The current study investigated the proportion of HBV screening in patients with rheumatoid arthritis prescribed antirheumatic drugs. No studies have been conducted using large Japanese medical information databases since 2014. The proportion of patients underwent tests for HBsAg, HBsAb, and HBcAb was only 9.7%, which was similar to the insufficient proportion identified in a previous survey [20]. It should be noted that the proportion of HBV screening (HBsAg, HBsAb, and HBcAb) in the most recent two-year period (2021–2022) had approximately doubled compared with the previous years (2016–2020).

Japanese guidelines recommend a combination of antigen and antibody testing and HBV screening prior to the initiation of immunosuppressive therapy, including chemotherapy and antirheumatic drugs, which are covered by insurance. However, the test for HBsAg is the most common, and the proportion of patients who tested for antigens and antibodies is low. In the US, the Centers for Disease Control and Prevention (CDC), American Gastroenterological Association (AGA), and American Association for the Study of Liver Diseases (AASLD) recommend testing for both antigens and antibodies, as in Japan. Specifically, in the USA, 43.4% of patients underwent a combination test [18]. In contrast, in Taiwan, the testing of both antigens and antibodies both is recommended; however, most tests have been performed for HBsAg only, and a combination test was only performed in 16.3% of patients [18]. In Taiwan, prophylactic administration of antiviral drugs is not covered by insurance for patients previously infected with HBV, which may be a reason for the low proportion of antibody testing [18]. In Japan, most of the tests have been performed for HBsAg only, despite the fact that the prophylactic treatment of previously infected patients with HBV-DNA detected above a certain amount is covered by insurance.

The proportion of patients with rheumatoid arthritis who were tested before immunosuppressive therapy was low, but the proportion of Japanese patients undergoing cancer chemotherapy tested for HBV antigen and antibodies was 41.3% in a 2014–2015 database [26]. The incidence of reactivation after chemotherapy was reportedly 20–50% in HBsAg-positive patients and 0.3–9.0% in previously infected patients (HBsAg negative/HBsAb or HBcAb positive) [27,28,29]. Despite a similar risk of HBV reactivation, patients with rheumatoid arthritis are undertested for HBV antigens and antibodies at a proportion of 9.7% and may be overlooked when compared with that in patients with cancer. Considering that rheumatoid arthritis is a chronic disease frequently treated with combinations of immunosuppressive drugs over a prolonged period, this is an important issue and needs to be addressed.

In our study, logistic regression analysis revealed that male patients, initial outpatient prescription, the prescribing department being rheumatology/collagen disease, and concomitant use of steroids were associated with HBV screening. Male patients and patients with concomitant use of steroids have been reported to increase the proportion of HBV screening [18], which is consistent with the results of the present study. In Japan, systemic administration of steroids for more than 2 weeks has been recognised as a risk factor for HBV reactivation, even if administered alone [30]. Although the history of hospitalisation affects the proportion of tests performed [18], the fact that the initial prescription is given in an outpatient clinic did not seem to substantially impact the proportion of tests performed. Comparing these results with previous reports was challenging because we had to determine whether the patients were treated as outpatients or inpatients at the time of their first prescription. Moreover, the fact that the prescribing departments were rheumatology and orthopaedics, which treat many patients with rheumatoid arthritis, affected the proportion of tests performed. A possible factor is that the JCR has issued an alert [5] regarding HBV reactivation, which is also described in the guidelines for treating rheumatoid arthritis in Japan [31]. By classification, patients prescribed tsDMARDs were more frequently tested for HBsAg, HBsAb, and HBcAb than those prescribed other drugs. tsDMARDs have been approved for use in patients who show an insufficient response to existing antirheumatic drugs. Although the actual reason is unknown, long-term safety has not been fully established, which may have an impact on the alerts regarding HBV screening tests in response to the guidelines. In addition, the differences between injectable and oral drugs did not affect the proportion of tests performed in the current study because most of the injectable drugs were subcutaneous formulations, and all tsDMARDs, which were associated with a high proportion of tests performed on patients, were oral drugs.

In the current study, although the proportion of HBV screening among patients with rheumatoid arthritis remained low, it has been rising since 2018. A survey on the prescribing trend of antirheumatic drugs in Japan [32] has revealed that several csDMARDs and approximately 20% of bDMARDs are prescribed as the first-line treatments; therefore, we consider that the lack of HBV screening for patients starting immunosuppressive therapy for the first time is most problematic. Among bDMARDs, IL-6 inhibitors that have stronger immunosuppressive effects than anti-inflammatory effects [33], had a higher screening proportion than those of TNF-α inhibitors. Additionally, the proportion of HBV screening at the initial prescription of tsDMARDs, a new class of antirheumatic drugs, was higher than that of current antirheumatic drugs (i.e. csDMARDs and bDMARDs); however, adequate screening was not performed. The usage of new biological agents for patients with rheumatoid arthritis is increasing, along with the number of immunosuppressive agents related to HBV reactivation. Immunosuppressive therapy for treating rheumatoid arthritis is used in several patients and administered over the long term. Furthermore, the proportion of patients tested for HBV before chemotherapy was much lower than that before immunosuppressive therapy, which needs to be addressed immediately to improve the proportion of patients with rheumatoid arthritis tested for HBV before immunosuppressive therapy.

As a solution, simplifying the process of test orders and education have increased the proportion of HBV screening [23], and the introduction of an automated decision support system within the electronic medical record system has resulted in more efficient HBV screening [34]. This suggests improvements in the process and system of educational intervention and test ordering for healthcare professionals.

Our study has several limitations. First, ascertaining all medical facilities attended by the patients was difficult owing to the characteristics of the database used in this study. The JMDC hospital-based administrative claims database was unable to identify whether the same individual had visited multiple hospitals. Therefore, there may be a history of previous HBV or RA treatment, or an HBV screening test may have been performed but not recorded in the database. Second, the accuracy of disease name as rheumatoid arthritis was not validated. Therefore, we used a previously reported definition of rheumatoid arthritis [20].