RSS Feeds
CLINICAL GUIDELINE: PART I: Suggested Technique for Fecal Occult Blood Testing and Interpretation in Colorectal Cancer Screening
- American College of Physicians* *This paper, written by David F. Ransohoff, MD, and Christopher A. Lang, MD, was developed for the Health and Public Policy Committee by the Clinical Efficacy Assessment Subcommittee: George E. Thibault, MD, Chair; John R. Feussner, MD, Co-Chair; Anne-Marie J. Audet, MD; Gottlieb C. Friesinger Jr., MD; Daniel L. Kent, MD; Keith I. Marton, MD; Valerie Anne Palda, MD; John J. Whyte, MD; and Preston L. Winters, MD. This paper was approved by the Board of Regents on 26 October 1996. Requests for Reprints: Customer Service Representative, American College of Physicians, Independence Mall West, Sixth Street at Race, Philadelphia, PA 19106-1572.
The numbers in square brackets are cross-references to the numbered sections in the accompanying paper, “Screening for Colorectal Cancer with the Fecal Occult Blood Test: A Background Paper,” which is part II of this guideline (see pages 811-822).
Strong evidence from three randomized, controlled trials indicates that screening with fecal occult blood tests reduces mortality rates associated with colorectal cancer. Other evidence shows a benefit for screening with sigmoidoscopy. One or both screening methods are almost universally endorsed today.
These guidelines are primarily concerned with the technique of screening by using fecal occult blood tests, with special emphasis on the interpretation of positive and negative test results and the subsequent work-up of persons with positive results. They do not address the question of whether to screen or the decision to use fecal occult blood testing alone, sigmoidoscopy alone, or both. These issues are addressed elsewhere.
This discussion reflects a synthesis of data published through 1996. Although some features of these recommendations are expected to remain stable for the next few years, others-particularly those on the type of fecal occult blood test used and the frequency of screening-may evolve substantially as new data become available. One purpose of the accompanying background paper is to provide a conceptual framework for assessing future developments.
Background
Colorectal cancer is a major cause of cancer-related death in the United States and many other countries. Because of its generally favorable clinical biology, it is an appropriate target for screening. About 6% of persons will develop colorectal cancer by 80 years of age, and half of these persons will die as a result. Screening for colorectal cancer has been recommended for many years, but the lack of a randomized, controlled trial showing proof of benefit was a barrier to widespread implementation. However, convincing evidence from three randomized, controlled trials now shows that mortality rates associated with colorectal cancer can be reduced by about 15% to 35% by screening with fecal occult blood tests. The true efficacy may be even higher because compliance in these trials was only about 50%. Screening with sigmoidoscopy may reduce mortality rates associated with colorectal cancer by about 60% for lesions that can be reached by the instrument or about 30% for the whole colon. Therefore, in 1997, we can expect the rates of screening for colorectal cancer to start increasing as this encouraging evidence is translated into stronger recommendations and practical plans for implementation.
Screening with fecal occult blood tests involves many specific decisions about technique. Persons who should not be screened must be identified; a type of fecal occult blood test must be chosen; and frequency of screening (annual or biennial), restriction of diet and medication during screening, number of specimens to be collected, whether to rehydrate samples in guaiac-based fecal testing during slide development, definition of a positive test result, and management of persons with positive test results must be determined (Table 1). Although many of these details may seem mundane, each can substantially affect the interpretation of test results and, ultimately, the benefit of and the effort involved in the screening process.
Suggested Technique for Fecal Occult Blood Testing and Interpretation
Relative Contraindications to Screening with Fecal Occult Blood Tests
Screening with fecal occult blood tests should not be done in persons who are likely to have misleading results, such as those with active hemorrhoidal bleeding or symptoms that already suggest colorectal cancer.
Because screening with fecal occult blood tests only identifies persons who need a complete colonic evaluation for such important neoplasms as early colorectal cancer or large adenomatous polyps, persons who are too frail to undergo complete colonoscopy, barium enema plus sigmoidoscopy, or both should not be screened. Moreover, because the benefit of screening is not fully realized until at least 3 to 5 years after the initial screening, persons with only a short life expectancy should generally not be screened.
Type of Test and Specimen Collection
The most practical and appropriate technique for screening with fecal occult blood tests currently involves a guaiac-based test. Two slides should be prepared from each of three consecutive bowel movements [2.2.3]; the person being screened should be encouraged to prepare for stool sampling by abstaining from aspirin in dosages greater than 325 mg/d, substantial doses of nonsteroidal anti-inflammatory drugs, red meat, poultry, fish, some raw vegetables, and vitamin C [2.2.3.2]. Slides should not be rehydrated and should be developed within 7 days of preparation [2.2.3.4].
These are the techniques that were generally used in the three published trials that reported a reduction in mortality rates associated with colorectal cancer [2.1, 2.2.1, 2.2.2]. In one trial, rehydration was done at the time of slide development to try to improve the sensitivity of the test. Rehydration, however, increases the rate of false-positive results to 10% or greater with each application of the test, prompting many additional, nonproductive colonoscopic evaluations [2.2.3.3]. Although some persons advocate the rehydration of slides for fecal occult blood tests because this practice was shown to be effective in one trial, it can be argued that rehydration makes screening impractical and should be discouraged.
The type of fecal occult blood test used in screening is likely to evolve in the next few years. The Hemoccult II test (SmithKline Diagnostics, San Jose, California), which has been used for years in the United States, may eventually be replaced by new approaches that improve overall sensitivity for important neoplasms without incurring an excessive burden of false-positive results. A sequential test strategy combines a more sensitive test, such as HemoccultSENSA (SmithKline Diagnostics), with a more specific test, such as HemeSelect (SmithKline Diagnostics). An office-based immunochemical test also may become practical.
Screening Test Frequency
No recommendation is made about the optimal frequency (that is, annual or biennial) for screening with fecal occult blood tests.
The optimal frequency of screening with fecal occult blood tests remains uncertain [2.1]. Both annual and biennial screening have been associated with reductions in mortality rates in published trials. Although the point estimates for the overall efficacy of screening seem to favor annual screening, the trials were different in many ways that might have contributed to the differences in efficacy. Detailed analyses would be useful for the assessment of the effort and cost of annual compared with biennial screening. A greater frequency of testing should confer more protection, but the cost and effort would be greater.
Definition and Evaluation of Positive Fecal Occult Blood Test Results
A positive result on a fecal occult blood test should be defined as positivity in one or more slide windows [2.2.3.1]; this definition maximizes the sensitivity of the entire screening procedure. A positive result should generally lead to a complete colorectal examination within 2 to 3 months for neoplasms [2.4].
A middle-aged person who has a positive result on an initial fecal occult blood test without slide rehydration has about a 7% to 14% probability of having early-stage (Dukes stage A or B) colorectal cancer. The probability of early-stage colorectal cancer or a large (≥ 1 cm) adenoma, which is an important precursor of colorectal cancer, is about 30%. With subsequent rescreening tests, these rates will decline. Nonetheless, the rates are high enough to warrant a complete evaluation of the colon and rectum when a person has a positive test result either initially or at rescreening.
In one trial, a somewhat complex strategy of repeated testing was applied if a test resulted in four or fewer positive slides out of six. The effect of this approach on efficacy and effort is difficult to quantify. Although the approach may be considered further, it seems sensible to proceed directly to a complete colonic evaluation after a positive test result as defined above.
Method of Complete Colorectal Evaluation
The best approach to a complete colorectal evaluation is to proceed directly to complete colonoscopy, assuming that high-quality colonoscopy is readily available [2.3.3, 2.4]. A possible alternative is an examination with flexible sigmoidoscopy and high-quality, air-contrast barium enema.
Colonoscopy is more sensitive than barium enema, and any lesions found on colonoscopy can be removed or sampled for biopsy [2.4]. However, even colonoscopy is not perfectly sensitive, and it may miss small lesions, especially behind folds. Colonoscopy incurs a higher risk for complications of sedation or perforation, but barium enema has a substantial false-negative rate for potentially important neoplasms, such as adenomas of 1 cm or larger. Inadequate or suspicious results of barium enema require subsequent colonoscopy about 15% of the time.
Interpretation of Negative Fecal Occult Blood Test Results
A negative result on a fecal occult blood test cannot rule out colorectal cancer. If symptoms that indicate possible colorectal cancer develop after a negative screening result, timely evaluation is warranted. (Such an evaluation of symptoms would be considered a diagnostic work-up rather than screening).
The sensitivity of fecal occult blood tests for important neoplasms is limited (probably between 30% and 50%); therefore, a negative result cannot rule out colorectal cancer. If symptoms develop that suggest colorectal cancer, only a test with almost perfect sensitivity would provide adequate assurance that an important neoplasm is not present.
Management after a Complete Colorectal Evaluation for a Positive Screening Test Result
If Colonoscopy Reveals No Clinically Important Colorectal Neoplasm
If complete colonoscopy reveals no colorectal neoplasm or only a single, small (<1 cm) tubular adenoma, further screening for colorectal cancer may reasonably be deferred for 5 or more years.
Information about the state of the colon provides strong prognostic information about colorectal cancer for 3 to 5 years or longer. If the results of high-quality colonoscopy are negative, screening with fecal occult blood tests for the next 5 years is not clinically useful [2.5, 2.6]. In making a decision about deferral, physician and patient must understand that the rate at which colonoscopy misses colorectal cancer is not zero, but it is small. The prognosticating power of negative results on sigmoidoscopy plus air-contrast barium enema is less certain.
If Colonoscopy Detects a High-Risk Lesion
If colonoscopy detects a lesion that is associated with a high risk for colorectal cancer, colonoscopic surveillance is warranted after the lesion is removed.
If colonoscopy reveals colorectal cancer, a large (≥ 1 cm) adenoma, an adenoma with villous features, or several small adenomas, the subsequent risk for colorectal cancer is high. Although the benefit of colonoscopy for surveillance has not been proven in a randomized, controlled trial, a strong biological argument can be made for periodic colonoscopic surveillance to reduce the increased risk for colorectal cancer [2.6].
- Copyright ©2004 by the American College of Physicians
