P3B411: Development of a near Real-time Data Stream to Communicate Perioperative Clinical Outcomes: Resourcing Quality Improvement and Clinical Outcomes Research at Scale

Background: Continuous improvement in the perioperative space is challenged by the lack of high-quality, specific, near real-time, validated data streams. Our team obtained hospital resources to unify perioperative reporting from the electronic health record (EHR) and report clinically important clinical outcome data for the purposes of quality improvement, registry reporting, and outcomes research work. Our aim was to create a dashboard to serve as a complete and accurate source of “truth” regarding clinical outcomes and safety events (both common and rare).

Methods: Prior to 2022, a limited number of clinical quality indicators were updated monthly on a spreadsheet and shared. Data collection for clinical quality, registry reporting and clinical outcomes research was abstracted largely using manual chart review. Recent labor shortages and a desire to improve the efficiency of data collection processes motivated our team to develop an effective data extraction infrastructure. The Departments of Surgery and Pediatric Anesthesiology hired a dedicated software engineer/quality analyst to lead efforts to improve reporting for the Performance Improvement and Patient Safety committee related to Children’s Surgery Verification. Project goals included: 1, create a dashboard to include safety metrics; 2, review/validate existing data; 3, harmonize definitions; 4, centralize perioperative data and index all events against a surgical case; 5, create new data streams backed by standard of care workflows to support real-time collection; and 6, create meaningful data visualizations for metrics, such as statistical process control charts or cumulative summation charts for exploratory data analysis and communication of clinical outcomes. The dashboard can be filtered by clinical service to resource local quality improvement teams and allow access to near real-time data for the purposes of clinical quality improvement work.

Results: All project goals have been met. Additionally, our team has developed discrete fields within the EHRs to facilitate reporting needs for rare events. The dashboard has highlighted important positive trends and areas for teams to focus their clinical quality improvement work while minimizing the burden of manual chart abstraction. The development of these data streams has allowed our team to automate reporting to multiple clinical registries and substantially decrease the volume of manual chart review. We are now able to leverage this data infrastructure for ongoing clinical outcomes research efforts. Figure 1 depicts the dashboard configuration with sample data visualization.

Conclusion: Standard definitions, centralizing data, and automating data collection has stimulated teams to engage in various quality improvement initiatives. Prioritizing efficient data access and creating a versatile system that can grow to accommodate future data needs increases team engagement and uses resources effectively in a busy hospital system.