Minnesota Colon Cancer Control Study
The effect of fecal occult-blood screening on the incidence of colorectal cancer
This is a structured summary of a research study originally published in 1993 in the [New England Journal of Medicine] [1]. This page is under development; please see the [discussion] of this page for details
Summary
A randomized controlled trial of 46,551 participants aged 50 to 80 years were randomized to one of:
- screening for colorectal cancer once a year using the stool guaiac test. Per the text of the paper, 'six guaiac-impregnated paper slides with two smears from each of three consecutive stools'.
- screening every two years using the stool guaiac test. Per the text of the paper, 'six guaiac-impregnated paper slides with two smears from each of three consecutive stools'.
- control group.
Results
| Deaths from colorectal cancer |
Alive | Totals | |
|---|---|---|---|
| Annual screening | 82 | 15488 | 15570 |
| Control | 117 | 15277 | 15394 |
| Totals | 199 | 30765 |
(this table was reconstructed by using the totals in Table 1 of the article and using the outcomes in Table 4 of the article)
According to the article, the 13-year cumulative rates of mortality from colorectal cancer were:
- 0.588% in the annually screened group (95 percent confidence interval, 4.61 to 7.15)
- 0.833% in the biennially screened group (95 percent confidence interval, 6.82 to 9.84)
- 0.883% in the control group (95 percent confidence interval, 7.26 to 10.40)
These numbers lead to the following measures of efficacy:
| Absolute risk reduction | 0.3% |
| Number needed to treat | 339 |
| Relative risk reduction | 33.4% |
Per the authors, 'the rate in the annually screened group, but not in the biennially screened group, was significantly lower than that in the control group'. The p-value is not in the article.
Are the results significant?
- To assess whether the results are statistically significant, not only the p-value is important, but factors such as publication bias that might influence the p-value are important. To alter the a priori estimate of the null hypothesis, click here. (this paragraph needs to be much more user friendly)
- Clinical significance must also be considered. For example, the results may be statistically significant, but the number needed to treat may reveal that too few patients will benefit for patients to be willing to accept expense and effort of the treatment.
How will these result be changed in patients at higher or lower risk of the outcome?
The risk of death from colorectal cancer in this study is 0.883% in the unscreened group. For patients who have risk factors that make their risk higher or lower than this number, their benefit will be higher or lower. Use [this link] to adjust the baseline risk.
Follow-up
The results of the is study after 18 years of follow-up have been published [2].
References
- ↑ Mandel J, Bond J, Church T, Snover D, Bradley G, Schuman L, Ederer F (1993). "Reducing mortality from colorectal cancer by screening for fecal occult blood. Minnesota Colon Cancer Control Study". N Engl J Med 328 (19): 1365-71. PMID 8474513. "
- ↑ Mandel J, Church T, Bond J, Ederer F, Geisser M, Mongin S, Snover D, Schuman L (2000). "The effect of fecal occult-blood screening on the incidence of colorectal cancer". N Engl J Med 343 (22): 1603-7. PMID 11096167.