How Do You Test Whether an Antibiotic Works? Measure the Clear Circle

Bacteria taken from a patient are spread evenly across the surface of a nutrient jelly in a dish, thickly enough that they will grow into a continuous lawn. Small paper discs, each carrying a measured quantity of a particular drug, are placed on the surface and the dish is incubated overnight. The drug diffuses outward from each disc, forming a concentration that is high next to the disc and falls with distance. Where the concentration is high enough to prevent growth, the bacteria do not grow, producing a clear circle around the disc, and the diameter of that circle is measured in millimetres.

The number is interpreted against published tables rather than read directly:

• •A larger circle indicates the organism is more sensitive to that drug
• •Each drug and organism combination has its own threshold values
• •Results are reported as sensitive, intermediate or resistant
• •Those categories predict whether treatment is likely to work
• •Several drugs are tested on one plate simultaneously
• •Control organisms of known behaviour are run to check the method

The size of the circle depends on how fast the drug diffuses as well as on how effective it is, so two drugs producing the same circle are not equally potent and the raw measurement cannot be compared across drugs. Thresholds are therefore established by correlating circle sizes against a more direct laboratory measurement of the lowest concentration that stops growth, and against what actually happens to patients treated with achievable doses. That calibration is done by international committees, is revised as resistance patterns change and as dosing practice changes, and differs slightly between the main standard-setting bodies, which is why a laboratory must state which standard it is applying.

The result depends on conditions that must be controlled tightly, and the standard procedures specify them for good reason. A lawn spread too thickly produces smaller rings and can make a sensitive organism look resistant, while too thin a lawn does the reverse. The jelly must be an exact depth, because a thin layer lets the drug spread further sideways and exaggerates the ring. Incubating too long lets slow-growing resistant individuals fill in the edge. Discs must be stored cold and dry, since the drug on them degrades. And colonies growing inside an otherwise clear ring are not contamination but resistant mutants, which is a finding rather than an error.

The technique was standardised by Kirby and Bauer in the 1960s, building on earlier work, and its virtue is that it is cheap, requires no specialised equipment and can be performed anywhere, which is why it remains in worldwide use sixty years later. Its limitations are that it takes a day or more, works poorly for slow-growing or fastidious organisms, and gives a category rather than a number. Automated systems using broth dilution give a numerical result faster. Genetic methods detect known resistance genes within hours without culturing anything, which is far quicker and detects only the mechanisms it is designed to look for, so the older method retains a role as a general screen.

A drug diffusing outward from a paper disc on a lawn of bacteria leaves a clear ring wherever the concentration is high enough to stop growth, and the diameter is measured and compared against published thresholds. Those thresholds exist because diffusion rate affects the circle as much as potency does, so raw sizes cannot be compared between drugs. The method is slow, cheap and works anywhere.