Background: Understanding real-world usage of high-cost drugs is crucial to support planning, adoption of innovation and reduce unwarranted variation in treatment. Hospital Episode Statistics (HES) contain diagnoses (coded by ICD-10) and procedures/treatments (coded by OPCS) for all daycase or inpatient care in England. However, OPCS codes are not specific for individual drugs, for example X921 (cytokine inhibitors band 1) includes rituximab (RTX) and 15 other drugs.

Objectives: We aimed to validate the accurate identification of patients treated with RTX for ANCA-associated vasculitis (AAV) using HES data.

Methods: We used data from the National Congenital Anomaly and Rare Disease Registration Service (NCARDRS) at Public Health England and their legal permissions (CAG 10-02(d)/2015). We extracted records from HES of all patients treated at two hospitals during financial year 2018/19 who ever had a coded diagnosis of granulomatosis with polyangiitis (GPA, M313), eosinophilic granulomatosis with polyangiitis (EGPA, M301), microscopic polyangiitis (MPA, M317), polyarteritis nodosa (PAN, M300) or arteritis unspecified (I776). Where people had multiple diagnoses of vasculitis, the most specific was considered their diagnosis. Enabled by data sharing agreements, we reviewed hospital records of those patients to validate diagnoses and whether X921 reliably identified RTX. We report the positive predictive value and sensitivity of the coding for X921 and GPA/EGPA/MPA for identifying people with AAV who are treated with RTX.

Results: At Trust 1 records ever coded with GPA/EGPA/MPA identified 74 people, 69 of whom had AAV confirmed in their medical notes. Among these 74 patients there were 59 episodes coded with X921 procedure codes, of which 56 correctly identified a RTX infusion given for AAV. A total of 64 RTX infusions were given to people with AAV – 3 missed infusions were X921 procedures in patients who had coded diagnoses of PAN or I776 but never GPA/EGPA/MPA and 5 infusions were not coded as X921.

The same analysis at Trust 2 identified 46 people, 44 of whom had AAV confirmed in their medical notes. Among patients identified with AAV there were 17 episodes coded as X921, of which 15 correctly identified a RTX infusion. A total of 23 infusions were given to people with AAV: 4 infusions were X921 procedures in patients who had coded diagnoses of PAN or I776 but never GPA/EGPA/MPA, and 4 infusions were not coded as X921.

Conclusion: HES data identified patients treated with RTX for AAV with a PPV of 93.4% (85.3-97.8) and sensitivity of 81.6% (71.9-89.1). This demonstrates the utility of national data to identify people receiving RTX for AAV. The RECORDER project, within the National Disease Registration Service plans to conduct real-world studies of the high-cost drug RTX, given for AAV, across the whole of England, and assess whether geography, demographics or socioeconomic factors influence frequency of prescription of this, and potentially other, high-cost drugs in line with the NHS long term plan.

References: [1]Ward-Platt M, Stevens S, Miller N. I18 The national congenital anomaly and rare disease registration service (NCARDRS): The first year

Acknowledgements: I have no acknowledgements to declare.

Disclosure of Interests: Cattleya Godsave: None declared, Mithun Chakravorty: None declared, Megan Rutter: None declared, Peter Lanyon Grant/research support from: P.C.L. is a recipients of a grant from Vifor Pharma. Vifor Pharma had no influence on the design, conduct or interpretation of this study., Jeanette Aston: None declared, Mary Bythell: None declared, Sarah Stevens: None declared, Fiona Pearce Grant/research support from: I have received a research grant from Vifor Pharma