In:
Rheumatology, Oxford University Press (OUP), Vol. 59, No. Supplement_2 ( 2020-04-01)
Abstract:
The National Congenital Anomaly and Rare Disease Registration Service (NCARDRS) records those people with congenital anomalies and rare diseases across the whole of England. The NCARDRS Rare Rheumatology Project is establishing methodologies to identify and register people with rare rheumatic diseases. Hospital Episode Statistics (HES), comprising data from all admitted-patient NHS care in England, provides a promising source for case ascertainment. We aimed to confirm the reliability of diagnostic ICD codes for rare autoimmune rheumatic diseases within HES. Methods We used NCARDRS’ legal permissions (CAG 10-02(d)/2015) to extract cases who had at least one inpatient or daycase admission recorded in HES during 2003 to 2017 with a primary or additional diagnostic code for Kawasaki disease, Takayasu’s arteritis, Adult-onset Still’s disease, Relapsing polychondritis, Polyarteritis nodosa, Granulomatosis with polyangiitis (GPA), Microscopic polyangiitis (MPA) or Eosinophilic granulomatosis with polyangiitis (EGPA). Using Data Sharing Agreements between NCARDRS and NHS Trusts, we reviewed medical records for all cases admitted to one NHS Trust to confirm diagnoses. For ANCA-associated vasculitis (AAV), up to 20 randomly selected cases admitted to 5 Trusts with a diagnostic code of GPA, MPA and EGPA were reviewed. Results Table 1 shows the positive predictive value (PPV) of each diagnosis at one Trust: Kawasaki disease 100%, Takayasu’s arteritis 91.7%, Adult-onset Still’s disease 42.8%, Relapsing polychondritis 40.0%, and Polyarteritis nodosa 5.0%. For AAV, which had a PPV at the index Trust of 92.0%, the PPV across the 5 Trusts varied from 92.9% to 74.0%. Combining all 5 Trusts PPV for AAV was 85.9%. It was higher in MPA (PPV=98.4%) than in GPA (PPV=83.3%) and EGPA (PPV=77.4). Conclusion This is the first study to validate the coding of rare rheumatic disease diagnoses within HES data. Our results highlight heterogeneity in coding accuracy both between conditions and between NHS Trusts. Further work is needed to develop and validate algorithms for conditions where HES coding alone has poor positive predictive value, and test sensitivity. National rare disease registration will enable robust, cost-effective and reproducible population-based epidemiology of rare rheumatic diseases across England which is needed to inform service provision and aid development of new treatments. Disclosures F.A. Pearce None. M. Rutter None. B. Griffiths None. C. Mukhtyar None. R. Al-Jayoussi None. R.A. Watts None. J. Aston None. M. Bythell None. S. Stevens None. P.C. Lanyon None.
Type of Medium:
Online Resource
ISSN:
1462-0324
,
1462-0332
DOI:
10.1093/rheumatology/keaa110.035
Language:
English
Publisher:
Oxford University Press (OUP)
Publication Date:
2020
detail.hit.zdb_id:
1474143-X
Permalink