Bowel cancer screening, the latest preventive medicine being urged on the public, provides an instructive case study of how a single interpretation in medicine quickly transforms into established fact and, before long, into justification for wholesa
The medical profession is frustrated by the fact that in the US 140,000 cases of bowel cancer appear and claim the lives of 60,000 every year. In Britain, more people die from bowel cancer than from breast and cervical cancer together. Understandably, medicine is desperate to be seen to be doing something.
Both The Lancet (20 March 1993) and the Journal of American Medical Association(10 March 1993) were recently aflutter about sigmoidoscopy. This test is conducted with a sigmoidoscope a 12 inch long hollow lighted tube which is inserted into your back passage. This erstwhile flashlight allows your doctor to view the inside of your rectum and the last foot of your colon to check for any suspicious lumps.
The reason for all the fuss was a single test conducted at Kaiser-Permanente in Oakland, California. The researchers simply matched the case records of 261 patients who'd died of bowel cancer (only 9 per cent of whom had undergone screening) with 868 control patients who were still alive (24 per cent of whom had been screened). These few facts led the researchers to conclude that exposure to sigmoidoscopic screening within the last decade was associated with a 70 per cent reduction in the risk of fatal bowel cancer.
This (and another non randomized, controlled study) were trumpeted by both The Lancet and JAMA as conclusive rationale for plunging into full scale bowel screening every three to five years in people over 50.
Any student of logic could pick apart the supposed conclusions of the Californian study in a minute. Just because a quarter of the survivors were screened doesn't mean that the test saved them. What were their diets like? The sizes of their tumours? All they were matched for was age and sex. What about the three quarters of survivors who weren't screened?
What worries me is that this 70 per cent figure will now be regarded as established fact. It may preempt any proper study of, say, whether harmless abnormalities in the colon like polyps always at some point convert into cancer. It will blind doctors to the not inconsiderable potential side effects of sigmoidoscopy: intestinal perforation, bleeding and, in heart patients, irregular heartbeat.
It is an apt example of the clouding of vision that occurs anytime somebody says: "We know what we're doing is right."