Table 8LOWER GASTROINTESTINAL CANCER: investigations

AuthorSettingDescriptionNo.InclusionExclusionResultsGold StdQuality
Church, 1991USAPatients presenting to a colorectal surgeon underwent colonoscopy, with the findings being related to the bleeding pattern.269, 115 having outlet bleeding, 59 suspicious bleeding, 27 heamorrhage, 64 occult bleeding--4% of outlet cases had carcinoma (all in the rectum or sigmoid), 22% with suspicious (10 left colon, 3 right colon), 19% with occult bleeding (12 of 13 proximal to splenic flexure), and 4% in haemorrhage (all 3 cancers right sided). The sensitivity of air contrast barium enema was 75%, specificity 43%, PPV 71% and NPV 47%colonoscopyA selected group of referred patients.
Duffy et al 2003Guidelines produced by the European Group on Tumour MarkersExtensive review of the evidence
Fijten et al 1995NetherlandsA further analysis of Fitjen et al, (1993){651}. The aim was to determine the diagnostic value of combinations of signs, symptoms and simple laboratory test results for colorectal cancer in patients presenting with rectal bleeding to the general practitioner. The tests were haemoglobin, erythrocyte sedimentation rate (ESR), white blood cell count (WBC), and faecal occult bloodIn a multiple logistic regression that included symptoms and signs, none of the tests were significant independent predictors of colorectal cancer in patients with rectal bleeding.
Meyer et al, 2000USAA retrospective cohort study167347 patients undergoing gastrointestinal endoscopiesMedicare beneficiariesUnder age 65, claims for reimbursement disallowed, services provided outside continental USOnly 8% of colonoscopies were performed by generalists; generalists performed 35.2% of rigid sigmoidoscopies and 42.7% of flexible sigmoidoscopies. Specialists were more likely to perform the procedure to investigate cancer.Medicare data
Muris et al 1995NetherlandsProspective observational study in 80 general practices. patients presented to their general practitioner with new non-acute abdominal complaints lasting two or more weeks. A structured history was obtained, an examination performed, and the following laboratory tests undertaken: haemoglobin, white blood cell count, ESR, faecal occult blood (three times, with peroxidase-free diet).933 patients24 (2.6%) of the sample of 933 were diagnosed to have cancer during the following year. Multiple logistic regression was used to estimate the odds of cancer given certain symptoms, signs and investigation results. Only an ESR greater than 20mm/hour was associated with a diagnosis of cancer (odds ratio 3.0 [95% CI 1.1–8.2]).
The paper did not report sufficient data to enable the sensitivity or specificity of a raised ESR to be calculated.
NHS centre for Reviews and Dissemination 1997Evidence based review to support the CSG on colorectal cancerEvidence based, but not focused on diagnosis
Pierzchajlo et al 1997USACase series of 751 colonoscopies performed by a single family physician555 patients with bleeding (49.9%), polyp follow-up (20.9%), pain (11.7%), diarrhoea (11.6%), abnormal findings on sigmoidoscopy (8.4%).--The adequacy of bowel preparation was excellent in 79.2%; completion (caecal intubation) was achieved in 91.5%. 407 biopsies and polypectomies were preformed, and three carcinomas discovered.histologyOnly a single physician included.
Rodney 1987USABefore and after study of impact of an education course on the use of flexible sigmoidoscopy114 physicians--The number of procedures performed by physicians increased after the course. 5467 procedures had been performed, and there had been one perforation. Physicians reported reaching an insertion depth of 52 cm by the 21st examination.-Study does not present information about the diagnostic utility of flexible sigmoidoscopy by family physicians.
Selvachandran et al 2002UKProspective study, the findings from a patient questionnaire that provided a malignancy risk score being compared to final diagnosis. The likelihood of cancer was graded by the surgeon on the basis of the GP’s letter into one of five groups.2268 patients referred by GPs to a surgeon--95 (4.2%) of the 2268 patients had colorectal cancer. The proportions with cancer were higher among those placed in high likelihood groups by specialists on the basis of the GP’s letter, and among those with higher patient questionnaire scores. The sensitivity and specificity of the malignancy risk score varied depending on the cut off point.Final diagnosis after investigation
Sorensen et al 1992DenmarkData from the cancer registry were associated with data from service records of general practitioners in one county in Denmark146 general practices; 95 patients diagnosed as having cancer of the rectum in 1986-Practices were excluded if they experienced a change in doctors during the year of the study.Proctoscopies were preformed in 128 (88%) of the practices, the number of examinations performed by GPs varied from 1 to 107. There was no relationship between the number of proctoscopies performed per year and the Dukes’ stage at time of diagnosis of cancer of the rectum.Cancer registrations.The study has no information on the findings of proctoscopy examinations and does not permit the calculation of sensitivity of specificity of the examination.
Steine 1994NorwayA survey of patients referred for barium enema; information on symptoms and tests were obtained from patients and referral letters.1852 referred patients, 1477 from GPs.--76% had had a haemoglobin test 37% faecal occult blood, and 16% proctosigmoidoscopy.noneThe study did not obtain data from clinical records, and did not relate the result of tests to final diagnosis
Tate et al 1990UKA prospective cohort study of patients referred for double contrast barium enema. Patients were randomly allocated to use one of three types of occult blood test before the enema.969 patients undergoing barium enema.Appointment within 7 days of request for examination by physician49 (5.1%) had colorectal cancer detected on barium enema. The sensitivity, specificity and PPV of the three tests were: Haemoccult 80.0, 88.8, 32.7; Fecatwin 93.3, 71.6, 13.3; E-Z Detect 57.1, 88.9, 19.0, respectively.Double-contrast barium enema

From: Appendix A, Evidence Tables

Cover of Referral Guidelines for Suspected Cancer in Adults and Children
Referral Guidelines for Suspected Cancer in Adults and Children [Internet].
NICE Clinical Guidelines, No. 27.
Clinical Governance Research and Development Unit (CGRDU), Department of Health Sciences, University of Leicester.
Copyright © 2005, National Collaborating Centre for Primary Care.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.