To our knowledge, this was the first cross-sectional study to report on regional disparities in breast cancer care by focusing on DCCHs in Japan. Our analysis revealed some variability in the outcomes among prefectures in Japan with our estimates suggesting that just over half (57.7%) of breast cancer patients received initial treatment at a DCCH in 2020, a value that varied considerably among the prefectures, ranging from 15.5% to 89.8%, as illustrated in Fig. 2.

The variability among prefectures may be due to two factors: the difference in the proportion of patients registered in the HBCR among all breast cancer patients registered in the PBCR, and the difference in the proportion of patients who visited a DCCH for their initial treatment (ESM Fig. 1 [Online Resource 1]). The median proportion of patients registered in the HBCR out of all breast cancer patients was < 50% in three prefectures where most patients visited facilities not participating in the HBCR. The median percentage of patients registered in the HBCR who visited a DCCH was < 60% in five prefectures. In Okinawa, one of these prefectures, although the percentage of patients registered in the HBCR among all breast cancer patients was 63.4%, the percentage of patients registered in the HBCR who visited a DCCH was just 24.4%. According to the annual HBCR report, this is due to the geography of Okinawa, which consists of multiple islands, and the fact that breast cancer is mainly treated at “breast clinics,” which cannot participate in the HBCR [12]. Given this context, any measures aimed at improving the quality of medical care in the future should consider the actual condition and clinical situation in individual regions. This is particularly important because Japan is a country with a rapidly aging population, especially in rural areas. Therefore, it is crucial to ensure that people have access to medical care, particularly in underpopulated or remote areas, through measures such as shuttle services, telemedicine, or funded accommodation. For this, we may look to other countries like Australia, which also exhibits marked variation and disparities in breast cancer care between regional and rural areas. In particular, rural patients frequently encountered delayed treatment and had lower rates of breast-conserving surgery, which was at least partly related to limited access to specialists, travel costs, and inadequate care coordination [13]. In an effort to address such disparities, Australia has implemented initiatives to ensure people living in rural or remote areas can access the same quality of care as people in metropolitan areas. These programs include facilitating finding cancer specialists and hospitals and offering subsidized accommodation and patient-assisted travel schemes [14, 15]. In summary, when interpreting the variability among prefectures, it is important to consider that breast cancer care is provided according to regional features, including geographical characteristics and demographic trends, and any prefectural measures should consider these factors.

We also looked at the trends in the proportions of breast cancer patients who visited DCCHs for their initial treatment over time and found a continuous increase between 2018 and 2020 overall (from 53.8% in 2018 to 56.6% in 2019 and 57.7% in 2020). This trend might be due to a variety of measures, including the implementation of the recommendations of the Third BPPCC, particularly enhanced capability and availability of DCCHs. While most prefectures demonstrated similar trends, a small decline over this period occurred in some prefectures. We believe that this was driven by the COVID-19 pandemic, which potentially interfered with breast cancer diagnosis and treatment in terms of resources or the capacity for delivering breast cancer care as part of COVID-19-related risk management in 2020. Supporting this possibility, a prior study using the HBCR revealed reductions in cancer cases, cancer screening, and cancer detection rates in 2020 compared with those in 2016–2019 [16]. Additionally, the rate of breast cancer diagnosis was reported to decline in 2020 compared with 2019, with reductions in services such as breast biopsy [17] and decreased cancer screening [18]. Another report showed that the COVID-19 pandemic was associated with a diagnosis of more aggressive and more advanced disease following the suspension of breast cancer screening services [19]. In view of these findings, it is possible that the proportion of patients who received initial treatment at a DCCH in 2020 could have been affected by the COVID-19 pandemic. Overall, we should interpret these trends cautiously.

Another objective of our study was to examine the function, staffing, and medical services utilized at DCCHs. We noted some variability among the prefectures. In terms of the functions of the DCCHs, although all 47 prefectures had at least one JBCS-certified institution and at least one JOPBS-certified institution, the numbers varied among prefectures. The proportions of hospitals classified as Cancer Genomic Medicine core, hub, and affiliated were 2.7%, 7.1%, and 34.8%, respectively. Notably, 21 prefectures lacked both core- and hub-designated hospitals, although all prefectures had at least one Cancer Genome Medicine-designated (core, hub, or affiliated) hospital. From a staffing perspective, the data from 2020 indicated there were 769 specialists certified by the JBCS, 1,159 certified by or affiliated with the JOPBS, and 231 breast cancer nurses certified by the JNA. The median numbers of specific staff in DCCHs per prefecture were 77.40, 51.20, and 237.50, respectively. These values varied greatly among the prefectures.

Certified institutions, specialists, and specialized nurses are given functions/qualifications according to the requirements established by each society. For breast cancer, two societies have established requirements related to hospital function, and three societies have established requirements related to personnel. Utilizing an independent breast cancer registry combined with the quality indicators set by the JBCS has contributed to improved patient outcomes [8, 20]. However, there appears to be limited cooperation among societies, DCCHs, and local governments in terms of certifying institutions and personnel with the aim of improving the quality of community medical care along with the BPPCCs. In Europe, the initiative “Transforming Breast Cancer Together” was implemented in 2017. In this initiative, the European Parliament and several academic societies involved in breast cancer care, including the European Society of Breast Cancer Specialists (EUSOMA), the European Society of Surgical Oncology, and the European Breast Cancer Coalition, are collaborating to improve breast cancer care [21]. Other activities driven by EUSOMA, as an example, include performance monitoring using quality indicators [22] and certification of breast units [23]. Differences in healthcare policies and financial resources may affect healthcare facilities and personnel support, and urban areas may attract more specialized staff due to better infrastructure. Therefore, further cooperation between cancer care policymakers and societies is desired to further support the functions of institutions and deploy specialized personnel related to breast cancer care. One effective approach can be seen in Scotland [24], which aims to make cancer services more efficient and effective by emphasizing patient feedback and user experience to maintain and improve the performance of standard cancer care across the country. The introduction of a similar approach in Japan could further increase the number of patients visiting DCCHs for breast cancer treatment.

Although the data only covered approximately 14% of the Japanese population and were coupled with other limitations, and the information gained should not be conclusive, we tried to identify potential issues that could be addressed from the health services perspective by analyzing the proportions of patients with relevant medical fees for services related to breast cancer in the study period. The percentages of patients who received cancer patient management and guidance fees, which correspond to the provision of care-related information or interviews by healthcare professionals, did not reach 100%. In fact, these proportions were 0% in three prefectures. The percentage of patients with a cancer patient management and guidance fee (Ni), which is a category related to genetic testing, was 22.20%. Meanwhile, the percentages of patients with the BRCA1/2 genetic test fee, genetic counseling, and cancer genome profiling test fees were 60.00%, 33.00%, and 11.10%, respectively. Although services related to BRCA testing were frequently accessed, relatively few patients received services related to genetic counseling. This may be because the number of institutions for Cancer Genomic Medicine covers less than half of DCCHs. Regarding the variability of all the health services examined here, further research is needed to better understand these findings. It may be necessary to review the calculation requirements to ensure that they are consistent with the medical reality because of differences in staffing availability between hospitals in rural and urban locations.

Our study captured the progressive implementation of the Third BPPCC using public data and claims data from the perspective of breast cancer. In an interim report [25], it was noted that overall cancer care has improved following the introduction of DCCHs, but the progress varied among regions and medical institutions [25]. In fact, we have revealed some variabilities in the function and staffing assignments related to breast cancer care, not just its treatment, among prefectures. Ongoing activities to improve the quality of cancer care include those driven by the PDCA Cycle Forum [26]. Additionally, discussions between local government departments and DCCHs provide actions to implement based on data reported to the Base Hospital Liaison Council and core hospitals for cancer care coordination [27]. Additionally, the JBCS independently monitors the quality of breast cancer care using a patient registration system and quality indices, including postoperative radiotherapy and the proportion of patients undergoing HER2 testing. The availability of these services at DCCHs/base hospitals and the number of certified healthcare professionals represent how well each prefecture has incorporated the clinical guidelines developed by the JBCS and other relevant societies [8, 9]. However, some issues remain. For example, the currently available information is not fully utilized, the municipality participants only have observer status, and societies have limited involvement. Okuyama and Higashi shed light on the usefulness of and challenges associated with linking the HBCR and diagnostic procedure combination-based information [28]. They mentioned that it was useful to analyze such information with the aim of supplementing cancer registry information and gaining a more detailed understanding of the clinical reality, but they also highlighted the need to improve the process for utilizing the information, such as simplifying the approval system and introducing digital tools [29]. Therefore, it is hoped that sharing disease-specific information that better reflects the regional conditions with local healthcare professionals, government administrators, and other relevant stakeholders, such as society committees, will help us to identify and resolve problems, and ultimately achieve uniform accessibility to reduce disparities. In addition, providing patients with appropriate information could help to reduce disparities from the perspective of patients.

There are some important limitations of this study that may impact the generalizability of this study. Based on 2020 data, the HBCR comprised 79,062 patients, including 59,891 (57.7%) of patients who visited a DCCH for initial treatment of breast cancer. However, 103,744 patients with breast cancer were recorded in the PBCR. Therefore, the HBCR does not include all patients registered in the PBCR. The JMDC claims database is a large database compiling administrative data (healthcare claims) for over 17 million people in Japan, approximately 14% of the total Japanese population [30]. However, it has some limitations that influence the generalizability. For example, it is primarily composed of working-age people with a mean age of 35.7 years [31]. It cannot trace patients whose insurance category changes (e.g. due to a change of employer or stopping work) and some patients may be duplicated following the change in the insurance category. Furthermore, only a small proportion of patients are ≥ 65 years and none are aged ≥ 75 years old, indicating an underrepresentation of elderly patients. Other possible limitations of our study include the inability to track patients across multiple data sources, which could prevent us from understanding the reason(s) for the disparities in regions. If the HBCR could be linked to the JMDC claims database, it may be possible to collate more information on the stage of cancer in relation to treatment, and more accurately evaluate the facilities available at DCCHs. Finally, we sought to elucidate the real-world situation in Japan and identify possible gaps, not to propose specific solutions or strategies.