Correspondence

Improving the Fecal Occult-Blood Test

N Engl J Med 1996; 334:1607-1608June 13, 1996

Article

To the Editor:

My colleagues and I reported (Jan. 18 issue) that HemeSelect and a combination test in which HemeSelect is used to confirm positive Hemoccult II Sensa results are an improvement over Hemoccult II in screening patients for colorectal carcinoma.1 In their accompanying editorial, Ransohoff and Lang2 limited their endorsement of the combination test, because they misinterpreted our methods and did not recognize that our study was conducted in a setting of routine patient care, not in a clinical-trial setting. Fecal occult-blood testing has been a part of the Kaiser Permanente personal- health appraisal since the late 1970s.

We did not, as Ransohoff and Lang suggest, ignore the fact that neoplasms bleed intermittently and that “the sensitivity may be lower if lesions are tested again and do not bleed.” All patients received three specially designed cards; each card contained all three fecal occult-blood tests: Hemoccult II, Hemoccult II Sensa, and HemeSelect. We requested samples from three separate stool specimens, which were simultaneously applied to the tests. Because all the tests were on one card, we could assess the sensitivity, specificity, and positive predictive value of all three and the combination of Hemoccult II Sensa and HemeSelect. A combination test was considered positive if a positive Hemoccult II Sensa result was confirmed by a positive HemeSelect result. All other combinations of results were considered negative.

Ransohoff and Lang state that a specificity below 95 percent results in an unacceptably high rate of workups and, on this basis, dismiss both new tests, Hemoccult II Sensa and HemeSelect, as impractical. We agree in the case of Hemoccult II Sensa; however, the specificity of HemeSelect for detecting colorectal cancer and large polyps was 95.2 percent, with a narrow 95 percent confidence interval (94.7 to 95.7 percent) and thus, by Ransohoff and Lang's own criteria, should not be dismissed as unacceptable.

As Van Dam et al. note,3 fecal occult-blood testing has its limitations, but it is the only practical screening strategy currently available for the detection of cancer at an early stage throughout the colon. Now that virtually every major U.S. health organization recommends annual fecal occult-blood tests in patients 50 years old or older, we believe our data provide the best available information for making rational decisions about the use of the available tests.

James E. Allison, M.D.
Irene S. Tekawa, M.A.
Laura J. Ransom, M.S.
Alyn L. Adrain, M.D.
Kaiser Permanente Medical Care Program, Oakland, CA 94611-5693

3 References

1. 1

   Allison JE, Tekawa IS, Ransom LJ, Adrain AL. A comparison of fecal occult-blood tests for colorectal-cancer screening. N Engl J Med 1996;334:155-159
   Full Text | Web of Science | Medline

2. 2

   Ransohoff DF, Lang CA. Improving the fecal occult-blood test. N Engl J Med 1996;334:189-190
   Full Text | Web of Science | Medline

3. 3

   Van Dam J, Bond JH, Sivak MV Jr. Fecal occult blood screening for colorectal cancer. Arch Intern Med 1995;155:2389-2402
   CrossRef | Web of Science | Medline

Author/Editor Response

The authors reply:

To the Editor: We agree that endorsement of the combination test can be stronger if all three specimens are collected at the same time. We had understood the collections to be separate, because different compliance rates were reported for the three tests: 93.5 percent for Hemoccult II and 91.4 percent for Hemoccult II Sensa, but only 62.2 percent for HemeSelect.

We agree that Hemoccult II Sensa is not specific enough for use as a practical stand-alone test; the number of workups for false positive results would be enormous. The problem with HemeSelect as a stand-alone test is not lack of specificity (which is necessarily similar to that of the combination test) but rather the cost and effort involved; indeed, one advantage of the combination approach is that HemeSelect is performed only in persons with a positive Hemoccult II Sensa test. Unlike Hemoccult II (or Hemoccult II Sensa), HemeSelect cannot be developed in a physician's office, because it is considered a “complex” test, like an enzyme-linked immunosorbent assay, and must be sent to a suitably equipped laboratory. Although most hospital or commercial laboratories could provide this service, very few currently do. These logistic problems may constitute a substantial barrier in routine clinical settings, especially small offices with limited facilities and staff.

Even though the whole colon can be examined with fecal occult-blood testing, this technique has limited sensitivity, so that mortality from colorectal cancer is reduced by only about 30 percent.1 With sigmoidoscopy, half the colon is examined, but much more thoroughly, so that mortality can be reduced by at least 60 percent for cancers in the part of the colon examined.2 Furthermore, sigmoidoscopy does not need to be performed yearly to be effective.2 This relative simplicity may provide an important practical advantage, especially if fecal occult-blood screening becomes more complicated. Indeed, the case for once-in-a-lifetime screening with sigmoidoscopy is strong enough to have warranted a clinical trial in the United Kingdom.3

The good news is that the combination approach to fecal occult-blood testing provides better results than conventional Hemoccult II. The combination approach, however, is neither widely available nor — more important — clearly practical in routine clinical settings. The next step is to see how big the practical problems are and how they can be overcome.

David F. Ransohoff, M.D.
University of North Carolina, Chapel Hill, NC 27599-7105

Christopher A. Lang, M.D.
Colorado Permanente Medical Group, Denver, CO 80205

3 References

1. 1

   Mandel JS, Bond JH, Church TR, et al. Reducing mortality from colorectal cancer by screening for fecal occult blood. N Engl J Med 1993;328:1365-1371[Erratum, N Engl J Med 1993;329:672.]
   Full Text | Web of Science | Medline

2. 2

   Selby JV, Friedman GD, Quesenberry CP Jr, Weiss NS. A case-control study of screening sigmoidoscopy and mortality from colorectal cancer. N Engl J Med 1992;326:653-657
   Full Text | Web of Science | Medline

3. 3

   Atkin WS, Cuzick J, Northover JM, Whynes DK. Prevention of colorectal cancer by once-only sigmoidoscopy. Lancet 1993;341:736-740
   CrossRef | Web of Science | Medline

Citing Articles (2)

Citing Articles

1. 1

   T. Laubert, J. K. Habermann, F. G. Bader, T. Jungbluth, H. Esnaashari, H.-P. Bruch, U. J. Roblick, G. Auer. (2010) Epidemiology, molecular changes, histopathology and diagnosis of colorectal cancer. European Surgery 42:6, 252-259
   CrossRef

2. 2

   Diana L. Miglioretti, Carolyn M. Rutter, Susan Carol Bradford, Ann G. Zauber, Larry G. Kessler, Eric J. Feuer, David C. Grossman. (2008) Improvement in the Diagnostic Evaluation of a Positive Fecal Occult Blood Test in an Integrated Health Care Organization. Medical Care 46:Supplement 1, S91-S96
   CrossRef