Colonoscopy for unexplained rectal bleeding

Is diverticular disease associated with colonic malignancy?

Colon cancer and diverticular disease have common characteristics; there are increases in the incidences in both disease entities and these diseases are more common in the westernized world. There is also an increase in the age-specific incidence with advancing age. Similar dietary features have been implicated for both diseases and already during the 1960s it was postulated by Burkitt that there is an association. Observational studies initially were able to demonstrate that patients with a history of diverticular disease of the colon had an increased risk of colon cancer, especially in the left side. However, the results from these studies have not been consistent, and problems like selection bias and confounding by indication have been major drawbacks in order to interpret the results and infer causality. Recent studies, which have had a better assessment of diverticular disease by new diagnostic methods, do not support such an association to the same extent as previously. Moreover, surveillance bias has become an increasing problem as patients with diverticular disease of the colon are subjected to a higher diagnostic intensity than other individuals in a population-based setting. A critical evaluation of the studies published so far therefore clearly indicates that the proposed association between diverticular disease and colonic malignancy is not evidence based, which should have an impact on clinical practice as well as on how to deal with these patient groups within the realms of a screening program.

Diverticulosis and colorectal cancer: between lights and shadows

The prevalence of diverticulosis and colorectal cancer (CRC) is markedly increased in the last century. Both diseases are highly frequent in Western countries and in aged people. Western diet--low in fiber and rich in dietary fat--has been largely regarded to play a major role in the pathogenesis of both conditions. A causal relationship between diverticulosis and CRC has been suggested in different studies. Epidemiologic series found a more frequent rectosigmoid localization of neoplastic lesions (advanced adenoma and CRC) in patients with diverticulosis as compared with controls, particularly in those with a previous diverticulitis episode or with an extensive disease. However, data are still controversial, with other studies failing to confirm this observation. Such discrepancy could be referred to the highly heterogeneous study design and setting in the different epidemiologic series. Pathologic studies showed that either macroscopic and microscopic chronic inflammation--which is regarded as risk factor for CRC development--is present in the colonic mucosa of some patients with diverticula. Moreover, alterations in the extracellular matrix, also involved in colorectal carcinogenesis, have been depicted in diverticulosis. In addition, an upward shifting of cell proliferation occurs in diverticular mucosa, and in nondiverticular patients with advanced adenomas. Finally, aberrant crypt foci--which are considered potential markers of CRC risk in ulcerative colitis--have been detected in colonic mucosa of patients with diverticulosis. Despite this substantial amount of evidence, however, the available data are not yet strong enough to suggest a more aggressive CRC prevention in diverticular as compared with nondiverticular subjects.

Endoscopy for hematochezia in patients under 50 years of age

The optimal strategy for evaluating rectal bleeding in young persons is unknown. This study examines the prevalence of adenomatous neoplasms identified at endoscopy for rectal bleeding. Retrospective cross-sectional analysis was made of patients under 50 years of age undergoing elective outpatient colonoscopy or flexible sigmoidoscopy for hematochezia. In all, 570 patients (309 F/261 M) met our inclusion criteria. The prevalence of neoplasms was 3.8% (95% CI 1.2-8.5%) among persons under 30, 2.8% (95% CI 0.9-6.3%) among persons age 30-39, and 10.9% (95% CI 7.4-15.4%) among persons age 40-49. The prevalence of neoplasms was higher in persons over age 40 (relative risk 3.43, 95% CI 1.70-6.94). Six of seven advanced neoplasms were identified in persons over age 40 (relative risk 7.4, 95% CI 0.89-60.7). In conclusion, the prevalence of colonic neoplasms in patients 40-50 years old with hematochezia is substantial. Among those persons younger than 40 years, the prevalence of colonic neoplasms is significantly lower.

Defining the role of fiberoptic sigmoidoscopy in the investigation of patients presenting with bright red rectal bleeding

OBJECTIVE

This study was done to determine whether sigmoidoscopy could theoretically constitute sufficient investigation for some patients with bright red rectal bleeding.

METHODS

One hundred and forty-three patients undergoing investigative colonoscopy for bright red rectal bleeding and whose source of bleeding was identified were studied. The investigation took place in a large urban hospital over an 11-month period. Data obtained included changes in stool pattern, characteristics of the bleeding, lesions identified, and the distance of the lesion from the anus.

RESULTS

In patients younger than 55 yr, all serious lesions except for one malignancy in a patient with massive bleeding lay within 60 cm of the anus and theoretically within reach of the fiberoptic sigmoidoscope. The mixing of red blood with stool was commonly due to distal lesions, especially hemorrhoids.

CONCLUSIONS

In young persons with bright red rectal bleeding, fiberoptic sigmoidoscopy may prove to constitute appropriate initial investigation.

Occurrence and clinical significance of overt blood loss per rectum in the general population and in medical practice

While the occurrence of blood loss per rectum in general practice is common and usually not serious, it may be the first symptom of a colorectal malignant disorder. To determine the occurrence and clinical significance of overt blood loss per rectum a Medline literature search was undertaken. The incidence of this symptom in the general population was reported to be approximately 20 per 100 people per year, the 'consultation incidence' in general practice approximately six per 1000 and the incidence of referral to a medical specialist was estimated to be about seven per 10,000 per year. The clinical significance of the symptom varied within different populations: the predictive value of overt rectal blood loss for colorectal malignancy was estimated to be less than one in 1000 in the general population, approximately two in 100 in general practice and up to 36 in 100 in referred patients. How the manifestation of the bleeding (and other signs and symptoms) contributes to the selection of patients at each of these stages, leading to increasing prior probabilities, is unknown.

Double-contrast barium enema and flexible sigmoidoscopy for routine colonic investigation

Over a 3-year period, all patients referred for barium enema examination had a double-contrast barium enema and flexible sigmoidoscopy performed on the same day. A total of 462 joint examinations were performed. Abnormalities were found in 193 patients by the use of barium enema, 164 patients by using sigmoidoscopy and 294 by the use of both methods of investigation. Sigmoidoscopy was superior to barium enema in the detection of polyps and inflammatory bowel disease but barium enema was more sensitive for diverticular disease. The presenting symptoms had no predictive value in distinguishing carcinoma, polyps and diverticular disease. Diverticular disease did not reduce the sensitivity of barium enema examination to polyps in the sigmoid colon. Fibreoptic sigmoidoscopy immediately before barium enema was well tolerated by patients. The investigations were complementary in the diagnosis of colonic polyps, inflammatory bowel disease and diverticular disease.

Rectal bleeding and polyps

Colorectal polyps are an important albeit uncommon cause of rectal bleeding in children. Colonoscopy promotes both rapid and accurate diagnosis and the opportunity for immediate therapeutic polypectomy. A 10 year audit of polyps diagnosed and treated endoscopically has been undertaken in the children's endoscopy unit. Twenty nine polyps were diagnosed from 730 colonoscopies; 24 were juvenile, two inflammatory, two Peutz-Jeghers, and one an adenomatous polyp. All but one of the juvenile polyps were solitary. All children had bleeding per rectum as one of the major presenting features. About two thirds of the patients were under the age of 5 years; the mean age was 5.6 years. Most of the juvenile polyps were on the left side of the colon; 41% were distal to the sigmoid colon. However polyps were found throughout the colon, indicating that total colonoscopy is wise and rewarding in any child with persistent and intermittent rectal bleeding.

Gastrointestinal endoscopy: an accurate and safe primary diagnostic and therapeutic modality

OBJECTIVE

To review the place of gastrointestinal endoscopy in the management of upper and lower gastrointestinal disorders.

DATA SOURCES

We reviewed articles on endoscopy reported over two decades. A Medline search complementing our experience and knowledge of the literature was used to identify the articles.

STUDY SELECTION

Papers were selected which focused on indications, comparison with radiology, including clinical outcome measures, and complications. One hundred papers, including those from radiology journals, were reviewed.

DATA EXTRACTION

Results of studies are referenced as appropriate.

DATA SYNTHESIS AND CONCLUSIONS

Endoscopy allows direct visualisation of the mucosa of the upper gastrointestinal tract, colon and terminal ileum. Subtleties of colour change, vascular pattern abnormalities and scarring are easily detected at endoscopy and are often of diagnostic importance. Endoscopy also provides access for tissue biopsy and allows a wide variety of therapeutic interventions. Traditionally barium studies have been the first step in the evaluation of many gastrointestinal symptoms and still retain cost advantages over endoscopy. However, endoscopy is frequently more sensitive and specific than barium studies. Costs associated with incorrect diagnoses may undermine the apparent cost benefits of barium studies. Advances in endoscopic design have allowed wider therapeutic options and increased safety. Gastrointestinal endoscopy should now be the first line of investigation where diagnostic precision is required or where therapeutic intervention is likely.

Endoscopic crescentic fold disease of the sigmoid colon: the clinical and histopathological spectrum of a distinctive endoscopic appearance

In this paper we describe an endoscopic appearance of the sigmoid colon characterised by mucosal swelling, erythema and haemorrhage strictly localised to the crescentic mucosal folds. In a 5-year period these changes were seen in 34 (1.42%) of 2380 colonoscopies and fibreoptic sigmoidoscopies. The majority of patients were middle-aged or elderly. Diverticular disease was present in most (82%) but the abnormalities were confined to the crescentic mucosal folds with sparing of the diverticular orifices. The majority of patients presented with a history of bleeding per anum. Histologically there was a spectrum of changes varying from minor vascular congestion to florid active inflammatory disease with crypt architectural abnormalities mimicking ulcerative colitis, but rectal biopsies were invariably normal. Three patients later progressed to typical distal ulcerative colitis and two other patients presenting to us with endoscopic crescentic fold disease had a previous histologically documented history of distal ulcerative colitis. In three patients the histological features were of mucosal prolapse. About half the patients required some form of therapy to control their symptoms. Steroids and/or sulphasalazine were of value although two patients subsequently underwent sigmoid resection, one to control bleeding and the second for a diverticulosis-associated stricture. Whilst endoscopic crescentic fold disease represents a specific endoscopic appearance the clinical and histological features indicate a wide spectrum of disease.

Colonoscopy in patients aged 80 years or older and its contribution to the evaluation of rectal bleeding

Colonoscopies performed in patients aged 80 years or older at the Sheba Medical Center were analysed according to the primary indication for the procedure: 101 colonoscopies were performed because of rectal bleeding of at least 2 months duration, and 335 for all other indications. Carcinoma of the large bowel was found in 29 (28.7%) bleeders, with the rectum being the most frequently involved site (12 patients). Among the non-bleeders, the prevalence of cancer was significantly lower (33 cases, 9.8%; P less than 0.001), and rectal carcinoma was less common (five patients, P = 0.04), but proximal tumours were more frequent. Of patients with cancer who had operations, the majority (72%) had a tumour confined to the bowel wall (Dukes A or B). The rate of adenomas was similar for both groups (34% vs 29%). The non-bleeders complained more frequently of abdominal pain or a change in bowel habits as compared to the bleeders, but both groups had similar rates for anaemia and weight loss. In all, 47% of these octogenarians with cancer, and 26% with adenomas were referred for colonoscopy because of rectal bleeding. This procedure was found to be safe in old age, as we recorded only four (0.9%) non-fatal complications among our series, a similar figure to the overall incidence of complications at our Institute. In conclusion, our data indicate that rectal bleeding in octogenarians warrants a complete colonic investigation, preferably by total colonoscopy.

Identification of the site of severe colon bleeding by technetium-labeled red-cell scan

In cases of severe colon bleeding, it is more important to know the site than the cause, in case surgery is required. Technetium-labeled red-cell scan (TLRCS) is known to identify the site of bleeding at the rate of 0.1 ml per minute or more. The aim of this retrospective study was to see whether TLRCS was a reliable indicator of the site of severe colon bleeding. A retrospective study was made of patients investigated in this way for acute, severe colon bleeding at St. Vincent's Hospital, Melbourne, from 1984 to 1988 (five years). TLRCS identified the site of bleeding in less than half of the cases but correctly identified the site in all nine patients in whom bleeding was so severe as to require emergency surgery, avoiding total colectomy in eight cases.

Analysis of the colonoscopic findings in patients with rectal bleeding according to the pattern of their presenting symptoms

Patients presenting with rectal bleeding were prospectively categorized according to the pattern of their presentation into those with outlet bleeding (n = 115), suspicious bleeding (n = 59), hemorrhage (n = 27), and occult bleeding (n = 68). All patients underwent colonoscopy and this was complete in 94 percent. There were 34 patients with carcinoma and 69 with adenomas greater than 1 cm diameter. The percentage of neoplasms proximal to the splenic flexure was 1 percent in outlet bleeding, 24 percent with suspicious bleeding, 75 percent with hemorrhage, and 73 percent with occult bleeding. Barium enema was available in 78 patients and was falsely positive for neoplasms in 21 percent and falsely negative in 45 percent. Colonoscopy is the investigation of choice in patients with suspicious, occult, or severe rectal bleeding. Bleeding of a typical outlet pattern may be investigated by flexible sigmoidoscopy.

Frank melena in a patient with fallopian tube pregnancy

We present herein a case of a young woman who experienced lower intestinal hemorrhage caused by ischemic necrosis of the small intestine induced by the compression of a Fallopian tube hematoma in an ectopic pregnancy. All accessible preoperative attempts to clear the site of the bleeding were unsuccessful and an exploratory laparotomy was indicated seventeen days after the development of melena. An ileocecal resection and right salpingectomy finally had to be performed.

Primary colonoscopy

There is good evidence that colonoscopy, in expert hands, is the most accurate investigation for the diagnosis of colorectal disease, and it also allows histological confirmation and therapeutic procedures to be carried out. Furthermore, by screening high-risk groups together with regular follow-up of patients with known colorectal neoplasia and surveillance of long-standing ulcerative colitis patients, it may be possible to reduce the incidence of colorectal cancer. However, at the present time, the lack of widespread availability and the variability in the quality of examinations precludes the employment of colonoscopy as the first-line investigation in colorectal disease. Flexible sigmoidoscopy combined with good quality double contrast barium enema is a reasonable alternative in the majority of cases, reserving colonoscopy for investigation of the elderly and high-risk patients, together with surveillance of patients with premalignant conditions. Technologically, colonoscopy has probably reached its peak and it is now necessary to make provision for more widely available colonoscopy services, provided by adequately trained endoscopists who can guarantee total colonoscopy in more than 90% of cases safely and rapidly. This requires structured training programmes for gastrointestinal physicians and surgeons and ultimately changes in patterns of working practice if adequate numbers of colonoscopy sessions capable of dealing with a steadily increasing workload are to be achieved.

Colonoscopy for investigation of unexplained rectal bleeding in a tropical country

One hundred and forty four patients underwent colonoscopy to detect the cause of rectal bleeding. This was successful in either identifying the lesion or excluding the presume of lesions up to the cecum in 88.8% patients. The source of bleeding was diagnosed in 106 (73.61%) patients. Predominant lesions were nonspecific colitis and ulcers (62.26%), polyps (17.92%), cancer (8.49%), rectal varices (3.77%) and tuberculosis (1.88%). The remainder had other colonic conditions such as radiation colitis, ischemic colitis, vascular malformation, diverticulosis, right sided ulcerative colitis and pseudo-pancreatic cyst communicating with the descending colon. The majority (94.33%) of these lesions involved the left colon. Thus colonoscopic examination was useful in localizing the cause of rectal bleeding and the predominant lesions were differed from those reported in the western hemisphere.

The benefit of colonoscopy

In a prospective study involving 833 consecutive outpatient and open-access colonoscopies, attempts were made to characterize the benefit of colonoscopy in terms of both predicted and unpredicted findings and therapeutic procedures. The endoscopist therefore predicted the endoscopic findings before the endoscopy. The results were compared for the different indications for colonoscopy. The overall agreement between the predictions and the colonoscopic findings was 61%. Clinically significant abnormalities were found in about half the examinations. The most frequent abnormal findings were benign polyps (24%), inflammatory bowel disease (17%), and malignancy (5%). In about half the patients with a malignancy the indication for colonoscopy was rectal bleeding, and half of the malignancies were not predicted. The greatest benefit of colonoscopy was found in patients referred because of overt rectal bleeding or occult faecal blood, and abnormal barium enema or endoscopy findings. The importance of complete colonoscopy in connection with operation for colorectal carcinoma is emphasized.

Lower gastrointestinal bleeding. A systematic approach to classification and management

Early classification of lower gastrointestinal bleeding as occult, minor overt, or major overt allows a practical approach to evaluation and management. Localization of the bleeding site is the next step. In occult and minor overt bleeding, the cause can usually be determined from results of conventional diagnostic tests; however, angiography and even intraoperative endoscopy may be necessary in some particularly difficult cases. In contrast, major overt bleeding may only allow time for angiographic localization before surgery. Using this approach, the primary care physician can successfully manage most patients with lower gastrointestinal bleeding.

Association of colonic diverticula with adenomas and carcinomas. A colonoscopic experience

A case control study of 150 individuals with colonic symptoms and diverticular disease diagnosed by total colonoscopy was performed to ascertain whether adenomas and carcinomas are detected with a higher frequency in these patients than in matched controls with symptoms but not diverticular disease. Adenomas and carcinomas were seen in 36 percent of the patients and in 17 percent of the controls (P less than .001); the overall odds ratio was calculated to be 3.0 (95 percent confidence interval +/- 1.8). When examined separately, adenomas maintained their significantly higher frequency (27 vs. 10 percent, P less than .001), while no difference was observed as regards carcinomas (9 vs. 7 percent). The odds ratios for adenomas and carcinomas were calculated to be 3.5 +/- 2.5 and 1.4 +/- 1.4, respectively. From the fifth to eighth decades there was a fourfold increase in premalignant and malignant lesions in the patient group and a twofold increase in controls. With relation to sex, a statistically significant difference was reached in men but not in women in the sample examined. These data show that symptomatic patients with colonic diverticula have more frequent adenomas, but not carcinomas, than symptomatic control matched by sex and age.

Open access colonoscopy for suspected colonic neoplasia

An open access endoscopy service for direct referral of patients suspected to have colonic neoplasia has been analysed. In the first year of the service 137 patients were referred by 52 general practitioners. The most common reason for referral was rectal bleeding (58%). Colonoscopy was done without outpatient consultation or barium enema examination in 130 patients: there were no complications. The diagnostic yield was 57%: 34 patients (25%) had neoplasia (cancer 14, adenoma 20), which was higher than the yield of neoplasia in patients referred for open access double contrast barium enema during the same period. Only nine patients (7%) needed further investigation because colonoscopy was incomplete and 21 (16%) were treated by polypectomy at the diagnostic colonoscopy.

Comparative evaluation of investigations for colorectal carcinoma in symptomatic patients

We studied 154 patients presenting with significant colonic symptoms and subsequently diagnosed to have colorectal carcinoma. They were investigated by faecal occult blood tests, fibresigmoidoscopy, double contrast barium enema (DCBE) and colonoscopy. Faecal occult blood tests (Haemoccult) alone were positive in 26% of patients with Dukes' A, in 69% with Dukes' B and in 64% with Dukes' C lesions. DCBE alone identified the lesion in 32% of Dukes' A, 79% of Dukes' B and 81% of Dukes' C carcinomas. Fibresignoidoscopy diagnosed colorectal malignancy in 84% of patients with Dukes' A, 90% with Dukes' B and 81% with Dukes' C stage. A diagnostic yield of 88% for Dukes' A, 96% for Dukes' B and 100% for Dukes' C carcinomas was seen with colonoscopy. Detection rate for all stages of carcinoma was greater than 95% when fibresigmoidoscopy and DCBE were used together. Faecal occult blood tests and DCBE alone are inadequate in diagnosing early malignancy in symptomatic patients. Fibresigmoidoscopy and DCBE used in conjunction compare favourably with the technically difficult procedure of colonoscopy and should routinely be undertaken in these patients before malignancy can confidently be excluded.

The impact of colonoscopy on the early detection of colonic neoplasms in patients with rectal bleeding

A retrospective analysis was made of 372 consecutive patients who underwent colonoscopy because of rectal bleeding. The three distinct patterns of bleeding studied were chronic (224 cases), recent major (93 cases), and acute bleeders (55 cases). In 50% of the cases, colonoscopy detected a lesion other than diverticula. These lesions consisted of several forms of colitis, arteriovenous malformations (AVMs), ulcers, and most importantly, neoplasms in 34% of the cases. In 13% of the cases, an invasive neoplasm was detected and 76% of them were early stage (Dukes A or B). A third of the neoplasms were located proximal to the splenic flexure. Among patients with a negative barium enema, 35% had a neoplasm detected on colonoscopy. These findings were similar for the three distinct patterns of rectal bleeding studied. These data support the need for colonoscopy in all types of rectal bleeders, regardless of the results obtained by BE.

Colonoscopy or barium enema as initial investigation of colonic disease

To determine whether double-contrast barium enema (DCBE) or fibreoptic examination should be the first-line investigation for colonic disease 76 consecutive patients presenting for the first time to the outpatient clinic with symptoms of colonic disease deemed to need a DCBE after negative rigid sigmoidoscopy were entered into a trial. All underwent flexible sigmoidoscopy, then DCBE, and finally colonoscopy. 66 patients completed the study. DCBE alone gave the final diagnosis in 42 (67%) and colonoscopy alone in 60 (91%) (p = 0.0004). A combination of flexible sigmoidoscopy and DCBE led to the diagnosis in 50 patients (76%). With DCBE alone 73% of polyps and 64% of patients with inflammatory bowel disease were missed. No complications arose from the investigations. 32 (48%) patients found DCBE distressing and 15 (23%) found colonoscopy uncomfortable (p = 0.004). Its high diagnostic accuracy and relative lack of discomfort for patients make colonoscopy the primary procedure for investigating patients with large bowel symptoms referred to the general surgeon.

Preoperative and intraoperative localisation of gastrointestinal bleeding of obscure origin

In the past six years, 37 patients with gastrointestinal bleeding of obscure origin had their bleeding sites localised preoperatively or intraoperatively. Preoperative investigations followed a regime consisting of endoscopy, barium meal and follow through, small bowel enema, 99mTc pertechnetate scan, 99mTc-labelled red blood cell scan and selective coeliac and mesenteric angiography. Bleeding lesions were localised preoperatively in 36 patients. In one patient, diagnostic laparotomy had to be carried out immediately before any investigation because the bleeding was severe. At operation, angiosarcoma of ileum was found. Unless preoperative investigations showed the lesions to be in anatomically fixed organs like the duodenum or colon, the lesions had still to be found at operation. Palpation and transillumination detected the lesion intraoperatively in 21 patients while only some lesions were found in three patients with multiple lesions. Sigmoidoscopy through enterotomies was required in one patient. Intraoperative enteroscopy was done for small lesions not found grossly at operation in nine patients, to detect additional lesions in three patients or to rule out suspicious lesion shown on preoperative tests in one patient. In another patient with diffuse lymphoma of small bowel with bleeding from only a small segment of jejunum, injection of methylene blue intraoperatively through a previously placed angiographic catheter stained the bleeding segment of jejunum blue. This segment was identified easily and resected. These preoperative and intraoperative localisation procedures were simple and effective and we recommend them to be used more freely.

Diagnostic accuracy of double-contrast enema and rectosigmoidoscopy in connection with faecal occult blood testing for the detection of rectosigmoid neoplasms

Four hundred and fifty-eight consecutive subjects with a positive faecal guaiac test when screened for early detection of colorectal neoplasms were investigated with double-contrast enemas and rectosigmoidoscopy (60 cm). The results of these two methods were compared. The radiologists and endoscopists were unaware of the result of each others' examination at the time of their own investigation. Altogether ten subjects with carcinoma in the rectosigmoid area were found. The radiologists and endoscopists each overlooked four of these ten carcinomas and only four of the carcinomas were diagnosed with both methods. Fifty-six of one hundred and seven adenomas were 1 cm or larger in diameter and located in the rectum or the sigmoid colon. Thirteen of the fifty-six adenomas were missed by double contrast enema and ten by endoscopy and only thirty-three adenomas were diagnosed with both methods. Neoplasms in the rectum and the sigmoid colon are sometimes difficult to diagnose with both radiology and endoscopy. Rectosigmoidoscopy (60 cm) should therefore be used as a complement to double contrast enemas if this method is chosen for investigation of a patient with rectal bleeding.

Colonoscopy findings in symptomatic patients without X-ray evidence of colonic neoplasms

Colonoscopy was carried out in 97 patients with persistent large-bowel symptoms in whom double-contrast barium enemas were either normal or showed diverticular disease alone. In 37 the X-ray was normal, but colonoscopy showed 9 polyps and 3 carcinomas. In 60 patients with barium enemas showing diverticular disease alone colonoscopy revealed 13 polyps and 1 carcinoma. Of the 4 carcinomas 1 was Dukes' A, 2 Dukes' B, and 1 Dukes' C. 23 of the 26 patients with a neoplastic lesion at colonoscopy had rectal bleeding, compared with only 40 of the 71 without a lesion. The frequency of neoplasia in those with a normal barium enema was 32%, compared with 23% in those whose X-ray showed diverticular disease. Colonoscopy, rather than double-contrast barium enema, should be the first line of investigation in patients with persistent large-bowel symptoms, especially rectal bleeding.

The importance of laparotomy in the diagnosis and management of intestinal bleeding of obscure origin

Thirty-seven patients who had had either a colonoscopy or a selective mesenteric angiogram while being investigated for severe or persistent gastrointestinal bleeding of obscure origin were reviewed. Failure to make a prompt diagnosis was partly responsible for the 16% hospital mortality in the series. Colonoscopy was diagnostic on 6 out of 38 examinations but detected 43% of lesions in the colon. Angiography achieved only 3 diagnoses in 17 examinations. Fourteen patients had an exploratory laparotomy which was diagnostic in 9. We believe that early laparotomy still has an important place in the diagnosis and treatment of intestinal bleeding of obscure origin.

How important is rectal bleeding in the diagnosis of bowel cancer and polyps?

Overt bleeding from the anus is a common symptom of colorectal cancer but most frequently arises from a benign anal source. The aim of this study was to determine how successfully general practitioners and gastroenterologists could differentiate anal from colorectal sources of bleeding before full colonic investigation. 145 consecutive patients aged 40 years and over who had presented to a GP with rectal bleeding of less than 6 months' duration were referred to a specialist for total colonic investigation. The source of bleeding was diagnosed as colorectal cancer in 15 patients (7 stage A, 3 stage B) and polyps in 11. Of 63 patients in whom GPs predicted an anal source of bleeding only, 11 were ultimately found to be bleeding from a colonic or rectal source. The gastroenterologists (after rigid sigmoidoscopy) predicted an anal source of bleeding in 97, in 5 of whom the source was ultimately found to be colorectal. All patients aged over 40 who present with recent rectal bleeding should be referred for full colonic investigation.

Haemorrhage from idiopathic annular ulcers of the small intestine

Annular small bowel ulceration is rare. We report three patients who presented with signs of chronic gastrointestinal blood loss. In two patients the main presenting symptom was intermittent abdominal pain while the third patient subsequently developed abdominal pain during the course of investigation. Preoperative localisation of the source of haemorrhage proved difficult in each case. However in all three patients an easily identifiable lesion was found at laparotomy. We believe that laparotomy should be undertaken more readily in the investigation of patients with the combination of abdominal pain and gastrointestinal bleeding as these patients are likely to have an anatomically identifiable cause for their symptoms.

Twenty year experience with familial polyposis coli in Cape Town

Forty-nine patients undergoing surgery for familial polyposis coli (FPC) over a twenty year period are presented. Forty-three patients underwent total colectomy with ileorectal anastomosis (TC + IRA) and five patients subsequently developed a rectal carcinoma. The incidence of rectal carcinoma following TC + IRA appears to increase with time. There were significant postoperative complications secondary to adhesive obstruction and desmoid tumour recurrence. Upper gastrointestinal pathology has been detected by endoscopy in only one of nine patients, but the importance of upper gastrointestinal pathology is now appreciated. The problem of screening a widely distributed and closely knit community has been considerable. TC + IRA remains the operation of choice for the majority of patients with FPC, but a total colectomy with ileal reservoir and ileo-anal anastomosis is appropriate in some cases.

Diverticular disease and minor rectal bleeding

The purpose of the study was to compare the history of rectal bleeding during the year before the study and the presence of faecal occult bleeding in patients with advanced diverticular disease of the large bowel and two sex- and age-matched reference groups, one consisting of patients with normal large-bowel barium enemas and the other of persons without any history of gastrointestinal disease. The number of persons with faecal occult bleeding or macroscopic anal bleeding during the year before the study was not increased in the group with diverticular disease as compared with the other two groups. The bleeding frequency and characteristics of the bleeding did not differ among the three groups. It is concluded that minor rectal bleeding is rather uncommon in persons with uncomplicated diverticular disease. Patients with a history of rectal bleeding in whom a barium study has only shown diverticular disease should be further investigated as though the diverticula were not present.

Technetium labelled red blood cell scintigraphy in the diagnosis of intestinal haemorrhage

99m-Technetium labelled red blood cell scintigraphy was used in the investigation of 15 adult patients with suspected small or large bowel bleeding requiring at least five units of blood (mean 14.3 units) and one neonate with rectal bleeding. Scintigraphy was found to be an accurate method of detecting the site of haemorrhage and was superior to angiography. This technique may be of particular value in patients with profuse colonic haemorrhage when the view at colonoscopy is poor.

Faecal occult blood testing in the detection of colorectal cancer

5012 asymptomatic patients over the age of 40 years were asked by their general practitioners to perform a Hemoccult faecal blood test over 3 days. 3422 patients on a normal diet completed the test (a compliance rate of 67 per cent). Ninety-nine patients (3 per cent) had a positive result. Those patients with a positive result then underwent a second Hemoccult test over 6 days on a restricted diet. Thirty-two patients had a positive result on repeat testing (1 per cent). Only those patients who were positive on the second testing were fully investigated (including double contrast barium enema and colonoscopy). Twenty-seven patients on investigation were found to have neoplastic disease (84 per cent). Twelve had invasive carcinomas (nine Dukes' A, two Dukes' B, one Dukes' C) and thirty-seven adenomas were identified and removed at colonoscopy (barium enema missed three carcinomas--all dukes' A and only identified sixteen of the thirty-seven adenomas, 43 per cent). Those patients found to be negative after the second Hemoccult test are being followed up.

Prospective evaluation of hospitalized patients with nonactive lower intestinal bleeding--timing and role of barium enema and colonoscopy

A prospective evaluation of 176 hospitalized patients with nonactive lower intestinal bleeding demonstrated that a barium enema is not a necessary prerequisite for a sensitive and safe total colonoscopic examination. Colonoscopy was better in detecting colonic polyps and colon cancers than barium enema. It was concluded that colonoscopy without prior barium enema provides a sensitive diagnostic approach to hospitalized patients with nonactive lower intestinal bleeding. This approach does not interfere with the early use of angiography and may shorten hospitalization.

Intestinal vascular ectasia. Improving diagnostic capability poses therapeutic dilemma

A case of gastrointestinal bleeding in a 73-year-old man secondary to vascular ectasia is presented. The use of the 99m technetium-tagged erythrocyte scanning to localize the bleeding site and arteriography to document a vascular malformation strongly suggested this diagnosis. Elective hemicolectomy failed to eliminate bleeding, however. Endoscopic and surgical therapies are available for this condition, but the relative frequency of postoperative rebleeding requires conservatism in recommending surgical excision.

Is colonoscopy necessary in diverticular disease?

Colonoscopy was carried out in 65 patients in whom double-contrast barium enemas for bowel symptoms had shown sigmoid diverticular disease. In 19 of these, barium enemas had shown neoplastic lesions--polyps in 17 and carcinoma in 2; but colonoscopy showed no polyps in 9 of the 17. In 1 a carcinoma not a polyp was found, and of the 2 carcinomas only 1 was confirmed. In 46 patients the barium enemas showed diverticular disease only, but colonoscopy revealed polyps in 8 and carcinoma in 3. Thus in 23 patients (35%) the barium enema was inaccurate. 13 of 20 patients with neoplasia but only 11 of 45 without had rectal bleeding, but the numbers in the two groups who had abdominal pain or change in bowel habit were similar in the two groups. Colonoscopy revealed neoplastic lesions in 20 of the 65 patients (31%), an incidence great enough to recommend routine colonoscopy in patients with symptomatic diverticular disease, especially those with rectal bleeding.

Recurrent gastrointestinal bleeding of obscure origin: report of 17 cases and a guide to logical management

Seventeen patients with chronic obscure gastrointestinal bleeding presenting over a 5-year period are described. While representing only 6 per cent of the total number of admissions due to gastrointestinal haemorrhage they required a disproportionate amount of medical attention, including 67 admissions, 73 radiological and 56 endoscopic procedures and transfusion of a total of 198 units of blood. The diagnoses included 5 cases of recurrent acute upper gastrointestinal erosions, 7 small bowel abnormalities and 4 colonic lesions, with 1 patient undiagnosed. Eight arteriovenous malformations (AVM) were demonstrated angiographically and one Meckel's diverticulum identified by a 99Tcm pertechnetate scan. There were 5 cases in which laparotomy yielded the final diagnosis and in one of these peroperative endoscopy of the small bowel was required. On the basis of our experience and a review of other series, a logical plan of investigation is described for such difficult cases.

Diagnosis of carcinoma of the colon by barium enema

Double-contrast barium enemas can detect 96% of colonic carcinomas. This is similar to the detection rate for colonoscopy. Single-contrast enemas compare badly. Small sigmoid tumours are most likely to be missed. Inadequate bowel preparation, single-contrast examinations and observer error account for most barium enema failures.

Accuracy and value of the Hemoccult test in symptomatic patients

Hemoccult faecal occult blood testing is widely advocated as a screening test for colorectal cancer but few studies have shown its correlation with conventional methods of investigation for colorectal disease. In a prospective study of 802 symptomatic patients with suspected colorectal disease there was good patient compliance (92.5%) and a high specificity for colorectal cancer (85.4%). The false positive rate was 8.6% (12 of 140 patients with positive results), and while the test result was positive in 22 of 26 colonic cancers the false negative rate for rectal cancer (45.4%) should not detract from its value as a screening test if proper digital anorectal and proctosigmoidoscopic examination are widely practised. A positive Hemoccult test result is a useful indicator for the need to proceed to full colorectal investigation, including colonoscopy.

Total colonoscopy in children

One hundred and twenty-three total colonoscopies were performed on 115 children with ages ranging from 3 months to 16 years. The major indications were suspected inflammatory bowel disease and unexplained rectal bleeding. Ninety-seven per cent of all procedures were carried out with sedation only. Adult colonoscopes were used in most of the patients but in babies and small children paediatric instruments were preferable. Total colonoscopy was possible in all patients with a patent colon. The terminal ileum was examined in 63 patients. Endoscopic snare polypectomy was successfully carried out in 8 children and multiple haemangiomas were electrocoagulated in one. Total colonoscopy in this paediatric series proved to be at least as easy, rapid, well-tolerated, and safe as in adults. In selected patients as single colonoscopy can give an accurate diagnosis with biopsy proof and sometimes the opportunity for definitive treatment.

Colonoscopic polypectomy in children

Five children presenting with chronic and intermittent rectal bleeding were diagnosed as having colorectal polyps by fibreoptic colonoscopy performed under sedation. Three of the children had had barium-enema films reported on as normal. Eight polyps were seen, of which six were proximal to the sigmoid colon. All were removed endoscopically (one by proctoscopy, one by snare-intussusception) without complication. Colonoscopic polypectomy is a safe and efficient procedure in children, and colonoscopy may be regarded as first-line management in those with rectal bleeding.

Colonoscopy in the diagnosis of unexplained rectal bleeding

Three hundred six patients with unexplained rectal bleeding were examined by colonoscopy. Significant lesions were found in 30% including polyps having a diameter of 0.5 cm or larger in 14%, carcinoma in 8%, and a small number of patients with inflammatory bowel disease, vascular formations, and radiation colitis. Colonoscopic findings in a large subgroup of patients with diverticulosis established by barium enema did not differ significantly from those of the group as a whole. Significant lesions were also identified in 22% of small subsets of patients with occult bleeding or unexplained melena.

Pyoderma gangrenosum and ulcerative colitis

The relationship of pyoderma gangrenosum and ulcerative colitis remains uncertain. We investigating 14 patients with pyoderma gangrenosum by colonoscopy with multiple biopsies. Six patients had ulcerative colitis and all of these had disease affecting the whole colon. There were no correlations between exacerbations of the colitis and the onset or course of pyoderma gangrenosum. The remaining eight patients with pyoderma gangrenosum had no other disease and they were found to be significantly older than those patients with coexisting colitis (P less than 0.002).