Who is to blame for the rising inappropriate use of antibiotics? A new report suggests that doctors believe that, in many cases, the overprescribing of these drugs may not be due to misdiagnoses, but to insistent requests and subtle hints from their patients.

Indeed, the manner in which the patient communicates his or her illness to the physician strongly influences whether or not the physician prescribes an antibiotic, say researchers in New Jersey.

In a report almost entirely uncritical of doctors, the investigators analysed data from a 1996-1999 observational study of nearly 300 outpatient visits to 18 family practices in the US. All of the visits were related to acute respiratory tract infections.

Doctors prescribed antibiotics in nearly 70 per cent of visits but, according to guidelines devised by the US Centers for Disease Control and Prevention, a prescription was uncalled for in 8 out of 10 of these visits.

The unnecessary prescriptions were often given as a result of subtle pressuring of physicians by the patients, and the doctors’ desire to 'maximise patient satisfaction' with their care. In a third of the cases, patients described their condition as so severe that they could not cope on their own. 'I just can’t shake it', one patient said - a statement the researchers say 'implies . . . the need for a prescription medication to alleviate his sore throat'. In 13 per cent of the visits, patients reported a previous positive experience with an antibiotic, which the doctors interpreted as an indirect suggestion that an antibiotic was again necessary to alleviate symptoms.

Other approaches were explicit requests for antibiotics and appeals to non-medical circumstances, such as a patient’s desire to be rid of an illness before an upcoming vacation, each of which was observed in 6 per cent of visits.

The findings, say the authors, show that 'the encounter between the physician and the patient with respiratory infection is an intricate dance in which both partners play important roles' (J Fam Pract, 2001; 50: 853-8).