co-morbidity was associated with a higher diagnostic yield (60% vs 45%, p = 0.01) and therapeutic yield (43% vs 17%, p < 0.0001), irrespective of age, possibly due to a higher proportion of vascular lesions in this group. On multivariate analysis, transfusion dependence predicted a positive yield (p = 0.04, CI 0.04–0.95). Duration of procedure (p = 0.2), sedation (p = 0.9) and tolerability (p = 0.3) did not make a difference. The median length of small bowel examined was greater for the anterograde route compared to the retrograde route (170 cm vs 100 cm; p < 0.001). The retrograde route took longer (75 mins vs 60 mins; p < 0.001), but was tolerated better (patient score >3 in 14% vs 24%; p = 0.02) and required lower doses of both midazolam (median: 5 mg vs 6 mg; p = 0.001) and fentanyl (median: 62.5 mcg vs 87.5 mcg p = 0.2). Change in management occurred in 41% of procedures and factors associated with this was age >70 years (p = 0.01). Small bowel tumour was diagnosed in 5% (n = 19) of patients, whilst 40% were subsequently treated for Crohn’s disease. The overall complication rate was 1.25% (pancreatitis n = 2, respiratory arrest n = 1, pulmonary oedema n = 1 and NSTEMI n = 1).

Conclusion

Transfusion dependence was associated with a positive diagnostic yield for DBE whilst the elderly were more likely to have a change in management post DBE. Cardiovascular comorbidity also positively influenced the diagnostic and therapeutic yield. Advocating CE as a first line investigation allows better patient selection. DBE remains a useful modality for small bowel pathology.

Disclosure of interest

None Declared.