In a commentary in the November 30, 1996 issue of The Lancet, David Lieberman, Oregon Health Sciences University, Portland, and Marvin H. Sleisenger of University of California, San Francisco, wrote: "Health-care policy makers are now confronted with clear evidence that screening can reduce mortality from the second leading cancer killer in Europe and North America," which is what cancer of the colon/rectum is. They were referring specifically to two papers now hovering on the edge of the Top Ten and to an older study, the Minnesota trial (New Engl. J. Med., 328:1365-71, 1993).

   The Nottingham, U.K. randomized study by Jack D. Hardcastle and colleagues (The Lancet, 348:1472-7, 1996; currently at #12 with 19 citations this period and 60 to date) recruited 150,000 people, half of whom were offered screening by testing for fecal occult blood. The simlilarly sized population base for Ole Kronborg’s team was in Funen, Denmark (The Lancet, 348:1467-71, 1996; at #13, with 19 cites this period and 63 to date), and the test method was also the commercial Haemoccult kit for blood in stools. The 18% reduction in mortality from this cancer in the screened group would, the Danish authors say, prevent 360 of that country’s annual mortality of 2,000 for colorectal cancer. In Nottingham the reduction in mortality specific for this cancer was 15% in the group randomized to screening (though not all agreed to the tests); total mortality was not affected, in part, perhaps, because of the higher mortality in those refusing screening.

   Transfer of impressive results--and, though both reports are preliminary and further data are even now accumulating, the studies are impressive--from the research setting to, say, a national screening policy is always difficult. Occult blood testing is cheap and simple but the compliance rate is far from perfect. There will always be false-positives, and that means anxiety for the patient and unnecessary (with hindsight) colonoscopies. Tests would have to be repeated every one or two years. Screening colonoscopy could be done less frequently but compliance with this strategy in the general adult population and the resource implications are largely unexplored. Another worry is the awkward habit of research results not translating exactly into day-to-day healthcare settings.

   In the wake of these papers, the U.K. Department of Health has held two working-parties to review the evidence, and two pilot projects to look at the service implications of such screenings are envisaged.

   Kronberg draws Science Watch readers to recent publications on the implications (Health Economics, 7:1-7, 9-20, 21-29, 1998). In the Danish trial the seventh biennial screen was done in July, 1998, and Kronberg notes that the trial continues to demonstrate "a possible reduction of incidence of colorectal cancer following removal of more large polyps (adenomas) in the test group." The latest cumulative figures show that, among the cancers, 21% in the screened group but only 11% in the controls were Dukes’ stage A (a more favorable prognosis).