BACKGROUND: In patients with obscure gastrointestinal bleeding (OGB), capsule endoscopy (CE) is the investigation of choice after negative conventional endoscopy. However, there is paucity of data on which clinical factors predict the ability of CE to detect pathology and alter management. AIM: To investigate clinical factors which predict a higher diagnostic yield and a change in management in patients presenting with OGB. METHODS: Data was collected in patients with OGB for type of OGB [overt (OB) or iron deficiency anaemia (IDA)], co-morbidity, use of anticoagulants, transfusion dependence, subsequent CE diagnosis and follow up data. Factors that predicted a higher yield and a change in management were identified using logistic regression. RESULTS: There were 427 patients (IDA 74%) with a mean age of 58 years (range 17-92).The diagnostic yield was 50% with a change in management in 30% of patients. The commonest finding was angiodysplasia in 52% of patients. There was no significant difference in the diagnostic yield between the two groups (OB: 56%, IDA 48%, p=0.15). On logistic regression, age (p=0.001), the use of warfarin (p=0.02) and liver co-morbidity (p=0.001) were significant factors which predicted a higher yield with CE. Factors which predicted a change in management included the presence of co-morbidity (p=0.02) and having a diagnosis of angiodysplasia on CE (p=0.001). CONCLUSION: This is a large series of CE in OGB. Increasing age, anti-coagulation and liver co-morbidity are significant predictors of a positive diagnostic yield whilst the presence of co-morbidity or a diagnosis of angiodysplasia may predict a change in management.
BACKGROUND: In patients with obscure gastrointestinal bleeding (OGB), capsule endoscopy (CE) is the investigation of choice after negative conventional endoscopy. However, there is paucity of data on which clinical factors predict the ability of CE to detect pathology and alter management. AIM: To investigate clinical factors which predict a higher diagnostic yield and a change in management in patients presenting with OGB. METHODS: Data was collected in patients with OGB for type of OGB [overt (OB) or iron deficiency anaemia (IDA)], co-morbidity, use of anticoagulants, transfusion dependence, subsequent CE diagnosis and follow up data. Factors that predicted a higher yield and a change in management were identified using logistic regression. RESULTS: There were 427 patients (IDA 74%) with a mean age of 58 years (range 17-92).The diagnostic yield was 50% with a change in management in 30% of patients. The commonest finding was angiodysplasia in 52% of patients. There was no significant difference in the diagnostic yield between the two groups (OB: 56%, IDA 48%, p=0.15). On logistic regression, age (p=0.001), the use of warfarin (p=0.02) and liver co-morbidity (p=0.001) were significant factors which predicted a higher yield with CE. Factors which predicted a change in management included the presence of co-morbidity (p=0.02) and having a diagnosis of angiodysplasia on CE (p=0.001). CONCLUSION: This is a large series of CE in OGB. Increasing age, anti-coagulation and liver co-morbidity are significant predictors of a positive diagnostic yield whilst the presence of co-morbidity or a diagnosis of angiodysplasia may predict a change in management.