Multiple sclerosis (MS) and neuromyelitis optica spectrum disorder (NMOSD) are immune-mediated inflammatory diseases in the central nervous system (Thompson et al., 2018; Wingerchuk et al., 2015). MS and NMOSD were previously thought to be similar disease entities, but in 2004, Lennon et al. revealed anti-aquaporin-4 immunoglobulin G (AQP4-IgG) as a highly specific antibody in patients with NMO (Lennon et al., 2004). The discovery of AQP4-IgG mainly provided two benefits: accurate NMO diagnosis and specific treatments. In clinical characteristics, not only optic neuritis and myelitis, but also diencephalic and dorsal medulla lesions are common in patients with NMO, and the diagnostic criteria, including these core clinical characteristics, were proposed in 2015 (Wingerchuk et al., 2015; Misu et al., 2005).

Until recently, a certain percentage of patients with NMO were misdiagnosed with MS and were ineffectively treated with a disease modifying therapy for MS (DMT) (Solomon and Weinshenker, 2013; Solomon et al., 2019). Additionally, DMT for MS was not only ineffective but are also harmful to patients with NMO (Tanaka et al., 2009; Min et al., 2012; Kleiter et al., 2012), who should have been treated with oral steroid, immunosuppressants, and monoclonal antibodies (Pittock et al., 2021). However, the AQP4-IgG assay was limitedly available to some institutes doing in-house cell-based assay in Japan before the national insurance system approved it in 2013 (Fujihara, 2019). Therefore, some diagnosis and treatment inequalities in NMO existed until recent years. Nationwide research on the impact of changes on the actual number of admissions, treatment, and disabilities in real-world remained unavailable, except for small cohort studies, although these dramatic changes in disease concepts and treatment occurred in patients with NMO (Hyun et al., 2016; Kanikannan et al., 2019; Rattanathamsakul et al., 2020; Hamid et al., 2017).

This study investigated the trend of the number of admissions, treatment, and disabilities from 2012 to 2017, using data from the national administrative database in Japan, named the Diagnosis Procedure Combination (DPC). The DPC database contains inpatient and administrative claim data of over 47 million inpatients from 1253 hospitals since 2012, representing >60 % of all acute-care hospitals in Japan (Yasunaga, 2019). The diagnoses are coded with the International Classification of Diseases and Related Health Problems 10th Revision (ICD-10). The diagnoses in DPC data were well validated, with a specificity of >96 % (Yamana et al., 2017). The database also contains detailed medical information, such as patients’ age, sex, body weight and height, drug, treatment, length of hospital stay (LOS), discharge status, including in-hospital death, emergency transport, and the Barthel index (BI) as activities of daily living (ADL) score at admission and discharge (Wada et al., 2016).