In a complex, dynamic system such as healthcare delivery, there is often a “gap” between work as imagined (how clinical workflow ideally occurs) and work as done (how clinical workflow occurs in real-time). Therefore, it is critical when implementing new health information technology to work within the constraints of hospital-specific workflow to understand the reality of work, rather than relying solely on how it is reported. The present research seeks to fully understand the current state of workflow as it relates to cardiovascular risk calculation, with the ultimate goal of implementing an atherosclerotic cardiovascular disease risk calculator within the clinicians’ electronic health record. By conducting both interviews and observations, researchers were able to compare work as imagined to work as done in order to implement the risk calculator to meet clinician needs.

Usability testing has become a necessary step to successfully implement health information technology, but despite the expanded variety of usability methods, systems still are not guaranteed to meet user needs in a hectic and dynamic healthcare environment. The present work discusses some of the metrics employed with the intent of usability testing clinical decision support alerts for the early identification of sepsis patients. By utilizing methods such as eye tracking, think aloud protocols, human computer interaction, and more, health information technology can be adequately developed to ensure effective, efficacious, and optimal integration.

Trauma care is complex as severe injuries happen unexpectedly, may not immediately present symptoms, and may require assessment by multiple providers to improve outcomes. This work aims to explore the complexities of trauma patient flow and documentation challenges through semi-structured interviews of both in-field and in-hospital providers. Ten trauma care providers were asked to explain their typical workflow and documentations related to trauma care. We found that trauma care flow is convoluted and non-linear depending on several factors, such as how the patient enters the system and the severity of injuries. Additionally, documentation practices vary, following no specific guidelines, partly due to the chaotic nature of incoming traumas. Participants mentioned using apps to aid workflow, but none were used to aid in documentation or clinical handover. Documentation usually takes place on paper or away from electronic devices, which has design implications for clinical decision support or artificial intelligence development tools.

In the best examples, clinical decision support (CDS) systems guide clinician decision-making and actions, prevent errors, improve quality, reduce costs, save time, and promote the use of evidence-based recommendations. However, the potential solution that CDS represents are limited by problems associated with improper design, implementation, and local customization. Despite an emphasis on electronic health record usability, little progress has been made to protect end-users from inadequately designed workflows and unnecessary interruptions. Intelligent and personalized design creates an opportunity to tailor CDS not just at the patient level but specific to the disease condition, provider experience, and available resources at the healthcare system level. This chapter leverages the Five Rights of CDS framework to demonstrate the application of human factors engineering principles and emerging trends to optimize data analytics, usability, workflow, and design.