Mapping of UK Biobank clinical codes: Challenges and possible solutions

Abstract

Objective The UK Biobank provides a rich collection of longitudinal clinical data coming from different healthcare providers and sources in England, Wales, and Scotland. Although extremely valuable and available to a wide research community, the heterogeneous dataset contains inconsistent medical terminology that is either aligned to several ontologies within the same category or unprocessed. To make these data useful to a research community, data cleaning, curation, and standardization are needed. Significant efforts to perform data reformatting, mapping to any selected ontologies (such as SNOMED-CT) and harmonization are required from any data user to integrate UK Biobank hospital inpatient and self-reported data, data from various registers with primary care (GP) data. The integrated clinical data would provide a more comprehensive picture of one’s medical history. Materials and methods We evaluated several approaches to map GP clinical Read codes to International Classification of Diseases (ICD) and Systematized Nomenclature of Medicine Clinical Terms (SNOMED CT) terminologies. The results were compared, mapping inconsistencies were flagged, a quality category was assigned to each mapping to evaluate overall mapping quality. Results We propose a curation and data integration pipeline for harmonizing diagnosis. We also report challenges identified in mapping Read codes from UK Biobank GP tables to ICD and SNOMED CT. Discussion and conclusion Some of the challenges–the lack of precise one-to-one mapping between ontologies or the need for additional ontology to fully map terms–are general reflecting trade-offs to be made at different steps. Other challenges are due to automatic mapping and can be overcome by leveraging existing mappings, supplemented with automated and manual curation.

Opaque Ontology: Neuroimaging Classification of ICD-10 Diagnostic Groups in the UK Biobank

“…Furthermore, these results are consistent with prior work highlighting challenges with UK Biobank clinical codes [20]. Specifically, Stroganov et al highlighted mapping issues of available hospital inpatient and general practitioner information onto the ICD-10 ontology [20]. The limited classification accuracy for most ICD-10 diagnostic groups observed in this study furthermore confirms previous work suggesting that the identification of biomarkers of psychopathology is not feasible without increased efforts to address suboptimal phenotypic reliability [11].…”

“…This conclusion corroborates past work on low inter-rater reliability in ICD-10 coding, which may be driven by many factors including patient-provider communication, administrative decision chains, and insurance incentives [8]. Furthermore, these results are consistent with prior work highlighting challenges with UK Biobank clinical codes [20]. Specifically, Stroganov et al highlighted mapping issues of available hospital inpatient and general practitioner information onto the ICD-10 ontology [20].…”

“…Furthermore, these results are consistent with prior work highlighting challenges with UK Biobank clinical codes [20]. Specifically, Stroganov et al highlighted mapping issues of available hospital inpatient and general practitioner information onto the ICD-10 ontology [20].…”

“…First, ICD-10 diagnostic groups in the UK Biobank were automatically mapped from available hospital inpatient and general practitioner data sources. These automated mappings are known to be imprecise [20], suggesting that the reliability of ICD-10 diagnostic information in the UK Biobank may be lower than in other research and clinical settings. Second, although we tested our 14 classifications across a range of feature sets and classifiers, it is possible that higher classification accuracy can be achieved using alternative feature sets or imaging modalities (e.g., clinically sensitive modalities such as susceptibility weighted imaging) and/or by using more sensitive classifier architectures (e.g., leveraging deep learning).…”

Development of a Method for Automatic Matching of Unstructured Medical Data to ICD-10 Codes

“…For instance, in method 1, errors occur during translation from ICD-10-CM to SNOMED, while in method 2, issues arise from incomplete coverage and accuracy of the WHO map from ICD-10-CM to ICD-11. Additionally, studies focus on translating Read codes [5]. The examination revealed that achieving a lossless correspondence between Read codes and the ICD10-CM disease coding system is unattainable due to disparities in ontologies and insurmountable ambiguity in the utilization of the Read code system by medical professionals [5].…”