The dreaded MRSA superbug has been making the headlines again. Yesterday an expert at the government's Health Protection Agency, Georgia Duckworth, together with other infection specialists, warned of a sharp rise in its prevalence in children, while the Sun newspaper reported apparently alarming results from an undercover investigation into the problem at a London hospital, already known to have one of the worst rates of MRSA in patients.

Less than a month ago the HPA was suggesting that the bacterium, resistant to frontline antibiotics, was implicated as the underlying cause or contributory factor in the deaths of 800 patients in England and Wales during 2002, a 15-fold increase on 1993. And in December last year the chief medical officer, Professor Sir Liam Donaldson, conceded that much more needed to be done in tackling the steep increases in all hospital-acquired infections, a phenomenon that might affect as many as one in 11 patients. This is a staggering statistic for a health service designed to cure or mitigate illness, not make it worse. Things may have got so bad that recently vets reported finding MRSA in cats and dogs. Normally such infections leap from ani mals to humans. It seems likely in this case that humans have passed it to their pets.

So how worried should we be that a bug, common on the skin and in the nasal passages of many healthy people, can turn into a potential killer in already sick patients, poisoning the bloodstream, threatening organ failure, sometimes leading to amputation and death? The answer is very. Not only because of the consequences of its higher than acceptable prevalence as a bug in itself, but in its totemic significance for hospitals' more general infection control. Specialists in the field believe the medical profession has become dangerously complacent, ignoring sim ple measures such as hand-washing, under pressure to shift as many patients through beds as quickly as possible to meet more pressing priorities such as cutting waiting lists. And infection is not only dangerous in dealing with emergency cases. Complications, for instance, in routine surgery, such as replacement hip operations, can leave people permanently disabled.

MRSA, or to give its full name, Methicillin-resistant Staphylococcus aureus, arrived in this country in the early 1980s, possibly from Australia. It is no more virulent than the Staph bug itself, one that can cause all sorts of infections once it has invaded parts of our bodies that our natural defences might usually protect.

It started developing resistance to penicillin almost as soon as that wonder drug heralded a brave new post-second world war world of antibiotics. Methicillin, however, was the antibiotic of choice in treating many infections that penicillin could no longer target, and later strains of MRSA are also proving resistant to other frontline drugs that have better penetrative power in areas such as the brain or heart valves.

Now more than 40% of Staph infections identified in British hospitals are methicillin resistant, when it was less than 5% in 1990. This country was once at the forefront of "search and destroy" checks against MRSA infection, but now compares abysmally with other European countries such as Denmark and the Netherlands, where resistance rates are just 1%, Austria (11%) or Germany at (19%), although Greece, Italy and Portugal have comparable problems. Among children in hospital, as many as 13% of Staph infections may be resistant - not as high as among older patients, but the first evidence of an emerging problem in paediatric wards.

Dr Duckworth, the Health Protection Agency's expert in the field, says: "Levels are very high. For a country like ours, it is embarrassing - if not worse - to have rates that high. It puts us in leagues we would not want to be in, and should not be, given our history." She insists there are signs that the steep rises of the late 90s, which also saw reports of blood poisoning by MRSA leap 25-fold, appear to be levelling off. Voluntary, then mandatory, reporting systems in hospitals may have helped concentrate minds that were once diverted elsewhere.

The importance of infection control was largely ignored in the new NHS managerialism, but not any longer. Ministers have ordered all hospital trusts to appoint managers responsible for monitoring infection, reporting directly to the chief executive and board. League tables of performance on MRSA are designed to "name and shame" hospitals into tougher action. The worst perfomers have records seven times worse than the best.

There is, however, a recognition that some of our best hospitals will appear to have some of the worst rates, sometimes because they have diverse surgical specialities, sometimes because they are taking cases from other hospitals, where the actual infection may have occured. League tables for other hospital-acquired infections may follow, eventually, allowing patients to consider a hospital's record on controlling infection, as well as its surgical and medical success rates, when they choose where to be treated.

Some hospitals - in York, Liverpool, Nottingham, Lincolnshire, Northamptonshire and London - have experimented with measures to improve things further. These repeat basic measures such as hand-washing and encouraging medical staff to carry small containers of anti-bacterial fluids on belts to clean their hands between patients. Patients themselves are also being asked to check with staff whether they have washed their hands to raise awareness. Officials are evaluating the success of these programmes before deciding whether they could be repeated nationwide.

Lessons might be learned from abroad. The Dutch have a good record on controlling MRSA, partly because they have far more single rooms to screen patients needing routine operations. Here they can isolate patients until results are available, and that can take up to three days.

Some British hospitals are already employing such practices themselves, but it may be some time before all medical centres can follow suit. And with people queueing up for beds even before they are empty, there is not always time to clean sufficiently thoroughly. It is in part an unavoidable price, Duckworth argues, for improved treatment. In intensive care units for instance, she says: "Most life support breaks through the normal defence barriers. You have lines going into every orifice and lines going where you don't have orifices."

Thus the gastric juices, hair-like cilia of the windpipes and other bodily secretions that protect the body naturally are bypassed, leaving a certain degree of hospital-acquired infection inevitable. A similar story could be told about treatments against cancer and other diseases. That does not mean that more could not be done to limit MRSA. "We have to do our damnedest," says Duckworth.

Peter Walsh, chief executive of Action Against Medical Accidents, believes hospitals are still complacent and insensitive when patients or their relatives complain. "It is so often said it is endemic and one of those things that one gets in hospital. It is not what one wants to hear. A standard line is, 'We have no way of knowing whether you contracted it in our hospital or not'. Some people have been told, 'For all we know, you may have brought it into hospital.'"

But he has some sympathy for staff too. "It is difficult when you are in a hospital where the nearest washbasins are at the other end of the ward. It would save a lot of time if the consultant did not have to run up to the end of the ward every time they have touched a patient.

"When you have people working very hard round the clock to make people better and then have people coming into a hospital and the hospital making them worse, it can be very sapping for morale."

Meanwhile, the North Middlesex hospital, London, where the Sun yesterday claimed to have found high MRSA readings at 24 sites, said that its own recent tests for the bug had found just two positive samples out of 40 taken. "MRSA is a bug common in the wider community and can be difficult to prevent in public environments such as hopsitals. Many people, including those in the community we serve, carry MRSA without causing any harm," a hospital statement said. One in 10 tested on admission for surgery, or within 48 hours of it, tested positive for the bug itself, but of 45,000 inpatients treated last year, only 50 or 0.01% actually had evidence of bacteraemia in their blood, usually the most dangerous form of infection.

The hospital, like many others, is working to try to win the fight against MRSA - with hand-washing awareness training, and individual gel handwash for all clinical staff. It may be some time before we know whether the battle has been won, or lost.