Gastro-intestinal Oncology Unit, Department of Gastroenterology, Tel-Aviv Sourasky Medical Center and Sackler School of Medicine, Tel Aviv University, Tel-Aviv, Israel
Received 17 September 2001; accepted 10 October 2001.
Key words: colorectal cancer, fecal occult blood test, screening
Introduction
Colorectal cancer (CRC) is a major health concern and a leading cause of cancer death in the Western world. The estimated lifetime risk of CRC is 5% to 6%, the incidence rate increases sharply after the age of 50. Approximately 75% of new cases occur in individuals at average risk [1, 2]. Worldwide, it is estimated that there will be over one million new CRC cases in the year 2001 (personal communication), and the annual age-adjusted incidence rate is 57 per 100 000. Survival is directly related to the extent of disease at the time of diagnosis. Those diagnosed at an advanced stage have an estimated 5-year survival rate of 7%, in contrast with a survival rate of 92% for individuals detected at an early stage, since advanced CRC is largely refractory to conventional therapy and is one of the least curable malignancies [3]. Despite continuing advances in diagnosis and therapy, long-term survival has not improved significantly over the last four decades, and almost 50% of CRC patients will eventually die of their disease [1, 2]. This situation mandates improvement in the early detection of this surgically curable disease and preventative interventions to reduce the incidence of the disease and its downstream morbidities and mortality. Indeed, given our understanding of the pathogenesis of CRC, current screening technologies, and effective preventative interventions, CRC should be highly preventable.
Colorectal cancer has become an important issue for physicians and the general population in the last decade, ever since preventative medicine became a key issue in health care. Indeed, CRC is preventable in up to 90% of cases [4, 5]. CRC fulfils the criteria a disease must have for mass screening. Indeed, intestinal tumorigenesis is exceptionally suited for screening, since the adenomacarcinoma sequence allows the detection and removal of pre-malignant lesions, and it is well established that patients who are maintained free of adenomas by endoscopic polypectomy are generally kept cancer free [4, 6, 7].
A study using Markov chain analysis has found that CRC decreases the life expectancy calculated for the entire US population by 292 days at age 5054 years and by 70 days at age 7074 years [8]. At the same time, an important controversial issue is the optimal screening approach for the average-risk population at the age of 50 years. Endoscopic screening is undoubtedly the most effective screening method, but its potential disadvantages, such as invasiveness and cost, should be considered. The simplest and most evaluated screening method available for CRC is the fecal occult blood test (FOBT). This periodic stool testing is considered a low-priced, non-invasive test that requires no cathartic preparation. It can be performed on mailed specimens without a health center visit, and may reflect the full length of the colon. On the other hand, it is a less sensitive method, since many cancers and the large majority of pre-malignant adenomas do not bleed and are therefore missed. Furthermore, high false-positive rates, due to dietary peroxidases and meat, or frequent and trivial sources of occult bleeding, lead to unnecessary colonoscopies that drive up programmatic costs.
The first national screening program with FOBT was introduced in 1977 in Germany. The evaluation of this program was limited because of confidentiality problems. Casecontrol studies have yielded conflicting results regarding the effectiveness of FOBT as a screening modality [9].
The rationale for the use of FOBT screening programs is based on three major long-term randomized trials. Overall, those trials using annual FOBT have reported a 15% to 33% reduction in CRC related death (for review, see reference 10). The Minnesota prospective casecontrolled study [11] is the study that is most cited. It included 46 000 subjects and showed a reduction in CRC mortality, after 1318 years of follow-up, of 33% for annual FOBT examination. Biennial testing resulted in only a minor non-significant mortality reduction (6%) after 13 years. It increased to a cumulative 21% reduction after 18 years [11, 12]. The benefit of FOBT alone in the Minnesota study was questioned, since 38% of the participants underwent colonoscopy, which may have influenced the mortality reduction rate. The authors have attributed 6% to 11% of the mortality reduction to the chance detection of lesions (especially small adenomas) by colonoscopy. Two other large randomized trials from the UK [13] and Denmark [14] have reported 15% to 18% decreases in mortality for biennial FOBT after 810 years. As in the Minnesota study, patients with positive results underwent colonoscopy; therefore, a similar debate regarding the incidental endoscopic detection of neoplasia was raised. Some better conducted prospective as well as casecontrol studies further support the efficacy of FOBT screening [1316].
Recently, Lieberman and Weiss [17] reported on their intriguing results assessing the sensitivity of one-time screening with FOBT, sigmoidoscopy and colonoscopy in 2885 asymptomatic subjects age 5075 years. All participants underwent a complete colonoscopy after returning three specimens for FOBT. Sigmoidoscopy was defined as examination of the rectum and sigmoid during colonoscopy. A positive FOBT was found in 23.9% of patients with advanced neoplasia (determined as an adenoma 10 mm or larger, villous adenoma, high grade dysplasia or carcinoma). Sigmoidoscopy alone or combined with FOBT failed to identify 30% and 24% of advanced neoplasia, respectively. The overall sensitivity of FOBT for any neoplasia was 11.7%.
Towler et al. [18] performed a systematic review of the literature and found that 1000 people would have to be screened for ~10 years in order to prevent one death from CRC.
Our review will present and discuss the pros and cons of FOBT as a screening tool for CRC.
Rationale of the test
The FOBT exploits the tendency of CRC and large polyps to bleed. Ideally, the subject uses a wooden stick to sample from two areas of a formed non-bloody stool, smearing the samples onto two windows on the test card; this procedure is repeated for two more bowel movements. The cards are then developed in a physicians office or a clinical laboratory. If any of the six windows on the three cards is positive, then the test is interpreted as positive, and subjects undergo evaluation of the entire colon [19]. It is important to emphasize that a single occult blood examination has no clinical significance in the screening of CRC. The detection rate of adenomas and carcinomas by one spot FOBT screening examination is 2/1000 and 0.7/1000 screens, respectively [20].
FOBT screening offers no benefit without appropriate follow-up diagnostic testing and treatment. It is an indirect screening test that identifies a subgroup of the average-risk, asymptomatic population sufficiently likely to have a clinically important CRC to justify more expensive, invasive diagnostic tests (e.g. colonoscopy). The screening test by itself neither rules in nor rules out colonic neoplasia [21].
Sensitivity, specificity and positive predictive value
Usually FOBT is based on the detection of peroxidase activity in the stool. Hemoccult II (Beckman Coulter Diagnostic, Palo Alto, CA, USA), the most popular test kit, employs guaiac-impregnated paper and developing solutions to detect the oxidative conversion of a colorless compound to a colored one in the presence of hemoglobin pseudoperoxidase activity. Hemoccult sensa improved the sensitivity of the test; it can detect increased peroxidase activity in the presence of smaller amounts of fecal blood [22, 23]. On the other hand, it reduced the specificity of the test because of an increased false-positive ratio [2224].
More sensitive assays are the immunochemical tests HemSelect (HS) and FlexSure (FS). These tests detect directly human hemoglobin in the stool by using anti-human hemoglobin antibodies [23]. In the HS test, agglutination of chicken erythrocytes coated with antibodies occurs in the presence of intact human hemoglobin. The FS test employs the binding of human hemoglobin to antibodies immobilized on a test strip. A color change appears when the antibodyhemoglobin complex migrates chromatographically along the strip [10, 25].
In comparative screening with guaiac and immunohistochemical tests, the sensitivity for CRC or large adenoma is 50%. The best results for carcinoma detection were achieved by a combination of the two methods, with a sensitivity of 66% and a specificity of 97% [10]. In one Japanese study, using a 3-day FOBT, the sensitivity and specificity for the detection of adenomas (
1 cm), in a set of five different immunochemical tests were 48% and 96%, respectively [26]. The positive predictive value of FOBT is
10%; it is negative in about half of patients with CRC [19]. Positive unhydrated FOBT has been associated with a 22% to 58% CRC detection rate [1114, 27].
Lieberman et al. [28] detected polyps (1 cm or larger) in 17% of average-risk patients referred for colonoscopy with a positive FOBT. The same author [17] recently assessed the sensitivity of one-time screening with FOBT, sigmoidoscopy and colonoscopy in asymptomatic subjects, aged 5075 years. A total of 2885 participants underwent a complete colonoscopy after returning three specimens for FOBT. Sigmoidoscopy was defined as recto-sigmoid examination during colonoscopy. A positive FOBT was found in 23.9% of patients with advanced neoplasia (determined as an adenoma 10 mm or larger, villous adenoma, high grade dysplasia or carcinoma). Sigmoidoscopy alone or combined with FOBT failed to identify 30% and 24% of advanced neoplasia, respectively. The overall sensitivity of FOBT for any neoplasia declined to 11.7%. Other authors have reported 7% to 12% detection rates for large adenomas and 3% to 11% for carcinomas [27, 29, 30]. A possible explanation for the heterogeneous results is the use of different types of tests for different periods of follow-up. The higher detection rates are usually reported in longer study periods. The need for at least three annual consecutive tests has been supported by the fact that results were significantly lower on the first screening and approached their maximal range only after several rounds [1114, 27].
The effectiveness of FOBT depends on stool rehydration (which increases sensitivity but reduces specificity), hemoglobin degradation (decreases sensitivity), and the presence of interfering compounds (such as animal meat and dietary peroxidases). False-positive results are 2% to 4% without rehydration and 8% to 16% in rehydrated samples, placing a heavy burden of diagnostic evaluation on a screening program [27, 29, 30].
Cost-effectiveness of FOBT
Screening the entire average-risk population for CRC would be very expensive. However, it is important for clinicians, health care payers and the public to realize that the cost of missing a curable cancer or of failing to prevent cancer by resecting pre-malignant polyps may be greater [21]. An average cost of US$71007800 per case of CRC detection was reported in several studies [29, 31].
Primarily using data from the Minnesota FOBT trial, health economists from the US National Cancer Institute and from the Office of Technology Assessment of the US Congress estimated that the cost of FOBT screening was less than US$15 000 per year of quality life gained [21]. On the other hand, the relatively low specificity and low predictive value of FOBT necessitates a large amount of upper and lower endoscopy. According to this calculation, it is equivalent to the accepted cost-effectiveness, to prevent a disease, with approxi-mately US$25 000 per year of life saved [10]. This is still substantially less than the cost of many well-accepted preventa-tive medical interventions. They concluded that screening for CRC is highly cost-effective and represents a good investment for US society. In part, because of these costbenefit considerations, the US Congress provided colorectal cancer screening benefits, for the first time, for all Medicare patients beginning January 1, 1998. At the same time, one of the most alarming findings from this survey was that most of those who had not been screened reported that no medical professional had ever recommended screening to them [21].
Lieberman [5] calculated the cost-effectiveness of several CRC screening programs. The analysis included the cost of preventing one CRC-related death during a 10-year screening period. In a model with a 100% compliance rate, FOBT alone had the lowest cost per death prevented as compared with sigmoidoscopy (with or without FOBT), colonoscopy and barium enema. None the less, for a more realistic compliance rate of 50% or less, the estimated cost per death prevented was similar for FOBT and endoscopic screening modalities, while a higher cost was shown for barium enema. Furthermore, the authors have found that if colonoscopy costs are US$750 or less, one-time colonoscopy is more cost-effective than any other screening program at every level of compliance.
Disadvantages
In order for a screening test to be successful and cost-effective in large populations, the physicians and the subjects being screened must be compliant with screening recommendations. This is not the case with FOBT.
FOBT is usually performed by the patients and rarely during a rectal digital examination by the physician. Positive testing rates based on self-reported data were found to exceed rates based on computerized laboratory records by 14%. These data suggest cautious use of self-reported results [32].
Compliance is a major barrier to the achievement of optimal results. Average-risk population screening programs have reported compliance rates of 15% to 40% [10]. These rates gradually decrease over the years, although the efficacy of FOBT depends on the long-term performance of at least three consecutive annual examinations. Niv et al. [33] reported a high compliance rate of 71.5% for a 3-year annual FOBT among 3548 average-risk individuals between 40 and 75 years of age. During the 3-year screening period, the CRC incidence was similar in the FOBT group compared with refusers and controls that were never offered the screening test, but CRC stage and mortality were lower in the FOBT group. The beneficial effect of FOBT on decreased mortality remained during the 8-year follow-up period. This unique study cohort, consisting of a very stable and highly compliant population of kibbutz settlements, may explain the high compliance rate achieved in this trial. At the same time, their compliance decreased dramatically thereafter.
Compliance rates are further limited by the recommendations of dietary restrictions to reduce the number of false positives as a result of ingested peroxidases. A 314 day delay of Hemoccult sensa (HOS) test development, which allows breakdown of dietary peroxidases, has been suggested to overcome the need for dietary restrictions [24].
An important medical and legal issue is the approach to the patient with a positive FOBT and a clean colon on colonoscopy. Should an upper gastrointestinal tract evaluation be performed? Although there are no clear recommendations, this has been suggested by many studies, and is routinely performed by many physicians [24, 30, 34, 35]. Rocky et al. [30] examined a group of patients with positive FOBT who were referred for further evaluation. Upper gastrointestinal lesions were identified in 29% of the patients, while colonic lesions were identified in only 22% of cases. In another study, by the same author, he assessed healthy asymptomatic volunteers who drank small amounts of their own blood for 35 consecutive days. At least one positive HOS test was present in 50% to 100% of participants, depending on the ingested amount of blood (1020 ml) [23]. Lurie and Welch [36] evaluated the pattern of diagnostic testing after an initial positive FOBT. Their analysis suggested that only a minority of subjects screened received an appropriate diagnostic work up. In another study it was suggested that up to one-third of people who tested positive did not respond to requests for follow-up [37].
Stool testing merits further consideration
Stool testing merits further consideration, as its theoretical potential has not yet been achieved. Better tailored screening tools are needed that would exhibit the combined features of high sensitivity and specificity for early-stage cancers and large premalignant adenomas. There should be broad acceptability by the general population, affordability and safety. Neoplasm-specific DNA alterations have been well characterized [12, 13], and represent intriguing candidate markers for stool screening. DNA appears to be stable in stool samples [26] and amplification techniques permit detection of minute amounts of analyte. Several investigators have recovered mutant DNA in stools from patients with CRC or adenomas [20, 2832]. A recently developed prototype stool assay system, the p-EXACT test (Exact Laboratories, Maynard, MA, USA), has been designed to target a spectrum of DNA alterations that occur with CRC and adenomas, with a sensitivity and specificity of almost 90%. This multi-component EXACT test assays point mutations at any one of 15 mutational hot spots on the K-ras, APC and p53 genes; BAT 26, a marker of microsatellite instability and highly-amplifiable or long DNA (L-DNA) [38].
Discussion
The criteria for an optimal screening program include a common preventable disease, and a relatively sensitive, specific, safe and inexpensive test. The effectiveness of FOBT for CRC screening has been a subject of controversy. It is simple, safe, and low-priced, but it is limited by disappointing rates of compliance as well as poor sensitivity and specificity. The consequence of frequent false-positive tests results in a significant increase in cost, due to the many unnecessary gastroscopies and colonoscopies that are conducted. Moreover, a preventative program is worth while when the nature of its benefits outweigh its harm, and a screening program can cause harm. The harm to society may come indirectly by diverting funds from other beneficial projects [39]. Towler et al. [18] have performed a systematic review of the literature. It was estimated that 1000 asymptomatic persons (age >50) would have to be screened for ~10 years in order to prevent a single death from CRC. Hence, a more accurate and effective screening modality, with a higher impact on mortality, should be implemented.
In order for a screening test to be successful and cost effective in large populations, the physicians and the subjects screened must be compliant with screening recommendations. This is not the case with FOBT. Lurie and Welch [36] evaluated the pattern of diagnostic testing after an initial positive FOBT. Their analysis suggested that only a minority of screened patients received an appropriate diagnostic work up. In another study, it was suggested that up to one-third of people who tested positive did not respond to requests for follow-up [37].
Three randomized controlled trials have demonstrated reduced CRC mortality with usage of FOBT screening [11, 13, 14]. A meta-analysis pooling these three studies with new, unpublished data from a Swedish trial estimated a 16% reduction in CRC mortality [RR 0.84 with a confidence interval (CI) of 0.77 to 0.93]. This reduction in mortality would increase to 23% if all patients were to adhere to the screening recommendations [18]. Much of the efficacy in the largest trial may reflect the application of colonoscopy to a large segment of the study population (Minnesota trial) [11]. Even more, the calculated cost per year of life saved is similar to that of the endoscopic screening of CRC [10].
Although compliance is a key factor in CRC mortality reduction by FOBT, the rates of compliance achieved in clinical trials are sub-optimal. The compliance in the community, outside the frame of a clinical study, is no doubt significantly lower. Data from the 1992 National Health Interview Survey indicated that only 17% of the population (50 years of age) performed FOBT within the past year [40]. In this range of less than 50% compliance rate, the calculated cost per CRC death prevented was similar for FOBT, sigmoidoscopy and colonoscopy. Moreover, if colonoscopy costs are $750 or less, one-time colonoscopy is more cost-effective than any other screening program at every level of compliance [5]. A recently published study [17] has found that even among compliant subjects who underwent a complete colonoscopy after returning three specimens for FOBT, the positivity rates of the test for advanced neoplasia and for any neoplasia were 23.9% and 11.7%, respectively. Moreover, sigmoidoscopy alone and combined with FOBT failed to identify 30% and 24% of advanced neoplasia, respectively.
In considering all the advantages and drawbacks of FOBT for CRC screening, we can conclude that this examination is undoubtedly better than no testing at all. However, an ~20% reduction in mortality is not satisfactory, in particular when compared with other screening modalities. FOBT is insufficient to be used alone for CRC screening. It might be used with sigmoidoscopy every 5 years. As of the year 2001, colonoscopy is still the best screening modality that significantly reduces CRC mortality. It seems almost unfair to offer FOBT when considerably higher rates of CRC prevention are achieved by colonoscopy carried out every 10 years (80% to 90%) [4, 5, 10]. A positive FOBT warrants colonoscopy and, if negative, gastroscopy followed by small bowel studies. A negative test does not give complete reassurance to the patient and physician, since most early colorectal neoplasia, especially small ones, are FOBT negative.
Recently, Inadomi and Sonnenberg [8] estimated the impact of CRC screening on life expectancy. They found that the potential extension of life through screening colonoscopy is 2- to 3-fold more than the extension achieved through flexible sigmoidoscopy or fecal occult blood testing.
Improved methods of cancer detection with newer technology look very promising. In particular, molecular biology tests such as the EXACT test are being tested in a large randomized clinical study headed by Ahlquist (Mayo Clinic) and sponsored by the National Cancer Institute (personal communication). It is hoped that further development of this technology could result in non-invasive screening tests with increased sensitivity and specificity.
The most important question from a public health standpoint is not which screening modality to use, but how we can improve access to screening for all members of society.
Acknowledgements
The excellent secretarial assistance of Mrs Baron and Ms Blanga is highly appreciated.
Footnotes
+ Correspondence to: 6 Weizmann Street, 64-239 Tel Aviv, Israel. Tel: +972-3-697-4968; Fax: +972-3-695-0339; E-mail: narber@post.tau.ac.il
References
1.
Parkin DM, Pisani P, Ferlay J. Global cancer statistics. CA Cancer J Clin 1999; 49: 3364.
2. Peeters M, Haller DG. Therapy for early stage colorectal cancer. Oncology 1999; 13: 307315.[Medline]
3. American Cancer Society. Cancer facts and figures. Atlanta, GA: American Cancer Society 1998.
4.
Winawer SJ, Zauber AG, OBrien MJ. Randomized comparison of surveillance intervals after colonoscopic removal of newly diagnosed adenomatous polyps. N Engl J Med 1993; 328: 901906.
5. Lieberman DA. Cost-effectiveness model for colon cancer screening. Gastroenterology 1995; 109: 17811790.[ISI][Medline]
6. Morson, B. Presidents address. The polypcancer sequence in the large bowel. Proc R Soc Med 1974; 67: 451457.[ISI][Medline]
7. Fearon ER, Vogelstein B. A genetic model for colorectal tumorigenesis. Cell 1990; 61: 759767.[ISI][Medline]
8. Inadomi JM, Sonnenberg A. The impact of colorectal cancer screening on life expectancy. Gastrointest Endosc 2000; 51: 517532.[ISI][Medline]
9. Scheitel SM, Ahlquist DA, Wollan PC et al. Colorectal cancer screening: a community casecontrol study of proctosigmoidoscopy, barium enema radiography and fecal occult blood test efficacy. Mayo Clin Proc 1999; 74: 12071213.[ISI][Medline]
10. Burt RW. Colon cancer screening. Gastroenterology 2000; 119: 837853.[ISI][Medline]
11.
Mandel JS, Bond JH, Church TR et al. Reducing mortality from colorectal cancer by screening for fecal occult blood. N Engl J Med 1993; 328: 13651371.
12.
Mandel JS, Church TR, Ederer F, Bond JH. Colorectal cancer mortality: effectiveness of biennial screening for fecal occult blood. J Natl Cancer Inst 1999; 91: 434437.
13. Hardcastle JD, Chamberlain JO, Robinson MHE et al. Randomized controlled trial of faecal-occult-blood screening for colorectal cancer. Lancet 1996; 348: 14721477.[ISI][Medline]
14. Kronborg O, Fenger C, Olsen J et al. Randomized study as of screening for colorectal cancer with faecal occult blood test. Lancet 1996; 348: 14671471.[ISI][Medline]
15. Faiver J, Tazi MA, Milan C et al. Controlled trial of faecal occult blood screening for colorectal cancer in Burgundy (France). Results of the first 9 years. Gastroenterology 1999; 116: A400 (Abstr).
16. Faiver J, Tazi MA, Mrini T et al. Faecal occult blood screening and reduction of colorectal cancer mortality: a casecontrol study. Br J Cancer 1999; 79: 680683.[ISI][Medline]
17.
Lieberman DA, Weiss DG. One-time screening for colorectal cancer with combined fecal occult blood testing and examination of the distal colon. N Engl J Med 2001; 345: 555560.
18.
Towler B, Irwig L, Glasziou P et al. A systemic review of the effects of screening for colorectal cancer using the faecal occult blood test, hemoccult. BMJ 1998; 317: 559565.
19. Levin TR. Colorectal cancer screening: new opportunities. Surg Oncol Clin N Am 1999; 8: 673691.[Medline]
20. Rasmussen M, Kronborg O, Fenger C et al. Possible advantages and drawbacks of adding flexible sigmoidoscopy to hemoccult-II in screening for colorectal cancer. A randomized study. Scand J Gastroenterol 1999; 34: 7378.[ISI][Medline]
21.
Bond JH. Improving the effectiveness of fecal occult blood screening for colorectal cancer. J Natl Cancer Inst 1999; 91: 16021603.
22. Rozen P, Knaani J, Samuel Z. Eliminating the need for dietary restrictions when using a sensitive guaiac fecal occult blood test. Dig Dis Sci 1999; 44: 756760.[ISI][Medline]
23. Rocky DC, Auslander A, Greenberg PD. Detection of upper gastrointestinal blood with fecal occult blood tests. Am J Gastroenterol 1999; 94: 344350.[ISI][Medline]
24. Rozen P, Knaani J, Samuel Z. Comparative screening with a sensitive guaiac and specific immunochemical occult blood test in an endoscopy study. Cancer 2000; 89: 4652.[ISI][Medline]
25.
Alison JE, Tekawa IS, Ransom LJ, Adrain AL. A comparison of fecal occult blood tests for colorectal cancer screening. N Engl J Med 1996; 334: 155160.
26. Nakama H, Fattah A, Zhang B et al. A comparative study of immunochemical fecal tests for detection of colorectal adenomatous polyps. Hepatogastroenterology 2000; 47: 386389.[ISI][Medline]
27. Greenberg PD, Bertario L, Gnaauck R et al. A prospective multicenter evaluation of new fecal occult blood tests in patients undergoing colonoscopy. Am J Gastroenterol 2000; 95: 13311338.[ISI][Medline]
28. Lieberman DA, de-Garmo PL, Fleischer DE et al. Colonic neoplasia in patients with nonspecific GI symptoms. Gastrointest Endosc 2000; 51: 647651.[ISI][Medline]
29.
Bini EJ, Rajapaksa RC, Weinshel EH. The findings and impact of nonrehydrated guaiac examination of the rectum (FINGER) study: a comparison of two methods of screening for colorectal cancer in asymptomatic average risk patients. Arch Intern Med 1999; 159: 20222026.
30.
Rocky DC, Koch J, Callo JP et al. Relative frequency of upper gastrointestinal and colonic lesions in patients with positive fecal occult blood tests. N Engl J Med 1998; 339: 153159.
31. Yamamoto M, Nakama H. Cost-effectiveness analysis of immunochemical occult blood screening for colorectal cancer among three fecal sampling methods. Hepatogastroenterology 2000; 47: 396399.[ISI][Medline]
32. Mandelson MT, LaCroix AZ, Anderson LA et al. Comparison of self-reported fecal occult blood testing with automated laboratory records among older women in a health maintenance organization. Am J Epidemiol 1999; 150: 617621.[Abstract]
33. Niv Y, Levi EM, Fraser G et al. Protective effect of fecal occult blood test screening for colorectal cancer: worse prognosis for screening refusers. Gut 2001, In press.
34. Bini EJ, Rajapaksa RC, Valdes MT, Weinshel EH. Is upper gastrointestinal endoscopy indicated in asymptomatic patients with a positive fecal occult blood test and negative colonoscopy? Am J Med 1999; 106: 613618.[ISI][Medline]
35. Nakama H, Kamijo N, Fattah A, Zhang B. Immunologic detection of fecal occult blood from upper digestive tract diseases. Hepatogastroenterology 1998; 45: 752754.[ISI][Medline]
36.
Lurie JD, Welch HG. Diagnostic testing following fecal occult blood screening in the elderly. J Natl Cancer Inst 1999; 91: 16411646.
37. Winchester DP, Shull JH, Scanlon EF et al. A mass screening program for colorectal cancer using chemical testing for occult blood in the stool. Cancer 1980; 45: 29552958.[ISI][Medline]
38. Ahlquist DA, Skoletsky JE, Boynton KA et al. Colorectal cancer screening by detection of altered human DNA in stool: feasibility of a multitarget assay panel. Gastroenterology 2000; 119: 12191227.[ISI][Medline]
39. Marshall KG. Prevention. How much harm? How much benefit? Physical, psychological and social harm. CMAJ 1996, 155: 169176.[Abstract]
40.
Vernon SW. Participation in colorectal cancer screening: a review. J Natl Cancer Inst 1997; 89: 14061422.