2005 Scientific Session Abstracts
A Workflow Utilizing a Blend of Narrative and Synoptic Reporting, Structured Data Capture, and Clinical Decision Assistance Generate the Most Profound Results
William O’Toole, M.D (colin.murphy@mtuitive.com) John W. Murphy, MBA, mTuitive, Inc, Centerville, Massachusetts
Context : Standardized terminology and pathology report structure help clinicians to clearly and quickly recognize the most important findings in determining treatment. Synoptic reports address these issues by using published, peer-reviewed checklists.
Handcrafted narratives of anatomical pathologists are of diagnostic importance due to the unique set of events that surround each patient’s case. Report standardization must not impede flexibility to offer individual diagnostic opinions. The ideal anatomic pathology report blends synoptic elements with narrative descriptions to provide the best assessment to the clinician.
By constraining a portion of the pathology report to structured data elements, it is possible for a computer to intelligently analyze reports for clinical and research purposes. Maintaining structure in synoptic reports can clarify findings for computers, resulting in meaningful aggregate analysis.
Computers can provide valuable decision assistance to clinicians through real-time, context-sensitive knowledge delivery and error prevention to ensure accuracy and completeness of data captured at the point of decision.
Technology :
- An authoring environment is used to create report templates and business rules for knowledge delivery and error prevention
- Intellectual property from the College of American Pathologists (CAP), their Division of SNOMED® International, and other public and private sources used to define standard reports and map data to ontologies and controlled vocabularies
- Simple user interface with expert system that is usable by physicians
- Services are used to send structured reports to multiple systems automatically.
Design : We imported structured checklists developed by the CAP Pathologists in their Cancer Protocols. Templates for each protocol have been created to capture the required data sets and automatically code results in SNOMED CT®. Pathologists access the templates through a software application, xPert for Pathology that provides context-sensitive knowledge and ensures compliance with requirements. Narrative descriptions can be included in the template. The resulting pathology report is sent to the pathology system and any other database. Certain results from the pathologic findings, such as the pathology stage and histologic grade, are automatically calculated and made available to the managing physician through a similar intuitive interface. This includes computer assisted quality assurance methods and guidance for managing physicians that are integral to oncology staging.
Results : N/A
Conclusion : Error prone and costly process of transcription has been virtually eliminated. Abstraction of the required data elements from text strings has been eliminated by capturing the structure up front. Communication from the pathology report to the managing physician is improved through standard reports that allow for descriptive narrations. SNOMED CT Codes are captured consistently and accurately at the point of diagnosis and delivered to the appropriate registries.
