Article Text

  1. J. Robinson2-1,
  2. P. Small2-1,
  3. G.D. Bell2-2,
  4. I. Crighton2-1,
  5. J. Corson2-1,
  6. J. Hancock2-2,
  7. D. Nylander2-2,
  8. H. Mitchison2-2
  1. 2-1Depts of Surgery and 2-2Gastroenterology , Sunderland Royal Hospital, Sunderland SR4 7TP, UK
  1. S.D. Johnston,
  2. A. McKenna,
  3. T.C.K. Tham
  1. Division of Gastroenterology, Ulster Hospital Dundonald, Belfast, N. Ireland, UK
  1. J. Gibson,
  2. P. Evans,
  3. J. Featherstone,
  4. J. Galloway,
  5. R. Spence,
  6. P. Willoughby
  1. Endoscopy Subcommittee PCSG
  1. A. Douglass,
  2. P. Cann
  1. South Cleveland Hospital Endoscopy Unit, Middlesbrough, UK
  1. H.C. Mitchison,
  2. M. Rutter,
  3. G.D. Bell
  1. Dept of Gastroenterology, Sunderland Royal Hospital, Sunderland, UK
  1. P. Mairs,
  2. P. Wheeler
  1. J. Meenan,
  2. T. Wong,
  3. G. Rottenberg
  1. Endosonography Centre, Guys' and St. Thomas' Hospital, London, UK
  1. U. Dave11-1,
  2. A. Williams,
  3. D. Larkman,
  4. D. Gilderdale,
  5. Z. Amin11-2,
  6. M.R. Thursz11-1,
  7. S. Taylor-Robinson11-1,
  8. N. DeSouza
  1. Robert Steiner MR Unit, Hammersmith Hospital;11-1Depts of Medicine and 11-2Radiology, Imperial College School of Medicine, London, UK
  1. M. Priest,
  2. A.J. Morris
  1. Dept of Gastroenterology, Glasgow Royal Infirmary, Glasgow, UK
  1. S. Mahadeva,
  2. M. Lynch,
  3. D. Mawer,
  4. S. Babu,
  5. M. Hull
  1. Division of Gastroenterology, St. James's University Hospital, Leeds, UK

Statistics from

Aims: The aim of this retrospective study is to identify whether there is a difference in the pattern of presentation of “ulcer-cancers” and “mass-cancers” and so identify any potential to improve the detection of gastric cancer at an early stage.

Methods: All upper GI adenocarcinomas diagnosed in South Tees Health District (population ∼ 350,000) were identified from the computerised pathology database (for the period April 1991- January 2000). The pathology records and GP records were then scrutinised and the data recorded and analysed using EpiInfo.

Results: The total number of patients identified was 618; to date 102 patients' notes have been analysed. 94% were adenocarcinoma of which 26% were oesophageal and 74% were gastric. 6% were squamous carcinoma (mis-coded on the pathology computer). The M:F ratio was 63.5:36.5 and the mean age was 70.2 years (range 43–91). For these the nature of diagnostic investigation was 98% OGD (routine list 69.8%, emergency 17.7%, open access 12.5%) and 2% laparotomy. The naked eye appearance at diagnostic OGD was ulcer in 49%, mass in 29%, stricture in 18%, mucosal abnormality in 2% and other in 2%. Of the “cancer-masses” (n=28) 10 were treated with acid suppressing medication (AST) compared with the “cancer-ulcers” (n=47) where 34 were given treatment (p=0.002). The mean time to diagnosis for the “cancer-masses” given AST was 9.9 weeks (range 0–26); if no AST had been prescribed it was 7.4 (range 0–45). For the “cancer-ulcers” given AST this time was 18.7 weeks (range 3–107) and if AST had not been prescribed it was 8 (range 0–26). Overall the mean time from 1st GP consultation to diagnosis for “mass-cancer” was 8.2 weeks (median 6, range 0–45) whereas for “ulcer-cancers” it was 15.7 weeks (median 10, range 0–107) (p=0.028 Kruskal-Wallis analysis). Of the 28 “mass-cancers” 4 (14%) had had a gastroscopy within 3 years of the diagnostic scope whereas of the 47 “ulcer-cancers” 15 (32%) had had previous investigations (p=0.09).

Conclusion: The early results suggest that “ulcer-cancers” receive more treatment (up to 2 years) and are more likely to have had a gastroscopy within 3 years of the diagnostic investigation at which the cancer was diagnosed. There is therefore more potential to diagnose “ulcer-cancers” earlier by avoiding AST prescription and performing gastroscopy earlier.


Introduction: Colonoscopy is acknowledged as the gold standard for examining the colon. The goal of colon cancer prevention is interrupting the adenoma-carcinoma sequence and effective management of polyps requires complete colonoscopy with complete polyp removal, retrieval and histological assessment to enable appropriate follow up. As part of the IBNC prospective audit of 8902 colonoscopies, the prevalence and management of polyps was assessed.

Results: Polyps were diagnosed in 2014/8902 (22.6%) of the colonoscopies. Polyps were the sole diagnosis in 1309/8902 (14.7%) of colonoscopies. In 84/8902 (0.9 %) colonoscopies, polyps were reported in the presence of obvious cancer, and in 621/8902 (7.0%) polyps occurred in the setting of other disorders. The caecum was successfully intubated in 1670/2014 (82.9%) colonoscopies where polyps were reported and in 54/84 (64.3%) colonoscopies where polyps were associated with cancer. No answer regarding intervention was given in 59/2014 (2.9%) of colonoscopies. Where polyps were reported, there was no attempt at polypectomy in 126/2014 (6.3%). All polyps were judged completely removed in 1401/2014 (69.6%) colonoscopies and incomplete removal was recorded in 331/2014 (16.4%). In 1239/2014 (61.5%) of cases information on polyp retrieval was not answered. Retrieval of all polyps was reported in 465/2014 (23.1%) of colonoscopies where polyps were discovered and incomplete retrieval was recorded in 310/2014 (15.4%).

Conclusion: Currently, colonoscopy is incomplete in almost 1:5 patients with polyps and it is likely that a significant number of proximal polyps remain undetected. A small but significant number of polyps are either left in situ or only partly removed, and retrieval is often incomplete. It is likely that improved performance will require a program of intensive training and repeated audit cycles.


Background: JAG guidelines state that colonoscopists need to have a caecal intubation rate in excess of 90% to be qualified to train but does not state if this success rate based on an intention to treat (ITT) or per protocol (PP) analysis. In a previous audit (Gut 2000;46 (suppl II) A15) our hospital's overall caecal or surgical anastomosis intubation rate (COSAIR) was only 67%(ITT). We therefore carried out a further more detailed prospective 5 month audit to try to identify factors effecting COSAIR (analysed both on an ITT and PP basis). Endoscopists graded reaching the caecum as 1) essential 2) desirable or 3) optional.

Results: 5 consultant surgeons and 3 medical gastroenterologists plus their junior staff including 4 SpRs (3 surgical and one medical) carried out a total of 534 colonoscopies. In the 67 patients who had had a previous colonic resection the COSAIR was 66/67 or 98.5% while in those with an intact colon the success rate was 388/467 or 83.1%. Our overall COSAIR[ITT] was 454/534 or 85.1%. The PP figures were significantly better than ITT (p=0.008) with 4 instead of 2 consultants achieving a >90% COSAIR. The failure rates for females (54/274) was significantly greater (p=0.0025) than men (26/260). Furthermore, in women, the COSAIR with the stiffer Pentax Shinya colonoscope (57/91) was significantly lower (p=0.0043) than with floppier Olympus endoscopes (109/129).COSAIR was significantly (p<0.001) influenced by perceived desirability to reach the caecum: with success rates of 93%, 86% and 45% respectively in the essential, desirable and optional groups.

Conclusion: JAG recommendations on colonoscopy should clearly state if the “being able to reach the caecum in over 90%” is a) assessed on a PP or ITT basis b) includes patients with colonic resection and c) takes into account the perceived need to reach the caecum.


Introduction: Colonoscopy is acknowledged as the gold standard for examining the colon. A caecal intubation rate of 90% is the recommended target for practising colonoscopists. As part of the IBNC audit caecal intubation rate was recorded in 8902 colonoscopies. In addition, commonly used method(s) for identifying the caecum were recorded (transillumination, tri-radiate fold, appendiceal orifice, ileo-caecal valve, terminal ileum intubation, finger indentation of right iliac fossa, screening with image intensifier). Visualisation of the ileo-caecal valve or terminal ileum intubation are recognised as the most reliable indicators of successful caecal intubation.

Results: Caecal intubation was reported in 6864/8902 (77.1%) colonoscopies. If terminal ileum intubation or ileo-caecal valve visualisation are considered as the only reliable methods for identifying a complete procedure, the ‘adjusted’ caecal intubation rate falls to 5085/8902 (57.1%). Methods used for identifying the caecum included; transillumination 2364/6864 (34.4%), tri-radiate fold 4809/6864 (70.1%), appendiceal orifice 2942/6864 (42.9%), ileocaecal valve 4661/6864 (67.9%), terminal ileum intubation 1368/6864 (19.9%), right iliac fossa indentation 3057/6864 (44.5%) and screening 42/6864 (0.6%). The most common reasons for incomplete colonoscopy were “patient discomfort” and “looping of the colonoscope”.

Conclusion: Currently, the caecal intubation rate falls far short of the recommended 90%. In addition, caecal intubation is often recorded in the absence of positive identification of the ileo-caecal valve. Colonoscopists should be aware of the best methods for identifying the caecum. Failure to complete colonoscopy is most often ascribed to patient discomfort and looping and formal training and assessment in colonoscopy is required to reduce the number of technique-related failures.


Background: Myocardial ischaemia (defined as ST depression > 1mm) may predispose patients to ischaemic complications, for example, arrhythmias or procedure-related mortality. However few studies have evaluated its frequency and cause during ERCP.

Aim: To determine the frequency of myocardial ischaemia as defined by ST changes on electrocardiograph (ECG) during ERCP.

Methods: Consecutive patients undergoing ERCP in a single centre were included. Continuous Holter ECG recordings were carried out during ERCP along with pulse and blood pressure recordings. ST depression was defined as > 1mm lasting for longer than 1 minute. The timing of the administration of sedation, cannulation of the bile duct and any therapeutic procedures were recorded.

Results: There were 40 patients in the study (mean age 65 years, range 21–88; 14 males, 26 females). 16 had bile duct stones, 6 had malignant biliary strictures. 20 had sphincterotomy, 12 had stent insertion. Sedation comprised of midazolam, buscopan ± pethidine. Nine patients (22%) experienced ST depression during the procedure, of whom 4 showed significant ST depression (> 2 mm). None of the patients had any ischaemic symptoms during these events. Of these, 5 had no previous cardiac history and a normal ECG, 3 had no previous cardiac history but had abnormal ECG and 1 had a previous cardiac history and an abnormal ECG. ERCP procedures during the episodes of ST depression were administration of sedation (n=6), endoscopic sphincterotomy (n=4), balloon trawl (n=1), basket trawl (n=3), stent replacement (n=1).

Conclusions: Myocardial ischaemia occurred in approximately one quarter of patients during ERCP and over half of these had no previous cardiac history and a normal baseline ECG. Myocardial ischaemia often accompanied the use of sedation for ERCP.


Background: Long hospital waiting lists and savings from fundholding have contributed to an increase in endoscopy services in primary care. Concerns about safety, supervision and cost effectiveness have been voiced but the DH has also floated ideas about intermediate care centres.

Method: PC endoscopy units were identified via the Primary Care Society for Gastroenterology and appeals in GP journals. A postal questionnaire was sent seeking details about the history of primary care endoscopy unit, throughput, case mix, endoscopists' experience, supervision, audit and CME, waiting times, complications, equipment and reporting strategies.

Findings: 27 (96%) of 28 units identified replied. 13 provided both OGD and lower GI examinations, 6 OGD only, 8 lower GI only. The units had been in existence on average for 5 years (range 2–18 years). Services were provided by a total of 41 GPs and 68 nurse assistants. The average experience of endoscopists was 11 years. 96% of units undertook audit. Average “urgent” waiting time was 1–2 weeks; “routine” 3–4 weeks (range 1–6 weeks).

Total procedures performed by all units, to date 34,959: of these 24,195 were OGD and 10,764 lower GI. Total annual throughput for OGD (12 units) 4,506 and lower end (10 units) 1,305.

Complications: For OGD (24,195 procedures) one perforation of crico-pharyngeal pouch, admission for drip and suck; one admission to A&E with chest pain - discharged within one hour (“trapped wind”); one delayed discharge due to slow recovery after IV sedation. No mortality reported. For lower GI examinations (10,764): 1 death (caecal blow-out, carcinoma) after flexible sigmoidoscopy; 6 hospital admissions (3 perforations after colonoscopy); 4, delayed discharge.

Discussion: Community endoscopy is well established, having developed over 18 years, with experienced endoscopists. The survey also indicated that OGD in the community is safe, and contrasted well with hospital practice. This good safety record may be related to careful case selection and the low use of intravenous sedation.


Trusts have been under increasing pressure to reduce waiting times for all specialities. In Gastroenterology, one solution has been to provide extra ‘out of hours’ sessions, which rely upon the goodwill of medical & nursing staff involved and raise a considerable extra cost. The first year waiting list initiative (WLI), in our unit, has been to direct the resource towards the open access Colonoscopy (OAC) service, but selecting patients 45 years or older. This was driven by the need to move towards and meet the “2 week rule” requirement.

Results: Over the first 12 months (Feb 99–Jan 00), 44 additional lists were performed on Saturday mornings. 256 of the 263 patients endoscoped were 45 yrs +. Only 7 failed to attend (DNAs). The commonest symptoms were rectal bleeding (71), diarrhoea (67), constipation (32) or both (8). 120 (46%) of colonoscopies were entirely normal. 23 (9%) had adenomatous polyps (19 tubular/4 villous), 10 (4%) had carcinoma and 6 (2%) had inflammatory bowel disease. 14 patients had ‘dual pathology’. 6 (2%) (out of 43) had haemorrhoids requiring injection at the time. We count, therefore, 45 (17%) as having a “useful” procedure. The estimated cost, in extra salaries alone, was £29,000. This means about £650 per “useful” diagnosis. The OAC waiting time has fallen to half of the peak level before the WLI, despite a steady referral load.

Abstract 39, Table 1

Conclusion: Our “Waiting List Initiative”, directed towards OAC patients above 45, gives a high diagnostic yield with modest costs. It has been associated with a very low DNA rate. It has also been effective at decreasing the overall OAC waiting time and the numbers waiting.


Background: Maintaining GI outpatient (GOP) targets and keeping clinic size manageable, requires the number of new patients (NP) plus follow up patients (FUP) to be balanced by clinic discharges. Royal College of Physicians Guidelines (RCPG) on a) consultant OP workload and b) SpR training recommend 15–20 “units” per clinic (i.e. 6–8 NP or 15–20 FUP) for consultants and no greater than half to 2/3 this number for SpRs.

Methods: In our DGH (population served 330,000), we prospectively categorised GOP patients by diagnostic/symptom group for 6 months. Retrospectively we examined workload plus compliance with a) RCPG and b) the Patient's Charter.

Results: The 2 consultants averaged 3 clinics per week, the single SpR two. There were 587 true NPs plus 246 NPs who were endoscoped and then seen in GOP (ENP). The NP+ENP/FUP and NP/FUP ratios were 3.7 and 2.5 respectively. In toto 604 patients were discharged. Of 2209 patients receiving appointments, 210 did not attend. The numbers receiving 1,2,3,4 and 5 appointments were 1889, 227, 39, 12 and 8 respectively .The discharge rates for IBD (n=425) and Liver Disease patients (n=237) were low at 3% and 10% compared with those with functional bowel symptoms (41%) or reflux (41%).In these latter two groups of NP 32% were discharged at their first GOP visit. Despite both Consultants and the SpR working unacceptably high “units” per clinic, 25% of patients were kept waiting over 30 minutes beyond their appointment time.

Discussion: The study reveals how the balance between NP and FUP in an under staffed GOP was maintained. Prompt discharge of those in most diagnostic categories was essential to provide adequate time for the long-term care of chronic conditions such as IBD and liver disease. We would have required a further 3.2–5.9 GOP sessions per week to meet RCPG. We have now appointed a third consultant but experience already suggests a fourth consultant will be needed to meet service and training requirements.


This retrospective study looks at cumulative experience from a district general hospital in the south east of England in order to determine if the present age cut off for open access endoscopy in patients with dyspepsia of 45 could reasonably be increased to 55 without the risk of missing any gastric malignancies.

Methods: Since 1995 all endoscopies carried out on the endoscopy unit at the above hospital have been entered into the computerised database for reporting and audit. This database was searched to determine how many patients were being diagnosed with gastric cancer at endoscopy. The age range of patients being gastroscoped was analysed. To ensure that no patients were overlooked if for example their malignancy was diagnosed at open operation or for some reason they had not been entered onto the database a separate analysis was made of the hospital pathology department database. Over the period January 1995 to April year 2000: 6951 patients underwent gastroscopy. 5824 of these were aged over 45. 2228 of these were aged less than 55. Review of the pathology data confirmed that no patients under 55 were diagnosed with cancer of gastric antrum and/ or body at gastroscopy. All these patients were aged over 55 years.

No patients aged less than 50 were diagnosed with any malignancy (Cancer of gastric antrum and/ or body, cancer of the cardia, cancer of the oesophagus). In the study period a total of 7 patients aged between 50 and 55 were found to have a cancer of the gastric cardia or lower oesophagus. Notes were available on four of these and all had presented with symptoms of reflux and weight loss. Two had presented with dysphagia. A shift in the age range for open access endoscopy from 45 to 55 would not have missed one malignancy of the gastric antrum or body in the 5 year period studied. This would be in keeping with previous data. Those patients in whom malignancies were found were invariably aged greater than 50 and would have been candidates for open access endoscopy on the basis of alarm symptoms. All had cancer of the cardia or distal oesophagus. Provided guidelines emphasize the importance of alarm symptoms then it seems reasonable to lift the age cut off for open access endoscopy to 55. Gillen D. McColl KE: Does concern about missing malignancy justify endoscopy in uncomplicated dyspepsia in patients aged less than 55?Am J Gastroenterol1999;94:75–9.


Introduction: “Submucosal” tumours (SMT) in the upper gastrointestinal tract (UGI) maybe hard to evaluate. Endoscopic ultrasound (EUS) can provide high quality information about the nature of SMT and assist in their management.

Aims: To assess the use of EUS in the evaluation and management of UGI SMT.

Methods: We evaluated 41 consecutive patients who were referred with UGI SMT for EUS. All examinations were performed by one of two experienced endosonographers.

Results: The average age of the patients was 62.2yrs, 54% were female. The indications for endoscopy were; UGI bleeding 22%, dyspepsia/abdominal pain 27%, dysphagia 12%, anemia 12% and presumed co-incidental finding in 27%. Most patients were referred for EUS with a presumed diagnosis of a gastrointestinal stromal tumour (GIST). The pre-EUS diagnosis, however, did not correlate with the EUS diagnosis in 51% of patients (table 1). Pathological correlation was possible in 41% of patients, EUS diagnosis was confirmed in all these patients. Based on EUS findings 13 patients underwent resection. Pathology confirmed EUS findings in these patients. None of these lesions was malignant. Three patients underwent EUS follow up and the rest were felt not to require further endscopic or EUS follow up.

Abstract 42, Table 1

Conclusion: EUS can assist more accurate diagnosis of SMT, increase physician certainty, and facilitate the management process.


Background: Endoscopic ultrasound (EUS) provides an accurate, safe and rapid method for visualizing the biliary tract. EUS obviates the risks of pancreatitis and inappropriate intervention associated with ERCP. The growing number of centres utilizing EUS raises the question of whether the risks associated with diagnostic ERCP can be justified. Consequently, previous assumptions about numbers presenting for ERCP must be questioned.

Aim: To determine the proportion of cases referred for ERCP that could be diverted to EUS imaging.

Methods: During the six months to October 2000, patient referrals for ERCP (n=417) were triaged with respect to the likelihood of therapeutic intervention being required. Those without jaundice, fever or, biliary pathology previously identified by trans-abdominal ultrasound/CT, underwent EUS (n=81, 19%) with conscious sedation (midazolam/ fentanyl) using Olympus GFUM20 and GFUM200 echoendoscopes. These cases represented 38% of the total 210 EUS performed in this six month period.

Results: The indications for EUS were: (a) Raised LFT's/pre-lap-chole, n=21; (b) Acute pancreatitis, n=16; (c) Post-cholecystectomy RUQ pain, n=10; (d) RUQ pain and raised LFT's, n=8 (e) RUQ pain, other tests normal, n=16 and (f) Abnormal LFT's, n=10. In two cases, EUS failed to identify the CBD (both normal on subsequent ERCP). In two further cases CBD stones identified at EUS were removed at ERCP. No complications occurred with EUS.

Conclusion: EUS is a safe and effective method for imaging the biliary tree. Approximately one-fifth of cases referred for ERCP can be assessed by EUS, raising the question of whether the risks associated with potentially diagnostic ERCP's can be justified. Furthermore, in light of case-number recommendations for centre recognition, the greater use of EUS imaging will have profound implications on ERCP tuition and increases the need for centres offering training in EUS.


Background: The current gold standard for pre-operative staging of oesophageal cancer, endosonography (EUS), has limited accuracy for TNM staging. MR endoscopy (Endo-MR) allows simultaneous visual information and high definition cross-sectional imaging using endocavitary coil mounted on a MR compatible endoscope. Preliminary experience has shown that Endo-MR is particularly useful for local staging of oesophageal and rectal carcinoma.

Aim: To study endo-MR prospectively in patients with suspected oesophageal cancer.

Method: Endo-MR was attempted in 11 patients (7 male, median age 61 yrs (range 30–78 yrs), using prototype MR endoscope (diameter 13 mm, biopsy channel 2.8 mm, umbilical cord length 5 meters) and 0.5 T MRI machine. Propofol was used for sedation in 9 patients. EUS and operative pathology findings were available in 3 and 5 patients respectively (after reporting of Endo-MR) and 2 patients are awaiting EUS.

Results: Symptoms included dysphagia in 6 patients. Histology findings included: invasive cancer (7), dysplasia (2) and glycogenic acanthosis (1 patient). All patients tolerated endo-MR well with no complications, but one patient failed to be intubated. Endo-MR findings on TNM staging (shown below) and tumour size matched well with EUS and operative pathology findings.

Abstract 44, Table 1

Conclusion: Endo-MR is safe and well tolerated. Endo-MR findings correlate well with EUS and operative pathology.


Introduction: Push enteroscopy is of proven value in detecting proximal small bowel disease. The diagnostic yield for small bowel neoplasia, in an unselected population referred for this procedure, has not been reported. We have undertaken a retrospective review of all push enteroscopies carried out in a single referral practice, over a 4 year period, to identify the clinical, pathological, and enteroscopic features of patients with small bowel neoplasia.

Results: 410 procedures were performed on 388 patients (146 male, 242 female). Small bowel neoplasia were detected in 18 (4.6%) patients (4 adenomas, 7 polyposis syndromes, 7 malignant tumours). The small bowel neoplasia (SBN) group (excluding those with known polyposis syndromes) were significantly older than the total population (TP) (mean SBN 64.4 (range 35 to 79) vs. TP 55.6 (range 10 to 92) years, p < 0.05). The indication for enteroscopy is shown in the table.

Abstract 45, Table 1

Seven patients in the SBN group underwent enteroscopy for the investigation of known polyposis syndromes (5 FAP, 1 Peutz-Jegher syndrome, 1 Cowden's syndrome). 7 malignant tumours were identified (3 lymphomas, 2 gastrointestinal stromal tumours, 1 carcinoid tumour and 1 adenocarcinoma), 4 (57.1%) of which were in patients with known coeliac disease. Only 3 lesions (27.2%) identified were within reach of the standard endoscope (2 adenomas, 1 stromal tumour), all other lesions were located in the proximal jejunum.

Conclusions: The overall yield of push enteroscopy for small bowel neoplasia was 4.6%. In contrast to other reports, patients with small bowel tumours were significantly older than the general population referred for enteroscopy. In our series small bowel tumours were more common in patients undergoing investigation of malabsorption or with coeliac disease than those with anaemia or occult gastrointestinal bleeding.


Background: Current endoscopic methods to image the small bowel are limited in patients with obscure GI bleeding (negative gastroscopy and colonoscopy), push-enteroscopy fails to find the source in up to 60%. Improved methods are required to visualise the entire small bowel which are less invasive and better tolerated.

Methods: A new wireless capsule endoscope (Given®M2A video capsule system) was used to assess 6 patients with obscure or uncontrolled GI bleeding after ethical committee approval had been given. The patients were fasted overnight and swallowed the capsule endoscope with a glass of water.

Case 1: Female aged with Hereditary Haemorrhagic Telangiectasia (HHT) and a transfusion requirement of 4 units/year despite hormonal therapy and endoscopic treatment. Capsule endoscopy (CE) revealed several telangiectasias in the stomach, duodenum and proximal jejunum. 2 large bleeding lesions were seen in the jejunum and ileum.

Case 2: Male aged 39 with HHT and a transfusion requirement of 5 units/month unresponsive to hormonal therapy or endoscopic treatment. CE revealed 8 moderately sized angiodysplasias in the duodenum and proximal jejunum. Good views of the large bowel to the sigmoid colon revealed 3 further telangiectasias not seen on a recent colonoscopy.

Case 3: Male aged 16 with a past episode of melaena presented with an episode of melaena, Hb 9 with normal endoscopy, colonoscopy and Meckel's scan. CE revealed no abnormality although very good images were obtained from the stomach to caecum.

Case 4: Male aged 78 with HHT and a transfusion requirement of 3 units/month despite treatment hormonal and endoscopic therapy. CE revealed multiple angiodysplasias affecting the proximal small bowel.

Case 5: Male aged 79 with HHT and a transfusion requirement of 4units/month. Multiple telangiectasias, including one actively bleeding in the stomach, were seen at CE and subsequently treated endoscopically.

Case 6: Male aged 65 with recurrent anaemia and melaena. CE revealed bleeding erosive gastritis probably due to aspirin ingestion.

Conclusion: CE provides good views of the GI tract. The CE was easy to swallow, completely painless and preferable to conventional endoscopy. Results have been helpful in directing the management in these patients.


Background and methods: Single centre RCTs suggest that endoscopic haemostasis (EH) can reduce rebleeding, surgery and mortality in patients with bleeding peptic ulcers. In order to determine whether such RCT results are applicable in general, we retrospectively examined endoscopic practice in a large UK Teaching Hospital over a 3 year period (1997–99). Patients were identified from an endoscopy database and the clinical records reviewed.

Results: 872 patients were identified with upper GI haemorrhage. 179 (21%) had an endoscopic diagnosis of bleeding peptic ulcer. 154 (86%) medical records were available for review. 72 (47%) patients had stigmata of bleeding at initial endoscopy (adherent clot, visible vessel or active bleeding). 56 (78%) patients had EH (75% Adrenaline, 25% Adrenaline plus other). 20/56 patients had active bleeding at endoscopy compared with 0/16 in the group that did not receive EH. The re-bleeding rate in the group which had EH was 36% (11/42 Adrenaline only, 9/14 Adrenaline plus other), compared with 31% if no EH was attempted. Of the EH group that re-bled, 10 patients had a further attempt at EH (4 later required surgery), 8 patients went immediately for surgery and 2 died within 24 hours. Amongst those that rebled without prior EH, 2 had EH (1 required surgery later), 2 had surgery immediately and 1 died early. In total, 15/72 (21%) patients with stigmata of peptic ulcer bleeding required surgery. The mortality rate (<30 days) in patients with stigmata of recent bleeding was 16%, while overall mortality for all bleeding peptic ulcer patients was 14%.

Conclusion: Only 78% of patients with stigmata of bleeding peptic ulcer had EH at initial endoscopy. The re-bleeding rate following EH was higher than available RCT data. Endoscopic management (such as repeat EH for re-bleeding) performed in experienced centres in the setting of a RCT requires validation in individual endoscopy units.

View Abstract

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.