ATCodeR

ATCodeR transforms free-text medication entries into the standardized Anatomical Therapeutic Chemical (ATC) Classification System to improve the standardization and analysis of systemic anti-cancer treatment data, particularly for German-speaking regions.

Key Features:

• Dictionary-based standardization: Applies a dictionary-based approach to convert unstructured medication free-text into structured ATC codes.
• Structured output format: Produces a structured data frame containing columns for antineoplastic medication, other medications, and supplementary information.
• Accuracy validation: Achieved 88.5% successful transformation of 935 tokens from an evaluation data set of 561 entries, with 8.9% failed transformations after manual review.
• Supplementary information extraction: Extracted additional information from 23% of data entries, while identifying 4.1% of entries with no usable information.

Scientific Applications:

• Oncology real-world data analysis: Facilitates research and analysis with real-world data by standardizing substance-related documentation in medical records.
• Treatment pattern analysis: Enables analysis of systemic anti-cancer treatment patterns and outcomes from unstructured clinical documentation.
• German-speaking clinical data standardization: Optimizes efficiency and maintains accuracy for anti-cancer treatment data in German-speaking regions.

Methodology:

ATCodeR applies a dictionary-based approach to transform free-text medication entries into the Anatomical Therapeutic Chemical (ATC) Classification System. The method was validated using 561 data entries comprising 935 tokens, with all tokens undergoing manual review.