STRUCTURED REPORTING - PATHOLOGY
There is an emerging demand for pathologists to move to a form of structured reporting as the advantages have been reported as enhancing the accuracy and completeness of reports and improving the efficiency of using the reports. Recently a Roundtable for Structured Pathology Reporting was held by Cancer Australia, the Cancer Institute NSW, and the Royal Australasian College of Pathologists to identify progress and pathways to increase the amount of structured reporting by pathologists. We currently have a technology that takes free text and extracts structured items such as SNOMED CT codes, socio-demographics, qualifiers, negated clauses, and weights & measures. This is the basis of an automatic system for computing structured reports from narrative reports. While some clinical specialties have already designed structured reports (breast cancer, colorectal cancer, melanoma) they do not have methods for enforcing their use, nor verifying their comprehensiveness or accuracy, and none of their historical data can be used for research until an automatic conversion mechanism is created.
The aim of this project is to convert data written in pathology reports as prose into data elements of a structured report.
1. to provide pathologists with a means of verifying that their reports contain all the necessary and sufficient data required by the consulting physician.
The methodology is:
1. Compile a collection of pathology reports for any specific conditions.
2. Review the reports to identify the textual content that is relevant for the structured report.
3. Implement a processing system that analyses the language of the reports and identifies the contents required for the structured report.
4. Run a series of tests to assess the accuracy of the computational process.
5. Provide an application that can be used by the pathologists to verify their structured reports and forward them to the appropriate person.
6. Test the system with trusted staff and develop methods for operationalising the system.7. Train all staff to use the system.