The average adult​ carries about two kilograms of bacteria on and in their body. That’s more bacterial cells than human ones, trillions of them making a home on our skin and in our guts (the ‘microbiome’). We need them to help digest food, to fine-tune our immune systems, and to protect us against harmful micro-organisms. If you get ill with one of the nasty ones, the prescription you’re given won’t specifically target the pathogenic bacteria, but will wipe out plenty of the good ones too. Treatment with antibiotics for a typical course of five days or so is known to diminish the number and the diversity of gut bacteria for at least a few weeks, and in some people for as long as six months. We don’t yet fully understand the implications for the rest of the body’s health of making such profound alterations to the populations of bacteria that live alongside us.

Every day, as a GP, I scroll through lists of test results: haematology, biochemistry, radiology and microbiology. The microbiology results tell me whether any viruses, fungi or bacteria have been found in the samples I have sent to the lab, most commonly urine, wound swabs and vaginal swabs. When I started out in medicine twenty-five years ago the results came on sheafs of colour-coded paper: haematology was the pink of watery blood, biochemistry the green of chlorophyll, radiology the black and white of X-rays, microbiology the yellow of pus. Lab results now arrive on a screen, but the software has retained the old colour codes. The yellow screens for fungi and viruses are usually straightforward – a single list of identified strains or species. The bacterial results come with an extra column: next to each entry there will be a short inventory of five or six antibiotics and a letter, S or R. S means that the organism is sensitive to that antibiotic, i.e. the presence of the antibiotic kills or restrains the growth of that particular species. R means it’s resistant: the antibiotic has no effect.

Anyone who’s ever suffered a strep throat, a septic cut, pneumonia or vaginitis will be only too glad to take antibiotics in order to speed recovery. Countless millions of lives have been saved by these medicines. Lab reports can take three or four days to come back – too long to wait before prescribing – and so, with the help of local guidelines, GPs usually make an informed guess at which drug is most likely to help. They start the patient on a course, then wait for the report with fingers crossed. Back in the days of paper reports it was usual to see an uninterrupted column of ‘S’s, but in the last few years, as antibiotic resistance has increased, I’ve seen more and more ‘R’s. If the lab finds bacteria resistant to the drug I’ve prescribed I contact the patient to find out if they’re improving; if not, we start again with a different medication.

Bacteria, viruses and fungi have been finding ways to attack one another for aeons. Viruses known as bacteriophages (or simply ‘phages’) breed in bacteria and can kill them. They are the most abundant organisms on earth. In the 1910s and 1920s, before the advent of antibiotics, there was some success using phage therapy to combat cholera, plague, conjunctivitis and dysentery. Penicillin, the first commercially available effective antibiotic molecule, was refined from a fungus, and its success led to the near abandonment of phage therapy research. That story is now legend: in 1928, in his laboratory in London, Alexander Fleming noticed that bacteria grew poorly on petri dishes contaminated with Penicillium fungi. Many bacteria have rigid cell walls, and penicillin breaks them down, causing the bacteria to spill their guts and die.

There are two main classes of antibiotic: bacteriocidal, which kill the germs outright, and bacteriostatic, which prevent them from multiplying. The latter can be just as effective as the former: stalling the growth of bacterial colonies is often all that’s needed, while the immune system does the rest. I distrust military metaphors in medicine but when talking about infection they can be difficult to avoid: ‘battle’, ‘defence’, ‘kill’, ‘defeat’. A bacterium can be imagined, then, as a well-defended fort, with officers in the citadel (the DNA in the nucleus) and infantry workers in the courtyard (the proteins in the cytoplasm). Many antibiotics work by battering at the walls of the fortress: penicillin, but also vancomycin, meropenem and cefalexin. Some antibiotics (rifampicin, ciprofloxacin) take out the bacterium’s command-and-control centre, blocking, unravelling or otherwise interfering with its DNA. Others leave the citadel and walls alone, but incite mutiny among the infantry (tetracyclines, clindamycin and chloramphenicol all work by disrupting proteins).

Antibiotic resistance develops through evolution by natural selection. Bacteria can double their numbers in minutes or hours, and some of them, through random mutations, acquire tricks that circumvent the action of the drug. Bacteria have little molecular pumps in their walls to expel toxins; these can undergo modifications that enable them to pump out antibiotics. In the case of antibiotics that have to latch onto a protein to be effective, a small mutation in that protein can be enough to confer resistance. Some bacteria develop alterations in the structure of their walls so the antibiotic can’t get any purchase. The action of penicillin and its derivatives is dependent on a molecular constituent called a β-lactam ring; repeated exposure to β-lactams preferentially selects those bacteria that can create an enzyme to destroy the ring before it destroys the bacterial wall. Bacteria can swap useful mutations on slips of DNA (called plasmids), meaning that one bacterium is capable of ‘teaching’ others its tricks. If an antibiotic is used for too short a time, or in too small a dose, it gives bacteria a greater opportunity to evolve a way of evading it.

The ubiquity of antibiotics in modern healthcare and animal husbandry is driving a rise in antibiotic resistance. There are more and more ‘R’s in those columns. We’re beginning to realise just how close we might be to running out of antibiotic choices. In the absence of new approaches, the 21st century could see a return to the kind of infection management that was practised before 1928 – a reliance on strengthening and resting the body to fight disease, collapsing infected lungs to starve bacteria of oxygen, sanatoriums in the Alps. As a consequence, there will probably be a much higher death rate from infectious disease.

Some of my elderly patients suffer continually from urine infections. In the pre-antibiotic era many of them would have died long ago from urosepsis, an infection of the urinary tract that spreads to affect the whole body. Now, a daily dose of antibiotics can keep them well. I have to rotate the prescribed antibiotic every two or three months or their urine fills with resistant organisms. The easy availability of antibiotics creates a dilemma: on the one hand, there’s a treatment that can save my patient’s life; on the other, the chronic use of antibiotics is driving the increase in antibiotic resistance. I once had a patient who asked me to take over a private prescription of daily antibiotics, not because there was any proven infection requiring treatment, but because they made him feel a bit better. In a case like this there are clear NHS guidelines, and although I tend to think of my job as the art of making people feel better, I refused. He later found a private doctor who was more accommodating (in much of the world, he could have just bought them over the counter). Because we all influence one another’s microbiomes, swapping bacteria back and forth among ourselves, when antibiotic resistance increases in the microbiome of one individual, it is likely to spread to others. In 2019, 35,000 people in the US died of antibiotic-resistant infections – about the same number as died of prostate cancer, and almost as many as died of breast cancer. It has been estimated that by 2050, antimicrobial resistance will cause ten million deaths per year. The World Health Organisation calls the problem ‘one of the biggest threats to global health, food security and development today’.

We already use a lot of these drugs, and as patient populations become older and frailer, we need to use more all the time: one European study, published in 2015, gauged that 30 per cent of the Spanish population received at least one course of antibiotics a year; in the UK it was 29 per cent. Women were given more antibiotics than men, and children and the elderly far more than those aged between twenty and fifty. It’s difficult to find good data from primary care, where the majority of prescriptions are issued (far better data is available for the small number of people who end up in hospital), but worldwide the proportion of people who have never had an antibiotic is tiny.

Hospital labs are now turning up extraordinary levels of antibiotic resistance. A colleague of mine who works in a hospital recently took over the care of a patient who had been transferred from a long stay in intensive care in a country where antibiotics are for sale over the counter, and before reaching the UK had already received several weeks of indiscriminate antibiotic therapy. It’s common for people on ventilators to develop pneumonia, but the organism found to be responsible for this patient’s infection had no sensitivity to any of the drugs at my colleague’s disposal.

The Covid pandemic has shown that there are no borders for micro-organisms: when an infection affects one part of the world, it will eventually affect us all. It has also demonstrated the power of pharmaceutical companies to innovate quickly when they are backed by the state and obliged to adopt a model of research and development that isn’t geared only to the pursuit of profit. To stay ahead of antibiotic resistance, it is a matter of urgency that these vital drugs are no longer sold over the counter without prescription, that reliable guidelines are agreed as to when they should and shouldn’t be prescribed, and that we stop promoting bacterial resistance by adding antibiotics to feed for livestock. We also need new and better ways of defeating infection.

Pharmaceutical companies make most of their money from drugs patients take daily (inhalers, antidepressants, cardiac or diabetic medications), not drugs that are needed for just a five or seven-day course. They need greater incentives to turn to antibiotic research. Instead of relying on new antibiotics, an alternative might be to return to bacteriophages, tweaking their genetic code so they are able to single out pathogenic bacteria and leave the helpful ones alone. As organisms, phages aren’t covered by the same kinds of patent protection as drugs, and mainstream pharmaceutical companies have so far shown little appetite to get involved. But as antibiotic resistance rises, we are already starting to look back on the 20th century as a golden age of medicine, when infections could be treated effectively and there were plenty of antibiotics to choose from. As we learn more about the microbiome and its contribution to health, it may come to be seen as an age of recklessness, too.

Send Letters To:

The Editor
London Review of Books,
28 Little Russell Street
London, WC1A 2HN

Please include name, address, and a telephone number.


Vol. 44 No. 9 · 12 May 2022

Gavin Francis writes about the misuse of antibiotics, arguing that it may soon make everyday infections untreatable (LRB, 7 April). It’s important to recognise that everything he says about the mechanisms of antibiotic resistance was known fifty years ago. In the early 1970s, I was a bacteriologist in the pathology department at the Middlesex Hospital in London. The evolution by natural selection of resistance to penicillin and other antibiotics had been common knowledge for years. The horizontal transfer of resistance between bacteria, as they exchange scraps of DNA called plasmids, was a more recent discovery, but its alarming significance for infectious disease was already well understood.

Scientists at the Middlesex and all over the country were shouting about the importance of using antibiotics only when they were properly indicated. Organisations like the British Society for Social Responsibility in Science were urging governments to restrict the unnecessary use of antibiotics, especially in the agricultural industry. Successive governments took no effective action. To take just one illustration of the consequences: gentamicin, an antibiotic that has some very nasty side effects, was then a treatment of last resort for use only when other antibiotics had failed; today, resistance to other antibiotics is so common that gentamicin is a go-to treatment for infections that one wouldn’t have dreamed of treating with it before.

It seems worth spelling this out because the pattern, whereby well documented scientific knowledge is ignored for decades by policymakers, is now being duplicated with climate change.

Michael Parsons

send letters to

The Editor
London Review of Books
28 Little Russell Street
London, WC1A 2HN

Please include name, address and a telephone number

Read anywhere with the London Review of Books app, available now from the App Store for Apple devices, Google Play for Android devices and Amazon for your Kindle Fire.

Sign up to our newsletter

For highlights from the latest issue, our archive and the blog, as well as news, events and exclusive promotions.

Newsletter Preferences