JMIR Publications

ABSTRACT

Background:

Gaining insights from patients has become an important topic in pharmacovigilance because it would reveal new findings that cannot be obtained from healthcare databases.

Objective:

Our objective was to show a use case of incorporating patient-generated data in pharmacovigilance: understanding the epidemiology and burden of illness in Japanese patients with systemic lupus erythematosus (SLE).

Methods:

Using two independent databases, we looked at data on SLE, an autoimmune disease that substantially impairs quality of life. To understand the epidemiology of SLE, we analyzed a Japanese health insurance claims database. To understand the SLE burden, we analyzed textual data collected from Japanese disease blogs (tōbyōki) written by patients with SLE. Natural language processing was applied to these texts to identify frequent patient-level complaints. Term frequency-inverse document frequency was used to explore patients' burden during treatment. We explored the health-related quality of life based on patient descriptions.

Results:

We analyzed 4,694 and 635 patients with SLE in the healthcare database and the tōbyōki blogs, respectively. Analysis of the healthcare database showed the prevalence of SLE and the treatment details. Textual data analysis showed that pain-related words became more important after starting treatment, suggesting patients' experienced burden. We also found a continuous increase in patients’ references to mobility and self-care over time, which indicated increased attention to physical disability due to disease progression.

Conclusions:

A classical medical database represents only a part of a patient's entire treatment experience, and analysis using such a database alone cannot infer patient-level symptoms or patient concerns about treatments. This study showed that analysis of tōbyōki blogs could add information on patient-level details for advancing patient-centric pharmacovigilance.