Using a computerized provider order entry system to meet the unique prescribing needs of children: description of an advanced dosing model.

BACKGROUND: It is well known that the information requirements necessary to safely treat children with therapeutic medications cannot be met with the same approaches used in adults. Over a 1-year period, Duke University Hospital engaged in the challenging task of enhancing an established computerized provider order entry (CPOE) system to address the unique medication dosing needs of pediatric patients. METHODS: An advanced dosing model (ADM) was designed to interact with our existing CPOE application to provide decision support enabling complex pediatric dose calculations based on chronological age, gestational age, weight, care area in the hospital, indication, and level of renal impairment. Given that weight is a critical component of medication dosing that may change over time, alerting logic was added to guard against erroneous entry or outdated weight information. RESULTS: Pediatric CPOE was deployed in a staggered fashion across 6 care areas over a 14-month period. Safeguards to prevent miskeyed values became important in allowing providers the flexibility to override the ADM logic if desired. Methods to guard against over- and under-dosing were added. The modular nature of our model allows us to easily add new dosing scenarios for specialized populations as the pediatric population and formulary change over time. CONCLUSIONS: The medical needs of pediatric patients vary greatly from those of adults, and the information systems that support those needs require tailored approaches to design and implementation. When a single CPOE system is used for both adults and pediatrics, safeguards such as redirection and suppression must be used to protect children from inappropriate adult medication dosing content. Unlike other pediatric dosing systems, our model provides active dosing assistance and dosing process management, not just static dosing advice.

Real-World Integration of a Sepsis Deep Learning Technology Into Routine Clinical Care: Implementation Study.

BACKGROUND: Successful integrations of machine learning into routine clinical care are exceedingly rare, and barriers to its adoption are poorly characterized in the literature. OBJECTIVE: This study aims to report a quality improvement effort to integrate a deep learning sepsis detection and management platform, Sepsis Watch, into routine clinical care. METHODS: In 2016, a multidisciplinary team consisting of statisticians, data scientists, data engineers, and clinicians was assembled by the leadership of an academic health system to radically improve the detection and treatment of sepsis. This report of the quality improvement effort follows the learning health system framework to describe the problem assessment, design, development, implementation, and evaluation plan of Sepsis Watch. RESULTS: Sepsis Watch was successfully integrated into routine clinical care and reshaped how local machine learning projects are executed. Frontline clinical staff were highly engaged in the design and development of the workflow, machine learning model, and application. Novel machine learning methods were developed to detect sepsis early, and implementation of the model required robust infrastructure. Significant investment was required to align stakeholders, develop trusting relationships, define roles and responsibilities, and to train frontline staff, leading to the establishment of 3 partnerships with internal and external research groups to evaluate Sepsis Watch. CONCLUSIONS: Machine learning models are commonly developed to enhance clinical decision making, but successful integrations of machine learning into routine clinical care are rare. Although there is no playbook for integrating deep learning into clinical care, learnings from the Sepsis Watch integration can inform efforts to develop machine learning technologies at other health care delivery systems.