> FAQ — METHODOLOGY & DESIGN DECISIONS

Cockcroft-Gault estimates kidney function from age, weight, sex, and serum creatinine — then feeds that estimate into a separate vancomycin equation. It introduces two layers of estimation before you even get a PK prediction. For non-obese patients (BMI < 40), Vancomyzer uses the Colin 2019 model, which takes serum creatinine directly as a covariate — no intermediate CrCl calculation required.

For patients with BMI ≥ 40, Vancomyzer activates a separate obesity model (Smit 2020 + Zhang 2023) that does use Cockcroft-Gault CrCl for clearance estimation. This is clinically appropriate because vancomycin clearance in obesity scales with total body weight via renal elimination, and CG with TBW is the established method in the obesity PK literature. The key difference: the obesity model uses CG intentionally for CL only, while volumes of distribution are scaled to Fat-Free Mass (FFM) — not total body weight.

It was developed in 1976 on 249 male patients to estimate creatinine clearance — not to predict vancomycin pharmacokinetics. It systematically underperforms in elderly patients, low muscle mass, obesity, and critical illness. The National Kidney Foundation no longer recommends it for clinical use because it was not expressed using standardized creatinine values — the assay used in its derivation was likely 10–20% higher than current methods, meaning CG-based calculations lead to higher drug dosing recommendations than originally intended.

Creatinine has been measured two ways in clinical labs:

The old way — Jaffe method (picric acid reaction, since 1886): A colorimetric reaction that measures creatinine but is non-specific. Glucose, bilirubin, acetoacetate, and certain drugs like cephalosporins all interfere and falsely elevate the result. At low creatinine concentrations — below 1.0 mg/dL — the Jaffe method reads approximately 7% higher than the enzymatic method, exceeding the acceptable bias threshold. This matters most in elderly, cachectic, and low-muscle-mass patients — exactly the populations most likely to receive vancomycin.

The current standard — Enzymatic method (IDMS-traceable): Modern labs use an enzymatic assay traceable to isotope dilution mass spectrometry (IDMS), the international gold standard. The enzymatic method is more specific — glucose, acetoacetate, and cephalosporins do not interfere — giving it better accuracy especially at lower creatinine concentrations.

Why this gap matters for dosing: Cockcroft-Gault was derived using Jaffe-measured creatinine. When you plug an enzymatic creatinine value into it — which is what modern labs report — you are using a number the equation was never calibrated for. Converting enzymatic SCr values back into Jaffe-equivalent values has been shown to significantly improve the performance of the Cockcroft-Gault equation for predicting vancomycin concentrations. That conversion step is rarely done at the bedside, creating a systematic mismatch baked silently into every Cockcroft-Gault-based vancomycin calculation in a modern hospital.

Serum creatinine is a byproduct of muscle metabolism — so it reflects muscle mass as much as kidney function. A frail 80-year-old with a creatinine of 0.8 mg/dL may have severely reduced kidney function masked by low muscle mass. Cockcroft-Gault partially adjusts for this using age and a sex factor, but it was not built for patients with sarcopenia, critical illness, or extreme body compositions.

Colin 2019 uses serum creatinine directly as a continuous covariate within a Bayesian framework — not to estimate CrCl, but to inform the model’s prediction of individual PK parameters. Combined with age and weight, the model accounts for muscle mass effects implicitly through the posterior update when a measured vancomycin level is entered. A bedbound geriatric patient’s low creatinine and low weight together shift the posterior estimate in a way Cockcroft-Gault cannot replicate.

Colin 2019 uses age, weight, and serum creatinine directly as covariates in a two-compartment Bayesian model — no intermediate CrCl calculation. It was built from pooled data across 14 studies and multiple patient populations from neonates to elderly, making it one of the most broadly validated vancomycin PK models published. The model includes a specific age-decline function (FDecline) that captures the natural reduction in vancomycin clearance after peak adulthood — something Cockcroft-Gault approximates crudely through age alone.

In traditional vancomycin dosing, determining the right dose requires answering a question that has no single correct answer: which body weight do you use?

Three options exist in clinical practice, each requiring its own calculation:

Total body weight (TBW): The patient’s actual weight. Simple — but in obese patients, dosing on TBW can produce dangerously supratherapeutic levels. In one study of patients receiving TBW-based vancomycin dosing, 48% of patients with a BMI ≥35 were supratherapeutic on their first trough level.

Ideal body weight (IBW) — Devine formula: The weight a patient “should” be based on height and sex.

Males:   50 kg + 2.3 kg per inch over 60 inches
Females: 45.5 kg + 2.3 kg per inch over 60 inches

This requires height — a variable Cockcroft-Gault does not even include — adding yet another manual calculation step. And IBW alone can underestimate dose requirements in obese patients.

Adjusted body weight (AdjBW): IBW + 0.4 × (TBW − IBW). Used when TBW exceeds IBW by more than 30%. Standard clinical dosing guidelines apply this formula when actual body weight is more than 30% above IBW.

The problem: This weight selection step — TBW vs IBW vs AdjBW — is debated, inconsistently applied between institutions, requires height that may not be reliably documented, and adds a third manual calculation on top of the CG estimate already in progress. The 2020 ASHP/IDSA guidelines moved toward recommending TBW for empiric dosing precisely because the IBW debate had no clear resolution.

How Vancomyzer handles body size: For non-obese patients (BMI < 40), the Colin 2019 model takes actual body weight directly as a covariate — no IBW, no AdjBW, no height required. The Bayesian model handles body size effects through its mathematical structure. When a measured level is entered, the posterior update further refines the individual PK estimate regardless of body composition.

For patients with BMI ≥ 40, Vancomyzer activates an obesity-specific model that uses a more physiologically appropriate metric: Fat-Free Mass (FFM). FFM is calculated from height, weight, and sex using the Janmahasatian 2005 equations. Volumes of distribution are scaled to FFM because vancomycin is hydrophilic and distributes poorly into adipose tissue. This is more precise than IBW/AdjBW approximations — it uses a validated pharmacometric equation rather than a height-based rule of thumb.

Independent evaluations consistently rank it among the top performers:

• Outperformed 6 of 7 literature models in a McGill University Health Centre validation • Second best in a Belgian multicenter study of 169 patients and 923 TDM samples • Identified as one of two best-transferable models in a head-to-head comparison of 7 vancomycin PopPK models • Validated across ICU, general ward, and outpatient settings in multiple countries • For patients with BMI ≥ 40, Vancomyzer supplements Colin 2019 with an FFM-based obesity model (Smit 2020 + Zhang 2023) that scales V₁ and V₂ to fat-free mass for more accurate volume estimation in morbid obesity

For non-obese patients (BMI < 40): No. The Colin 2019 model takes age, weight, and SCr directly as covariates — no CrCl, no IBW, no AdjBW. This removes multiple sources of estimation error and keeps the input set minimal.

For patients with BMI ≥ 40: Yes. The obesity model uses Cockcroft-Gault with total body weight to estimate CrCl for the clearance equation (CL = 0.0571 × CrCl + 0.0158 × TBW). This is standard practice in the obesity PK literature — renal elimination of vancomycin scales with total body weight, and CG-TBW is the established estimator in this population. The CrCl value is shown in the clinical advisory panel when the obesity model activates.

Volumes of distribution in the obesity model are NOT based on CrCl or TBW — they use Fat-Free Mass (FFM) from the Janmahasatian 2005 equations, because vancomycin distributes into lean tissue, not adipose.

Fat-Free Mass is the portion of body weight that is not adipose tissue — it includes muscle, bone, organs, and water. FFM is calculated using the Janmahasatian 2005 equations, which require weight, height, and sex:

FFM (male)   = (9270 × TBW) / (6680 + 216 × BMI)
FFM (female) = (9270 × TBW) / (8780 + 244 × BMI)

Vancomycin is a hydrophilic glycopeptide — it distributes into plasma and interstitial fluid but poorly penetrates adipose tissue. In patients with BMI ≥ 40, total body weight dramatically overestimates the volume available for drug distribution. Scaling V₁ and V₂ to FFM instead of TBW produces more accurate volume estimates and prevents overdosing.

The Vancomyzer obesity model (derived from Smit 2020 and Zhang 2023) uses FFM for volumes of distribution while retaining TBW-based CrCl for clearance — reflecting the physiological reality that drug distribution is lean-tissue-limited but renal elimination scales with total body mass.