PrecisePK Uses A.I. and Bayesian Dosing to Bridge the Gap between Research and Clinical Care

Anh Ta, PharmD
October 20, 2020

PrecisePK is bridging the critical gap in medical literature and bedside patient care.

A 15-year-old boy with a history of leukemia, MRSA bacteremia, and acute kidney injury (AKI) needs Vancomycin stat. What loading dose would you recommend to target 400 mg*h/L AUC? Twenty-four hours later, after a loading dose of 1000 mg IV, the measured level is 12 mg/L. What should the next dose be adjusted to? What resources would you use to help answer these questions?

PrecisePK bridges this gap between drug reference guides and precise dosing recommendations for medications with narrow therapeutic index.

The widely used drug references such as Clinical Pharmacology, UpToDate, PubMed, etc. provide excellent raw information to pharmacists but fall short of translating that information into usable dosing regimen recommendations. A pharmacist must still use a calculator to compute the recommended dose, however, even the manually calculated values often cannot account for all the nuanced factors that influence patients' pharmacokinetics for a specific drug. Also, for a truly precise dosing calculation, a Bayesian algorithm must be utilized to tailor the delivered dose for that specific patient. PrecisePK is built to address this critical gap in clinical care and academic research. It combines the latest drug research, published guidelines, and machine learning algorithms to compute individualized dosing regimens and predictive serum levels. This helps users pick the right dose and make well-informed clinical decisions.

PrecisePK uses Bayesian algorithms to calculate precise dosing regimen, predict future AUC or serum levels, thus help pharmacists provide dosing recommendations in a timely manner.

  • Use a wide library of clinical guidelines and references to build drug models for different patient populations such as pediatrics, neonates, obese, patients with kidney injury, critically ill, generally healthy adults, etc.
  • Account for patient-specific factors such as amputations, pregnancy, other drugs, and genotypes that influence the patient pharmacokinetics.
  • Automatically select the most appropriate pharmacokinetic model, known as Bayesian priories, to compute the population pharmacokinetic parameters and initial dosing recommendation.
  • Combine the observed serum levels and the Bayesian algorithm to compute patient-specific pharmacokinetic parameters, known as Bayesian posteriori, thus generate a more individualized dosing regimen.
  • Continuously learn the pharmacokinetic models specific to a hospital's local population to further optimize vancomycin exposure and discover new parameters that influence dosing calculations.

PrecisePK is an indispensable dosing software in point of care according to peer-reviewed articles

In a recent paper published by Turner, R.B., et al. in the journal Pharmacotherapy, PrecisePK was shown to be the most accurate in predicting the Vancomycin AUC in critically ill patients. The graph below shows the ratio between the empirically measured AUC (using the trapezoid rule) and the AUC predicted by the software. Therefore, a value of 1.0 would imply the highest accuracy. PrecisePK was able to predict the AUC most accurately due to the comprehensiveness of the pharmacokinetics models and the two-compartment kinetics models used by the program in its computations.

Optimize Your Clinical Experience with PrecisePK.