Abstract: FR-PO0022
Design, Deployment, and Early Adoption of an Electronic Health Record (EHR)-Integrated Longitudinal Visualization Tool for Inpatients with AKI
Session Information
- Artificial Intelligence: Clinical Prediction, Decision Support, and Digital Kidney Care
October 23, 2026 | Location: Exhibit Hall A, Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Artificial Intelligence, Digital Health, and Data Science
- 300 Artificial Intelligence, Digital Health, and Data Science
Authors
- Thiessen Philbrook, Heather, The Johns Hopkins University School of Medicine, Baltimore, Maryland, United States
- Bitzel, Jack, The Johns Hopkins University School of Medicine, Baltimore, Maryland, United States
- Menez, Steven, The Johns Hopkins University School of Medicine, Baltimore, Maryland, United States
- Baughman, Derek J., The Johns Hopkins University School of Medicine, Baltimore, Maryland, United States
- Parikh, Chirag R., The Johns Hopkins University School of Medicine, Baltimore, Maryland, United States
Background
Acute kidney injury (AKI) complicates more than 20% of hospital admissions and is associated with poor outcomes. Inpatient nephrology consults require rapid synthesis of comorbidities, laboratory trajectories, nephrotoxin exposures, fluid balance, and procedural events — information that is fragmented across multiple EHR screens and time windows. No readily available visualization tool is purpose-built for inpatient nephrology workflow.
Methods
Patient Insight, an interactive longitudinal viewer embedded in the Johns Hopkins EHR was deployed for nephrology providers in November 2025. A single dashboard displays (a) a patient summary panel — comorbidities, KDIGO AKI stage, estimated baseline and peak serum creatinine, and dialysis history; and (b) configurable longitudinal panels for AKI-relevant medication classes (RAASi/SGLT2i, diuretics, NSAIDs, antimicrobials, pressors, immunosuppressants, chemotherapy, immunotherapy), key patient events, kidney function, electrolytes, blood counts, blood pressure, and intake/output (see figure).
Results
In the first 5 months, 20 nephrology providers (55% faculty, 45% fellows) opened Patient Insight 299 times for 246 unique patients. The tool was accessed in 10% to 15% of nephrology consults. Adoption was higher among fellows than faculty (69% vs. 55%). The median user opened the tool 8.5 times.
Conclusion
Patient Insight is, to our knowledge, the first EHR-integrated longitudinal visualization tool purpose-built for inpatient nephrology. Early uptake supports feasibility, while heterogeneous adoption highlights opportunities for targeted onboarding and tighter workflow integration. Ongoing evaluation will quantify the tool's impact on chart-review time, and patientcare outcomes. The harmonized FHIR + analytics framework is designed for portability to other EHR systems.