
Improving Provider Documentation Using Pediatric AutoDx
0
18
0
The Problem: Coding Queries and Provider Frustration
Provider documentation is the cornerstone of hospital operations and effective clinical communication. However, the process of ensuring accurate documentation often involves Clinical Documentation Improvement (CDI) specialists reviewing charts and sending coding queries to providers. These queries, while essential, can consume a significant amount of provider time and contribute to frustration and dissatisfaction.
Our Goal: Streamlining Documentation and Improving Provider Experience
Driven by the desire to enhance the provider experience, we set an ambitious goal: to decrease applicable coding queries by 50% and improve the ease of practice for 25% of pediatric inpatient providers at Comer Children's Hospital by June 30, 2024.
The Solution: Tailoring AutoDx for Pediatric Care
To achieve our goal, we turned to AutoDx, a system already in use for adult patients. However, adapting it for pediatric use required careful consideration of the unique needs of this population. We focused on three key areas:
Pediatric-Specific Rules:Â We developed clinical rules that accounted for pediatric diagnoses that differ from adults, ensuring accurate and relevant documentation.
Optimized Lookback Windows:Â We adjusted lookback windows to streamline information and focus on the most pertinent data for daily pediatric care.
Documentation Variation:Â We adapted the system to accommodate the specific ways pediatric services document patient information.
Implementation and Evaluation
We deployed AutoDx to inpatient pediatric primary services and conducted pre- and post-intervention surveys to assess provider documentation burden and ease of practice. Additionally, we measured coding queries before and after implementation and used PDSA cycles to optimize tool usage.
Results: A Significant Impact
The results were remarkable! AutoDx reduced applicable coding queries by an impressive 58%. Nearly half of the providers agreed that AutoDx had improved their ease of practice. Furthermore, AutoDx usage steadily increased and then stabilized, indicating consistent adoption.
Table 1. Coding queries per 1000 discharges before and after AutoDx implementation
 | Pre-AutoDx | Post-AutoDx | |
Condition | Total Queries per 1000 Discharges | Total Queries per 1000 Discharges | % Change |
Anemia | 11.53 | 4.75 | -59% |
Coagulopathy | 0.66 | 0.00 | -100% |
Electrolyte Disturbances | 29.00 | 5.43 | -81% |
Malnutrition | 21.42 | 14.25 | -33% |
Mineral Metabolism Disorders | 5.27 | 0.68 | -87% |
Obesity | 12.85 | 8.82 | -31% |
Total | 80.72 | 33.92 | -58% |
Figure 1. AutoDx usage over time in provider notes
Conclusion and Future Directions

This project highlights the power of clinical informatics to transform healthcare processes and improve provider satisfaction. AutoDx has proven to be a valuable tool in streamlining documentation and reducing the burden of coding queries. Moving forward, we plan to expand the number of diagnoses included in AutoDx and continue exploring innovative ways to enhance the provider experience.
Acknowledgments
I extend my sincere gratitude to Robert Strong and Benjamin Petro for their invaluable IT support in building AutoDx, Amy Kratochvil for providing coding query data, and Drs. Riley Boland, Matthew Cerasale, and Cheng-Kai Kao for their guidance and expertise.
By Kevin Smith, MD
ACIF Chair of Communications
Pediatric Critical Care
University of Chicago