Previous studies have highlighted the importance of cancer registries and the growing role of RWD in oncology research [19, 20]. However, these studies mainly focused on the application of RWD and its contribution to the generation of clinical evidence rather than on systematically characterizing the underlying data sources. In contrast, the present review provides a comprehensive overview of oncology-related RWD sources in China and shows that registries remain the main source of RWD.

In this context, this review thoroughly examines sources of RWD for oncology in China, emphasizing their applications in epidemiology and HEOR. The number of oncology-focused RWD studies has significantly increased, primarily addressing cancer epidemiology, treatment patterns, and clinical outcomes. Registries were the most common RWD sources, offering critical data for RWS and clinical evidence generation. Notably, coastal regions and Sichuan Province have played key roles in developing RWD and contributing to related RWS publications. This study reviewed oncology RWD studies conducted in China between 2015 and 2025.

Dominance of Registries and EHRs

Registries are the primary sources of oncology RWD in China, aligning with global trends [18] in which cancer registries dominate [21]. The national cancer center registry (NCCR) is China’s key oncology database; however, its utility is limited by the lack of a nationwide open-access system, center-based operations, and integration challenges [22].

Internationally, oncology RWD ecosystems are more mature and highly integrated. In the USA, the Surveillance, Epidemiology, and End Results (SEER) Program is a large-scale cancer surveillance system, and such programs are widely established and routinely linked with Medicare claims and other administrative datasets, enabling comprehensive longitudinal analyses spanning incidence, treatment patterns, and survival outcomes [20]. In Europe, oncology RWD systems are increasingly characterized by multi-country collaboration, standardized data governance, and efforts toward interoperability and harmonization across national datasets [3]. In contrast, China remains at an earlier stage of development, with ongoing challenges in standardization, accessibility, and cross-institutional integration. This comparison helps situate China’s progress within the broader global RWD landscape and highlights future directions for development.

Claims databases are also important for evaluating disease prevalence, patient characteristics, treatments, and the economic burden of cancer. Importantly, they are increasingly used in regulatory evaluation, HTA, and reimbursement decisions, particularly for real-world effectiveness assessment. However, the generalizability of data covering the Chinese population remains limited. A national-level claims database, based on the Urban Employee Basic Medical Insurance (UEBMI) and the Urban Resident Basic Medical Insurance (URBMI), covers insured treatments but lacks long-term patient data [21, 23]. Regional claims databases can provide longitudinal data across several years; however, in certain regions, data are incomplete and often derived from random samples [21]. Other databases, such as the EHR database, also face limitations, including incomplete data on various variables and unstructured fields. The scarcity of longitudinal healthcare data in China results from fragmented systems, gaps in outpatient data, and increasing use of online healthcare services [24]. These challenges reflect limitations in availability, usability, and quality.

These findings have important implications for regulatory and HTA decision-making in China [3]. As RWE becomes increasingly integrated into clinical, regulatory, and reimbursement pathways, its value extends beyond expanding the evidence base to supporting decision-making in contexts where randomized controlled trial evidence is limited or not fully generalizable [3]. In this context, RWD can support regulatory assessment of treatment effectiveness, HTA evaluations of comparative effectiveness and cost-effectiveness, and reimbursement decisions related to resource allocation and budget impact. However, the policy value of RWD depends not only on its existence but also on its fitness for purpose.

An important implication of this review is the need to distinguish among data availability, usability, and quality [4]. Data availability refers to whether a dataset exists and can be accessed; usability refers to whether a dataset is suitable for a specific research purpose, including interoperability, linkage potential, and completeness of key variables; and data quality refers to the reliability of the data, including accuracy, consistency, traceability, and completeness. These dimensions are interrelated but not interchangeable, and explicitly distinguishing them is essential for appropriately interpreting the evidentiary value of existing RWD resources in China.

Strengthening oncology RWD systems in China will therefore require not only expansion of data sources but also improvements across the full data lifecycle, including standardized data collection, harmonization, linkage, cleaning, and governance [24]. Clearer regulatory expectations regarding acceptable data sources and appropriate RWE use cases will further improve consistency in evidence generation. In addition, multi-institutional collaboration and interoperable data-sharing mechanisms will be essential to improve representativeness, reduce single-center bias, and enable external validation of findings [21]. Collectively, these developments will enhance the reliability and policy relevance of oncology RWE for regulatory, HTA, and reimbursement applications.

Most Commonly Investigated Neoplasm Types

Our study found that digestive system, thoracic, and breast neoplasms were the most frequently investigated, consistent with the most commonly diagnosed cancers in China [18]. Similar trends have been observed in other studies. For instance, a Taiwanese study reported the highest publication rates for breast (16.3%) and lung (11.4%) cancers [25]. Thus, our findings provide a comprehensive overview of the utilization of oncology RWD in China. However, the focus on specific cancer types varies by region, influenced by factors such as genetics, lifestyle, environmental exposures, and geographical location. Data accessibility also affects cancer research priorities, potentially contributing to publication bias.

Geographical Disparities in RWD Sources

Oncology publications and databases in China are concentrated in coastal regions, particularly in southern China, because of higher economic development, population density, cancer prevalence, and healthcare resources [26,27,28]. This geographic concentration of RWD sources in coastal regions has important implications for national equity in oncology research. This pattern likely reflects broader disparities in healthcare infrastructure, digital maturity, and research capacity across regions in China. However, when RWD sources are disproportionately concentrated in economically developed coastal areas, the resulting RWE may underrepresent patients from central, western, rural, and other under-resourced regions. Such underrepresentation may limit the generalizability of findings and reduce the relevance of the evidence for national policy and reimbursement decisions. Addressing this issue will require more geographically balanced data networks, cross-regional collaboration, and harmonized governance frameworks to support broader participation in the generation of oncology RWD across China [26,27,28].

Sichuan Province, the only western province with significant progress in RWD and RWS research, benefits from its level of development and the contributions of West China Hospital of Sichuan University. This hospital attracts patients from neighboring provinces and has established numerous specialized databases, including the DACCA and the Western China Lung Cancer Database. DACCA continues to be updated and has produced multiple RWD studies published by the hospital’s gastrointestinal surgery team [29,30,31,32]. In addition, West China Hospital actively promotes research, encouraging physicians and scientists to engage in RWD studies. In contrast, economically developed coastal provinces, such as Jiangsu, Zhejiang, and Guangdong, despite their strong healthcare capacity, have fewer oncology databases and RWS publications because of the relatively recent development of RWD infrastructure, leading to a lag in related research output.

Limitations in Data Quality, Aggregation, and Access

RWD databases in China face persistent challenges related to data aggregation and accessibility. In addition, database coverage is limited, with 90 (71.43%) oncology databases providing no information on the number of hospitals they cover.

Collaboration and data linkage among registries, claims, and EHR databases in China remain insufficient. Although registry databases provide highly specific data, they lack longitudinal follow-up, a gap that could be addressed by integrating EHR databases. Standardized data-upload protocols and database linkage could enhance data granularity, supporting epidemiologic and HEOR analyses. Moreover, integrating diverse data sources would improve data completeness, thereby benefiting health policy development, patient care, and healthcare management [33].

Database accessibility in China is further hindered by stringent privacy, data security, and compliance requirements, which create barriers to data sharing and reuse. Data ownership and governance are often unclear and vary across regions and institutions, leading to lengthy approval processes. Unlike the US and the European Union, where data access is streamlined through clear regulations and user-friendly feedback mechanisms, the data access procedures in China require ethical approval and/or approval from the Human Genetic Resource Administration of China, a lengthy process often exceeding 6 months [21, 24]. In addition, researchers must obtain approvals from multiple hospital ethics committees, further delaying access [21]. However, recent national guidelines on RWS, RWE, and RWD are expected to introduce clearer regulations on variable definitions, data evaluation, availability, and protection, potentially improving access in the future [12,13,14].

Real-world oncology research in China also has data source and regional disparity limitations, similar to those observed in studies of developmental delay. For example, an RWD study in China found that very few oncology studies use EHR databases, and many datasets have limited hospital coverage and limited years of follow-up, raising concerns about representativeness and longitudinal validity [34]. Geographical disparities in treatment uptake and healthcare costs for patients with cancer approaching end of life in China have been well documented, revealing substantial differences between urban/rural and eastern/western provinces in treatment uptake and cost. However, the study acknowledged that hospital- and record-level inclusion biases, as well as incomplete capture of socioeconomic and clinical variables, may limit the generalizability of its conclusions [35]. These oncology examples underscore the importance of clearly defining data source categories, specifying coverage, and assessing completeness, especially when making regional comparisons.

The Road Ahead: Strengthening RWD Infrastructure

Strengthening regulatory frameworks and fostering healthcare partnerships are essential for improving data quality, integration, and utility. This process requires standardized data collection, governance, cleaning, integration, security, and continuous quality monitoring. Standardized protocols ensure consistency, whereas clear governance policies uphold data integrity and privacy. Routine validation helps eliminate errors, and integrating data from multiple sources provides a more comprehensive patient profile while maintaining security. Continuous quality assessment enables timely issue resolution, and collaboration among stakeholders enhances data reliability and patient outcomes.

At present, RWD in China represent a nascent but rapidly evolving ecosystem, requiring collective efforts from researchers, information technology vendors, hospitals, and government agencies [21]. Data-sharing policies should clearly define property rights, establish compliance frameworks for data authorization, and delineate responsibilities for data providers, users, and managers. A robust data infrastructure, including secure data spaces, high-speed networks, and privacy-enhancing technologies such as anonymization, federated learning, and multi-party secure computing, can strengthen data security, transparency, and accessibility.

Guidelines from organizations such as the China RWD and Study Alliance and the Chinese Thoracic Oncology Group provide technical support for RWS design and quality control [21, 36]. Future efforts should focus on simplifying data access, standardizing database formats, promoting structured data uploads, and strengthening research collaborations. Despite existing limitations, Chinese databases contain vast amounts of valuable data, particularly on rare cancers, offering unique research opportunities.

Limitations of this Review

This review has several limitations. The inclusion criteria for RWD and RWS involve a degree of subjectivity. Single-center studies without established databases were excluded, potentially introducing selection bias. In addition, only studies published in English or Chinese were included, which may have led to the omission of relevant research in other languages. Identifying RWE and RWD studies remains challenging because of the recent emergence of these terms, potentially leaving some relevant studies undiscovered. The study screening, data extraction, and classification were performed by one reviewer. Necessary discussions were held with another reviewer to address some uncertainties. Besides, 11.4% (500/4385) of the full-text literature was randomly selected in batches for re-review by the second reviewer during the screening and extraction process. This process may have introduced potential bias because of the incomplete duplicate review. No formal quality assessment was conducted, as this review focused on the overall distribution of oncology RWS and data sources in China. Publication bias may also exist because only published literature was included. Furthermore, this review included research published up to July 2025, and the rapid advancements in this field may have resulted in the exclusion of more recent studies. Additionally, this review roughly categorized the included studies based on the nationalwide registries' names, obscuring the diversity of the actual data sources to some extent.

Although the present mapping highlights the availability and diversity of RWD sources for oncology research in China, inherent challenges, including data heterogeneity, incomplete capture, and variable quality, underscore the need for standardized methodologic approaches to enhance the reliability and interpretability of RWE. To address these challenges, established best-practice recommendations aligned with International Society for Pharmacoeconomics and Outcomes Research (ISPOR) guidelines can be implemented. Key strategies include transparent study design and reporting, systematic standardization of data collection across sources, rigorous validation of data quality, and application of robust analytical methods to mitigate confounding and bias [37]. Furthermore, ISPOR emphasizes reproducibility through detailed protocol registration and sensitivity analyses to evaluate the robustness of findings [38]. Incorporating these principles into future oncology research in China can strengthen the credibility of RWD-based studies, facilitating their utility for regulatory, payer, and clinical decision-making. Collectively, database accessibility constraints, publication bias, and regional disparities may introduce systematic bias, affecting the representativeness and generalizability of oncology RWD evidence in China.