BetaEntity Annotation Prototype
← Back to diseases

Annotated abstract

1-016 Evaluating the concordance of ATTR amyloidosis recording between primary and secondary care electronic healthcare data using general practice survey feedback

heartjnl · 2025-08-13 · canonical JSON source

29 visible annotations · policy: published · automated confidence ≥ 75.00%

Document resource

Introduction Accurate recording of amyloidosis is crucial for patient care and research. Typically, general practices (GPs) receive discharge letters or information from specialists and record them into the GP system; only formally coded data is available in the Clinical Practice Research Datalink (CPRD) database for research. Discrepancies in coding between primary and secondary care lead to inconsistent records and unreliable data. This study evaluates the concordance and accuracy of transthyretin amyloidosis (ATTR) in CPRD by comparing them with GP-held records using surveys.Methods We conducted a retrospective analysis by comparing CPRD data linked to Hospital Episode Statistics (HES) with GP-held records using GP survey responses. We assessed records of patients with coded amyloidosis (ICD-10 E85*) from 2000–2021 who were still alive and registered at a GP as of October 2024. Queries on 455 patients were obtained from 170 practices through the survey: Q1: Has this patient been diagnosed or treated for amyloidosis by a hospital specialist at any time? Q1A: If yes [to Q1], which specific amyloidosis information is recorded in primary care? (descriptive diagnoses by keywords from hospital correspondence, see figure 1 for options) Q2: Which amyloidosis diagnosis is assigned in the GP-held record? (coded by GPs using SNOMED-CT/Read codes, see figure 1 for options)Results Of the 455 patient queries received, 97 (21%) were excluded due to either a lack of access to patient records at the time of completing the survey or missing responses to Q1. Among the 358 valid responses, 293 (92%) had a record of diagnosis or treatment for amyloidosis by a specialist documented in GP-held records. A total of 89 (30%) records included references to cardiac amyloidosis (CA), with 29 (10%) specifically mentioning ATTR-CM (ATTR-cardiomyopathy; figure 1). Among the 89 GP-held CA records, 99%, 49%, 24%, and 23% had corresponding entries for unspecified CA, specific CA, unspecified ATTR-CM, and specific ATTR-CM codes, respectively, in the CPRD database. For the 29 GP-documented references to ATTR-CM in response to Q1A, 100% had a corresponding unspecified CA record. However, only 28% and 28% had a record for either unspecified ATTR-CM or specific ATTR-CM codes, respectively in the CPRD database. Among the 89 and 29 GP-held records for CA and ATTR-CM, 66% and 62%, respectively, also contained entries for unspecified amyloidosis polyneuropathy (PN) (figure 2). This may suggest the presence of mixed phenotypes.Conclusion The findings show inconsistencies in amyloidosis coding, especially with specific codes; some of the codes are now obsolete, while new codes are added over time. This highlights the need for better coding practices. Standardising the recording and data integration between hospital and GP practices could enhance patient care coordination and provide more accurate disease tracking across healthcare systems.Abstract 1-016 Figure 1Distribution (in percentages) of GP-recorded amyloidosis codes versus specific amyloid clinical scenariosNote: **Cells contained a small number of responses (≤5). Q1A is from specialist or hospital information; Q2 is based on SNOMED-CT/Read codes recorded by the GP.Abbreviations: ATTR, transthyretin amyloidosis; GP, general practice; PN, amyloidosis polyneuropathy.Abstract 1-016 Figure 2Distribution (in percentages) of amyloidosis subtypes recorded in CPRD-HES linked data versus specific amyloid clinical scenariosNote: **Cells contained a small number of responses (≤5). An algorithm was used to define unspecified and specific codes. The unspecified codes encompass a combination of selected ICD-10 E85.x codes for suspected cases of:• Cardiac amyloidosis, identified by E85.x combined with cardiology-related codes (I43.1, I42.0, I42.2, I42.5, I42.8, I42.9, I50, I50.0, I50.1, I50.9, I11.0, I25.5).• Amyloid polyneuropathy, identified by E85.x combined with neuropathy-related codes (G61.9, G62.9, G63.3, G63.5, G63.6, G63.8).• Broadly defined amyloidosis subtypes, based on SNOMED-CT/Read codes.Specific codes refer to specific SNOMED-CT/Read codes with clear descriptions of the phenotype.Abbreviations: ATTR, transthyretin amyloidosis; ATTRv, hereditary transthyretin amyloidosis; ATTRwt, wild-type transthyretin amyloidosis; CM, cardiomyopathy; PN, amyloidosis polyneuropathy.