TL;DR
This paper introduces MyHeartPass™, a novel clinical decision support system (CDSS) designed to aid emergency department (ED) physicians in managing critically ill children with congenital heart disease (CHD). By employing a Decision-Centered Design (DCD) framework, the system translates complex clinical decision requirements into actionable guidance, aiming to improve patient outcomes in high-stress, time-constrained situations.
What Problem Does This Paper Address?
Children with congenital heart disease (CHD) often present with fragile hemodynamics that can rapidly deteriorate, especially during common childhood illnesses. These children frequently end up in local emergency departments (EDs) where expertise in CHD is limited. ED physicians face significant challenges in accurately assessing these complex cases, determining the underlying causes of deterioration, and selecting appropriate interventions. This gap in expertise can lead to delayed or incorrect treatment, potentially resulting in severe consequences. While previous research has identified these unmet decision-making needs, there's been a lack of actionable guidance to bridge this gap. This paper aims to address this by developing a CDSS that specifically supports ED physicians in managing these high-risk pediatric patients.
Methods
The study utilized the Decision-Centered Design (DCD) framework, a methodology focused on supporting decision-making in time-constrained, challenging, and uncertain environments. The DCD process involved several key stages:
- Knowledge Elicitation: Cognitive Task Analysis (CTA) methods, specifically Critical Decision Method (CDM) interviews, were used to elicit decision-making processes from both CHD experts and ED physicians. This involved recalling specific clinical encounters to understand how they acquire data, interpret it, and make decisions.
- Analysis and Representation: A modified framework analysis was applied to the interview transcripts to identify key decision requirements and associated design concepts. This involved inductively coding the data to uncover themes related to the cognitive processes involved in managing CHD patients.
- Application Design: Based on the identified decision requirements and design concepts, a prototype CDSS named MyHeartPass™ was developed. This phase involved iterative design and refinement, including a design workshop with experts and ED physicians, and user interface (UI) optimization.
The core innovation lies in the DCD approach, which moves beyond simply providing information to actively supporting the cognitive processes of decision-making in complex scenarios.
Datasets
Not discussed in the paper. The study relied on interview transcripts from CHD experts and ED physicians for analysis.
Key Insights and Ideas
- The "Sensemaking" Gap: ED physicians struggle to visualize a patient's unique cardiac anatomy and translate it into an understanding of their specific physiology. This "sensemaking" deficit is a major hurdle in their decision-making process.
- Anticipation is Key: Effectively anticipating potential deterioration requires explicitly considering CHD-specific differential diagnoses. ED physicians often lack the specialized knowledge to generate these crucial differential diagnoses.
- Managing Complexity is Crucial: Selecting appropriate interventions for CHD patients is highly dependent on their unique physiology. Standard treatments for non-CHD patients can be detrimental to CHD patients, highlighting the need for tailored guidance.
- Physiology-Based Decision-Making: CHD experts consistently base their decisions on a deep understanding of the patient's cardiac physiology, a skill that is less developed in ED physicians.
- Bridging the Expertise Gap: The CDSS aims to democratize CHD expertise by providing ED physicians with readily accessible, synthesized information that supports their cognitive processes.
Key Results
The study identified three primary decision requirements for ED physicians managing pediatric CHD patients:
- Distinguish the patient's unique physiology based on their unique cardiac anatomy: This involves visualizing the heart's structure and understanding how it affects blood flow and function.
- Explicitly consider CHD specific differential diagnoses to allow a more structured reflection of diagnosis: This helps ED physicians anticipate potential causes of deterioration.
- Select CHD appropriate interventions for each patient: This ensures that treatments are tailored to the patient's specific condition and physiology.
These requirements were translated into design concepts incorporated into the MyHeartPass™ prototype. The prototype includes features like anatomical diagrams, descriptions of cardiac anatomy and surgical history, latest echocardiography reports, patient-specific vital signs and ECG tracings, descriptions of physiological states, potential signs/symptoms, diagnostic considerations, and physiology-based therapeutic recommendations.
Limitations and Future Work
- Limitations:
- The study was conducted at a single, major cardiac center, which may mean the ED physicians had more exposure to CHD patients than those in general EDs.
- The reliance on recall for CTA may introduce some bias, although CDM is designed to focus on key decisions rather than precise details.
- Future Work:
- A case-based simulation study is underway to evaluate the effectiveness of MyHeartPass™ on treatment-related decision-making.
- Future work should focus on integrating MyHeartPass™ with existing Electronic Health Records (EHRs) for automatic data updates and exploring optimal integration into ED workflows.
- Further studies are needed to assess decision-making among general ED physicians, as their experience with CHD may differ.
Summary updated