The degree of commonality and interoperability among data was assessed across 3 levels: clinical trial protocols, metadata, and data systems and management.

The process began with generating a list of the main points and parameters of general interest (eg, type of trial, randomization, primary and secondary aims, and safety) across the 3 specific clinical trial protocols: EU-COVAT-1_AGED, EU-COVAT-2_BOOSTAVAC, and EU-COVPT-1_COVACC [9,10]. The second step involved comparing the 3 trial protocols to identify interoperable data points based on this list of key parameters. In step 3, the outcomes derived from the adult trial protocols were compared against the VACCELERATE adult master protocol (Figure 1). This comparison focused on 4 core areas of the master protocol template’s core outcome set (COS): (1) immunogenicity, (2) safety, (3) efficacy/clinical/physiological, and (4) other outcomes of interest. Differences and similarities among the protocols were identified manually.

All data management (DM) processes, including the design of study databases for all 3 VACCELERATE clinical trials, were carried out by the Clinical Trials Centre Cologne (CTCC). The DM team at the CTCC developed the Study Database Specification (DB Spec) using an Excel (Microsoft Corporation) workbook. This workbook included metadata for defining the study database, case report forms (CRFs), edit checks, and a data dictionary. For each CRF study, the workbook contained an individual worksheet specifying database details such as field names, field response types, and formats. Additionally, the detailed metadata for all 3 vaccine trials were systematically developed using internal standards, terminology, and the general approach of Clinical Data Acquisition Standards Harmonization (CDASH) [12] for data collection, such as on electronic CRFs (eCRFs; Table 1). For instance, questions were categorized into relevant Clinical Data Interchange Standards Consortium (CDISC) [13] domains, along with prefixes for adverse events (AEs), concomitant medications, vital signs, protocol deviation, etc, while standard CDISC suffixes were applied to indicate question types.

Although the databases were generated by a single center and coordinated by the same individual, it was valuable to evaluate the degree of standardization, commonality, and data interoperability among the DB Spec of all 3 trials. To achieve this, a manual comparison of the Excel DB Spec from all 3 VACCELERATE clinical trials was conducted (Figure 1). This comparison focused on the main core areas—immunogenicity, safety, efficacy/clinical/physiological—emphasizing the development of the master protocol template.

The other variables were not considered due to differences in demographic details and treatment approaches among the 3 groups (older adults, adult, and pediatric). Metadata items such as data format, data dictionary name, response type, item text, required field indicator, and review groups were compared for consistency and discrepancies. Once identified, the next step was to analyze the rationale behind the observed consistencies and discrepancies among the metadata items.

Interoperability in DM and systems was assessed through a questionnaire (Multimedia Appendix 1) distributed to the central laboratories: the University College Dublin (UCD) Centre for Experimental Pathogen Host Research (CEPHR, Dublin), Instituto de Salud Carlos III (ISCIII, Madrid), and the central biobank (UZA/UAntwerpen Biobank, Antwerp University Hospital). The questionnaire aimed to gather their experiences working with the VACCELERATE DM systems (Figure 1). UZA/UAntwerpen provided an overview of samples received from local sites (shipment manifest) and transferred to the research laboratories (transport manifest) and the clinical trial center. CEPHR and ISCIII, as research laboratories, performed the specified analyses on samples received from the central biobank. After completing the analyses, the research laboratories sent the results to the CTCC.

The questionnaire was designed to maximize accessibility and ease of use for site staff by categorizing the questions into broad areas while allowing flexibility through free-text response options: (1) laboratory procedures and documentation, (2) sample handling and storage data, (3) data transfer, and (4) general laboratory DM. Based on the information gathered through the questionnaire, a descriptive analysis was conducted to examine how each site handled, stored, and transferred data, as well as how their processes evolved within the VACCELERATE network. Furthermore, the analysis evaluated potential factors and new concepts required to harmonize and standardize the VACCELERATE DM systems.

The DM documents—external DM plans, edit check plans, CRF completion requirements, and DB Specs—developed for the clinical trials by the CTCC were reviewed. Guided by the questionnaire, a semistructured interview was conducted with the CTCC data manager.

The 3 VACCELERATE COVID-19 vaccine trials (EU-COVAT-1_AGED, EU-COVAT-2_BOOSTAVAC, and EU-COVPT-1_COVACC) received favorable ethics opinions and approval from the national competent authorities in the participating countries (State Medicines Control Agency, Norwegian Medicines Agency, Spanish Agency of Medicines and Health Products, Health Products Regulatory Authority Ireland, and Paul-Ehrlich-Institut for the EU-COVAT-1 AGED trial; Health Products Regulatory Authority Ireland, Norwegian Medicines Agency, Swedish Medical Products Agency, The Spanish Agency of Medicines and Medical Devices, Federal Agency for Medicines and Health Products, and Paul-Ehrlich-Institut for the EU-COVAT-2 BOOSTAVAC trial; and Netherlands – Competent Authority, Norwegian Medicines Agency, Swedish Medical Products Agency, Hellenic National Organization for Medicines, and Paul-Ehrlich-Institut for the EU-COVPT-1 COVACC trial). Written informed consent was obtained from all participants before any trial procedures. These trials were conducted in accordance with the International Conference on Harmonization for Good Clinical Practice (ICH-GCP) guidelines and complied with the Declaration of Helsinki at all times. Furthermore, the provisions of the EU General Data Protection Regulation (GDPR) 2016/679 were strictly observed to ensure data protection. All investigational materials and data were pseudonymized in compliance with data protection regulations before undergoing scientific processing.

Our harmonization and standardization efforts began with a comparison of the 3 clinical trial protocols against the COS of the master protocol. The COS includes all appropriate measurements necessary to ensure standardization, efficiency, and uniformity in the conduct of clinical trials. Four main COS areas were adopted for phase II and III studies: (1) immunogenicity, (2) safety, (3) efficacy/clinical/physiological, and (4) other outcomes of interest.

The harmonization process resulted in the creation of Tables 2-5, which illustrate the interoperability between the master protocol and the 3 clinical trial protocols based on the COS areas. A summary of the outcomes is presented in Figure 2 (see Table S1 in Multimedia Appendix 2). Humoral and cellular responses to COVID-19 vaccines were assessed in a similar manner across all trials, considering both vaccine and host characteristics (Table 2), although minor adjustments to time points were incorporated into the pediatric protocol.

The early detection of serious AEs (SAEs) can be vital for the protection of the participants and via the interoperability of safety data, an overall better general view of the patient’s clinical data could be achieved, resulting in the reduction of any potential risk of harms or adverse impacts and higher clinical safety. The developing safety profile in the trials is in a sufficiently similar motive and includes all the required information about reporting solicited, unsolicited, serious, and medically attended AEs, as shown in Table 3. By contrast, the definitions of the COVID-19 disease (eg, asymptomatic infection, confirmed and reinfection cases, the severity of signs and symptoms) are not reported in the protocols, even though it is referred to as part of the trial dataset outcomes (Table 4). Finally, the profile of immunophenotyping of immune responses (at the cellular level) and genotypic testing of breakthrough infections (vaccine immunogenicity) were also determined in the protocols, enhancing interoperability among them (Table 5). It should be noted that we did not present herein the comparison of pediatric protocol versus pediatric master protocol because the former is currently under review process.

Several parameters were considered to enable and enhance interoperability across the developed COVID-19 vaccine trials and master protocols (Figures 3 and 4). It is well-known that the lack of standardization in data processing exacerbates inhomogeneity in raw data. This often results in errors and significant time delays, reducing the reusability and interoperability of data within protocols. To address this, the various protocol formats were homogenized by manually identifying discrepancies and commonalities among selected parameters (see Tables S2-S13 in Multimedia Appendix 2). The selection of parameters was based on their relevance to the trials. Parameters present in only 1 protocol were excluded, while efforts were made to promote the consistent use of the same parameters across adult trials.

Concealment (blinding) of treatment group allocation is not anticipated for the current trials, while both adult trials support vaccine booster strategies. All clinical trials assess reactogenicity symptoms and the immunogenicity (efficacy) profile of COVID-19 vaccines (following a fourth dose) to inform potential optimized vaccine strategies. The secondary objectives include evaluating immune response for up to 12 months (see Tables S2-S4 in Multimedia Appendix 2).

The safety aims and the increase in antibody titer were found to be similar across all trials, as were the primary end points in the adult population protocols, due to the nature of the BOOSTAVAC trial (see Table S5 in Multimedia Appendix 2). By contrast, stratification/subpopulation, inclusion, and exclusion criteria among volunteers are directly influenced by the participants’ age and other trial-specific factors (eg, vaccine scheme, gender, immune status; see Table S6 in Multimedia Appendix 2). Additionally, patients’ insurance and consent forms are addressed in all protocols, and the sample size calculation was reported based on the specifics of each protocol (see Table S7 in Multimedia Appendix 2).

We also conducted a comparison of the 2 adult trials with the master protocol (see Tables S8-S13 in Multimedia Appendix 2). The primary aim of the adult trials is to evaluate the immunogenicity and reactogenicity (safety) of vaccine booster strategies and homologous mRNA vaccination for COVID-19 at different time intervals (see Tables S8 and S9 in Multimedia Appendix 2).

Both the adult and master protocols include the relevant subprotocols, cohorts, and descriptions of the randomization process, while blinding is not anticipated in the adult trials. All protocols support phase II trials and share similar primary objectives. The primary and secondary objectives/end points (including exploratory ones) constitute a nearly identical core dataset. The secondary trial objectives for immunogenicity and reactogenicity focus on the immune response and associated signs and symptoms over up to 12 months (see Tables S9 and S10 in Multimedia Appendix 2).

The reporting periods for unsolicited and solicited AEs, as well as the rates of SAEs, are identical across the adult and master protocols. However, discrepancies in inclusion and exclusion criteria are directly influenced by the trial participants (see Tables S11 and S12 in Multimedia Appendix 2). The treatment duration and follow-up period for the studies are identical in both the adult trials and the master protocol, extending up to 12 months. All clinical trials are conducted in accordance with the ICH-GCP, including the use of appropriate consent forms and patient insurance, based on the allocated treatment group (see Table S13 in Multimedia Appendix 2).

The points of consistency and discrepancies across the main core areas (eg, immunogenicity, safety, efficacy, clinical, and physiological aspects) among the metadata items of the 3 VACCELERATE clinical trials are presented in Tables 6-8 and Figure 5. The noncommonalities identified within the CRFs were attributed to differences in the 3 populations (aged 75+, adult, and pediatric), variations in protocol design, and differing vaccination patterns (Figures 3 and 5). Additionally, differences in the coding and labeling of variables, as well as in the item groups, were identified. These discrepancies required clarification to enhance interoperability.

VACCELERATE received a significant boost through the selection of the CTCC as the sole DM provider, with system database development coordinated by a single individual. This eliminated the need for coordination across multiple trial units. For the CTCC, the standardized format adopted within the VACCELERATE project utilized a data system featuring an installation of the Anju [16] (formerly OmniComm) TrialMaster clinical database management system, accompanied by metadata.

Data transfer using tried-and-tested methods common to all sites (spreadsheet format) enabled a quick and relatively secure exchange, which proved highly effective. This process was further facilitated by the implementation of the TrialMaster system, which generated templates that could be adapted for use with spreadsheets. The responses received from the laboratories and CTCC are shown in Figures 3 and 6, while the questionnaire results are presented in Table S14 in Multimedia Appendix 2.

To the best of our knowledge, this is the first report to assess the degree of data commonality and interoperability between vaccine clinical trial protocols, the master protocol template, metadata, and data systems and management. In this study, we considered multiple parameters for the first time to enhance and improve interoperability across the developed COVID-19 vaccine clinical trial and master protocols, based on the selected COS areas: (1) immunogenicity, (2) safety, (3) efficacy/clinical/physiological outcomes, and (4) other outcomes of interest. This is significant because the lack of standardization in raw data processing is a well-known issue that exacerbates data inhomogeneity. This often results in errors and significant time delays, reducing the reusability and interoperability of data within protocols. To address this, the various protocol formats were homogenized by manually identifying discrepancies and commonalities among the selected parameters.

The database systems prepared by the CTCC already demonstrate substantial use of CDASH, a key enabler for the implementation of other CDISC standards. CDASH plays a crucial role in supporting the Study Data Tabulation Model (SDTM) [17]. Transforming study data to SDTM could also prove beneficial in the context of data sharing, especially as data interoperability becomes increasingly important. The adoption of CDASH standards [18] within the VACCELERATE DM systems establishes an environment conducive to generating FAIR data throughout the life cycle of the 3 trials. This approach ensures that the data will be available for collaboration and future use.

Pooling the metadata from the VACCELERATE studies would create a valuable metadata resource, independent of the underlying data. This resource could potentially be made accessible (eg, via the VACCELERATE official website and the ECRIN data-sharing repository for COVID-19 trials, EOSC-Life) [19], offering a unique opportunity for other sponsors to design related studies, particularly those utilizing the VACCELERATE infrastructure. The VACCELERATE metadata, as a shared resource, could play a pivotal role in standardizing data items and the controlled terminologies used within categorized questions, thus contributing to greater harmonization and interoperability in future research.

Education and training are critical for the successful implementation of FAIR DM in clinical data. Therefore, additional CDISC training for DM staff could be beneficial, both within VACCELERATE and more broadly, to support the adoption of data standards. Metadata harmonization should be completed before database development.

Although the DM staff at the CTCC are highly experienced and have already implemented international standards, there is a lack of planning for long-term data storage and reuse. This occurred because it was not anticipated during the database setup phase and due to the lack of a long-term data sharing budget, either linked to VACCELERATE itself or allocated to a data-sharing repository. Our report suggests that metadata harmonization planning should begin before database development and be supported by requirements (eg, CDASH) and oversight throughout the entire database setup phase, as well as any subsequent amendments. All subcontractors should be informed and included in this planning process. Furthermore, generating metadata and SDTM data for secondary use can be done under less time pressure compared with other aspects of standards implementation, such as the initial development of eCRFs. Conversely, while the final application of both is likely to occur at the end of the study, this does not mean the work must be confined to that time frame. Transformations can be planned and tested at any stage, using dummy data if necessary, as can the generation of metadata.

VACCELERATE highlights some of the challenges in making data standards a practical reality. At the same time, it provides opportunities to address these challenges, developing methods that can help embed data standards within noncommercial clinical research.

VACCELERATE received a substantial boost from the selection of the CTCC as the sole DM provider, with system database development coordinated by a single individual. This approach eliminated the need for coordination among different trial units, resulting in a high degree of uniformity within the system. The DM systems used in VACCELERATE are relatively new and have the potential to perform well with the limited number of sites currently using them. To effectively test these DM systems, it is necessary to have multiple sites utilizing them, with additional trials planned for later stages of the rollout. While the central Electronic Data Capture (EDC) system (TrialMaster) was used to manage sample processing data, there were concerns from the laboratory sites and the central biobank regarding the time required to enter sample data, resolve errors, and reconcile site datasets with the EDC. Transferring data using tried-and-tested methods common to all sites (spreadsheet format) enabled a harmonized, rapid, and relatively secure exchange. Before a scanning system was adopted, allowing sites to upload processing data directly into the EDC could significantly reduce the time constraints associated with manual data entry.

As only 3 clinical studies were funded under the VACCELERATE project, these results were not applied to any newly registered studies. However, as the objective of the VACCELERATE project has now been extended to other vaccine-preventable infectious diseases with pandemic potential, there will be a broader opportunity in the near future to implement the results presented here. Nevertheless, these trials provide valuable insights and a scalable framework for future vaccine studies, especially in the context of epidemics and pandemic preparedness. The methodology and protocols of these trials address critical questions regarding vaccine immunogenicity, reactogenicity, and optimal dosing strategies across diverse populations, including older adults (≥75 years), the general adult population, and pediatric patients. By using adaptive, multinational protocols, these trials simulate the real-world conditions of pandemic response, where speed, adaptability, and inclusivity are critical.

The methodologies from these trials are highly generalizable and provide a robust blueprint for future vaccine clinical studies within research networks. The inclusion of underrepresented populations, such as older adults and children, ensures that future vaccine trials will generate comprehensive, inclusive data applicable to all demographics. Additionally, the adaptive design of the protocols used in these studies allows for future modifications in response to emerging data and infectious diseases—a critical feature for addressing the dynamic nature of pandemics and evolving infectious threats.

Beyond the outcomes of individual trials, these studies contribute to the establishment and strengthening of international vaccine research networks. The collaborative, multicentric nature of the VACCELERATE network demonstrates the feasibility of harmonized vaccine research across borders, promoting shared protocols, centralized data analysis, and the dissemination of findings. This approach not only accelerates the vaccine development process but also strengthens global preparedness by creating investigator-initiated clinical research infrastructure and expertise that can be rapidly mobilized for emerging epidemics.

Moreover, the lessons learned from these trials can inform the design and implementation of studies for vaccines targeting other infectious diseases with pandemic potential, such as influenza or novel zoonotic pathogens. The emphasis on immunogenicity, dosing strategies, and booster efficacy can be adapted to various pathogens, ensuring that the evidence generated remains both relevant and actionable.

In summary, the VACCELERATE trials serve as a cornerstone for advancing and standardizing vaccine research methodologies. They underscore the importance of inclusivity, adaptability, and international collaboration, setting a precedent for future studies aimed at mitigating the impact of pandemics and safeguarding global public health.

This paper presents a novel approach to evaluating the degree of data commonality and interoperability between vaccine clinical trial protocols, the study master protocol template, metadata, and data systems and management. It did not benefit from preexisting methodologies, and its scope may therefore be limited by the specific cases examined. Nonetheless, it is anticipated to provide valuable insights that contribute to the improvement of clinical research data sharing and reuse.

This study demonstrates that using master protocols can significantly enhance trial operational efficiency and data interoperability when similar infrastructure and DM procedures are consistently adopted across multiple trials. To achieve data interoperability while preserving meaning in interpretation and analysis, shared data should be structured, described, formatted, and stored using widely recognized data and metadata standards.