Understanding Complex Clinical Decision Tasks for Better Health Information Technology System Design

Clinical Decision Support Systems (CDSS) have been shown to improve healthcare quality, safety, and value. However, most of the studies demonstrating the benefits of CDSS were conducted in healthcare organizations that built their own electronic health record (EHR) systems and CDSS capabilities. Conversely, typical commercial EHR systems have only basic CDSS, such as drug-drug interaction alerts and preventive reminders. This kind of decision support fails to account for factors that complicate decision making tasks, resulting in widely reported issues such as alert fatigue and limited use. Prior research on task complexity in other domains has shown the potential to improve the quality of decision making support by incorporating task complexity in the system design. Less is currently known about complexity in the context of clinical decision making.

This study investigated infectious disease clinicians’ coping strategies to deal with complexity factors for better decision making and enhanced patient safety, and identified specific complexity factors to support high-level reasoning. Complexity-contributing factors (CCF) may occur when 1) the overall clinical picture does not match the pattern, 2) there is a lack of comprehension of the situation, and 3) physicians are dealing with social and emotional pressures. Using CCFs to design and develop a CDSS visual display, the study compared similar complex patient cases for improved clinical reasoning.

The specific aims of the project were as follows:

• Characterize leverage points of complex clinical decision tasks. 
• Identify factors that contribute to complex decision tasks. 
• Design and evaluate a prototypical CDSS that supports high-level reasoning for decision tasks. 

Researchers conducted cognitive task analysis interviews, identified clinician strategies to adapt to their information environment, and mapped relevant decision support tools. They subsequently developed an integrated clinical complexity model and used it to identify 20 specific CCFs. CCF findings informed the development of a population information display, which was evaluated by querying a Veterans Administration clinical database to extract information from patients who were similar to the complex simulated case. Outcomes indicated that the population information display helped to change 60 percent of patient treatment plans positively. The study also found that experts processed the population-based information faster than non-experts, thus validating the display content.