A systematic literature search was performed across the CNKI, Wanfang, and VIP databases to identify clinical trial quality management CTQM-related publications. The search string applied was: (“drug clinical trial” OR “pharmaceutical clinical trial” OR “good clinical practice”) AND (“quality control” OR “quality management” OR “quality assurance” OR “Risk Management”). A total of 1,333 articles were retrieved, with the following distribution: 280 from CNKI, 520 from Wanfang Data, and 533 from VIP Information.

To capture internationally published works, supplementary searches were executed in Scopus, Web of Science (WOS), and PubMed for CTQM publications authored by Chinese scholars. The three database search terms used were as follows: (1) TITLE-ABS-KEY [(“drug clinical trial” OR “pharmaceutical clinical trial” OR “good clinical practice”) AND (“quality control” OR “quality management” OR “quality assurance” OR “Risk Management”)] AND [LIMIT-TO (AFFILCOUNTRY, “China”)]; (2) {[“drug clinical trial”(Mesh) OR “pharmaceutical clinical trial” OR “Good Clinical Practice”(All Fields)] AND [“quality Assurance”(Mesh) OR “quality Control”(Mesh) OR “Risk Management”(Mesh) OR “Quality Management”(All Fields) OR “risk based monitoring”(All Fields)]}; (3) TS = [(“drug clinical trial” OR “pharmaceutical clinical trial” OR “good clinical practice”) AND (“quality control” OR “quality management” OR “quality assurance” OR “risk management”)]. This search yielded 264 articles published by Chinese authors. However, after two researchers (XN and LJ) reviewed them individually, only six were found to be relevant to the research topic. Given the small number of articles and their limited impact on the quantitative results of this study, we only included only these six articles in the discussion to ensure the completeness of our findings.

The inclusion criteria were as follows: (1) Study types: empirical research, systematic reviews, or policy analyses; (2) The title/abstract contains a clinical trial or drug clinical trial and at least one CTQM core term from the predefined lists: monitoring, data management, ethics review, risk management, quality control, auditing, protocol deviations and good clinical practice (GCP). (3) ≥ 50% of the results/discussion sections address CTQM processes (e.g., QC procedures for electronic data capture and risk-based monitoring workflows). (4) Literature form the establishment of the database to 30 December 2024. The following documents were excluded: (1) Duplicate publications and studies with incomplete information. (2) other types of publications (such as meeting abstracts, editorial materials, letters and early access, etc.).

Figure 1 presents the data processing and analysis. The literature obtained from each database was imported into NoteExpress for deduplication. Two researchers (XN and LJ) independently reviewed the literature; in cases of disagreement regarding inclusion, a third researcher was consulted to resolve the differences and reach a consensus. Of the literature obtained, 210 articles were retained from CNKI, 219 from Wanfang Data and 99 from VIP Information for analysis. The filtered literature was then imported into COOC 20.5 for further data cleaning. This included supplementing missing fields (e.g., keywords, institutions and authors), batch merging synonyms and deleting meaningless terms. COOC 20.5 was then used for the following analyses: publication statistics and frequency analysis: collaborative network analysis of institutions, authors and journals, and thematic evolution analysis based on keywords. CiteSpace 6.1.R6 was used to conduct keyword co-occurrence analysis, keyword clustering analysis and keyword burst detection. The parameters were set as follows: Time Slicing (Year per Slice) = 1 year; TopN = 50.

Figure 2 illustrates the annual and cumulative publication trends in the field of CTQM between 1997 and 2024. The results demonstrate that Chinese research originated from Yao’s (4) pioneering work in 1997, when his team first systematically proposed the idea of achieving scientific and objective evaluation of clinical efficacy through the implementation of internationally standardized GCP. This laid the theoretical foundation for subsequent studies.

From a temporal evolution perspective (Figure 2), research has progressed in three distinct phases. Emergence phase (1997–2003): annual publications averaging fewer than 1, with research themes concentrated on foundational concept introduction and policy framework exploration (5–8). Slow development phase (2004–2010): The number of annual publications increased to 9. Rapid growth phase (2011–2024): Annual publications surged to 33.8, with dual peaks observed in 2015 (52 publications) and 2022 (49 publications), during which research hotspots became prominently concentrated.

According to the analysis shown in Figures 3, 4, the two peaks are both concentrated in the areas of clinical trial quality management, risk management, information management and clinical trial institutions. Additionally, research on data management and pharmaceutical regulation was highlighted in 2015 (Figure 3), whereas studies on artificial intelligence and protocol deviations rose in popularity in 2022 (Figure 4). Remarkably, following China’s 2017 accession to the ICH, publication output rose at an average annual rate of 24% (2017–2024). While this temporal association suggests that international alignment policies may have contributed to the increase, we cannot rule out the influence of other concurrent factors.

The author collaboration network (Figure 6A) shows that 272 nodes with 360 connecting lines formed a cooperative cluster with significant clustering characteristics (modularity Q = 0.59, mean profile coefficient S = 0.86). The author collaboration rate was 0.92, and the collaboration degree was 2.89 (Figure 6C). These findings indicate that the academic community exhibited a highly structured collaboration pattern.

Based on Price’s (19) Law, the threshold for core authors was calculated as M = 0.749Nmax≈ 3. The maximum number of papers published by an author among the 1,661 authors was 12. A total of 172 core authors (with ≥ 3 published papers) were identified, accounting for 10.36% of the total number of authors. Twenty-two highly productive scholars (with ≥ 7 papers published) formed the core research group (Figure 6B), such as Chen Yongchuan and Gao Rong, who typically collaborate closely with other authors (Figure 6A), indirectly indicating that strengthening collaboration can increase research output.

A total of 528 articles were scattered across 127 journals. Sixteen titles contributed six or more papers, collectively accounting for 360 publications (68.2%). As shown in Table 2, Chinese New Drugs published the largest share (n = 74; IF = 1.908), followed by the Chinese Journal of Clinical Pharmacology (n = 58; IF = 1.851), Chinese New Drugs and Clinical Remedies (n = 53; IF = 1.529), and China Pharmacy (n = 39; IF = 2.414).

Table 3 presents the frequency, betweenness centrality, and year of the first appearance of high-frequency keywords, revealing the development process of CTQM research. Betweenness centrality analysis revealed that clinical trials (0.62), drug clinical trials (0.50), quality control (0.25), good clinical practice (0.21), and drug clinical trial organizations (0.19) form the core framework of the CTQM knowledge network, establishing long-term connections between standard formulation and implementation research. In contrast, emerging themes such as risk management (0.05), data management (0.01), ethical review (0.02), and informatization (0.01), which emerged after 2015, exhibit low centrality, indicating that they remain on the periphery of the network and rely on traditional core terms for access. This suggests the need to establish stronger cross-domain bridges in the areas of risk, information, data, and ethics in the future.

Keywords clustering analysis based on the log-likelihood ratio (LLR) algorithm and mutual information (MI) metrics identified 11 valid clusters (modularity Q = 0.5139, average silhouette coefficient S = 0.8225) (Table 4). Six clusters presented S values > 0.7, indicating strong internal keyword associations and rational classification. In clustering, the higher the LLR or MI is, the better the metric reflets the theme content of the cluster.

The drug clinical trial cluster (Cluster 0) was the largest cluster (48 nodes, S = 0.701), which originated in 2013 and focused on core themes such as “drug clinical trials” (LLR = 66.02), “quality control” (LLR = 55.25), and “normative standards” (MI = 1.17), and “quality assurance system” (MI = 1.17). This once again demonstrates the central role of basic standard setting and management systems in CTQM research. Cluster 1 includes “data management” (LLR = 23.1) and “change of production site” (MI = 1.07), whereas Cluster 2 includes “data verification” (LLR = 13.11) and “information technology management” (MI = 0.42). Both clusters are highly associated with data management and informatization technologies in CTQM. Cluster 3 aggregates research on drug clinical trial institutions, featuring “management model” (LLR = 11.97), “clinical research coordinator” (LLR = 7.18), and “quality management evaluation system” (MI = 0.21), and the research focus lies primarily on institutional operational models and personnel management. Cluster 4 (quality control) highlights “antitumor drugs” (LLR = 8.51). The risk-control theme (Cluster 5), formed in 2019, centers on ethical review (14.04).

CiteSpace was used to generate a keyword-burst map (Figure 7A), In contrast, COOC produced a hotspot-shift diagram (Figure 7B) and thematic-evolution maps (Figures 7C, D) to explore the evolutionary trajectory and shifting hotspots of clinical-trial quality control in China.

The evolutionary trajectory of CTQM research delineates three sequential phases (Figure 7): 2003–2012 was dominated by foundational studies, as evidenced by active themes such as “drug clinical trials” and “quality control” (Figure 7B) and the emergence of burst keywords such as “regulation” (burst = 2.56), “hospital” (2.10), and “three-level quality control (20)” (1.89) (Figure 7A). From 2013 to 2019, the field shifted toward a technologization trajectory, with “risk management” reaching its peak (Figure 7B) and concurrent bursts of “drug” (2.57), “data management” (3.55), and “risk management” (3.54). From 2020 onward, the paradigm experienced an intelligent leap, as “informatization management,” “ethical review,” and “protocol deviation” rapidly ascended to the research frontier (Figures 7A, C, D).

The results of this study show that China’s CTQM research followed a three-stage path of “basic specification→technology integration→intelligent transformation,” which is an evolutionary trajectory defined by both policy regulations and literature visualization.

An analysis of institutional and author collaboration networks (sections “3.2 Analysis of publishing institution characteristics” and “3.3 Author collaboration network analysis”) indicates that government-academia-hospital partnerships are the primary drivers of CTQM research in China. This collaborative framework has significantly increased research output, as evidenced by the positive correlation between collaboration levels and publication rates. These findings are consistent with the principle of promoting regulatory science through “government-industry-research” synergy, which is endorsed domestically (28).

However, compared with international standards, the scope of CTQM development in China is relatively insufficient in terms of cross-system collaboration. The subordinate agencies of the NMPA played a leading role in the early stages of CTQM development (Table 1). In subsequent stages, however, mechanisms for deeper integration with industry, academia and research institutions have remained inadequate and most collaborations have been limited to bilateral partnerships. In contrast, Europe and the United States emphasize establishing a multistakeholder collaboration network that includes sponsors (pharmaceutical companies), contract research organizations, academic medical centers and regulatory agencies (29). In order to advance CTQM research and enhance its global influence, China should prioritize expanding cross-system collaboration, strengthening deep cooperation with the industrial sector and actively integrating into global multicenter clinical trial networks.

Technologies such as artificial intelligence (LLR = 10.48) and blockchain (burst intensity = 7.5) have driven progress in smart monitoring and ethical governance. However, they have also exacerbated challenges such as data silos and privacy risks. Vallée (30) highlighted that, although digital twin technology has the potential to transform precision medicine and patient outcomes, it has also sparked significant controversy regarding data privacy and ethics. Harvey and Gowda (25) further elaborate on the complex regulatory balance that the FDA must maintain when regulating AI-based medical applications. Chinese research has recognized this global challenge, demonstrating foresight through preliminary exploration of ethical governance frameworks. In 2022, Liu et al. (31) from Central South University addressed the ethical challenges of AI-driven clinical trials by developing strategies to optimize performance while ensuring ethical integrity. Li and Yang (32) advocate resolving ethical conflicts through human-centered design to prevent technological alienation, and bridging the data divide through people-centered governance. Although China has achieved localized development in ethical governance, it still needs to address the global issue of “technological and ethical imbalance” by coordinating technological empowerment with regulatory frameworks. To address the widespread issue of “technology–ethics mismatch,” the international community must urgently establish comprehensive ethical standards and data governance mechanisms that are in line with the pace and scale of technological development.

This study has several limitations. Not all foreign databases were covered, which may have resulted in incomplete literature inclusion. Additionally, the results of the study exhibit significant regional characteristics, which limits the replicability of policy systems. China’s CTQM mechanism, which is driven by the government, fundamentally differs from enterprise-driven models in Europe and the United States. Policy transplantation teams must be aware of the risks associated with contextual adaptation. When applied in other countries, it must be adapted to the unique characteristics of local systems.

We intend to conduct thorough research into the status of drug clinical trials worldwide, establish a “global drug clinical trial quality benchmarking system,” and integrate China’s drug clinical trials further with national policies. This will promote the integration of Chinese practices with international developments.