Antibiotics changed the meaning of infection. Before penicillin was produced in quantity in the 1940s, bacterial infections that now seem minor - a cut that became infected, streptococcal throat illness, pneumonia after a bout of flu - were leading causes of death. Within two decades of their widespread introduction, a range of antibiotics had made most common bacterial infections readily treatable. The same two decades, however, also saw the emergence of resistance: bacteria whose descendants were no longer killed by the drugs that had previously destroyed them. Antibiotic resistance has grown since then into one of the most pressing medical challenges of our time.

Resistance arises because bacteria evolve quickly. A population of bacteria exposed to an antibiotic consists of perhaps a billion individual cells, with small genetic differences among them. Any cell whose genome happens to confer some ability to survive the antibiotic - a mutated enzyme that destroys the drug, a transport protein that pumps it out, a modified target molecule the drug cannot recognise - has an enormous advantage. The vast majority of cells die, the survivors reproduce, and within days the population is dominated by resistant variants. If the same antibiotic is then used again, it has much less effect.

Several further mechanisms speed the process. Bacteria do not inherit genetic material only from their own line; they can exchange short pieces of DNA, called plasmids, with cells of unrelated species. A resistance gene that arises in one species can therefore pass to others, sometimes in combinations that confer resistance to several antibiotics at once. Sub-therapeutic doses of antibiotic - too low to kill bacteria but high enough to create a selective pressure - are particularly conducive to resistance, because they allow partially resistant cells to survive and accumulate further resistance mutations.

The overuse of antibiotics in medicine and agriculture has provided large-scale selective pressure. In human medicine, antibiotics were for decades prescribed liberally, sometimes for viral infections that they cannot treat. In agriculture, antibiotics have been added to animal feed not only to treat disease but, at lower doses, to promote growth. Microbiologist Dr. Nikhil Bhattacharjee, working at the Indian Institute of Public Health, has documented the rise of resistance genes in soil bacteria from regions where intensive livestock farming is combined with weak regulation. His work shows that resistance detected today often has a history that extends back through agricultural use years or decades earlier.

The most concerning resistant bacteria are those carrying resistance to multiple drugs simultaneously. Methicillin-resistant Staphylococcus aureus (MRSA) became a major hospital problem in the 1990s and 2000s, though stricter hygiene protocols have brought infection rates down in many countries. More recent concerns include carbapenem-resistant Enterobacteriaceae, a family of gut bacteria that have acquired resistance to one of the last-line classes of antibiotics, and extensively drug-resistant tuberculosis, which is difficult and expensive to treat even in well-resourced health systems. Untreated, these infections carry high mortality rates.

Responses to the problem operate at several levels. Clinicians can prescribe antibiotics more carefully, using narrower-spectrum drugs where possible and stopping treatment when clear evidence of effectiveness is in hand. Hospitals can isolate resistant carriers and improve hand hygiene. Regulators can restrict the use of antibiotics in agriculture, an area where the European Union has moved considerably faster than many other regions. Dr. Bhattacharjee has argued that all of these measures help, but that none alone is sufficient. Resistance that has developed in one country can spread globally through travel and trade, making international coordination essential.

Research on new antibiotics has been slow. The pharmaceutical industry largely withdrew from antibiotic development in the 1990s and 2000s, because the economics - short courses of treatment, pressure to hold new drugs in reserve to delay resistance - were unattractive compared with those of drugs for chronic conditions. Several public initiatives have attempted to fill the gap, with new mechanisms such as advance market commitments that guarantee a minimum return for developers who bring useful antibiotics to market. A handful of new drugs have emerged from these programmes, but the pipeline remains thin relative to the scale of the problem.

Alternative approaches are under active study. Phage therapy uses viruses that naturally infect bacteria; it was developed in the early twentieth century, largely abandoned in the West in favour of antibiotics, and retained in the Soviet Union and some Eastern European countries. Recent clinical trials in Europe have shown promise for specific infections. Antibody treatments, vaccines that prevent bacterial infection before it starts, and drugs that disable bacterial virulence factors without killing the bacteria - and so apply less selective pressure for resistance - are all under investigation. Dr. Bhattacharjee argues that a mixed strategy will be required, because no single approach is likely to replace the extraordinary utility of antibiotics as they were during their mid-twentieth century peak.