M stands for methicillin, a chemical derivative of penicillin, first called BRL 1241 because it was developed during the 1950s in the Beecham Research Laboratories at Betchworth in Surrey. R stands for resistant; the development of methicillin resistance in a hospital was first detected in October 1960 in Guildford, also in Surrey. And SA stands for Staphylococcus aureus, the bacterium that causes boils, carbuncles, abscesses, osteomyelitis and most wound infections after surgery. It was discovered in the late 1870s by Alexander Ogston, a surgeon at the Aberdeen Royal Infirmary.
Alexander Fleming was studying Staphylococcus aureus when he discovered penicillin in 1928, and the first patient to be treated in the first clinical trial of the new antibiotic at Oxford was infected with it. Albert Alexander, a 43-year-old policeman, was suffering from a spreading infection of his face that had started with a rose thorn scratch. He had lost an eye and the infection had spread to his lungs and his shoulder. On 12 February 1941 he was injected with penicillin made by Howard Florey and his team. Alexander’s condition improved dramatically. Treatment continued for five days. Penicillin was extracted from his urine and used again. But ten days later he relapsed, dying of staphylococcal septicaemia on 15 March: the supplies of antibiotic had run out.
Penicillin revolutionised the treatment of staphylococcal infections. But its power over them began to wane soon after its general introduction. The first naturally occurring penicillin-resistant staphylococci were noted by Fleming in 1942. Between April and November 1946, 12.5 per cent of Staphylococcus aureus strains isolated at the Hammersmith Hospital in London were penicillin-resistant. By early 1947 the percentage had tripled. The bacteriologist Mary Barber showed that this rise was not due to the development of resistance while patients were being treated, but to the spread of a penicillin-resistant strain in the hospital. Some staphylococci had the ability to make penicillinase, a penicillin-destroying enzyme. The introduction of penicillin gave them an evolutionary advantage over strains killed by the antibiotic.
Methicillin was developed in response. It was resistant to penicillinase. ‘Resistance to BRL 1241 comparable to that which exists to penicillin G would require the ability to inactivate BRL 1241 by a new penicillinase,’ methicillin’s discoverers wrote in the Lancet in 1960. ‘Since cultures have not been encountered showing this property, it seems unlikely that the selection and proliferation of resistance strains will take place rapidly, if at all.’ The logic was impeccable – but only if the destruction of penicillin by penicillinase was the sole way a staphylococcus could become penicillin-resistant. It was not. Penicillin works by binding to and destroying the function of proteins which are essential components of the machinery that bacteria use to build their cell walls. MRSA strains acquired a gene (mecA) that codes for a wall-building protein resistant to methicillin. Less than a year after the publication of the optimistic prediction in the Lancet, the first MRSA were described. Two years later, in 1963, there was the first outbreak in a British hospital. An MRSA strain at Queen Mary’s Hospital for Children at Carshalton in Surrey spread to eight of the 48 wards, infecting 37 patients and killing one.
The evolution of MRSA continued and it spread internationally. New strains became resistant to more and more antibiotics. There were hospital outbreaks in the 1970s; by the early 1980s MRSA epidemics were being described in Australia, Ireland and London. Epidemic strains have gone from strength to strength in the UK ever since.
Surrey was often enough the site of developments in the early history of MRSA for common sense to identify it as the source of all our current crises. But it was not. The MRSA of today have no recent common ancestor. Worldwide there are at least 11 different kinds, which have evolved independently in different places at different times, all descendants of methicillin-sensitive Staphylococcus aureus strains good at spreading in hospitals, and which became MRSA when they acquired a chunk of DNA containing the mecA gene, the staphylococcal cassette chromosome mec (SCCmec). There are four different SCCmec. Geneticists describe these as mobile elements because their DNA is configured to stitch them into staphylococcal DNA.
British bacteriologists know the epidemic MRSA present in the UK by number; EMRSA 15 and EMRSA 16 currently dominate. EMRSA 16 is different from most other EMRSA types in having SCCmec II. It appeared in April 1992, when it infected 400 patients and 27 staff in three hospitals in Kettering in Northamptonshire – one acute, one elderly care, one psychogeriatric. It spread quickly, first to 15 hospitals and 845 patients in neighbouring counties, and then to London and more distant parts of the UK. By September 1994 it had caused problems in 21 hospitals in the capital and four elsewhere. In 14 instances, hospital to hospital transfer of patients caused its spread (in one case, to the Channel Islands); staff transfer was responsible for another. By 2000 it was common everywhere in Britain. It now occurs in Ireland, Denmark, Sweden, Finland, Norway, Switzerland, Spain, Greece, Cyprus, Turkey and the US.
So Staphylococcus aureus is good at getting about. But as soon as it touches an inanimate surface, such as the floor, wall or ceiling of a hospital ward, it stops growing and starts to die. Its natural home is the front of the human nose, the part that is picked. Most of us have it there at some time in our lives. About 20 per cent of the population carry it throughout their lives, 60 per cent intermittently, and 20 per cent never. In some people it inhabits other places, such as the armpit or the perineum (the region between the anus and the genitals). Although the staphylococcus can break out to cause boils, it does it so infrequently that carriers are unaware of its presence. This long-term cryptic relationship reflects the immune response, which is vigorous but does not eradicate the bacteria, instead keeping it at bay. Infections of surgical wounds are often the same. The nastiness gets walled off, the patient does not necessarily die, but takes a very long time to get better. In particular, if the staphylococcus gets into a bone it can be out of antibiotic reach and amputation may be the only remedy.
The nature of the anti-staphylococcal immune response explains why there are no successful vaccines. Enthusiasm for them in the 1920s and 1930s was a result of the rhetorical skills of Alexander Fleming’s boss at St Mary’s Hospital, Sir Almroth Wright, rather than any protective power. It is said that one of his products, the Anti-Catarrh (Public Schools) Vaccine was particularly popular, not because parents were impressed by the claim that its strength came from the 100 million dead staphylococci it contained per millilitre, but because they had heard a rumour that one of the components was Streptococcus regius, so called because it had been grown from pus taken from George V’s chest. All the staphylococcal vaccines from St Mary’s were useless. But they brought an antimicrobial benefit: the profits from their sale funded the department where Fleming worked and where he discovered penicillin.
In evolutionary time MRSA are brand new. It is less than two human generations since they became a cause of hospital-acquired infections. But from the point of view of the patient with a septic wound, the fundamentals haven’t changed much in hundreds of years. The kinds of disease caused by 19th-century Staphylococcus aureus strains are not very different from those caused by MRSA. And staphylococci infested hospitals then, just as they do today. As causes of pyaemia at the Munich General hospital in the 1870s they were described as having a ‘terrible dominance’ by Geheimrat von Nussbaum in his 1879 Guide to Antiseptic Wound Treatment. His introduction of antisepsis in 1875 was a big setback for Bavarian staphylococci. Swamping wounds and dressings and steeping surgeons’ hands and instruments in carbolic, with strict attention to detail, had an almost miraculous effect: ‘During the 16 years in which I have had charge of this hospital pyaemia has not been absent a single month, and yet it suddenly disappeared on the introduction of the Listerian method.’ It was the same across Europe: Scottish, Danish and French as well as German hospitals became less dangerous. But carbolic was not safe. It was absorbed through the skin and poisoned patients and surgeons, causing kidney damage (black urine was a tell-tale symptom). And laboratory tests showed that it was much less lethal for bacteria than expected. The notion that it was no good in theory, even if it worked in practice, caused serious damage to its reputation.
Sterilising dressings and instruments with superheated steam and the aseptic rituals of hand-washing, gowning and masking replaced antiseptics in the 1890s. They are still key methods more than a century later. They work. But Staphylococcus aureus still infests our hospitals. What has gone wrong?
Mathematical modellers of the spread of infection can get carried away by the elegance of their formulae, and politicians are easily impressed by their double differential equations, which caused millions of animals to be needlessly killed during the 2001 foot and mouth disease outbreak. But the work of B.S. Cooper and his colleagues is simple, biologically plausible, and persuasive. In an article in the Proceedings of the National Academy of Sciences of the USA, ‘Methicillin-Resistant Staphyloccocus aureus in Hospitals and the Community: Stealth Dynamics and Control Catastrophes’, they show that carriers are the problem. Many infected patients are repeatedly admitted to hospital. So the control of single hospital outbreaks is an incomplete victory: ‘When carriage can persist for a long time, the secondary cases caused by each case may be distributed over several hospital admissions … An endemic state can occur by stealth: all battles against MRSA may initially be won but there will be some probability of losing the war.’
There is a striking historical hospital infection parallel:
It is no question of unsanitariness or of overcrowding … or of the other factors that have been proposed. They are only side issues, important in their way, but side issues all the same. It is the actual infection that matters; it is the chronic cases, the ‘carriers’, who keep the … infection going … They form the keystone of the problem, and must be detected and isolated before any permanent good can be done.
Harold Gettings wrote this in 1913 about dysentery in mental hospitals. At his institution, the West Riding Asylum in Wakefield, the first case was diagnosed in 1818, a week after it opened. They were still occurring in 1929, the last time attempts to eradicate dysentery were discussed in the scientific literature. ‘Asylum dysentery’ was caused by a close relative of E.coli, shigella. It occurred in every mental hospital. In a bad year it ranked as the third most common cause of death in them, after syphilis of the brain and tuberculosis. Patrick Manson, a classmate of Ogston’s at Aberdeen University and the first to prove that mosquito bites could spread disease, called it ‘the very fatal type of dysentery, euphemistically called “colitis”, which is the scourge and disgrace of lunatic asylums’.
In his account of dysentery at Wakefield, Gettings said that ‘this article is not brought forward merely as a historical recital but because I believe the lessons it teaches are of value – as the lessons of history always are.’ He was right about the value, but wrong to think that his lessons would be learned. In 1984, his hospital, by then euphemistically called ‘Stanley Royd’, was the locus of a particularly vicious hospital infection outbreak when Salmonella-contaminated beef infected 436 patients and 106 staff, killing 19. Little attention has ever been paid to his words. He was an early member of a distinguished group of 20th-century British microbiologists whose opinions have been ignored. Mary Barber wrote in 1959 that
the control of antibiotic-resistant staphylococcal infection is an urgent problem and is not simple. It requires the active co-operation of doctors, nurses, physiotherapists, porters and domestics. Scrupulous asepsis and barrier-nursing mean extra work for all concerned. It is readily obtained in a ward where a severe outbreak of infection has recently occurred, but at other times, in the words of Sir Alexander Ogston, ‘human nature forgets unseen foes.’
And Gordon Stewart’s 1963 account of the first hospital outbreak of MRSA concluded: ‘Lastly, and most important, patients harbouring these rare strains must be isolated, vigorously treated, and preferably should be sent out of hospital as soon as possible.’ The words of Jean Bradley and her colleagues, describing in 1985 a two-year-long MRSA outbreak at the Royal Free Hospital in London, were prophetic: ‘Several authors have reported failure to contain MRSA infection without an isolation unit: hospitals without such facilities or, as at this hospital, unable to finance the staffing of a unit, may find that this epidemic MRSA will pose a considerable threat to their clinical practice.’
When EMRSA appeared, the Dutch decided to keep them out by a policy of ‘search and destroy’. Infected and colonised patients are segregated in strict isolation, and colonised staff treated and kept away from patients. Vigorous screening and properly staffed isolation facilities have been key features of this policy. Its success has shown that Barber and Stewart and Bradley were right. But search and destroy has never been adopted countrywide in the UK, and where MRSA is concerned, Britain has become the sick man of Europe. The National Audit Office report on Improving Patient Care by Reducing the Risk of Hospital-Acquired Infection opens with a map showing the proportion of Staphylococcus aureus bacteraemia (bacteria in the blood) isolates resistant to methicillin in various European countries. At 43.9 per cent the UK has the highest: in France it was 32.8 per cent, in Germany 18.7 per cent and in the Netherlands 1 per cent.
Only in the 21st century has Staphylococcus aureus, as MRSA, attracted much attention from British politicians. Michael Howard’s mother-in-law died from a hospital infection and he made its control an issue in the 2005 general election campaign. Risk theorists say that top-level support for safety spending is much greater for air travel than for oil refineries because chief executives use the former a lot but never go near the latter. Personal involvement counts. Staphylococcus aureus has not had the attention it deserved because its list of famous victims is so short. Even famous illnesses caused by it have escaped bacteriological attribution. Queen Victoria’s armpit abscess in 1871 poured pus for days after it was cut open by Joseph Lister. All that is remembered is the carbolic spray getting into her eyes. Rommel’s swollen nose caused him to leave North Africa just before the battle of el Alamein to recuperate. Monty gets all the credit still. Failure to recognise the importance of Staphylococcus aureus has not been confined to hospital chief executives, journalists and members of the public. Their hand hygiene habits show that many hospital staff have been unaware of it too.
Controlling an organism that evolves in real time obliges an organisation to do the same – to learn from experience. Important lessons such as the spread of antibiotic-resistant Staphylococcus aureus in the Hammersmith Hospital in 1946 and 1947 antedate the establishment of the NHS. It never caught up. And the bacterium continues to evolve: community-acquired MRSA have appeared. They are good at causing skin infections and, occasionally, pneumonia, which quickly kills healthy young adults. Like EMRSA there are different kinds that have evolved from different progenitors, but all have SCCmec IV and a gene coding for the Panton-Valentine leucocidin toxin (PVL). PVL was discovered by two bacteriologists (Philip Panton and Francis Valentine) at the London Hospital in 1932. ‘Leucocidin’ denotes that the toxin kills white blood cells. Bacteria that make it are more aggressive. One has spread to the UK. Molecular genealogy shows that it has evolved from a notorious penicillin-resistant Staphylococcus aureus strain, phage type 80/81. First seen in Australia in 1953, type 80/81 spread worldwide and was particularly virulent for babies, children and young adults. Methicillin and related antibiotics may have been responsible for its apparent disappearance in the 1960s. But it was exiled, not eradicated.
Vancomycin is one of very few antibiotics that will cure an MRSA infection. It has never been used on a grand scale. But MRSA resistant to it have appeared. And the golden age of antibiotic development is long past. Cynics say that Big Pharma is reluctant to fund the discovery of new antibiotics because drugs that cure in a few days are poor for profits. Realists say that scientists have run out of ways of targeting bacteria. Both are probably right.
Since 2000, many initiatives to counter MRSA have been announced by the government and the NHS. ‘Learning’ features a lot, so they are going in the right direction, but their titles are often too optimistic: ‘Saving Lives’, ‘Learning from the Best’, ‘Winning Ways’. In contrast, the current target of reducing MRSA bacteraemias by half by 2008 is very modest. It smacks of pessimism. This is right. If the mathematical modellers are correct, winning the war against MRSA requires the effective detection and management of carriers and the microbiological isolation of the infection. ‘Winning Ways’ lists 86 things to do. Carriage control is not one of them, and to say that ‘NHS Trust executives will ensure that, over time, there is an appropriate provision of isolation facilities within their healthcare facilities’ is neither a ringing endorsement nor a call for action.
The number of MRSA bacteraemias is a surrogate measure for non-trivial infections. Reducing them by half means the same for lethality. Even if this target is achieved MRSA will still be killing more people than all the microbial causes of food poisoning put together. A slaughterman killing a cow and preparing its carcass for the butcher is far more regulated by the Meat Hygiene Service, by the official veterinary surgeon, and by lots of rules with penalties for infringement than anyone handling the living in a hospital. European Commission inspectors can call at an abattoir at any time. Hospital infection control has no equivalents. Abattoirs are probably over-regulated. The same cannot be said for hospitals.
There is one very important historical exception to this generalisation: Victorian lunatic asylums. From 1845 they were inspected by the Lunacy Commissioners. This was a powerful incentive for cleanliness. Cleaning was done by inmates with military thoroughness and routine regularity because it was superintended by matrons ruling with rods of iron. My grandfather was chief attendant in one, my grandmother worked for years in an asylum’s dysentery wards, and my own employment in two left an abiding memory of their smell: strong soap and floor polish, with only a fleeting hint of faeces. For all this they were the seat of the longest ever recorded epidemic of hospital-acquired infection.
The end of asylum dysentery came when the hospitals closed. But diarrhoea in hospitals hasn’t gone away. In August this year, the Office for National Statistics announced that 44,488 cases of Clostridium difficile colitis were recorded in people older than 65 in England in 2004. The microbe is different: Clostridium difficile usually causes illness after antibiotic treatment which has killed the normal bowel bacteria that otherwise would outcompete it. But in many cases it now spreads from patient to patient in hospitals in exactly the same way as asylum dysentery. So history repeats itself.
Send Letters To:
London Review of Books,
28 Little Russell Street
London, WC1A 2HN
Please include name, address, and a telephone number.