This application addresses broad Challenge Area (10) Information Technology for Processing Healthcare Data, 10-LM-102: Advanced decision support for complex clinical decisions. The lack of prescribing information and meaningful dosing guidance places an incredible burden on our pediatric caregivers with the end-result often the sub-optimal use of many medicines in children [1]. It is widely believed by the Institute of Medicine and American Association of Pediatrics that prescribing errors can be reduced via intelligent decision support systems that facilitate pharmacotherapy decision making [2]. We are developing drug-specific Pediatric Pharmacotherapy Decision Support Systems (PPDSS) at The Children's Hospital of Philadelphia (CHOP) in an attempt to address these issues. Our goals for this system include the ability to: 1) provide dosing guidance consistent with formulary standard of care, 2) examine patient-specific pharmacotherapeutic indices with target agents relative to historical controls derived from the hospital data warehouse, 3) explore diagnoses - drug correlation in conjunction with utilization and 4) educate physicians on clinical pharmacologic principles specific to population and drug combinations of interest. Within this system, static compendial information (e.g., Lexi-Comp) can be searched, indexed and summarized for easy viewing;real-time, forecasting of relevant drug exposure or clinical markers (lab values, pharmacodynamics, adverse events) is made available based on interface to current patient data from our electronic medical records (EMR) system. Our long-term vision is to facilitate the safe and effective administration of drugs used in the treatment of children with an evolving collection of PPDSS providing improved management of patient pharmacotherapy. The objective of this challenge grant application is to create a patient-based informatics system that contains the relevant guidance concerning dosing of specific agents to various pediatric subpopulations. This guidance will have the opportunity to grow (artificial intelligence) as patient diversity expands the historical experience (population priors) with an agent or combination of agents. This research application provides compelling evidence that clinical pharmacology driven decision support systems can interface with hospital EMR to provide individualized pharmacotherapy guidance. Our central hypothesis is that dosing guidance can be improved when the caregiver responsible for pharmacotherapy is informed in an expedient manner while in the process of patient care. The integration of well-characterized models that account for sources of variation in pharmacokinetics, pharmacodynamics and/or relationships with clinical outcomes is married with the most relevant clinical data associated with the management of a particular or combination of drug and disease states. The incoming clinical data streams can of course expand as new information becomes available (e.g., genomic data and guidance). Drug dashboards are designed for and by the physician therapeutic area in collaboration with Clinical Pharmacology and IT team members. Current prototype dashboards provide forecasting of drug exposure at select time points consistent with clinical protocols used to manage toxicity and efficacy. Forecasting tools permit dosing scenarios to be explored via a user-friendly interface that front-ends a pediatric population-based PK/PD model. By mobilizing the pediatric caregiver, clinical pharmacology and health IT sectors we envision a heightened appreciation for pharmacology-based decision support solutions and an evolving strategy to use such solutions to improve outcomes in children. The project brings together drug-specific decision support generated by our clinical pharmacology experts and clinical caregivers with predictive models generated by our pharmacometrics and informatics team. Drug dashboard prototypes for methotrexate (chemotherapeutic agent) and tacrolimus (used to prevent organ rejection) are under development and will be clinically evaluated prior to a production launch. The prediction engines and forecasting routines are transparent to the end-user;emphasis on clinical outcomes including reduced medication errors and length of hospital stay will be used to demonstrate the ROI for individual dashboards. Funding this application will ensure the generalizability of our data integration solution beyond CHOP boundaries, provide the appropriate test data to qualify/validate the prediction algorithms and generate the clinical experience with the PPDSS that will encourage other therapeutic areas and institutions to utilize the system. The solution (PPDSS dashboards and installation guidance) will be provided to all who request as part of our data-sharing plan.
The proposed research provides a mechanism to incorporate the most relevant prior knowledge into dosing guidance for an often understudied population - children. An important feature of the research is the incorporation of predictive models that provide real-time forecasting of patient response to existing therapy recommendations conditioned on prior patient experience with the potential to provide clinical scenario testing if the caregiver requests [3]. The initial dashboards developed as part of this proposal can easily be extended to accommodate other populations (elderly, pregnancy, etc) and expanded to include participation from additional caregiver stakeholders, clinical pharmacologists, bioinformatics scientists and pharmacometricians.
