Document text
Principal Investigator: Masayuki Nigo
Organization: METHODIST HOSPITAL RESEARCH INSTITUTE
Fiscal Year: 2024
Award: $470,454
Funding agency: National Institute of Allergy and Infectious Diseases
Abstract: (30 lines)
Vancomycin is one of the most commonly used antimicrobial drugs in inpatient settings. National guidelines
recommend Bayesian models to monitor the therapeutic drug concentration of vancomycin, especially for
methicillin-resistant Staphylococcus aureus (MRSA), to minimize drug toxicity while maintaining its efficacy.
Existing Bayesian models, despite being claimed as patient-specific pharmacokinetic (PK) models, use simple
patient features and are studied in limited patient populations for the population-based PK parameters (the
Bayesian prior). Increasingly available real-world electronic health records (EHR) provide a wide range of
patient-specific data, including data on vancomycin dosage and serum levels. However, the limited flexibility of
the Bayesian model structure prohibits the full use of these rich data. Deep-learning models, such as recurrent
neural network (RNN), are particularly attractive for PK of vancomycin in EHR, compared to Bayesian models
and other traditional machine learning models, because deep-learning models enable more flexible patient-
specific inputs and possess a higher latent capacity. Thus, they deliver better predictions for a diverse
population. Our deep-learning model for vancomycin (PK-RNN-V) outperforms publicly available Bayesian
models but can be improved on various aspects. In Aim 1, we will improve PK-RNN-V model architectures and
add more patient-specific data and a finer timestep. We will construct two-compartment PK-RNN models to
increase predictive power in patients with unsteady states. We will augment PK-RNN-V with Med-BERT to
improve the embedding of categorical data. We will also develop multi-track ordinary differential equations
models for simultaneous prediction of serum creatinine and vancomycin levels. In Aim 2, we will use EHR from
different sources to validate our PK-RNN-V model and improve the data-extraction flow and pre-processing to
harmonize data from healthcare systems. We will use EHR from Houston Methodist Hospital and Memorial
Hermann Hospital System/The University of Texas Health Science Center in Houston, TX, the University of
Arizona in Phoenix, AZ, and the publicly available MIMIC-IV database (Boston, MA). These databases contain
data from more than 121,007 patients who received at least one dose of intravenous vancomycin. In Aim 3, we
will add dosing recommendations based on PK-RNN-V model predictions as a feature and validate our model
in specific subgroups with challenging vancomycin PK to predict PK levels. This project will deliver substantial
model improvements, leading directly to the optimization of vancomycin use in hospitals, increased in patient
safety by minimizing adverse events, and reduced healthcare costs, which align with NIH’s research mission.
Terms: <Adverse Experience><Adverse event><Architecture><Arizona><Bayesian Modeling><Bayesian adaptive designs><Bayesian adaptive models><Bayesian belief network><Bayesian belief updating model><Bayesian framework><Bayesian hierarchical model><Bayesian network model><Bayesian nonparametric models><Bayesian spatial data model><Bayesian spatial image models><Bayesian spatial models><Bayesian statistical models><Bayesian tracking algorithms><Blood Serum><Boston><Clinical><Complex><Connectionist Models><Creatinine><Critical Illness><Critically Ill><Data><Data Bases><Data Set><Data Sources><Databases><Differential Algebraic Equation><Differential Equation><Dose><Drug Kinetics><Drug toxicity><Drugs><Electronic Health Record><Engineering / Architecture><Future><Goals><Guidelines><Health Care Costs><Health Care Systems><Health Costs><Health Sciences><Healthcare Costs><Healthcare Systems><Hemodialyses><Hemodialysis><Hospitals><Human><Inpatients><Intensive Care Units><Intravenous><Kidney Replacement Therapy><Laboratories><MRSA><Medication><Methicillin Resistant S. Aureus><Methodist Church><Methodists><Mission><Modeling><Modern Man><Monitor><NIH><National Institutes of Health><Neural Network Models><Neural Network Simulation><Output><Patients><Pediatric Intensive Care Units><Perceptrons><Peripheral><Pharmaceutical Preparations><Pharmacokinetics><Population Heterogeneity><Publishing><Recommendation><Renal Replacement Therapy><Research><Sampling><Serum><Source><Structure><Subgroup><System><Techniques><Testing><Texas><Therapeutic><Time><Toxic effect><Toxicities><United States><United States National Institutes of Health><Universities><Update><Validation><Vancomycin><anti-microbial><anti-microbial agent><anti-microbial drug><antimicrobial><computer based prediction><data base><data harmonization><deep learning><deep learning based model><deep learning method><deep learning model><deep learning strategy><demographics><diverse populations><dosage><drug/agent><electronic health care record><electronic health data><electronic health medical record><electronic health plan record><electronic health registry><electronic medical health record><flexibility><flexible><harmonized data><heterogeneous population><improved><machine learning based model><machine learning model><methicillin resistance Staphylococcus aureus><methicillin resistant Staphylococcus aureus><methicillin resistant strains of Staphylococcus aureus><model generalizability><patient population><patient safety><patient subclass><patient subcluster><patient subgroups><patient subpopulations><patient subsets><patient subtypes><pharmacokinetic model><population based><population diversity><predictive modeling><pregnant><prototype><recurrent neural network><success><trend><validations>