In this study, we analyzed real-world patient claims data and showed the latest trends of medical costs, relapse rates, and rates of each category of medical treatment for MS in Japan. The MDV database is one of the largest claims databases in Japan, and it enabled efficient analysis of the nationwide trends in the treatment of MS. We extracted data for 4,374 MS patients from the dataset, and this is a much larger study population than that included in our previous study [12]. In 2014, the study population MS patients cover approximately 15% of the total number of MS patients in Japan [13]. The total medical cost for MS increased rapidly from 2012 to 2016, mainly due to increases in DMT cost. Concomitantly, the relapse rate as well as rate of hospitalization almost constantly decreased during the study period. Although our data did not reveal a direct causality between DMT and reduction of the relapse rate, it is demonstrated that disease progression in MS has decelerated, with longer time in remission correlated with the prevalence of DMT use.
Increase of medical costs for MS in Japan
Our previous analysis showed that the medical cost PPPM for MS patients from October 2013 to September 2014 was \93,542 (US$840.8 or €771.6), of which \45,284 (US$407.0 or €373.6) was outpatient DMT cost [12]. This study updated these data and showed the latest trends for categories of medical costs.
The total medical cost for MS did not dramatically change from 2008 to 2012, but increased rapidly thereafter. The total increment in cost during the study period was approximately \15,200 (US$137 or €125) PPPM, corresponding to 17% of medical cost in 2008 (Table 3). Further, outpatient DMT cost sharply increased after 2012 and was more than twice as high by 2016. Since other medical costs did not show such drastic changes, the increment in total medical cost was mostly attributed to the growth in outpatient DMT cost. In addition, the percentage of MS patients treated with DMT rapidly increased after 2012 (35% in 2012 to 44% in 2016; Table 2), suggesting that DMT is more commonly prescribed to MS patients, including those in early disease stages due to increased recognition of its benefit [14]. In addition, the first oral DMT drug, fingolimod, was launched in November 2011 in Japan. Fingolimod now sees widespread use, comprising nearly one third of the total DMT prescription in Japan [15]. The rapid increase in DMT cost from 2012 may, at least partially, be attributable to the propagation of fingolimod.
The ratio of DMT cost to total MS-related cost in Japan increased from 27% to 51% during the study period (Figure 2). In the United States, the cost of DMT has increased substantially over the last decade and currently constitutes more than 60% of the total MS-related cost [7, 8]. More than 67% of all MS-affected patients in the United States are estimated to be treated with DMT [8]. The prescription of DMT to patients with MS is likely to increase further in Japan. In addition, the number of DMT drugs available in Japan is expected to increase; for example, dimethyl fumarate, a new oral DMT, was approved in December 2016. Therefore, the market size as well as economic impact of DMT is likely to grow further in Japan and approach the situation in the United States.
Relapse and medical treatment rates
In Japan, the cost of DMT has markedly increased since 2012. As the unit price of each DMT did not significantly change during the study period, either or both the supply of total DMT or/and the ratio of DMT with higher price have expanded. Although the number of DMT prescriptions did not substantially change during the study period (Table 4), days supplied per prescription increased by nearly 11 days, leading to the increase of total DMT supplies (Table 4). This partly accounts for the increase in DMT cost. In Japan, newly launched drugs cannot be prescribed for more than 14 days at a time. For fingolimod, this limitation was lifted in November 2012, and long-term prescription was permitted. In addition, natalizumab, which is normally administered every 4 weeks, was launched in June 2014. As prices of fingolimod and natalizumab are higher than prices for other conventional DMT drugs, such as IFN-β, increased prescription of these new drugs may also account for the increase of DMT cost without any change in the prescription frequency.
Concurrent with the increase of DMT cost and usage, the rate of relapses per month declined by approximately 40% (from 0.032 to 0.019) during the study period (Table 4, Figure 3). Together with the decreased relapse rate, rates of hospitalization and relapse treatment also decreased. Although our data do not show any direct causality, the observed relationship suggests that DMT drugs reduced the incidence of relapse, which led to decreased hospitalization and relapse treatment. However, we cannot exclude the possibility that other factors, such as early diagnosis of MS, may have contributed to the reduction of the relapse rate. The DMT cost as well as supply increased sharply after 2012 (Table 2, Table 4), whereas the relapse rate constantly decreased from 2009 to 2016 (Figure 3). This suggests that DMT alone may not have affected the relapse rate, and that other factors, such as rate of patients with mild MS symptoms, are also likely to be involved.
The reduction of inpatient cost (Table 3 and Figure 2) is explained by the fall in relapse rate and consequent reduction of the number (or days) of hospitalizations. Despite significant decrease in inpatient cost, the increase in DMT cost exceeded this decrement, leading to a substantial increase in total MS-related medical cost.
Interestingly, we found that adherence of MS patients to DMT increased after 2011 (Figure 3), probably due to the launch and propagation of fingolimod and natalizumab that were more convenient for patients. This may have contributed to success in treatment, leading to longer remission and less hospitalization of MS patients. Although these new drugs are costly, their utilities are recognized, and increased usage of these drugs may have inflated the recent growth of DMT cost. In the United States, expansion of DMT costs especially due to newer DMTs is a serious social problem. Although an increase in the total medical costs for MS patients is not large so far under universal health insurance in Japan, we need to continuously watch cost-effectiveness of newer DMTs awaiting clinical launch.
Medical costs for patients with relapsing MS
Medical costs PPPM in the relapse month were 3.6-fold higher than the average medical cost PPPM for all MS patients (Table 5). This difference was mostly attributable to the difference of inpatient cost. The inpatient cost in the relapse month was approximately 10 times higher than that of the average monthly medical cost for MS. As relapse is often associated with functional impairment of patients, many patients require hospitalization for administration of acute relapse treatment. Therefore, reduction of the relapse rate not only contributes to improved quality of life for the patient but also helps reduce hospitalization costs [16].
Limitations
Since the database included data from DPC hospitals, treatment outside the hospital were not captured. The size of the database has expanded rapidly due to the increase in DPC hospitals included in the MDV database. Therefore, the patient background may not be consistent in each calendar year; for example, the average age of the patient varied greatly before 2010 (Figure 1B). As patient data were anonymized before receipt, the history of each patient cannot be tracked either before the first visit to the hospital or after changing hospitals. For calculating costs or numbers PPPM, we defined the observation period as the length of time from the first to the last claim data for each patient and used this as the denominator. As this observation period is shorter than the insurance coverage period for each patient, the calculated value PPPM may have been an overestimate. Since only the relapses treated at the hospital that were in the database could be captured, the frequency and costs for relapses might have been underestimated.