WASHINGTON, Jan. 4—The free samples of prescription drugs many doctors hand out to their patients are coming under heavy fire from critics who charge that the drugs are simply a marketing tool used to encourage doctors to prescribe—and patients to demand—the newest, most expensive medicines.

The pharmaceutical industry contends that the free samples—worth billions of dollars—that sales representatives give to doctors each year especially benefit their poorer patients. But a new study shows the poor and uninsured are less likely to get the free medicines than well-off patients.

The study, by researchers at Harvard Medical School and the Cambridge Health Alliance, a network of nonprofit health care centers affiliated with Harvard, examined data on 33,000 U.S. patients and found that those with the highest incomes were the most likely to get free samples. The pharmaceutical industry was quick to respond to the study, defending the practice and calling the research flawed.

The study, the first of its kind on the nationwide distribution of free drug samples, will be published in the American Journal of Public Health in February.

It's not that doctors don't try to help their needy patients, the researchers say, it's that many needy patients can't afford to visit a doctor's office very often in the first place.

"Doctors like to give free samples, and patients like to get them, and no one wants to think of them as a powerful marketing tool of the drug industry," says Sarah Cutrona, M.D., of the Cambridge Alliance, an instructor of medicine at Harvard and the lead author of the study. "But that's what samples are. And what they really are not is a safety net for the needy."

The study, using data collected in a 2003 federal survey of consumer health spending, found that 12 percent of all patients got at least one free drug sample. About 13 percent of the insured got a sample, and 10 percent of the uninsured. Only 28 percent of those receiving samples had income below the poverty level.

"The places the drug reps are targeting are not where the needy patients go," says Cutrona.

Moreover, she says, there is mounting evidence that free samples may influence a physician's prescribing habits and lead to safety problems. For instance, Cutrona says, "we found that the most widely distributed sample in 2002 was [the painkiller] Vioxx, with Celebrex being number three. These drugs turned out to have lethal side effects," and Vioxx was withdrawn from the market.

Overall, she says, "our study shows that samples are potentially dangerous and do little for the needy."

"Providing physicians with free samples...clearly benefits patients and advances health care throughout the United States," Ken Johnson, senior vice president of the Pharmaceutical Research and Manufacturers of America (PhRMA) said in a press release.

"Clearly, free samples often lead to improved quality of life for millions of Americans, regardless of their income," he said.

Some drug experts argue that the practice of free samples often has unexpected consequences.

"When they receive free samples, doctors tend to rapidly increase their prescriptions of these new drugs and yet, unlike older drugs, we really don't know what side effects these medicines have, " says Susan Chimonas, co-director of research for the Prescription Project, a national campaign funded by the Pew Charitable Trusts to advocate for evidence-based prescriptions.

Chimonas says newer medications often have side effects that aren't discovered until several years after the drug is introduced. Because older, frailer patients are usually underrepresented in drug trials, new medicines can be particularly hazardous for them. Older patients taking a number of drugs are usually safer with an older medicine, she says, because doctors can’t be sure of how a new drug will interact with the other medications.

And yet, Chimonas says, "studies show that 90 percent of doctors would provide a medicine that was not their first choice for a patient, simply because they have a free sample of that drug."