Annotation Guideline for Free-Text Activity Functioning Information: Mobility

Creation of the guideline was driven by a use case involving a collaborative agreement with the United States (U.S.) Social Security Administration (SSA) to leverage clinical natural language processing to improve the disability determination process. The research purpose was to develop tools that automate the detection of relevant functional information in electronic health records (EHR) to support medical record review by adjudicators. This guideline aims to standardize mobility information documented in free-text EHRs and provide a process for annotating this information. These efforts resulted in the creation of a gold standard corpus, which has been used to develop computational models of functioning. Contents of the guideline include an (1) Introduction that recommends how the guideline should be used, a description of the ICF as the conceptual framework, and an explanation of how mobility functioning is conceptualized; (2) Annotation data description; (3) Annotation tool description; (4) Annotation schema, including the entity, attribute, and value definitions, ICF codes and SNOMED clinical terms codes used, and their definitions and operationalization with examples of annotated text; (5) General annotation considerations; (6) Domain specific considerations; (7) Edge or difficult cases; (8) Acknowledgements; (9) References, and (10) four Appendices including an example of an annotated note and a troubleshooting guide. This guideline was developed by clinical rehabilitation domain experts with backgrounds in occupational therapy, physical therapy, and primary care medicine, and individuals with public health and data science expertise. The International Classification of Functioning, Disability, and Health (ICF) is used as a framework for this work and aims to develop analytic tools for clinical natural language processing and vocabularies of function.