NEW YORK – Screening smokers for cancer with lung scans can lead to a high rate of false alarms, unneeded tests and biopsies, a new study suggests.

Some hospitals are promoting lung cancer screening with a special X-ray called a CT scan. And some advocacy groups encourage current or former smokers to get tested.

However, there’s no convincing evidence that such tests save lives, and no doctors group recommends routinely screening smokers who don’t have symptoms of lung cancer. Government researchers are studying whether the scans could save lives.

While researchers are still waiting for the answer to that, they do have results on just how often scans are wrong. Scientists with the National Institutes of Health say doctors and people considering lung scans should take into account the high risk of false alarms in their study of 3,200 people.

For those who got CT scans, the risk of a “false-positive” — finding a harmless spot — was 21 percent after one scan and 33 percent after two. For chest X-rays, the risk was 9 percent after one, and 15 percent after two.

Anything suspicious detected through screening often needs to be followed up with more tests, biopsies and even surgery to find out if it really is cancer. Complications can include lung collapse, bleeding and infection.

“You need to know ahead of time” about the risks, said the study’s lead author, Dr. Jennifer Croswell, of the NIH’s Office of Medical Applications of Research. “Once you have the test, you just can’t back out of that.”

The study’s findings are published in today’s Annals of Internal Medicine.

Lung cancer is the deadliest cancer. Last year, an estimated 159,000 Americans died of the disease. It usually isn’t diagnosed before symptoms appear. that time, the cancer has often spread outside the lung and is harder to treat. Researchers have been looking at ways to catch lung cancer sooner and to find out whether that’s worth doing. Since the 1990s, the focus has been on CT scans, which can pick up smaller abnormalities than X-rays.

The latest research is from a pilot study for the larger research under way. Current and former smokers ages 55 to 74 were recruited for the study. The 3,190 participants were randomly assigned to get a CT scan or a chest X-ray. They got a second test a year later, and were followed for another year.

In the CT group, there were 506 false-positives; in the X-ray group, there were 216. More than half had follow-up scans or X-rays. About 7 percent in the CT group and 4 percent in the X-ray group had an unnecessary invasive procedure, including surgery or a tube snaked down the throat, researchers found.