Another recent failure of conventional wisdom involved diabetes patients. Over the past few years, many doctors have tried to lower diabetics' high blood sugar levels as much as possible - as close to those of normal patients as medication, diet and exercise allowed. But federal health officials this month halted one part of a major study testing this hypothesis after an abnormal number of patients died as they closed in on "normal" blood sugar levels.
So it might be with lung cancer screening. In the 1970s, the practice began to catch on, based on the apparently reasonable notion that chest X-rays would find cancers at a smaller, more curable stage. There was pressure to make such screening routine - just as there is in some circles today with CT scans - until a major Mayo Clinic study showed it did not save lives.
Most lung cancers caught before patients show symptoms are found accidentally when a patient comes into a doctor's office or hospital complaining of something else.
Routine screening has been proved to work for other cancers - such as mammography and colonoscopy - but it's not widely used for lung cancer, which kills more people each year than breast, colon and prostate cancers combined.
The newer, more sensitive technology for screening lungs - called spiral CT, or computed tomography - has been tested but not in the rigorous fashion considered the scientific gold standard: a randomized clinical trial.
The largest study of lung cancer CT scans was conducted by the International Early Lung Cancer Action Program (I-ELCAP). Results published in The New England Journal of Medicine in October 2006 showed that the technology caught a large number of early-stage lung cancers - and found that of those who had surgery within one month of diagnosis, the survival rate was 92 percent.
The eight patients who were diagnosed with early-stage lung cancer and who did not receive treatment died within five years of diagnosis.
But critics don't consider this a definitive test, mainly because there was no control group with which to compare the results. Others criticize the methodology, saying many cancers were "cured" that didn't need to be, and that early diagnoses did not translate into fewer late-stage cancers or deaths.
Critics said some patients appeared to live longer only because they were diagnosed years before they would have been otherwise. Most died of lung cancer anyway.
