If a doctor believes an antibiotic or sedative is not needed and instead provides a placebo to a patient, is the physician protecting public health or subjectively disregarding the ethics of full disclosure? You decide.

Many rheumatologists and general internal medicine physicians in the US say they regularly prescribe "placebo treatments" including active drugs such as sedatives and antibiotics, but rarely admit they are doing so to their patients, according to a study on BMJ.com today.

The use of placebo treatments in clinical practice has been widely criticized because it is claimed that the practice by its very nature is deceptive and therefore violates patients' autonomy. But advocates of placebo treatments argue that they could offer effective treatment for many chronic conditions without necessarily deceiving patients. Despite the controversy, there has been little data to-date on doctors' attitudes towards and the use of placebo treatments in the US.

Dr Jon Tilburt and his colleagues from the National Institutes of Health as well as collaborators at Harvard and the University of Chicago examined the attitudes and behaviors to placebo treatments in a national sample of general internal medicine physicians and rheumatologists in the US.

The researchers sent a confidential survey to 1200 randomly selected practising general internal medicine physicians and rheumatologists (a group of doctors who commonly treat patients with debilitating chronic conditions that are notoriously difficult to manage medically).

The authors report that among the 679 physicians (57%) who responded to the survey half of them said they prescribed "placebo treatments" on a regular basis. Most physicians (62%) believed the practice to be ethically acceptable and were happy to recommend or prescribe placebo treatments.

The most commonly used placebo treatments prescribed in the past year were over the counter painkillers (41%) or vitamins (38%). Some of the physicians reported using antibiotics (13%) and sedatives (13%) as placebos, only 3% reported using sugar pills.

Interestingly, among those who prescribe placebo treatments, most doctors (68%) said they typically describe the placebo treatments to patients as "a potentially beneficial medicine or treatment not typically used for their condition", only rarely did they admit to explicitly describing them to patients as "placebos".

Although there was only a moderate response rate to the survey (57%), even if all the non-responders never gave placebos, placebo prescribing is still surprisingly common, say the authors.

The authors say that while the use of placebos has been controversial, the physicians in the study did not believe they were behaving unethically by either using placebos or not being upfront with their patients about doing so.

They point out that understanding the role of placebo treatments in contemporary medicine is complex. They conclude that prescribing harmless treatments like vitamins or over the counter painkillers to promote positive expectations without full disclosure of motivations might not raise alarm bells, but prescribing antibiotics and sedatives when there is no clear medical indication could have serious adverse consequences for both patients and public health.