Current issues of ACP Journal Club are published in Annals of Internal Medicine


Diagnosis

Review: Positive test results identify early colorectal cancer and save lives

ACP J Club. 1997 Nov-Dec;127:73. doi:10.7326/ACPJC-1997-127-3-073


Source Citations

Ransohoff DF, Lang CA. Clinical guideline: part II. Screening for colorectal cancer with the fecal occult blood test: A background paper. Ann Intern Med. 1997 May 15; 126:811-22.

American College of Physicians. Clinical guideline: part I. Suggested technique for fecal occult blood testing and interpretation in colorectal cancer screening. Ann Intern Med. 1997 May 15;126:808-10.


Abstract

Objective

To make recommendations for interpreting colorectal cancer (CRC) screening test results using fecal occult blood tests (FOBTs) and follow-up of positive test results.

Data sources

Studies were identified using MEDLINE (1984 to 1996) with the terms fecal occult blood and CRC screening and bibliographies.

Study selection

Studies about the biology of CRC related to screening with FOBTs, validation and clinical evaluation of FOBTs, and cost-effectiveness were selected.

Data extraction

Patient characteristics, test properties and characteristics of various tests and procedures, mortality, data on rescreening and evaluation of positive test results, and costs.

Main results

3 studies found reduced mortality after FOBT (Hemoccult/Hemoccult II) (relative risk reductions of 33%, 18%, and 15%). The 33% reduction represents the number needed to screen over 13 years to prevent 1 additional death (approximately 300). In patients who have a workup after a positive test result, the probability of finding early-stage CRC ranges from 3% to 14%, any CRC ranges from 5% to 18%, and early-stage cancer or large adenoma ranges from approximately 20% to 40%. The rate of false-positive results ranges from 2% to 4% for nonrehydrated samples and 8% to 16% for rehydrated samples. Guaiac-based tests (Hemoccult II and HemoccultSENSA) can be done in offices. Immunochemical tests, including HemeSelect, are currently laboratory based. Hemoccult II testing usually has 2 window slides from 3 separate fecal specimens with a positive result defined as ≥ 1 positive window.

Conclusion

Follow-up of positive test results after screening for colorectal cancer with fecal occult blood tests identifies early-stage colorectal cancer and reduces mortality.

Source of funding: American College of Physicians Clinical Efficacy Assessment Program.

For article reprint: Dr. D.F. Ransohoff, CB#7105, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599-7105, USA. FAX 919-966-2274.


Table. Test characteristics and costs of fecal occult blood tests*

Test Relative sensitivity (%) Specificity (%) +LR -LR Cost† (U.S. $)
Hemoccult II 37 98 18.5 0.6 3.31
HemoccultSENSA 79 87 6.1 0.3 3.82
HemeSelect 69 94 11.5 0.4 19.23
HemoccultSENSA plus HemeSelect 66 97 22 0.3 5.22

*+LR = likelihood ratio for presence of disease if the test is positive; -LR = likelihood ratio if the test is negative.
†Cost of 3 samples.


Commentary

Ransohoff and Lang and the accompanying American College of Physicians' guideline make several practical recommendations on the use of FOBTs as part of a screening strategy to reduce death from CRC. Their recommendation to encourage but not overemphasize dietary and medication restrictions is reasonable. A positive test result warrants a complete evaluation of the colon rather than repeated testing, given the 30% incidence of a large adenomatous polyp or cancer. The authors correctly point out that, although common in clinical practice, the use of a stool specimen obtained during rectal examination should be discouraged as part of a screening strategy because insufficient data exist to support the interpretation of a positive result.

The authors' recommendation to avoid rehydration of stool specimens is controversial. Annual screening and rehydration may increase the effectiveness of FOBTs from a 15% to 18% reduction to a 33% reduction in death from CRC. The major drawback to rehydration is the 4-fold increase in the rate of false-positive results. A recent cost-effectiveness analysis indicates the cost per year of life saved with FOBTs (assuming a specificity consistent with rehydration of stool specimens and 100% compliance) is less than U.S. $20 000 (1). Given the apparent cost-effectiveness of FOBT with sample rehydration, it seems difficult to firmly recommend a less-effective strategy.

Various options are available for CRC screening, including FOBTs, flexible sigmoidoscopy, combined FOBTs and sigmoidoscopy, and periodic or 1-time colonoscopy. The efficiency and cost-effectiveness of these strategies depend greatly on patient compliance (2) and the availability and cost of procedures. As efforts are made to reduce the cost of screening colonoscopy, it may become the most cost-effective strategy (3).

Robert A. Gluckman, MD
Providence-St. Vincent HospitalPortland, Oregon, USA


References

1. Winawer SJ, Fletcher RH, Miller L. Gastroenterology. 1997;112:594-642.

2. Lieberman DA. Gastroenterology. 1995; 109:1781-90.

3. Bhattacharya I, Sack EM. Lancet.1992; 347:1744-5.