Enter the dose, dosing interval, half-life, volume of distribution and MIC to see steady-state Cmax/Cmin, 24h AUC and the PK/PD index (%T>MIC, Cmax/MIC or AUC/MIC) update in real time, with a verdict against the class target for β-lactam, AG, FQ, Vancomycin and Linezolid.
Parameters
Antibiotic class
Sets the PK/PD index and target
Dose D
mg
Dosing interval τ
hr
Half-life T1/2
hr
Volume of distribution V_d
L/kg
Patient weight
kg
Target MIC
mg/L
CLSI/EUCAST breakpoint or clinical isolate MIC
Reference AUC/MIC target
Display only; class target is automatic
Results
—
Volume V_d (L)
—
Steady Cmax (mg/L)
—
Steady Cmin (mg/L)
—
AUC 24h (mg·h/L)
—
PK/PD index value
—
Target met
—
Concentration-time curve with MIC and T>MIC band
Blue line: plasma C(t). Red dashed: MIC. Shaded: AUC over one interval. Green band: T>MIC (time concentration is above MIC).
PK/PD indices. β-lactams and Linezolid optimize %T>MIC, AG optimizes Cmax/MIC, FQ and Vancomycin optimize AUC/MIC. When Cmin stays above MIC, %T>MIC reaches 100%.
Antibiotic PK/PD — MIC, AUC/MIC and %T>MIC dosing design
🙋
I was taught "if the level is above MIC the antibiotic works", but why do different drugs use different indices — %T>MIC, Cmax/MIC, AUC/MIC? Isn't "above MIC" enough?
🎓
Great question. Antibiotic killing actually splits into three patterns, and each pattern correlates best with a different metric. β-lactams (penicillins, cephs, carbapenems) are time-dependent: the longer the plasma level stays above MIC (%T>MIC), the more kill. Aminoglycosides (GM, TOB) are concentration-dependent: a single high Cmax kills more. Fluoroquinolones and Vancomycin sit in between and respond to the total 24-hour exposure (AUC). "Above MIC" alone doesn't capture that distinction.
🙋
OK. With the default β-lactam, 1000 mg q8h, MIC=2, the tool shows %T>MIC = 100%. Is that good?
🎓
Yes — β-lactams aim for ≥ 50%, so 100% is well above target. Cmin is 3.2 mg/L, above MIC=2, so the concentration stays above MIC the whole interval. Now slide MIC up to 8 (typical Pseudomonas). %T>MIC drops and the verdict turns red. That's the moment to either shorten the interval to q6h or switch to a 3-4 hour prolonged infusion.
🙋
Selecting Vancomycin sets the target to AUC/MIC ≥ 400. I thought the old standard was trough 15-20 mg/L — did that change?
🎓
Spot on. The 2020 IDSA/ASHP MRSA guideline replaced trough-only monitoring with direct AUC/MIC ≥ 400 mg·h/L for severe infections. Two reasons: (1) trough alone misses the AUC variability between patients, and (2) trough >20 mg/L tracks AKI, while keeping AUC inside 400-600 limits the AKI risk while maintaining efficacy. Clinically, Bayesian software (PrecisePK and similar) estimates individual AUC. In this tool, picking Vancomycin and letting AUC climb past 600 trips the toxic-range flag.
🙋
Switching to aminoglycoside sets the target to Cmax/MIC ≥ 10. At 1000 mg q8h, Cmax is 50 — isn't that toxic?
🎓
Sharp catch. AG can damage hearing and kidneys when Cmax stays too high. The clinical standard is 5-7 mg/kg every 24 hours (ODD, once-daily dosing); the default 1000 mg q8h (which is a β-lactam regimen) is clearly excessive for AG. Try dose 350 mg, interval q24h. Cmax falls, and the long interval gives a deep trough — and AG toxicity comes from sustained low-level exposure, so the deep trough during ODD is actually protective.
🙋
One more. During empiric therapy you don't know the MIC yet — how do I use this then?
🎓
Good practical point. In empiric therapy, plug in the CLSI/EUCAST breakpoint of the most-likely organism. For pyelonephritis use the E. coli Cefepime breakpoint of MIC=2; for HAP use the Pseudomonas Pip/Tazo breakpoint of MIC=16, and so on. If "worst-case MIC still gives %T>MIC ≥ 50%", you're less likely to miss resistant strains during the 48-72 hours before culture returns. Once susceptibility comes back, re-enter the actual MIC here and move from empiric to targeted therapy.
Frequently Asked Questions
The PK/PD index depends on the killing pattern. β-lactams (penicillins, cephalosporins, carbapenems) and Linezolid are time-dependent: the fraction of the dosing interval with plasma concentration above MIC (%T>MIC) drives efficacy; targets are 50-70% for β-lactams and around 85% for Linezolid. Aminoglycosides (Gentamicin, Tobramycin) are concentration-dependent: Cmax/MIC ≥ 10 with once-daily large doses. Fluoroquinolones and Vancomycin are AUC-dependent: AUC/MIC ≥ 125 for FQ, AUC/MIC ≥ 400 (AUC 400-600 mg·h/L) for Vancomycin against MRSA.
The 2020 IDSA/ASHP guideline recommends AUC/MIC ≥ 400 mg·h/L for severe MRSA infections. Two reasons: (1) trough alone does not capture AUC variability between patients well, and (2) trough >20 mg/L correlates with acute kidney injury while AUC kept under 600 limits the AKI risk. In practice, Bayesian estimation (PrecisePK, RxKinetics, DoseMeRx) or two-point sampling estimates AUC and the regimen is adjusted to keep AUC in 400-600. With MIC=1 mg/L a trough of 15-20 roughly maps to AUC 400-600, but higher MICs make trough alone unreliable.
Because β-lactams are %T>MIC-driven, holding concentration above MIC longer improves kill. When 30-minute intermittent dosing leaves %T>MIC at only 40-50% — typical against Pseudomonas with MIC 8-16 — switching to a 3-4 hour prolonged infusion or continuous infusion lifts %T>MIC to 70-100% and improves clinical outcomes. Piperacillin/Tazobactam, Meropenem and Cefepime are the usual candidates. The shorter the half-life, the larger the benefit, so prolonged infusion is first-line in septic shock or against MIC>4 organisms. Shortening the interval in this tool shows how %T>MIC rises.
Aminoglycosides kill in a concentration-dependent way (higher Cmax = faster kill) and have a post-antibiotic effect of 1-3 hours after the drug falls below MIC. Bundling the daily dose into one large infusion (once-daily dosing, ODD) maximizes Cmax. Nephro- and oto-toxicity track sustained low-level exposure, so the deep trough during ODD actually reduces toxicity. Gentamicin 5-7 mg/kg every 24 h is standard. Conventional 2-3 doses/day are reserved for renal failure or endocarditis where synergistic killing is wanted. A sustained Cmax above 30 mg/L is the toxic range.
Real-World Applications
Initial dosing in ICU sepsis: Mortality in septic shock rises roughly 7% for every hour appropriate antibiotics are delayed. When starting empiric broad-spectrum therapy (Meropenem, Pip/Tazo, Vancomycin), enter the suspected organism's breakpoint MIC into this tool and confirm that the very first dose meets the %T>MIC or AUC/MIC target. If it doesn't, add a loading dose — for Vancomycin, 25-30 mg/kg loading is standard practice.
Therapeutic Drug Monitoring (TDM): Vancomycin, aminoglycosides and voriconazole call for blood-level monitoring. Measured trough and peak are used to back-fit V_d and k_e (Bayesian estimation) in the same 1-compartment model as this tool, and the next dose/interval is set accordingly. Hospitals deploy software such as PrecisePK, RxKinetics or DoseMeRx, and pharmacist-led AUC-guided monitoring is increasingly the norm.
Antimicrobial Resistance (AMR) and Stewardship: Under-dosing that fails the PK/PD target applies selection pressure and breeds resistant organisms (ESBL, CRE, VRE, MRSA). Stewardship programmes pair "de-escalation of broad-spectrum agents" with "smallest effective dose that meets the PK/PD target". EUCAST and CLSI breakpoints are set with PK/PD achievability in mind, so when this tool cannot meet the class target the safer move is to switch agents rather than blindly push the dose.
Working with the clinical microbiology lab: Automated susceptibility platforms (BD Phoenix, bioMérieux Vitek 2, MicroScan) report MICs in 6-24 hours. Combined with MALDI-TOF species identification (minutes), the time from sampling to targeted therapy can drop below 24 hours. When the MIC exceeds the breakpoint, switch class — or use this tool to see whether high-dose, shorter-interval or prolonged-infusion regimens can still reach the target.
Common Misconceptions and Pitfalls
The first pitfall is applying the 1-compartment model regardless of renal function and body composition. The formulas here assume a uniform V_d, but the real V_d swings with dehydration, oedema, obesity and sepsis. Acute sepsis raises capillary permeability and inflates V_d by 1.5-2× for Vancomycin and aminoglycosides — the first-dose Cmax is then lower than predicted. Renal function must also be folded into k_e via CrCl from Cockcroft-Gault or eGFR from CKD-EPI before adjusting the half-life slider. Dialysis patients need an entirely separate two-phase (on-dialysis / off-dialysis) model.
Second, treating MIC as a single fixed number. Even within one species, clinical isolates spread their MIC by 4-16×. When the reported MIC sits near the breakpoint ("intermediate"), routine doses often fall short, and this tool is exactly where you back-calculate the dose needed to recover the target. MIC creep — slow upward drift of MIC over years, as seen for MRSA against Vancomycin (1→2 mg/L through the 2000s) — also matters. Refresh the local antibiogram annually and revise the assumed MIC for empiric therapy.
Finally, "PK/PD target met = clinical cure" is wrong. This tool computes plasma kinetics, but the penetration into lung, CSF, bone or biofilm differs drug by drug. Daptomycin, for instance, is deactivated by lung surfactant and is useless in pneumonia; Vancomycin's CSF penetration is only about 20%, so meningitis demands a higher AUC. Highly protein-bound drugs (Ceftriaxone, 95%) act only in their free form, so hypoalbuminaemia can paradoxically raise free-drug exposure. Treat the numbers here as a starting point, then layer on site penetration, protein binding and host immunity before making the final call.
How to Use
Enter antibiotic dose in mg (e.g., 500 mg for amoxicillin, 2000 mg for piperacillin)
Set dosing interval in hours (standard: 6h, 8h, 12h, or 24h depending on agent)
Renal clearance affects t½ significantly; adjust half-life estimates in renal impairment (e.g., vancomycin t½ extends from 6 h to 30+ h in ESRD)
MIC interpretation varies by institution and EUCAST/CLSI breakpoints; always cross-reference local antibiogram data for your pathogen and site of infection
Non-linear PK (saturable absorption, protein binding changes) may reduce simulator accuracy at extreme doses; validate results against TDM where available