Antimicrobial susceptibility: MIC versus disk diffusion, and what each tells you

One method gives you a number with a range attached, the other a category from a zone of inhibition. The choice affects what your antibiogram can do later, particularly if breakpoints change.

4 August 2026 9 min read

Antimicrobial susceptibility testing comes in two broad forms. Dilution methods produce a minimum inhibitory concentration, a numeric value. Disk diffusion produces a zone of inhibition measured in millimetres, which is converted to a category. Both end in S, I or R; only one leaves you a number to re-analyse later.

Disk diffusion, still the most transparent method

A paper disk containing a fixed amount of antibiotic is placed on an agar plate inoculated with the organism. The drug diffuses outward, and the diameter of the zone where growth is inhibited is measured after incubation. The diameter is compared against a table for that organism and drug, giving a category.

It is inexpensive, it needs no instrument, it is flexible about which agents you test, and it fails visibly: a technologist can see a contaminated plate, a mixed culture or a strange colony inside the zone. That last property is worth more than it sounds, because an automated result that is wrong usually looks exactly like a result that is right.

Dilution methods and the MIC

Broth microdilution tests the organism against doubling concentrations of the antibiotic and reports the lowest concentration that inhibited visible growth. Most laboratories run this on an automated system that reads growth optically and returns a MIC over a limited panel of concentrations.

The number is more informative than a category, but it is not a precise measurement. Testing in doubling dilutions means the true value lies somewhere between the last concentration that grew and the first that did not, and panels are truncated at both ends, so results are commonly reported as less than or equal to the lowest concentration tested or greater than or equal to the highest.

Gradient strips, the practical middle ground

A plastic strip carrying a predefined antibiotic gradient is placed on an agar plate, and the MIC is read where the ellipse of inhibition meets the strip. It gives a MIC with the flexibility and visual feedback of an agar method, which is why it remains the common approach for the awkward organisms and the agents an automated panel does not cover.

What this means for your antibiogram

For a cumulative antibiogram, both methods work, because the report counts categories rather than averaging numbers. Where the choice matters is what you can do afterwards.

With MIC values retained, you can reinterpret a whole year against a revised breakpoint table and see what actually changed.
With MIC values retained, you can watch a distribution drift upward inside the susceptible range, before any percentage moves.
With categories only, you have the report and nothing behind it. That is workable, and it is a one-way door.

Mixing methods in one report

Most laboratories do mix them: an automated panel for the routine agents, disk or gradient testing for the rest. That is normal and it does not invalidate the antibiogram, because the category is the unit of analysis. It does mean the method should be recorded, and it means a sudden change in one drug's percentage is worth checking against a change in how that drug was tested before it is treated as epidemiology.

Practical guidance for the export

Include the MIC column even if you never use it. It costs nothing in the export and it is the only thing you cannot reconstruct later.
Keep the censoring characters. A stored value of 32 that was really greater than or equal to 32 is a silently wrong number.
Record the method where your system carries it.
Keep the reported interpretation as your laboratory issued it. It is the authoritative result, and any reinterpretation should sit next to it rather than replace it.