Effect of fecal occult blood testing on mortality from colorectal cancer. A case-control study

Ann Intern Med. 1993 Jan 1;118(1):1-6. doi: 10.7326/0003-4819-118-1-199301010-00001.


Objective: To estimate by case-control methods the effect of screening using the fecal occult blood test (FOBT) on mortality from colorectal cancer and to examine the relation of that effect to the interval since the most recent screening test.

Design: A case-control study.

Setting: The Kaiser Permanente Medical Care Program of Northern California.

Patients: A total of 485 persons who developed fatal colorectal cancer after 50 years of age and 727 age- and sex-matched controls.

Measurements: History of screening FOBTs during the 5 years before case diagnosis.

Results: After adjustment for potentially confounding factors, an odds ratio of 0.69 (95% Cl, 0.52 to 0.91) was observed for exposure to at least one screening FOBT during the 5-year interval. The odds ratio was lowest for the first year after the most recent FOBT and rose to 1.00 three years after the last screening examination. False-negative results among cases in the 1 to 2 years before diagnosis contributed substantially to lowering the estimate of efficacy.

Conclusions: These data suggest that a program of annual or biennial screening using FOBTs might lower population risk for mortality from colorectal cancer sufficiently to have important public health implications. However, the confidence intervals around our odds ratio estimates were wide. We therefore believe that additional data will be needed before making recommendations that FOBT screening be expanded.

Publication types

  • Research Support, U.S. Gov't, P.H.S.

MeSH terms

  • Adenocarcinoma / mortality
  • Adenocarcinoma / prevention & control
  • Aged
  • Case-Control Studies
  • Colorectal Neoplasms / mortality
  • Colorectal Neoplasms / prevention & control*
  • Confidence Intervals
  • Confounding Factors, Epidemiologic
  • Female
  • Humans
  • Logistic Models
  • Male
  • Middle Aged
  • Occult Blood*
  • Odds Ratio
  • Sensitivity and Specificity