A Principal Component Analysis Approach to Estimate the Disability Status for Patients with Multiple Sclerosis Using Japanese Claims Data

Introduction Claims databases are preferred for research on multiple sclerosis (MS) as this condition is characterized by low prevalence and long disease course. However, Japanese claims databases contain no information on disease severity or disability status of MS. Here, we aimed to explore the possibility of utilizing a principal component analysis (PCA) to estimate MS severity using a Japanese claims database. Methods An MS severity score was developed using a PCA. Factors related to functional systems for Expanded Disability Status Scale (EDSS) and higher disease severity (74 diagnoses, 68 drug prescriptions, and 77 procedures) were extracted from the claims database (April 2008–August 2018). The score (PC1 score) was developed for each patient-year—each year from the first diagnosis (excluding the year of the first diagnosis), based on the first principal component of the included factors. Finally, the patient-years were classified into quartiles based on the PC1 score, and demographic information and medical status were analyzed. Results The database contained 7067 patients with MS. The highest score group had a higher mean age (55.4 ± 0.2 [mean ± standard error] years), lower percentage of women (64.4 ± 0.7%), and longer mean disease duration from first diagnosis (8.1 ± 0.1 years) than the lowest score group (43.3 ± 0.2 years, 68.4 ± 0.8%, and 6.0 ± 0.1 years, respectively). In addition, the PC1 score of each patient positively correlated with disease duration from diagnosis. Conclusion We developed a PC1 score to indicate MS severity using information from a Japanese claims database. Since changes in demographic features we observed are consistent with findings of previous research, this score might represent MS severity to some extent. Further research is necessary to validate this score with clinical measurement of disability such as the EDSS.