A woman patient is examined in a west London clinic ... a doctor in Belfast makes the diagnosis

Sometimes when people arrive at Parsons Green Health Centre in south-west London, they are told that they need to be diagnosed by a doctor in Belfast. It's a problem, but one which the centre intended, because it is one of two in London experimenting with "telemedicine".

Showing the problem to the Belfast doctor is as easy as picking up a hand-held television camera and dialling a phone number. The picture is sent down a phone line to a screen in a Belfast hospital, where a doctor can study the pictures and use the phone to ask questions before making a diagnosis.

In this way, the nurses at the centre - which has no doctor on the premises - are able to make diagnoses that would otherwise involve referral to a hospital. "It's a very useful technology," said one of the nurses. Cheap, too: using the video link, on average 12 times a month, means the centre doesn't have to employ a doctor - saving about pounds 50,000 annually.

While much has been made in the past few years about the high-technology, high-price items in medicine - such as magnetic resonance imaging (MRI) or Positron Emission Tomography (PET) scanners - less attention has been paid to telemedicine, which takes advantage of the falling costs of communications, computers and video and audio technology.

Yet in some ways, telemedicine could change health care more radically than the expensive scanners that attract so much attention.

Both Parsons Green and South Westminster Health Centre are designed as "low intensity" centres: seriously ill patients would be sent immediately to a hospital; those with minor problems can be treated by the nurses, who can also prescribe a limited range of drugs. But some cases are harder to diagnose, which is when they use a telemedicine link to Belfast's Royal Victoria Hospital.

Similar trials have linked pregnant women on the Isle of Wight with specialists at Queen Charlotte Hospital in London, who can analyse live ultrasound pictures sent over the phone line and decide whether a trip to the mainland is needed or not.

"It's got great potential," said Dr Richard Wootton, director of the Institute of Telemedicine and Telecare at Queen's University, Belfast. "It has the potential to revolutionise the way that health care is delivered. But the problem is that we don't know if it's cost-effective, in strict accounting terms."

Telemedicine has been around only for the past 10 years, because it requires both computers able to compress the video signals, and falling telecommunications costs. It began in the US, where the large distances in remote areas made it sensible to be able to make informed diagnoses so that patients would know if it was worth making the full journey for a face-to-face consultation. In 1991, there were just four full-scale telemedicine programmes in the US; by 1996, there were 110.

The UK has been catching up rapidly. Later this week, a conference will take place in London to discuss the possibilities of "telepathology" - in which specialists would be able to study samples taken from patients without physically being present. Other studies may produce "teledermatology" - remote study of skin problems - and even telesurgery. The US Army has already tried the latter out, though only on a pig.

Part of the problem with telemedicine, though, is that under standard accounting systems it is hard to justify. "There are clear clinical benefits, in terms of getting instant referral," said Dr Wootton. "Patients feel more reassured, and it saves them the cost of travelling to the doctor for the consultation. But those are intangible benefits in accountancy terms. It might work to reduce waiting lists, but maybe by improving access to hospitals, it won't. I think nobody will know until we do it."