TOWARDS the end of 1990 Dr Andrew Leigh Brown, director of the Centre for HIV Research at Edinburgh University, took a call from the Maryland State Health and Hygiene Department in the US. They wanted his help in an 'HIV whodunnit'. A patient was found to be infected with the Aids virus, and the prime suspects were a surgeon from the Johns Hopkins Hospital in Baltimore, and a blood donor.

The surgeon, Rudolph Almarez, died from Aids in November 1990. He had operated on hundreds of patients - many of them women with breast cancer - for several years while HIV-positive. He had not told anyone about his condition. When the news broke, Johns Hopkins was inundated with calls from panic-stricken former patients.

In several towns in the UK, similar fears were being expressed last week after details emerged of three HIV-positive doctors who had treated patients while infected. Terence Shuttleworth, a gynaecologist from Kent, had an estimated 17,000 patients and performed 6,000 operations over the 10 years that he may have had the virus. Dr Peter Clayton, a trainee GP from Bridgend, worked in casualty departments and obstetric units in several hospitals in south Wales. He died from Aids in October 1992. Another trainee GP, who has yet to be named, worked in casualty at Bolton Royal Infirmary, in general practice in the area, and for other health authorities around the country, before dying from an Aids-related illness last summer.

Given the number of patients involved in these three cases who may opt for blood tests, it is statistically likely that at least one case of HIV infection will emerge. The incidence of infection outside London is 1 in 10,000 cases and may be higher in the Medway towns, where Mr Shuttleworth practised, as they are close to the capital, with its high concentration of HIV-infected individuals.

But even if HIV-positive patients are found, it is likely that the doctor involved will not be the source of the infection. Proving that fact beyond all reasonable doubt will be crucial. Should the virus in a patient be shown to be related to that carried by the doctor, it would be the first case of a doctor-patient transmission. Such a finding would have far-reaching implications for the ways that doctors work in hospitals worldwide and would strengthen calls for compulsory testing of health-care workers.

If, however, the viral strain proved to be unrelated, health officials would be able to reiterate last week's most popular quote - that there is 'minimal risk' of transmission between doctor and patient - with more confidence.

In the Baltimore case, just one previously unknown HIV-positive case was identified among 954 patients who agreed to take HIV antibody tests. This was a woman who had no obvious risk factors for the disease, and had one heterosexual partner who was healthy. But in 1984, just before the screening of blood donations was introduced in the US, she had received a blood transfusion. Medical records revealed that she had been given blood from a donor who was found to be HIV- positive.

'We were asked to see if we could shed any light on who was the most likely source of the virus,' says Dr Leigh Brown. The Edinburgh centre is internationally renowned for genetic studies of HIV. At the request of the Centers for Disease Control in Atlanta, scientists in Edinburgh had worked on the now infamous case of the Florida dentist who infected at least three of his patients with HIV. This is the only known case to date of HIV transmission by a health worker to patients.

'Using a particular area of the HIV virus, known as the gag gene, we had to demonstrate the extent of the relationship between samples of the virus from the individuals,' Dr Leigh Brown explains.

'The sequence data (the sequence of chemicals, known as bases, making up the genetic material) from the donor was more closely related to that of the patient. With the information we obtained we concluded that the woman's infection was consistent with infection from the donor, and quite inconsistent with infection from the surgeon.' A paper describing the centre's detective work will be published shortly in the Journal of Infectious Diseases.

Dr Leigh Brown has offered the expertise of the centre to the Department of Health should a patient of one of the HIV-positive doctors also test positive. The publicity surrounding the Baltimore and Florida cases led the American Medical Association and the American Dental Association to advise HIV-positive practitioners to tell their patients, or to give up surgery. The outcome of the recent British 'Aids doctor scares' may determine whether that guidance needs to be strengthened.