DataWeaver: An Interface to Complex Clinical Data

DESCRIPTION (provided by applicant): Pediatricians routinely need to make rapid decisions based on large amounts of complex and potentially unreliable data. However, this decision-making process has proved difficult to emulate, or even to support, using software systems. An important feature existing systems lack is a user interface allowing pediatricians to quickly and effectively understand and make use of complex interrelated and time dependent data. This proposal seeks to address this deficiency by using an innovative and patented graphical user interface methodology called "DataWeaver." The basis of the DataWeaver display is a hierarchical structure called a "weave" showing intersections of different types and pieces of data in a readily comprehensible way. The DataWeaver methodology is designed to provide the pediatrician with a complete view of the patient's overall status while supporting rapid access to the underlying detailed information, thus saving clinicians time and increasing the quality of care. Despite the attractiveness of the DataWeaver methodology, it has neither been implemented nor field tested. One reason is that mapping the human-oriented DataWeaver interface to the underlying database and to other components has until recently been deemed infeasible. However, recent advances in software engineering make constructing these mappings currently feasible. Specifically, by using model-driven architecture (MDA) methods to specify various components, such as data sources, decision support rules, and graphical components via an abstract model. This in turn allows much of the system's code to be automatically generated instead of needing to be hand-written, greatly reducing coding time and increasing code reliability. Because of the broad scope and complexity of the field of pediatrics, and because of strong interest from the Pediatrics Department at Johns Hopkins University, sickle cell disease content has been chosen for Phase I testing.