A new study casts doubt on the potential of lung cancer screenings to save lives.

Patients screened with spiral CT scans are three times more likely to be diagnosed with lung cancer. But they're no less likely to die from the disease than if they were never tested, according to an analysis in today's Journal of the American Medical Association.

Although smokers and ex-smokers may worry about their risk of lung cancer, Peter Bach of New York's Memorial Sloan-Kettering Cancer Center suggests that they wait for the results of the National Lung Screening Trial - a study of 50,000 patients - whose results are expected in 2009.

ON THE WEB:Calculate your risk of lung cancer

Bach's results differ from those of an October study in The NewEngland Journal of Medicine, led by Claudia Henschke of Weill Medical College of Cornell University. Henschke estimates that screening could boost 10-year survival rates to 80% - much higher than the current rate of 11%.
FIND MORE STORIES IN: Trial | Patients | Bach | CT scans | Claudia Henschke

Bach notes that Henschke's trial did not include a comparison group of unscreened patients. So it's possible that unscreened patients could have lived just as long, especially if the CT scans picked up slow-growing tumors that would have never caused symptoms.

New research suggests that small lung tumors found by CT may be very different than those found through older methods, such as X-rays, says William Black, a professor at Dartmouth-Hitchcock Medical Center in New Hampshire, who was not involved in Bach's study. Tumors found through CT can take five times longer to double in size than other lung tumors, for example. A 1-centimeter tumor like this could take 14 years to prove deadly, he says.

And Bach notes that even painless exams carry risks, especially if they lead to unnecessary and risky surgeries. On average, 1 in 20 people who have a lung removed die within 30 days of the operation, he says.

Bach acknowledges that his analysis, which involved 3,246 patients, is "small and preliminary." Because researchers have not yet completed large trials comparing screened and unscreened patients, Bach pooled the results of three studies and created a "synthetic" comparison group using a statistical model.

Henschke questions Bach's methods and says his study has several major problems. Bach followed patients for 3.9 years, for example, which may not be long enough to notice a real difference in lung cancer death rates.

The American Cancer Society's Robert Smith notes that - for now - there's no definitive answer about the value of lung screening. The society neither recommends nor discourages the tests.

Black says that patients should think carefully about whether they really need screening. He notes that ex-smokers may face a much greater risk from heart disease than cancer and could gain more from exercise and a healthy diet than a lung test.

"What a physician should tell people is, 'Don't smoke,' " Black says. "Not smoking would have a much bigger effect than getting a CT screening."