Minimum Inhibitory Concentration (MIC): How Broth Microdilution Determines It

Minimum inhibitory concentration (MIC) is the lowest concentration of an antimicrobial agent that prevents visible growth of a microorganism after a fixed incubation period. It is a laboratory number, not a treatment recommendation — on its own, an MIC value says nothing about whether a patient’s infection will respond to that drug. That translation from number to clinical category happens only when the MIC is read against a published breakpoint, and broth microdilution is the reference method that produces the number in the first place. This guide covers how the assay is built, what actually counts as the endpoint, how MIC connects to susceptible/intermediate/resistant reporting, and where the method commonly goes wrong.

What Broth Microdilution Actually Measures

Broth microdilution exposes a standardized inoculum of a test organism to a range of antimicrobial concentrations in liquid growth medium, then reads which wells show growth after incubation. It is one of the dilution methods described in CLSI M07, Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria That Grow Aerobically (12th edition, published March 2024), alongside broth macrodilution and agar dilution. CLSI M07 notes that the microdilution method it describes is the same underlying methodology as ISO 20776-1, so a plate run to either standard is testing the same thing the same way.

The test organism is grown to a standardized density, most commonly matched to a 0.5 McFarland turbidity standard, then diluted into cation-adjusted Mueller-Hinton broth so that each well receives a final inoculum on the order of 5 × 10⁵ colony-forming units (CFU) per mL — light enough that a single surviving organism can still multiply into visible turbidity by the time the plate is read, but standardized enough that two labs running the same isolate get comparable results. Getting that inoculum density right is the single most consequential setup step in the whole assay; too heavy an inoculum inflates the apparent MIC, too light one deflates it.

Setting Up the Serial Two-Fold Dilution

The antimicrobial is prepared as a serial two-fold (doubling) dilution series across the wells of a 96-well microtiter plate — each well holds exactly half the concentration of the one before it (for example 0.25, 0.5, 1, 2, 4, 8, 16, 32 µg/mL). This is why MIC values only ever land on specific doubling-dilution rungs rather than an arbitrary number: an MIC is reported as one of the tested concentrations, not interpolated between them. The plate also carries a growth-control well (inoculated broth, no drug, confirming the organism grows normally) and a sterility-control well (uninoculated broth, confirming no contamination) — without both controls reading as expected, the MIC result for every other well on that plate is uninterpretable. After inoculation, plates are typically incubated at 35–37°C in ambient air for 16–20 hours for most rapidly growing aerobic bacteria; fastidious organisms and specific drug-organism combinations (CLSI M07 lists several) call for longer incubation, added CO₂, or supplemented media, and using the wrong condition for a given organism/drug pair is a real source of an inaccurate result rather than a cosmetic protocol deviation.

Related lab technique: if serial dilution mechanics themselves aren’t second nature, see the serial dilution technique guide for how the two-fold dilution factor is actually calculated and pipetted.

Reading the MIC Endpoint

The MIC endpoint is the lowest concentration in the series that shows no visible growth (no turbidity, no visible pellet/button) after the specified incubation time — not the lowest concentration with reduced growth, and not an extrapolated value between wells. Reading is done against the growth-control well as the visual reference for “what growth looks like” on that plate. Some drug/organism combinations are prone to a trailing endpoint — faint, persistent haziness across several wells above the true MIC rather than a clean growth/no-growth cutoff (classically seen with sulfonamides, trimethoprim-sulfamethoxazole, and some azole antifungals) — and CLSI M07 gives organism/drug-specific guidance on how to call the endpoint in those cases rather than defaulting to the first clear well. Automated broth microdilution readers (turbidimetric or photometric) are widely used in clinical labs precisely because visual trailing-endpoint calls are one of the more operator-dependent steps in the whole method; where an automated system disagrees with a technologist’s visual read, that discrepancy is itself a quality signal worth investigating, similar in spirit to the multirule logic covered in the Westgard rules guide for other quantitative lab assays.

MIC vs. Clinical Breakpoints: Why the Number Alone Isn’t an Answer

An MIC of, say, 4 µg/mL is meaningless on its own — it only becomes a susceptible, intermediate, or resistant call when compared against a published breakpoint for that specific organism-drug combination. Breakpoints are not derived from the biology of a single isolate; they are set by expert committees using population-wide MIC distributions, pharmacokinetic/pharmacodynamic modeling of achievable drug exposure, and clinical outcome data, then published and periodically revised as resistance patterns and dosing evidence change. The two breakpoint-setting bodies referenced throughout U.S. and European clinical microbiology are:

• CLSI (Clinical and Laboratory Standards Institute) — publishes breakpoints annually in CLSI M100, Performance Standards for Antimicrobial Susceptibility Testing, the companion interpretive document to the M07 methodology standard.
• EUCAST (European Committee on Antimicrobial Susceptibility Testing) — publishes its own breakpoint tables, calibrated to the same underlying reference broth microdilution method (ISO 20776-1/ISO 20776-2). Since its 2019 revision, EUCAST reports results in three categories: susceptible, standard dosing regimen (S); susceptible, increased exposure (I); and resistant (R) — a deliberate reframing of the older “intermediate” category to make explicit that the I category means the drug can still work, at a higher dose or with a longer infusion, not that it sits in an ambiguous middle ground.

CLSI and EUCAST breakpoints for the same organism-drug pair are not always identical, and a lab that reports susceptibility must state which set of breakpoints it applied. Because breakpoints move over time as resistance mechanisms spread and dosing evidence accumulates, the same raw MIC value tested on the same isolate can be reported as susceptible under one year’s breakpoint table and intermediate or resistant under a later revision — the isolate hasn’t changed, the interpretive criterion has. This is also why raw MIC data has value beyond same-day patient reporting: labs that retain and aggregate MIC results over time can re-interpret historical data against revised breakpoints and track real resistance trends, which is the underlying logic behind a cumulative antibiogram.

Common Sources of Variability and Error

Broth microdilution is a reference method, but it is not error-proof in routine use. The factors that most often shift a reported MIC away from the true value:

• Inoculum density outside the target range. The single most common source of MIC drift. An inoculum denser than the standardized ~5 × 10⁵ CFU/mL target artificially raises the apparent MIC (more organisms to inhibit); a lighter inoculum artificially lowers it.
• Medium composition. Cation-adjusted Mueller-Hinton broth is specified because divalent cation content (calcium and magnesium) materially affects the activity of specific drug classes — aminoglycosides against Pseudomonas and daptomycin susceptibility testing are both cation-sensitive; uncontrolled cation concentration is a documented, specific source of erroneous results for those combinations, not a generic quality nicety.
• pH drift. Broth pH outside the specified range shifts the ionization and activity of several drug classes, again skewing the apparent MIC without any change in the organism’s true susceptibility.
• Incubation time and temperature deviation. Reading too early can miss slow-growing resistant subpopulations; reading late can let a marginally susceptible well cross into visible growth.
• Trailing-endpoint misreads. Covered above — calling the first faintly hazy well instead of following the organism/drug-specific CLSI guidance for that combination.
• Quality-control organism drift or omission. Every run should include CLSI-designated QC strains with known, published MIC ranges (commonly E. coli ATCC 25922, S. aureus ATCC 29213, and P. aeruginosa ATCC 27853, among others depending on the drug panel) run in parallel with patient isolates. A QC result outside its published acceptable range invalidates that day’s patient results for the affected drug until the cause is identified — skipping or under-frequency QC is a compliance gap as well as an accuracy one; see the CLIA quality control requirements guide for how QC frequency obligations apply to microbiology benches specifically.

Broth Microdilution vs. Other Susceptibility Methods

Broth microdilution is the CLSI/ISO reference method, but it is not the only susceptibility technique in routine use. Disk diffusion (Kirby-Bauer) reads a zone of inhibition around an antibiotic-impregnated disk on agar and is cheaper and faster to set up but reports a categorical result off a correlated breakpoint table rather than a true MIC. Gradient diffusion (commercially, Etest) uses a strip with a continuous antimicrobial gradient to read an actual MIC value off an agar plate, trading some of broth microdilution’s throughput for a more familiar disk-diffusion-style workflow. Agar dilution, the oldest of the three dilution methods in CLSI M07, incorporates the drug directly into agar plates rather than broth wells and remains a reference method for a small number of organism-drug combinations where broth methods perform poorly. All of these ultimately answer the same clinical question broth microdilution answers — how much drug does it take to stop this organism — but only a true dilution method produces the quantitative MIC value that trending, breakpoint-revision analysis, and pharmacokinetic dosing calculations actually require.

Frequently Asked Questions

Is a lower MIC always better?

Within the same organism-drug combination, yes — a lower MIC means less drug was needed to stop visible growth, indicating greater in vitro susceptibility. MIC values are not comparable across different drugs against the same organism, though, because different drugs achieve different concentrations at the site of infection; a lower MIC for drug A versus drug B does not by itself mean drug A is the better clinical choice.

Can an MIC value be reported without a susceptibility category?

Yes, and it sometimes should be — when no validated breakpoint exists for that organism-drug combination, or for organisms tested primarily for surveillance or research purposes, labs report the raw MIC without an S/I/R interpretation rather than force a categorical call the breakpoint literature doesn’t support.

Why do CLSI and EUCAST breakpoints sometimes differ for the same drug and organism?

The two committees can weigh population MIC distribution data, dosing regimens common in their respective regions, and pharmacokinetic/pharmacodynamic targets somewhat differently, and they don’t always revise on the same schedule. A lab’s susceptibility report should state which breakpoint set was applied so the result is interpretable downstream.

What does it mean when EUCAST reports “I” instead of “S” or “R”?

Since EUCAST’s 2019 revision, “I” means susceptible, increased exposure — the organism is expected to respond if the drug is dosed higher, given more frequently, or infused longer than the standard regimen to raise exposure at the infection site. It is not a statement that the result is ambiguous.

How is broth microdilution different from a minimum bactericidal concentration (MBC) test?

MIC measures the lowest concentration that inhibits visible growth; MBC measures the lowest concentration that kills a defined proportion of the original inoculum, determined by subculturing wells from the MIC test onto drug-free agar. MBC testing is far less common in routine clinical practice and is reserved for specific clinical scenarios, such as endocarditis or other infections where bactericidal rather than bacteriostatic activity is considered clinically important.

Broth microdilution’s output is only as useful as the breakpoint it’s read against and the QC discipline behind the run that produced it. Labs building or auditing their antimicrobial susceptibility testing program should pair this methodology with the interpretive and quality-management layers around it — from breakpoint-driven reporting and antimicrobial stewardship program integration, through to the day-to-day individualized quality control plan requirements CLIA imposes on the bench running the assay.