Validation of methods to identify people with idiopathic inflammatory myopathies using hospital episode statistics

摘要

Objective Hospital episode statistics (HES) are routinely recorded at every hospital admission within the National Health Service (NHS) in England. This study validates diagnostic ICD-10 codes within HES as a method of identifying cases of idiopathic inflammatory myopathies (IIMs). Methods All inpatient admissions at one NHS Trust between 2010 and 2020 with relevant diagnostic ICD-10 codes were extracted from HES. Hospital databases were used to identify all outpatients with IIM, and electronic care records were reviewed to confirm coding accuracy. Total hospital admissions were calculated from NHS Digital reports. The sensitivity and specificity of each code and code combinations were calculated to develop an optimal algorithm. The optimal algorithm was tested in a sample of admissions at another NHS Trust. Results Of the 672 individuals identified by HES, 510 were confirmed to have IIM. Overall, the positive predictive value (PPV) was 76 and sensitivity 89. Combination algorithms achieved PPVs between 89 and 94. HES can also predict the presence of IIM-associated interstitial lung disease (ILD) with a PPV of 79 and sensitivity of 71. The optimal algorithm excluded children (except JDM code M33.0), combined M33.0, M33.1, M33.9, M36.0, G72.4, M60.8 and M33.2, and included M60.9 only if it occurred alongside an ILD code (J84.1, J84.9 or J99.1). This produced a PPV of 88.9 and sensitivity of 84.2. Retesting this algorithm at another NHS Trust confirmed a high PPV (94.4). Conclusion IIM ICD-10 code combinations in HES have high PPVs and sensitivities. Algorithms tested in this study could be applied across all NHS Trusts to enable robust and cost-effective whole-population research into the epidemiology of IIM. Lay Summary What does this mean for patients? Information about every hospital admission in the National Health Service in England is collected in a database called hospital episode statistics (HES). The main diagnosis and any co-existing medical problems are recorded by professional coders, according to a set of rules called ICD-10 codes. Myositis is a group of rare conditions causing muscle inflammation, which are complex to diagnose and to code. There are eight possible ICD-10 codes that a coder might use to record an admission in someone with myositis. In this study, we looked at how accurately these codes are used. We compared lists of patients identified by HES with patients' hospital records. We developed methods such as combining ICD-10 codes to improve accuracy. We found that overall, 89 of people identified in HES genuinely had myositis, and 84 of people known in their local hospital to have myositis were picked up in HES. Research in rare diseases, such as myositis, is challenging because it is hard to find enough people to study and to obtain enough funding. This research shows how automatically collected data can be used to identify potential myositis cases, which will provide new opportunities for cost-effective research in this rare group of patients.