Reengineering Approaches for Learning Health Systems: Applications in Nursing Research to Learn from Safety Information Gaps and Workarounds to Overcome Electronic Health Record Silos

Abstract

Health systems engineering is an approach to effectively implement a learning health system to drive more efficient and safer care by adapting and aligning individual structures (e.g., applications) and processes (e.g., workflows) to optimize outcomes within a “system of systems”. Nurses have been described as particularly adept at identifying and utilizing workarounds to overcome poor system design. Workflows and workarounds are not limited to directly observable patient care activities; they also occur within documentation activities and can be modeled using metadata (data about data) from clinical information systems. This chapter will outline 3 broad approaches that can be triangulated within a systems engineering framework to reengineer nursing and patient care workflows and overcome information silos by actively learning from safety information gaps and workarounds within a health system: (1) “In the lab” participatory design and usability evaluation, (2) “In the wild” observations, and (3) “In the metadata” models of health care processes. Systems engineering methods can be applied to a broad range of healthcare processes to model workflow, data and information flow to support the development, integration, and optimization of health IT applications leveraging a 5-phase approach: problem analysis, design, development, implementation, and evaluation. In this chapter we present use cases of pragmatic applications grounded in theoretical and methodological approaches within a systems engineering framework that demonstrate the iterative nature of health IT evaluation. This chapter highlights the complexity of nursing and patient care workflows and the fact that even well-designed systems that adequately address socio-technical dimensions as part of the development process, require continued attention to workflow during and after implementation. Post-implementation attention to end-users’ concerns and feedback provides an opportunity for system optimization. Secondary analysis of EHR data after implementation can provide important clues about alignment with documentation workflows, effective information flow, and accurate data interpretation. These activities are needed to achieve a learning health care system and can be achieved if nursing domain experts work closely with data science experts throughout the system lifecycle to contextualize clinical analyses and help to successfully convert data into knowledge.