PICU CDS Redesign:
Reducing Diagnostic Error in the PICU

Children in the PICU are uniquely vulnerable to diagnostic errors due to their inability to articulate symptoms, varying clinical presentations, and the high stakes of clinical action or inaction. Diagnostic errors affect 12% of high-risk PICU patients, leading to harm in 60% of their cases. These errors are often linked to diagnostic uncertainty and atypical presentations upon admission.

As part of the Live Well Collaborative, I collaborated with a multidisciplinary team of 12, including physicians, nurses, professors, and students, working closely with Cincinnati Children's Hospital.

Over the course of 18 hours and 3 sessions, we shadowed doctors, nurses, and respiratory therapists to observe the intricacies of PICU workflows in real-time. Participated in 2 CODE simulations to analyze how medical teams communicate, share information, and coordinate to diagnose and develop effective care plans.

Conducted 13 interviews across five distinct PICU roles to uncover workflows, communication systems, and pain points during key activities such as rounds, hand-offs, and huddles. Expanded research to include patient perspectives by interviewing a former PICU patient turned advocate and a parent of a long-term critical care patient, gaining valuable insights into both clinical workflows and personal experiences within the PICU.

Through interviews and observations, we mapped the interactions among PICU team members, uncovering communication methods, challenges, and opportunities to reduce diagnostic uncertainty.

Developed an FMEA to identify potential failure modes and successes in PICU admission processes, covering unscheduled non-emergent, unscheduled emergent, scheduled non-emergent, outside hospital transfers, and CODE admissions.

Utilized a digital-twin VR simulation to observe how primary care facilitators (PCFs), registered nurses (RNs), and respiratory therapists (RTs) perform during a diagnosis-intensive event. Each simulation involved diagnosing a baby with an abdominal protrusion and was standardized across participants. This setup allowed us to analyze how each role accesses and communicates critical information. The findings will inform future testing of CDS tool prototypes in the same simulated environment.

By testing existing CDS tools in a VR environment and mapping how each role typically gathers information, we identified key patterns in tool usage. This enabled us to generate hypotheses about the optimal placement and role-specific access of CDS features. These insights were then synthesized into actionable design requirements for the development of their new tool.

Through 8 months of research, we uncovered 65+ unique pain points and ideated specific solutions for each. Through intensive analysis and synthesis, we were able to boil it down into 5 failure modes and opportunity areas.

Developed informational animations to provide a deeper understanding of the PICU: one outlines a typical day, another details teamwork during a CODE, and the last explores patient-care team dynamics. Originally created to onboard future Live Well Collaborative teams, Cincinnati Children's Hospital will also use them to onboard new nurses and RTs.

As part of a 4-year research grant, The Live Well Collaborative contributed to initial research and later usability testing. Our role began with ideating and defining the vision for the new CDS tool. The development and grant teams will now use our findings to create multiple prototypes tailored to different use cases. Live Well Collaborative will reengage in two years to test and validate the new tools using our original data.