Quality assurance in anatomic pathology: automated SNOMED coding

Area of Science:

• Medical Informatics
• Anatomic Pathology
• Health Information Systems

Background:

• Anatomic pathology laboratories face challenges with efficient and accurate coding.
• Manual coding methods can be time-consuming and prone to variability.
• Implementing automated coding systems requires careful setup and optimization.

Purpose of the Study:

• To detail the implementation of the PathNet autocoder in a high-volume anatomic pathology setting.
• To identify and address software coding errors to improve system performance.
• To evaluate the efficiency of electronic case retrieval compared to manual methods.

Main Methods:

• Case study methodology focusing on the PathNet autocoder (Cerner Corporation).
• Development of procedures to enhance system performance and identify coding error classes.
• Creation of a Cases Between Errors (CBE) index to measure coding accuracy.
• Comparative analysis of manual versus electronic case retrieval for efficiency.

Main Results:

• Initial setup challenges were overcome through procedural improvements.
• Four classes of software coding errors were identified and addressed.
• The Cases Between Errors (CBE) index showed a sharp increase, indicating improved accuracy.
• Electronic case retrieval demonstrated significant time savings and reduced variability compared to manual methods.

Conclusions:

• The PathNet autocoder implementation was successful in improving coding accuracy and efficiency.
• The developed techniques for error identification and performance improvement are transferable.
• Automated coding systems offer substantial benefits for anatomic pathology laboratory operations.
• This study provides a model for other institutions implementing similar coding systems.