SBIR Phase II: Applying Language Understanding at the Point of Care to Enhance Clinical Documentation and Realize Quality Improvements

点击翻译按钮获取中文摘要

The innovation presented in this proposal is a new paradigm for capturing critical quality information at the point-of-care using advanced technologies in an intuitive workflow. Healthcare quality and cost represent top national priorities. An increasingly common strategy to improve outcomes and value of care is performance assessment to promote best practices. Over decades, the most successful quality improvement (QI) programs have been heavily data-reliant. These programs require identifying patients that fit specific quality measures and assuring their care meets national guidelines. Linking patients to quality measures is the rate limiting step, involving an overwhelming amount of manual labor to review narrative notes one at a time and link them to an appropriate subset of hundreds of known quality measures. Leveraging a robust platform proven in Phase I, the proposed Phase II solution offers an automated approach to capturing a set of quality measures in real-time. The output will provide rich and compliant documentation enhanced with quality measures that feed the electronic health record (EHR) and downstream clinical, operational, and financial hospital systems through standard protocols. The goal is a disruptive change that will fast-track national initiatives and enable a safer and more efficient healthcare system.The broader/commercial impact of this program is to further national healthcare goals of reducing cost and improving quality in care. The approach leverages increased breadth, depth, and accuracy of patient data captured at the point of care. The most aggressive national initiatives encourage capturing a small portion of the hundreds of known quality measures. Accelerating capture and use of quality measures is an opportunity to meaningfully improve a healthcare system that lags in quality and cost. Impact must also be considered at a personal level. There is a cost to care within a system where quality is not documented and tracked. A typical example out of the hundreds of defined measures is ventilator associated pneumonia (VAP). Multiple studies on manual programs to document VAP and leverage care algorithms demonstrate greater than 40% reduction in mortality and 20% reduction in cost. VAP, though common, did not make the top 15 list of measures required by the government in 2014 because it is too difficult to capture. There is currently no automated approach to capture this quality measure. Addressing technical limitations in documenting quality measures will expand QI reach and save lives.