Serial two-fold antibiotic dilutions pinpoint the lowest concentration that prevents visible bacterial growth
Broth microdilution begins by constructing a geometric (two-fold) dilution series of antibiotic across a row of microtiter wells. Each well halves the concentration of its neighbor, spanning a clinically relevant range so that the true MIC of the test organism falls somewhere inside the tested window. Standardization of broth composition, well volume, and dilution accuracy is what makes MIC values reproducible and comparable across laboratories worldwide.
Antibiotic susceptibility varies across several orders of magnitude between isolates and drug classes, so a linear dilution scheme (e.g., 1,2,3,4 µg/mL) would either miss high-level resistance or waste wells testing indistinguishable low concentrations. Two-fold serial dilution compresses a wide dynamic range into a small number of wells: doubling from 0.25 to 32 µg/mL requires only 8 wells to span a 128-fold concentration range.
Preparation technique: • A stock antibiotic solution is prepared at high concentration and serially halved by transferring a fixed volume into an equal volume of fresh broth in the next well. • Cation-adjusted Mueller-Hinton broth (CAMHB) is the CLSI-recommended medium because divalent cations (Ca²⁺, Mg²⁺) affect activity of aminoglycosides and polymyxins, and pH/thymidine content affect sulfonamides and trimethoprim. • Automated dispensing systems (e.g., Sensititre, MicroScan panels) pre-dose and lyophilize antibiotic into wells for reproducibility, reconstituted at the time of testing.
Range selection considerations: • The tested range must bracket the epidemiological cutoff value (ECOFF) and the clinical breakpoint for the drug-bug combination, or the result is reported as off-scale ("≤" the lowest or ">" the highest concentration tested). • Standard CLSI/EUCAST panels are validated against reference strains (e.g., E. coli ATCC 25922, S. aureus ATCC 29213) to confirm dilution accuracy before patient testing.
Because the entire MIC result depends on accurate concentrations, each new lot of antibiotic-containing broth or each week of testing is validated against reference ATCC quality-control strains with well-established expected MIC ranges. A result falling outside the published QC range signals a dilution, potency, or pipetting error that must be corrected before patient isolates are tested.
A single pipetting error early in the series propagates through every subsequent two-fold dilution, so CLSI mandates QC strain testing to catch systematic dilution drift before it reaches a patient report.
The dilution series is only meaningful if every well receives the same number of viable organisms. Broth microdilution standardizes the inoculum using the McFarland turbidity method, ensuring the final bacterial density in each well is close enough to a reference standard that observed differences in growth reflect antibiotic effect, not variation in starting inoculum size.
Isolated colonies from an overnight agar culture are suspended in sterile saline or broth and adjusted, by comparison against a 0.5 McFarland turbidity standard (a barium sulfate suspension or electronic nephelometer), to approximately 1–2×10⁸ CFU/mL. This suspension is then diluted further — typically 1:200 in broth microdilution — so that after inoculation of each well the final bacterial density is ~5×10⁵ CFU/mL, a concentration low enough to allow antibiotic effect to be distinguished but high enough to produce reliably visible turbidity if growth is not inhibited.
Inoculation of the plate: • A small, precise volume of standardized suspension is added to every well of the dilution series, to the growth-control well (broth + bacteria, no antibiotic), and is deliberately withheld from the sterility-control well (broth + antibiotic diluent only, no bacteria). • The growth-control well confirms the organism is viable and grows normally in the absence of drug — a negative growth-control invalidates the entire run. • The sterility-control well confirms the broth and antibiotic diluent were not contaminated — any turbidity here invalidates the run.
Some drug-organism combinations, most notably beta-lactams against beta-lactamase-producing organisms, show a pronounced "inoculum effect": as starting bacterial density rises, more enzyme is present to hydrolyze the drug, and the measured MIC increases sharply. This is why CLSI specifies the inoculum density so precisely — a ten-fold heavier inoculum can shift a susceptible result into the resistant range purely as a testing artifact, not a true change in organism biology.
A too-heavy inoculum is one of the most common sources of falsely elevated MIC results in the clinical microbiology laboratory — standardization against the 0.5 McFarland standard exists specifically to prevent this artifact.
Once inoculated, the plate is incubated under controlled temperature and atmosphere for a defined window of time. During this period, bacteria in wells with sub-inhibitory antibiotic concentrations divide logarithmically and accumulate into visibly turbid suspensions, while wells at or above the organism's MIC remain clear because growth is arrested at or near the inoculum density.
Temperature and duration directly affect both bacterial growth rate and antibiotic stability, so deviation from CLSI-specified conditions changes the measured MIC. Incubating too briefly may leave slow-growing organisms below the visible turbidity threshold even in growth-control wells (an invalid test); incubating too long can allow resistant subpopulations or beta-lactamase accumulation to produce artifactual growth in wells that would otherwise show inhibition (trailing endpoints).
Special cases requiring extended or modified incubation: • Oxacillin/cefoxitin testing against Staphylococcus aureus for mecA-mediated resistance requires a full 24-hour read at 33–35°C to detect heteroresistant subpopulations that grow slowly. • Vancomycin MIC testing against staphylococci similarly benefits from full 24-hour incubation to detect subtle elevated-MIC phenotypes (hVISA/VISA). • Fastidious organisms (Streptococcus pneumoniae, Haemophilus influenzae) require enriched media and 5% CO₂ atmosphere to support adequate growth in the control well.
Below the MIC, the antibiotic concentration is insufficient to overcome the bacterial population's replication rate, and cells proceed through normal binary fission, roughly doubling every 20–30 minutes for common Enterobacterales, producing a visibly turbid suspension (>10⁷–10⁸ CFU/mL) within the incubation window. At the MIC and above, the drug either kills bacteria (bactericidal mechanism) or halts replication (bacteriostatic mechanism) fast enough that the population never crosses the visible turbidity threshold, leaving the well grossly clear even though a small number of viable organisms may persist (this residual viability is what a subsequent minimum bactericidal concentration, MBC, test would detect).
A well that appears clear is not necessarily sterile — it only demonstrates that visible growth was inhibited. Confirming a true kill (MBC) requires subculturing clear wells onto antibiotic-free agar.
After incubation, every well is inspected for turbidity — either by unaided eye against a dark background/reading mirror, or by an automated instrument measuring optical density or light-scatter kinetics. The MIC is defined as the lowest antibiotic concentration in the dilution series at which no visible (or instrument-detected) growth occurs, always validated against a growth-positive control well and a sterility-negative control well.
Visual reading is performed by gently agitating or viewing the plate against a dark, non-reflective background, sometimes aided by a reading mirror or magnifying lens placed beneath the plate. The reader scans from the highest concentration well toward the lowest, noting the first well (highest to lowest) that shows any visible turbidity — the well immediately below that transition is the MIC. Because human perception of faint turbidity is subjective, some drug-organism pairs show a "trailing endpoint" (a light haze persisting across several wells above the true MIC), for which CLSI provides specific reading rules (e.g., ignore trailing haze for sulfonamides/trimethoprim and read the well showing ≥80% growth inhibition relative to the growth control).
Automated/instrument reading: • Semi-automated readers use a photometer to measure light transmission or scatter through each well and apply a validated algorithm to classify growth versus no-growth, removing subjective interpretation. • Continuous kinetic readers (e.g., some Sensititre/Vitek platforms) capture growth curves over the incubation period and can flag atypical kinetics (e.g., slow initial growth followed by late breakthrough) that a single endpoint read would miss.
No MIC result is reportable unless both control wells behave as expected: the growth-control well (antibiotic-free) must show clearly visible turbidity, confirming the inoculum was viable and adequate; the sterility-control well (bacteria-free) must remain clear, confirming no contamination of broth or antibiotic diluent. If either control fails, the entire panel is invalid and must be repeated — the "clean bracket" of a growing control and a sterile control is what gives meaning to every well read in between.
A microtiter row can look correct at a glance, but an MIC read without a valid growth-control and sterility-control bracket cannot be trusted for patient reporting.
A raw MIC value is clinically meaningless without context. Laboratories translate the MIC into a categorical interpretation — Susceptible, Intermediate, or Resistant — using breakpoints set by CLSI or EUCAST that integrate the distribution of MICs across wild-type populations, achievable drug exposure at approved doses, and clinical outcome data. Clinicians and pharmacists then use pharmacokinetic/pharmacodynamic (PK/PD) indices to select a dosing regimen that achieves the exposure needed to make that specific MIC a susceptible target.
Antibiotics kill or inhibit bacteria by different time courses of action, and the PK/PD index that best predicts clinical success reflects that mechanism:
• Time-dependent killing (%T>MIC): beta-lactams (penicillins, cephalosporins, carbapenems) kill most effectively when free drug concentration remains above the MIC for a sustained fraction of the dosing interval; increasing the peak concentration above the MIC adds little extra benefit. This is why beta-lactams are increasingly dosed by extended or continuous infusion rather than large intermittent boluses.
• Concentration-dependent killing (Cmax/MIC): aminoglycosides (gentamicin, tobramycin, amikacin) kill more bacteria as peak concentration rises well above the MIC, favoring once-daily high-dose regimens that maximize Cmax/MIC while allowing a trough-free interval that reduces nephrotoxicity/ototoxicity risk.
• Concentration-dependent killing with time dependence (AUC24/MIC): fluoroquinolones (ciprofloxacin, levofloxacin) and vancomycin correlate best with the total drug exposure over 24 hours relative to the MIC, integrating both peak and duration effects.
In this simulator, the "Organism MIC" slider represents the true susceptibility of the modeled organism, and the "Determined MIC" metric is what the broth dilution assay reads back from the well pattern — by design they match exactly, illustrating a perfectly performed, contamination-free assay. The "PK/PD Target Met" badge compares that determined MIC against an illustrative susceptible breakpoint of ≤4 µg/mL: at or below that value, standard dosing is expected to achieve the necessary %T>MIC, Cmax/MIC, or AUC24/MIC exposure; above it, the standard regimen is unlikely to reach target exposure and the isolate would typically be reported intermediate or resistant, prompting dose escalation, an alternative agent, or infectious-disease consultation.
A susceptible MIC on the report is not by itself proof of adequate dosing — the prescribed regimen must actually achieve the PK/PD index target for that specific drug class at the patient's renal/hepatic function and site of infection.