We quantified variability of cardiac responses to lipid emulsion infusions by estimating individualized parameters for a pharmacokinetic–pharmacodynamic model applied to rats.

We used the variability observed across a small population to simulate a large virtual population experiencing bupivacaine systemic toxicity followed by an intravenous lipid intervention.

We thereby demonstrated that a predictive virtual population model for lipid emulsion therapy can be developed by integrating very limited physiological data with mechanistic physiologically based pharmacokinetic modeling. Median lethal dose predictions for treated versus untreated bupivacaine toxicity agreed well with prior experimental observations, and causal analysis supported a key role being played by accumulation of bupivacaine in muscle tissue rather than the liver.