Pan-intestinal capsule endoscopy (PIC) examines both the small bowel and colon noninvasively in a single test. PIC has been proposed as a first line diagnostic examination in the setting of suspected mid-lower gastrointestinal bleeding (MLGIB). as an alternative to performing a colonoscopy first and then a small bowel capsule. MLGIB is defined as those with iron-deficiency anemia (IDA) and either suspected occult (FIT) or overt gastrointestinal bleeding after a nondiagnostic gastroscopy. 

Aim to evaluate the efficacy of a PIC first approach in patients with suspected MLGIB. 

Following local ethics committee approval consecutive adult patients with IDA and overt or occult GI bleeding were prospectively enrolled. Study participants provided a faecal immunological test (FIT, results expressed in ng/mL) and baseline bloods prior to undergoing a gastroscopy followed by PIC on the same day for outpatients and after a short interval for inpatients. Basic demographics , clinical data , endoscopy and capsule findings were recorded. Key outcomes analysed included diagnostic yield (potential haemorrhagic lesions ( PHLs) – small bowel and colon) , safety , capsule performance measures including transit and image quality. Analysis of potential predictors of performance was also performed. Those with hematemesis, fresh anorectal type bleeding, known diagnosis of inflammatory bowel disease, decompensated chronic liver disease, known small bowel disease, those who had undergon gastroscopy or colonoscopy within 1 year, those with any contraindication to capsule endoscopy were excluded. 

In all 77 patients have been consented to date. Of these 10 did not proceed to either gastroscopy or PIC ( 2 failed patency capsules ,3 withdrew consent for PIC, 1 declined bowel preparation , 2 have had OGD/Colonoscopy in different institution, 1 lost follow up and 1 patient did not attend for their procedure). A further 21 are awaiting either procedure. A total of 46 have completed both procedures and results are available for analysis in 43 patients.    

Of these 36 (84%) presented with IDA and 7 (16%) with overt bleeding. Patients had a mean age of 64.8 years (range 32-88), and 24 (56%) were female. Of these 17/43 (40%) were receiving blood thinners and use was similar in both overt and occult groups. Mean haemoglobin level = 9.8 g/dL ( range 5.6-13.3 g/dL). Haemoglobin levels were similar in overt and occult groups; 8.9g/dL and 9.9g/dL respectively. A FIT was available in 28/36 (78%) of patients with occult bleeding. The mean FIT was 102 ng/mL (median 13, range 0 - ³1000). 

PIC performance was as follows; complete transit rate 67% (29/43), adequate image quality rate also 67%( 29/43) and 56% (24/43) had both adequate preparation and complete transit. 

PHLs were detected in 21% (n=9) of patients; 2/7 (29%) in  the overt group vs 7/36 (19%) in the occult group, p=0.6. PHLs were detected in the small bowel in 5 (12%) patients, (3 angiodysplasias , 1 bleeding jejunal diverticulum and 1 small bowel stricture) and in the colon in 4 (9%) patients (2 colorectal cancers, 1 large >2cm polyp and 1 significant radiation proctitis). Of note 15/43 (35%) required either completion endoscopy or therapeutic endoscopy either to obtain histological diagnosis or treat PHLs. 

There was a trend for a higher positivity in older subjects and those on a blood thinner which did not reach statistical significance. There was a moderate correlation between FIT level and positive PIC, r = 0.47, p=0.01, 95% CI 0.13-0.71. ROC analysis suggest a FIT level of ³ 24 ng/mL, has a sensitivity of 100% and specificity of 62.5% with AOC 83 for a positive PIC. 

Early analysis of our ongoing study of PIC in suspected MLGIB has found a high potential haemorrhagic lesions detection rate (21%), including colorectal cancer, despite issues with overall performance. PHLs were evenly distributed between the small bowel and colon highlighting the potential benefit of this modality. FIT appears to strongly predict significant disease.