Leveraging Automated Data Flows, AI, and Analytics for Chart Abstraction

Discover how to enhance registry participation by streamlining chart abstraction and reducing costs. This webinar will explore the latest technology innovations for cardiothoracic and oncology registries, showcasing AI-driven solutions and automated data flows. Learn how these advancements provide actionable insights and how Tech-Enabled Managed Services (TEMS) can significantly improve efficiency, allowing data abstractors to focus on process improvement and driving value from clinical registries.

1. Reducing the Cost and Increasing the Value of Registry Participation Automated Data Flows, AI, and Analytics for Chart Abstraction Allie Coronis, Sheila Fairless, and Melanie Rogan

2. Reducing the Cost, Increasing the Value of Registry Participation ARMUS: AI Chart Abstraction Tool and Analytics CRStar : Automated Data Integrations, NLP, Accreditation Management, and Analytics TEMS: Bundled Service Offering to Reduce Costs Insights to Impact – A Case Study

3. Two Types of Value in Registries Regulatory, Accreditation, and Reputation Insights, Improvement, and Better Performance

4. Reducing the Cost, Increasing the Value ARMUS for Hospitals Suite CV, GI, OB, and spine registries Streamlined data abstraction Analytics to provide insights CRStar Cancer registries More efficient case finding and abstraction Analytics to provide insights Data and Analytics Platform Data integration and automation Tools to support efficiency and process improvements TEMS (Technology- Enabled Managed Services) Health Catalyst takes on the responsibility for your data abstractors, trains them in process improvement, and provides them tools to gain efficiency

5. ARMUS for Hospitals Suite

6. ARMUS by Health Catalyst Technology and services combined into one clinical registry solution to increase the value of clinical registry participation for healthcare institutions

7. Poll Question #1 What is the average time to abstract an STS Adult Cardiac CAB Only record manually (roughly 500 fields)? 60 minutes 90 minutes 120 minutes 240 minutes

8. ARMUS Automation Technologies ADT Interface : Templated data extract to automatically import admission, discharge, and transfer data from Hl7 messages Clinical Data Importer : Custom data extract from EMR to prepopulate fields for abstraction cases (e.g., demographics and discreate clinical data points.). The number of fields that can be prepopulated varies based on data availability in the EMR. Ideal for importing discrete data that can be directly populated. Abstraction AI : Suggests an answer for each registry field based on patient note data. Ideal for populating fields that are not discretely documented in the electronic medical record (EMR). Basic Automation ADT Interface Advanced Automation ADT Interface Clinical Data Importer Abstraction AI

9. AI Chart Abstraction Tool Completed STS Registry Data PATIENT NOTES Fine-tuned Generative AI chart abstraction model* + New Patient Admission ^ LLM-based search:   for relevant notes or note segments related to a registry question PATIENT NOTES Suggested Answer with Supporting evidence * Fine-tuned on ~6% of available data, retraining on full data set Faster abstraction with higher quality ^ LLM- Large Language Model

10. ARMUS Chart Abstraction Tool with AI The-AI Suggested Answer is intended to supplement your review of all relevant records related to the episode. Definitions & Suggestions Definitions: Keyboard Shortcuts Registry Definitions Coding Instructions AI Suggestions: Suggested Answers to supplement abstractors chart review Validations Warnings for missing or out of range data Illegal for critical missing or out of range data 

11. ARMUS Chart Abstraction Tool with AI RF-Hypertension Missing The-AI Suggested Answer is intended to supplement your review of all relevant records related to the episode.

12. Analytic Insights to Drive Improvement High-value analytic insights in the HYBRID Platform, including benchmarking, to identify and drive outcomes

13. CRStar for Cancer Registries

14. CRStar: Who We Serve Patient Serving Staff Oncology Data Specialists (ODS-C) Oncology directors  Oncology service line partners Clinicians and Research Administrators Hospital-based Cancer Registries Large health systems Academic facilities Smaller community hospitals Small state reporting-only facilities

15. Integration and Interoperability Intelligent Abstracting and Casefinding Accreditation Tracker On-demand Education and Support Enterprise Reporting and Analytics Reporting to Support Oncology Service Line Initiatives Driving Improvement Transforming cancer data into patient experience to improve outcomes

16. Automated Data Integrations Improved Casefinding, Data Abstraction, and Follow-Up Interfaces Inbound Interfaces EMR Pathology Medical Oncology Data Warehouse Navigation Tumor Boards Precision Medicine Real time / Pre-scheduled HL7 API SFTP Outbound

17. NLP Reportability Solution Natural Language Processing (NLP) for more efficient and productive casefinding Why? Organizations spend an average of 15% of their time on casefinding Concurrent abstracting and reporting is needed for regulatory requirements and clinical research, so efficiency is vital What? Uses NLP to analyze free-text pathology reports and extract information to identify a case’s reportability Replaces manual methods Provides queues to manage workflows and case assignment

18. Poll Question #2 How many of your Oncology Programs are accredited by or have plans to seek accreditation from: Commission on Cancer  (COC) National Program for Breast Centers (NAPBC) National Program for Rectal Cancer (NAPRC) All of the above

19. Accreditation Management NAPBC NAPBC Chart Audits Continuum of Care Monitoring Quality Measures Data Analytics capabilities NAPRC NAPRC Chart Audits Continuum of Care Monitoring Quality Measures Data Analytics capabilities CoC CoC Chart Audits Continuum of Care Monitoring Tumor Board Management Clinical Trial Management Quality Measures Data Analytics NAPBC NAPBC Chart Audits Continuum of Care Monitoring Quality Measures Data Analytics NAPRC NAPRC Chart Audits Continuum of Care Monitoring Quality Measures Data Analytics National Accreditation Program for Breast Centers National Accreditation Program for Rectal Cancer Commission on Cancer

20. Insights to Drive Improvement Site Distribution Treatment Analysis Automated Survival Statistics Registry Management Custom Data Visualizations Data Quality Metrics Dashboard KPIs

21. Benchmarking with State and National Standards

22. Quality and Research

23. Technology-Enabled Managed Services (TEMS) A bundled services offering that further reduces the cost of registry participation

24. How TEMS Works A TEMS partnership transfers ownership and responsibility for data abstraction and registry submission to Health Catalyst. People. We “rebadge” team members to be Health Catalyst team members, assuming responsibility for pay, benefits, and training. Technology. We utilize tools to enable more efficient and accurate extraction and submission of data. Processes. Our project management, analytics, and improvement frameworks gives our team members a proven methodology as their time is freed up to focus on outcomes improvement.

25. Insights to Impact Community Health Network participates in the National Cardiovascular Data Registry. Quality metrics from the registry revealed an area for improvement: reducing the risk of bleeding after coronary intervention. Analytics and chart abstractors from Health Catalyst brought insights on how to improve . Insights and improvement Impact Implemented standard order sets to meet and document key requirements for PCI patients, with peer-to-peer education provided by physician champions Clarified the patient inclusion standards and engaged with the registry to clearly define the criteria for bleeding events 42.3% Relative increase in PCIs using a radial approach for lower bleeding risk $1.8 million Cost savings as a result of these improvements 58% Relative reduction in bleeding complications over 3 years and 18% reduction in LOS over 24 months Formed an interdisciplinary work group that used registry data insights to identify and resolve key care gaps, including time capture issues for point-of-care testing

26. Sheila Fairless | Clinical Services Director, ARMUS Melanie Rogan | Solutions Architect, CRStar Allie Coronis | SVP of Tech Enabled Managed Services Dan Samarov | VP of Data Science Solutions Alora Martin | Webinar Program Manager [email protected]

27. Thank You!