Marked improvements in childhood cancer survival are uneven across and within countries [1,2]. The World Health Organization’s (WHO) Global Initiative for Childhood Cancer targets at least 60% survival globally by 2030 and frames information systems as core enablers of planning, quality improvement, and accountability [3]. Cancer registries, clinical networks, and patient-facing documentation together build the measurement and feedback loops required for learning health systems [4-6].

Japan has a unified National Cancer Registry (NCR) mandated by law and coordinated national cancer information services [7-10]. Designated childhood cancer core hospitals have concentrated expertise and clarified referral pathways [11]. Nevertheless, the pediatric oncology ecosystem can feel like a black box to international collaborators. Population-level incidence and survival can be monitored, but detailed treatment exposure, risk-based survivorship, and real-world coordination are not consistently visible at the national scale. These blind spots relate less to technology than to governance—custodianship, consent, standards, and aligned incentives [12].

This paper offers a formative viewpoint aimed at stakeholders who can operationalize change—national cancer control and registry leaders, designated childhood cancer core hospitals and survivorship clinics, cooperative groups, and health IT and standards implementers.

This study aims to propose a governance-first, implementable design that connects registries, clinical systems, and survivorship services for childhood cancer in Japan, using open standards and proportionate privacy and consent.

The key messages of this viewpoint are as follows: (1) the main bottleneck is governance rather than software; (2) a clearly mandated Pediatric Data Steward function can align purposes, consent, agreements, and audit; (3) a harmonized pediatric data dictionary implemented through HL7 Fast Healthcare Interoperability Resources (FHIR) profiles and application programming interfaces (APIs) can reduce burden and improve data quality; and (4) a patient-facing Digital Survivorship Passport (DSP), with explicit provenance and layered consent, can return value to families while strengthening the learning loop.

We synthesized international guidance on registry quality and survivorship documentation and health data interoperability [4-6,13-16]. We also reviewed Japanese policy instruments and technical reports relevant to national cancer control, registration, and data governance [7-9,11,17,18]. Our practical experience with registry operations and pediatric oncology networks informed feasibility considerations.

We identified a minimal pediatric dataset spanning diagnosis, treatment summary, long-term follow-up (LTFU), and outcomes; mapped it to FHIR resources; and bound it to International Classification of Diseases for Oncology (ICD-O) and the International Classification of Childhood Cancer (ICCC) with pragmatic pathways to Systematized Nomenclature of Medicine–Clinical Terms (SNOMED CT) and Logical Observation Identifiers Names and Codes (LOINC) for clinical decision support and international comparability [15,16]. For FHIR implementation, we propose a reuse-first profiling strategy: starting from widely used international oncology and survivorship artifacts where applicable, and defining Japan-specific FHIR profiles only where required by Japanese registry items, consent models, and linkage workflows—while maintaining mappings to support international comparability. We delineated a governance role—the Pediatric Data Steward—to harmonize purposes and agreements under the Act on the Protection of Personal Information (APPI) and the Next Generation Medical Infrastructure Act (NGMIA) [17,18]. We did not conduct new empirical analyses or a formal consensus process; rather, we framed a design for pilot testing and iterative refinement.

Japan contributes to the global evidence base on cancer epidemiology and survival [2]. The NCR aggregates incident cases nationally and, with linkage to the resident registry and death certificates, enables survival analyses; the hospital-based cancer registry (HBCR) provides richer clinical granularity [8,19,20]. Incidence patterns for childhood, adolescent, and young adult cancers broadly align with other high-income settings [21]. Access to timely diagnosis and protocol-based therapy is generally strong in designated centers [11]. Nevertheless, barriers persist. Pediatric-specific variables in the NCR have not routinely captured treatment exposure at levels sufficient for risk-based survivorship planning. HBCR data are richer but variably connected to survivorship clinics and patient-reported outcomes. To make these blind spots concrete for international readers, examples of survivorship-critical data elements that are often missing or unevenly captured across systems include:

Regional variation in pediatric oncology workforce density and referral timeliness remains, as seen in other specialized services [12,22]. These gaps underscore the need for pediatric indicators in national dashboards and a governance mechanism that aligns registries, clinical networks, and patient-facing documentation.

The Act on Promotion of Cancer Registries mandates standardized reporting for national planning and research [7]. The APPI and associated guidance define duties for controllers and processors handling “special care-required” personal information [17]. The NGMIA provides a framework for certified operators to generate and manage anonymized medical data for research [18]. In practice, pediatric oncology data governance must navigate these statutes when linking cooperative group databases to registries and when enabling secondary use. Institutional variation in deidentification thresholds, consent renewal at transition to adult care, and cross-institutional sharing can delay projects. International experience suggests that clear, proportionate rules and privacy-by-design consent processes help sustain the social license for data reuse [23,24]. Within this envelope, a national Pediatric Data Steward could standardize templates for data-sharing agreements, consent language, and role-based access and coordinate audits and public reporting.

Children cured of cancer face heterogeneous risks of late effects across cardiac, endocrine, neurocognitive, fertility, and second-malignancy domains. International guidelines and tools convert cumulative treatment exposure into risk-based surveillance plans, enabling shared care between specialists and primary care [13,14,25]. In Japan, LTFU practices have advanced in several childhood cancer core hospitals and cooperative groups, but national visibility remains limited because survivorship documentation often sits in hospital silos and is not consistently linked to registries [11]. Transitions to adult services risk information loss. Interoperable, patient-facing summaries built with HL7 FHIR can travel across systems and embed granular consent preferences [15,16,26,27]. A minimal survivorship dataset embedded in the NCR and HBCR, paired with a bilingual DSP aligned to international terminologies, would support continuity for families and create a feedback loop for service planning.

Japan’s pediatric oncology workforce is experienced and organized in mature cooperative groups such as the Japan Children’s Cancer Group (JCCG) [11]. As in other high-income settings, competing demands on clinician time, limited registrar capacity, and scarce clinical informatics support can hinder routine, high-quality data entry and use. Adoption depends on perceived value and ease of use at the front line as much as on policy mandates [28-30]. National instruments already support clinical research and quality improvement, but the next step is to close the loop between registries, cooperative group data, and LTFU clinics so that a single pediatric data dictionary serves care, research, and patient-facing summaries. Financing models that recognize data completeness and LTFU adherence as quality indicators would help [9,11,19].

We propose a 4-layer architecture with explicit governance responsibilities and standardized data flows (Figure 1). We propose a 4-layer architecture with explicit governance responsibilities and standardized data flows. Source systems include electronic health records, laboratory and radiology information systems, pathology, and cooperative group databases. These feed an HL7 FHIR gateway that exposes validated, profile-conformant resources. Terminology and coding are handled in a distinct service that binds ICD-O and ICCC codes to relevant FHIR elements and maintains pragmatic mappings to SNOMED CT and LOINC to support clinical decision support and international comparability [16,27]. A national Pediatric Data Steward oversees standards and interoperability (FHIR profiles and APIs), terminology and coding, consent and privacy aligned to APPI and NGMIA, data-use agreements, data quality, and audit. Outputs are routed to the NCR and HBCR and to a patient-facing DSP. Bidirectional exchange should be governed by explicit provenance and validation rules. Clinician-verified treatment summaries and key exposures (eg, chemotherapy and radiotherapy details) should originate from clinical systems and cooperative group databases and remain the source of truth for registry-grade data. By contrast, patient or caregiver inputs—such as patient-reported outcomes and selected self-entered updates (eg, vaccination history and follow-up attendance and adherence information)—should be stored and exchanged as patient-reported data, clearly labeled as such, with audit trails. Patient-reported information should not overwrite clinician-verified registry elements and should be incorporated into analytics or learning workflows only under defined governance, including clinician review or attestation where clinically relevant. In particular, patient-reported inputs should not be written back into the NCR or HBCR as registry-grade fields unless and until they are reviewed and attested by an authorized clinician. Linkage to the resident registry and vital statistics supports survival analysis; linkage to claims databases (eg, the National Database) can be staged as a future extension. Security, audit logs, and public reporting span all layers and are core features rather than add-ons [8,19].

We outline five practical steps:

A staged road map over 3 years can move from pilot to national scale while maintaining attention to workload, security, and acceptability. Table 1 summarizes these priority governance actions by domain and aligns them with international exemplars.

This viewpoint reframes Japan’s childhood cancer information landscape as a digital health governance problem. The NCR and HBCR are strong foundations, but richer pediatric variables and LTFU data need to flow with less friction [8,19]. The proposed architecture intentionally separates standards and interoperability from terminology and coding, acknowledging that FHIR is a transport and representation standard, whereas ICD-O and ICCC are classification systems [16]. Embedding survivorship as a first-class output through a DSP changes the value proposition: data are not only for registries and research but must return to children and families in an immediately usable form [13,14].

We use “Pediatric Data Steward” to denote a mandated function (not necessarily a new standalone organization) that can be operationalized through an existing national body with clear accountability. Because the steward function includes standard-setting, data-use agreements, quality audits, and public reporting, its institutional anchoring determines whether the model can be enforced and sustained. In Japan, locating the steward solely within the Ministry of Health, Labour and Welfare could provide policy authority and convening power but may be less suited to day-to-day technical operations. Locating it solely within the National Cancer Center could ensure operational proximity to registry infrastructure and technical capacity but would require an explicit national mandate to convene hospitals and cooperative groups.

We therefore propose a hybrid model: a nationally mandated Pediatric Data Steward function under the cancer control framework, with policy leadership aligned with the Ministry of Health, Labour and Welfare and an operational and technical secretariat aligned with National Cancer Center registry and cancer information functions, governed through a formal multistakeholder structure (designated childhood cancer core hospitals, cooperative groups, such as JCCG, survivorship services, and patient and family representation). This arrangement aims to combine authority, implementation capacity, and legitimacy while avoiding fragmentation of responsibility.

International experience is consistent, but it also shows that technical interoperability alone rarely dissolves institutional silos. Registries work when purposes are clear, financing is stable, and quality is audited [5,6]—and when contributing sites receive visible value back for sustained participation. Survivorship passports diffuse when they are co-designed with survivors and families and embedded in routine workflows [13,14].

In Japan, a governance-first approach should therefore pair standards with reciprocity and credible support. Our proposed “performance-linked support” is best understood as targeted operational support—such as funded registry staffing time, shared technical services, or implementation assistance—linked to measurable indicators (eg, timeliness and completeness of pediatric core items and LTFU coverage), and implemented alongside workload-reducing measures (eg, FHIR-enabled reuse of existing clinical data rather than duplicate manual entry). Equally important is returning timely, usable outputs to hospitals (benchmarked dashboards, case-finding feedback, and survivorship care tools) so data sharing is experienced as an enabler of local quality improvement and legitimate research, not as a loss of institutional assets. Our proposal adapts these lessons to Japanese law and infrastructure, with a light but accountable governance layer and an implementation path that can be evaluated and adjusted.

Evaluation should be formative. Early pilots ought to track process metrics—case ascertainment, timeliness from diagnosis to registry entry, completeness of pediatric core items, LTFU attendance—and outcome metrics—stage distribution, relapse detection intervals, guideline-concordant surveillance, and early detection of late effects—alongside equity gradients by region and socioeconomic markers [4,19,28-30]. Workload, usability, and cybersecurity also require attention, especially for FHIR-based integration and patient-facing tools [26]. Where terminology licenses or system heterogeneity create hurdles, national support and shared services can mitigate site-level costs [16].