Ethics review or compliance check? an empirical analysis of 6740 requests for information in Belgian clinical trial evaluations (2017–2024)

Abstract

The EU Clinical Trials Regulation (CTR) was introduced to harmonize clinical trial evaluations across Member States while upholding participant protection and ethical integrity. This study analyzes 6740 Requests for Information (RFIs) issued by Belgian Medical Research Ethics Committees (MRECs) across 266 trial dossiers evaluated between 2017 and 2024, spanning both the CTR pilot phase and the initial CTIS implementation. Using framework content analysis, we examined the number and content of RFIs in relation to trial outcomes, sponsor type (commercial vs. non-commercial), and the MREC’s role as Reporting Member State (RMS) or Member State Concerned (MSC).

Results show a decline in total RFIs over time, mainly due to a reduction in typographical and linguistic remarks, yet significant variability persists in the formulation and scope of ethical feedback. While statistical and methodological concerns remained central in Part I evaluations, RFIs increasingly addressed newer challenges such as decentralized trials, e-consent, and data collection on ethnicity. Part II RFIs continued to focus heavily on informed consent documents. We further observed that MSCs raised fewer RFIs than RMSs for Part I, prompting reflection on the necessity and efficiency of full multi-state review in this section.

The study also highlights a growing emphasis on regulatory compliance—sometimes at the expense of ethical deliberation—and the limited authority of policy advisors to correct inconsistencies, despite their expertise. We recommend clearer guidance, formalized roles for policy advisors in quality control, improved pre-submission processes, and limited direct communication between MRECs and sponsors. These findings support ongoing efforts to improve ethics review efficiency and quality under the CTR, with broader relevance for harmonization across Europe.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12910-025-01296-0.

Background

The path from experimental treatment to approved therapy is paved with good ethical considerations meant to protect patients—but do they truly? Each year, approximately 2400 to 3700 clinical trials receive ethical approval in the EU [1, 2], with around 60% funded by the pharmaceutical industry and 40% by academic institutions. Since 2022, all trials have been regulated under the EU Clinical Trials Regulation (CTR), which aims to safeguard participants’ rights, safety, and well-being while ensuring reliable results [3]. The CTR introduced a centralized review process, with national contact points verifying regulatory compliance with the CTR in each member state. Depending on national regulations, Medical Research Ethics Committees (MRECs) evaluate both clinical and non-clinical elements, such as study design, data protection, participant information, and the use of human tissue, with the overarching goal of safeguarding patient welfare [4]. Previous research highlighted the potential benefits of such a centralization process, with developing timely guidelines, ensuring uniform application across countries, and improving enforcement. However, centralization must not come at the cost of research participant safety [5].

The extent of MREC involvement in the review process remains a subject of debate, even after the introduction of new evaluation schemes under the CTR. This discussion continues both at the policy level and among MRECs themselves. Existing literature presents varying perspectives, ranging from viewing ethics review as unnecessary [6], or merely an administrative procedure [7], to concerns that MRECs engage in excessive oversight, functioning as a second scientific review process, often referred to as ‘ethics creep’ or ‘mission creep’ [8]. Conversely, others argue that MRECs provide essential guidance to researchers, facilitating efficient approvals while ensuring ethical standards for protecting research participants are upheld without hindering scientific progress [9]. Some even suggest that MRECs not only protect participants but also strengthen the integrity and credibility of research [5].

Unlike directives, the CTR applies directly across All member states, ensuring legal uniformity while Allowing national differences in ethics evaluations to persist. The structure and organization of MRECs, as well as certain aspects of the evaluation process, remain under national jurisdiction. Some member states have centralized ethics review systems, where a single MREC or a small number of committees oversee all reviews with strong central support. Other countries maintain decentralized systems with minimal coordination. Belgium follows a hybrid Model in which 15 accredited MRECs conduct evaluations with partial coordination. Efforts to promote greater harmonization are supported by government offices and an unofficial national ethics association composed of volunteers from MRECs and their supporting staff. With the transposition of the CTR into national law, Belgium adopted a strict interpretation of MREC independence, as outlined in Article 9 of the CTR, and implemented a system of site-independent evaluations. Under this system, an Algorithm at the government level assigns ethics evaluations to one of the 15 MRECs that are not involved in the trial. This approach ensures that the evaluating committee has no direct affiliation with the trial site. Additionally, direct contact between the clinical trial submitter and the assigned MREC is not permitted. To facilitate this process of communication with the submitters, intermediate government organizations were established [10, 11]. However, the strict independence requirements have introduced multiple administrative steps before the ethics evaluation can begin or before requests for information (RFIs) can be relayed to the submitter. These additional procedures consume valuable time within the strict CTR timelines and often result in a review process limited to a single round of RFIs or hinder timely discussion of important ethical concerns during the trial’s admissibility assessment. Given these constraints, the quality of RFIs is crucial, as they must be sufficiently clear and comprehensive to allow MRECs to make a well-founded decision on the ethical acceptability of a trial based on the submitter’s responses [8].

Numerous studies have examined REC evaluations to better understand their functioning and review procedures [7–9, 12–18]. These studies have contributed to important recommendations for improving the ethics review process, some of which have been adopted in regulatory frameworks. Key proposals include centralization [14, 19], standardization of practices [7, 9, 16, 20, 21], monitoring and evaluation [13, 20, 22], collaboration and resource sharing [19]. Other research suggested improvements, such as enhanced communication between RECs and submitters [7, 8, 14, 17], standardized informed consents and procedures [16], a more structured review process with a single point of contact [14] and increased transparency in decision-making [7, 9]. Many of these proposals have been incorporated into the EU CTR and the corresponding Belgian legislation [10, 11], aiming to further harmonize and enhance the efficiency of ethical evaluations in clinical trials.

Despite extensive research on MREC operations, communication practices, and ethics review procedures, little empirical evidence exists on how MREC evaluations have evolved [23]. Most available studies are retrospective, focus on periods predating the CTR, or are limited to individual or a small number of MRECs. Moreover, there is limited insight into how MRECs interpret and apply current ethical and regulatory frameworks in their RFIs to clinical trial submitters. This study addresses that gap by examining how Belgian MRECs evaluated clinical trials between 2017 and 2024, covering both the five-year pilot phase and the first two years of the CTR, to assess national ethics review practices over time.

Methods

This analysis was designed to address two main issues: the number of RFIs issued by Belgian MRECs and to examine their content to determine the main focus of ethical considerations. Using framework content analysis [24, 25] we explored trends in relation to final decisions (approval, conditional approval, or refusal), the role of MRECs as Reporting (RMS) or Concerned Member State (MSC), and differences between academic and commercial trials. Exploratory aims included assessing whether ethics reviews under the CTR remain focused on substantive concerns or have shifted toward compliance, and how they reflect the CTR’s goals of efficiency and harmonization—such as centralized procedures, site-independent evaluations, limited direct contact, and a single RFI round. This was done in a second round of qualitative analyses.

In drafting and structuring the results and discussion, ChatGPT (OpenAI) was used as a writing support tool under close supervision to improve clarity, consistency, and conciseness.

From government files to structured dataset: building the empirical base

This research focuses on two key periods in the evaluation processes of Belgian MRECs: the pilot phase (2017–2021) and the period from 2022 to mid-2024, when the CTIS system became operational. During the pilot phase, MRECs no longer assessed trials solely from their own institutions but engaged in a site-independent, two-part review process involving additional government intermediaries and the production of a joint assessment report. The CTIS phase built on these foundations while introducing an EU-wide, multi-state evaluation process. This shift enabled a more detailed analysis of evolving review dynamics under the CTR, particularly with regard to Belgium’s role as either a MSC or the RMS. An Excel file provided by the Belgian government listed all trials submitted for ethics assessment during this period, including basic trial information and the conclusions from the National Contact Point (NCP), MREC, and the consolidated final decision. The pilot-phase dataset did not include information on whether trials were commercial or non-commercial, and roles such as RMS or MSC could only be identified in the CTIS phase. In addition, several variables were derived from the assessment reports (ARs) and used to construct codebooks, which served to classify the RFIs issued by MRECs (Table S2-S4). The following section explains how RFIs were categorized to allow consistent analysis of their type, focus, and origin across both evaluation periods.

Mapping ethical concerns: a three-tiered coding system

The CTR restructured clinical trial evaluations into two parts: Part I, covering clinical aspects, and Part II, focusing on participant-related aspects. While Part I involves a coordinated assessment across Member States where the trial is conducted (unless mono-national), Part II remains a national competence. RFIs were categorized using fixed variables in an Excel file, based on the assessment report categories for Part I and designated categories for Part II, in line with Articles 6 and 7 of the CTR and Belgian regulations defining the division of responsibilities between the NCP and MRECs. In Belgium, MRECs assess selected topics within Part I such as benefit-risk assessment, methodology and statistics, investigational product details, recruitment and consent procedures, inclusion and exclusion criteria, endpoints, safety monitoring, and vulnerable populations while the NCP is responsible for non-clinical and safety-related aspects. Part II, fully reviewed by MRECs, covers written information quality in the information form and consent form, participant compensation, biological samples, insurance, and medical care. Only RFIs issued by MRECs were included in this analysis. Categorization followed a three-tiered approach: RFIs were first classified as Part I or II, then into subcategories, and finally into detailed sub-questions aligned with the assessment reports (Table S2-S4). A separate category captured Typographical, Linguistic, and Coherence (TLC) issues, covering language-related remarks ranging from typographical errors to inconsistencies in spelling and phrasing. The functioning of this categorization was evaluated during the Belgian Association for Research Ethics Committees (BAREC) symposium, where intermediary results were presented and member checking was conducted [26, 27].

Descriptive statistics

Descriptive statistics were conducted using SPSS, focusing on frequency distributions and medians to explore trends and patterns in the data. No inferential statistical tests were performed. Nevertheless, the analysis provides valuable insights into recurring themes in MREC ethics reviews and highlights areas where further harmonization and procedural improvements could be pursued.

Qualitative analysis and categorisation of MREC RFIs in CTIS

In the quantitative phase of this research, a total of 3,045 RFIs submitted through CTIS were analyzed. The researcher identified a subset of particularly notable RFIs based on their relevance to harmonisation efforts among MRECs and the extent to which they aligned with existing guidance, legislation, or prior inter-MREC agreements. These RFIs were classified into six categories: These RFIs were classified into six categories: (A) relevant and suitable for inclusion in shared guidance, (B) contradictory to existing templates or MREC agreements, (C) non-compliant with legal requirements, (D) typographical, language, or coherence-related comments that were overly detailed or incorrect, (E) mutually contradicting RFIs, and (F) RFIs where the question or comment was unclear or ambiguous. A preliminary list of 172 selected RFIs was compiled, reviewed, and cross-checked with experts from UZ Brussel to refine interpretations and identify broader themes. A final set of 141 RFIs was then validated by members of the BAREC board.

The qualitative analysis followed a generic qualitative research approach. The RFIs were analyzed and interpreted by the researcher using reflexive thematic analysis, as described by Braun and Clarke [28], allowing for a flexible and iterative exploration of patterns of meaning within the data. Themes were identified inductively, and analytical memos were maintained throughout the process to support transparency, traceability, and reflexivity in interpretation.

To assess the accuracy of the classification system, a peer-checking session was held during the BAREC symposium, with participation from representatives of All 15 recognized MRECs [29]. Using the interactive tool Wooclap, participants were invited to assign these preclassified RFIs to the proposed categories and discuss their reasoning. Feedback and notes from this session, including input from CT-College and BAREC board members, were incorporated into the final qualitative analysis and conclusions.

Results

To explore trends in the ethics review process under the CTR framework, we analyzed a representative sample of dossiers from both the pilot and CTR phases. For the pilot phase (2017 to 2021), a total of 125 dossiers were analysed, representing approximately 25% of All available pilot dossiers. To ensure comparability, a similar proportion was applied to the CTIS phase covering the period from 2022 to mid-2024. From the 524 eligible CTIS dossiers, we selected those involving initial clinical trial applications with completed or withdrawn assessments. 25% of these dossiers, or 138 in total, were included in the analysis. Sampling was stratified to ensure proportional representation across key variables relative to the full set of 524 dossiers. These variables included Member State role, distinguishing between reporting and concerned Member States, sponsor type, and outcome. Within each subgroup, dossiers were selected at random. The yearly distribution varied according to dossier availability. All dossiers from 2017 were included, while later years, such as 2023, accounted for a larger share of the sample—38%—due to the higher volume and completeness of dossiers submitted during that year (Table 1).

Table 1.

Yearly oversight of processed CT reviews

Year	N Processed CT reviews	Percentage of Total Processed Reviews in this study (%)	Percentage of all CT Reviews in that Year (%)	Total number of RFI
2017	9	3.4%	100%	338
2018	32	12.0%	94%	849
2019	28	10.5%	30%	1092
2020	32	12.0%	15%	759
2021	27	10.1%	20%	641
2022	18	6.7%	22.2%	493
2023	100	37.5%	26.04%	2219
2024	20	7.5%	34.4%	349
Total	266			6740

Quantitative analysis of MREC considerations

A total of 6740 RFIs were analyzed, including 3679 from the pilot phase and 3061 from the CTR phase (Table 1). The highest number of RFIs in a single MREC assessment was 179 during the pilot phase and 98 during the CTIS period. In the pilot phase, only 2 out of 128 evaluations had no RFIs, compared to 18 out of 138 during the CTIS period. Most dossiers without RFIs during the CTIS phase were commercial trials. Figure 1

Fig. 1.

Median RFI Pilot versus CTIS

As shown in Fig. 2, the median number of total RFIs per dossier decreased between the pilot and CTIS phases. This decline appears mainly driven by a reduction in TLC remarks. Median values for Part I and Part II RFIs remained relatively stable.

Fig. 2.

Total RFI Median

To determine the focus of ethical considerations during the pilot phase, we analyzed the content of RFIs issued by MRECs. In Part I, the most frequent RFIs concerned statistical and methodological aspects (23%), focusing on sample size calculations, study design, and justification of methods. Additional RFIs addressed data protection measures (11%) and the clinical trial rationale (9%). The “Other” category (9%) captured issues not directly covered in the assessment report, such as missing translations of the protocol synopsis or incomplete sponsor details. In Part II, 77% of RFIs focused on the written information in the information form. Within this, key issues included unclear descriptions of trial conditions (14%), missing or incomplete data processing information (13%), inconsistencies between the protocol and the ICF (13%), and vague details about participant compensation (5%). RFIs concerning the consent process (6%) addressed missing placeholders for pregnant partners or legal representatives, as well as inconsistencies between ICF titles and protocol titles. Additionally, 4% of RFIs related to GDPR compliance, particularly regarding data transfers outside the EU and the alignment of data practices with national regulations.

To assess the main ethical focus during the CTIS period, we analyzed the content of MREC-issued RFIs. In Part I, the most frequent RFIs were classified under the ‘Other’ category (32%), which includes diverse concerns that do not fall neatly within the predefined headings of the assessment report. These RFIs addressed issues such as the inclusion of ethnicity as a study variable, inconsistencies within the protocol, elements related to decentralized trial conduct, and requirements concerning contraception. Statistical and methodological issues followed (26%), with common requests for clarification on randomization, sample size justification, and proposed analyses. RFIs related to the clinical trial rationale accounted for 8%. In Part II, the information form provided to participants remained the predominant source of RFIs (76%). Within this, the most frequent issues concerned unclear descriptions of study procedures (22%), followed by “Other” concerns such as missing references to Belgian legislation and questions about decentralized elements (16%). Protocol-ICF inconsistencies made up 14%, while 13% focused on data protection and GDPR compliance. A further 8% addressed unclear or insufficient information on participant compensation. Additional RFIs targeted the consent process (7%), with specific concerns about missing placeholders for the representatives that need to sign the ICF. Investigator suitability (4%) and recruitment details (3%) were also raised, primarily due to outdated CVs, missing GCP certificates, or lack of detail on recruitment steps.

Trends in RFIs

The role of the MREC as MSC or RMS

A notable trend in our dataset is the influence of Belgium’s role as RMS or MSC on the number and type of RFIs issued for Part I. When acting as RMS, the median number of Part I RFIs was 2, compared to 0 when serving as MSC. To assess the content of RFIs more closely, we compared evaluations based on the MREC’s role as RMS or MSC (Fig. 3). In the MSC role, RFIs most often fell into the “Other” category (40%), followed by statistical concerns (27%) and clinical trial rationale (13%), broadly reflecting overall trends. When acting as RMS, MRECs issued similar types of RFIs but in different proportions: “Other” (28%), statistical concerns (26%), and safety and monitoring (8%). Safety and monitoring issues were rarely raised in the MSC role, accounting for only 0.7% of RFIs. As Part II is reviewed nationally, it was excluded from this comparison.

Fig. 3.

CTIS Median RMS vs. MSC

The outcome of the ethics evaluation: accepted, accepted with conditions or a refusal

Across the combined CTIS and pilot datasets, trial outcomes—accepted, conditionally approved, or not accepted—were associated with clear trends in the number and type of RFIs. Outright refusals were rare, occurring in only 11 of the 266 ethics evaluations included in our dataset. However, this figure reflects a representative proportion of the total submissions due to the purposive sampling strategy. In contrast, 117 trials received conditional approval, 5 were withdrawn after receiving RFIs but before a final decision, and 133 were fully accepted following ethics evaluation. Conditionally approved trials had the highest median number of RFIs, with a median of 27. Trials that were not accepted had a median of 19, while fully accepted trials had a median of 12. This pattern was consistent across RFI categories. TLC RFIs showed a median of 7 in conditional approvals, 6 in not accepted trials, and 3 in accepted ones. Part I RFIs were most frequent in not accepted trials, with a median of 7, compared to 2 in conditionally approved trials and 1 in accepted trials (Fig. 4). Part II RFIs were highest in conditionally approved trials, with a median of 13, compared to 7 in not accepted trials and 6 in accepted trials. For Part I, non-accepted trials most frequently raised issues in the “Other” category (38%), followed by clinical trial rationale (25%) and statistical concerns (12%). Conditionally approved trials showed a similar distribution, with most RFIs related to statistical concerns (28%), “Other” (27%), and clinical trial rationale (9%). Fully accepted trials mainly raised RFIs in the “Other” category (26%), followed by statistical concerns (23%) and inclusion (7%) or exclusion criteria (8%). In Part II, the content of the information form remained the dominant issue across all trial outcomes. Non-accepted trials raised RFIs mainly concerning the information form (76%), consent form (15%), and data protection (3%). Conditionally approved trials showed a comparable pattern, with the information form accounting for 76% of RFIs, consent form 8%, and recruitment 4%. Fully accepted trials had an even higher proportion of RFIs concerning the information form (89%), while RFIs about the consent form and data protection were low (3% and 2%, respectively).

Fig. 4.

Differences in outcome of ethics evaluation by MREC

Commercial versus non-commercial trials

Across the dataset, non-commercial trials generated notably more RFIs than commercial ones, with a median of 27 compared to 17. This trend was reflected across all categories: TLC RFIs had a median of 8 in non-commercial versus 3 in commercial trials; Part I RFIs had a median of 4 versus 1; and Part II RFIs reached 11 in non-commercial trials, compared to 8 in commercial ones (Fig. 5). Content-wise, both groups showed similar patterns in Part I, with the majority of RFIs falling into the “Other” and “Statistical and Methodological” categories. However, in non-commercial trials, RFIs related to inclusion and exclusion criteria (9% and 8%) were more frequent than in commercial trials (5% and not ranked for inclusion). In Part II, the information form remained the top concern in both groups (78% for commercial, 70% for non-commercial). However, RFIs targeting the suitability of investigators and facilities were more frequent in non-commercial trials (9% and 4%) than in commercial ones (3% and not listed).

Fig. 5.

CTIS Median Commercial versus Non-Commercial

Qualitative results: variability in regulatory interpretations and quality of MREC RFIs

Despite the reduction in considerations over time, notable variability remains in how MRECs approach dossier reviews. Although the specific MREC was not a variable in this study, clear differences in review practices were observed. Some committees provided detailed feedback addressing both minor and major issues, while others focused primarily on critical safety and compliance concerns. During the pilot period, evaluations were often communicated via emails and letters, and several RFIs displayed typographical errors, duplicate questions, and poorly formulated English. However, following the introduction of CTIS and the centralized EU-wide submission process, these issues appear to have decreased.

Among the 6740 RFIs analyzed, the clarity of the questions remained a challenge in both periods. It was often unclear what specific revisions were expected from submitters. Language issues, likely due to non-native English usage, further complicated interpretation. Many RFIs lacked explicit instructions, such as requests to replace terms like “racial” in ICFs without suggesting alternatives. Inconsistencies were also apparent between MRECs. For example, one MREC objected to the use of “teaspoons” to quantify blood volume, requesting a different comparison, while another encouraged using measures like “teaspoons”. Similarly, views on abstinence as a contraception method varied, with some committees recommending its inclusion and others insisting on its removal.

Several RFIs contained incorrect or incomplete interpretations of regulations. Examples included questioning the legally required retention period of trial data and the removal of participant consent regarding sample anonymization. Some RFIs also incorrectly stated that collecting data on race or ethnicity is per definition prohibited by law.

Typographical, language and coherence in informed consent reviews

As shown earlier, TLC-related RFIs constituted an important portion of MREC feedback, particularly in the review of ICFs. Practices varied: some MRECs issued detailed lists of TLC corrections, up to 161 TLC RFIs in a single review, while others flagged only a few examples or broadly mentioned TLC concerns without specification. Common TLC RFIs involved correcting abbreviations, simplifying technical language, and clarifying vague terms. Examples include replacing “oral” with “via the mouth” or explaining technical terms like “sedation”, “intravenous”, “catheter”, or “ultrasound”. MRECs also requested clarification of terms such as “significant” and flagged formatting inconsistencies, like errors in numbering, inconsistent fonts, text colors, and missing headings or appendices. Punctuation mistakes and minor phrasing suggestions (e.g., “ask more information” corrected to “ask for more information”) were frequently noted. Additionally, sloppy errors, including untranslated sections or inconsistencies in basic instructions, were highlighted.

Discussion

This study offers empirical insight into how Belgian MRECs conduct ethics reviews under the CTR and whether harmonization efforts have shaped their decision-making processes. Contrary to the Belgian government’s assertion in its ‘Recommendations of the College Board to Reduce the Number of RFIs Raised During the Assessment of a Clinical Trial Study’ [30], the total number of RFIs has actually decreased over time as shown in Fig. 2 with a consistent decline since 2019. However, when analyzing the first-tier categories (Table S2), this downward trend appears to be driven by a reduction in TLC RFIs.

Better preparation and pre-submission support: strengthening ethics review efficiency

As also shown in other research [16, 18], our findings could indicate that improving the quality and completeness of submissions could prevent a substantial number of MREC RFIs. Non-commercial trials had 27 median number of RFIs and commercial ones 17 (Fig. 5), possibly due to the absence of dedicated regulatory departments or limited internal capacity for thorough pre-submission review. In contrast, commercial sponsors typically have specialized teams and multiple internal checks in place before dossier submission. Trials that were not accepted had the highest number of RFIs in Part I, pointing to unresolved methodological or ethical concerns as key reasons for rejection (Fig. 4). However, most trials received conditional approvals rather than outright refusals, suggesting that MRECs sought to enable adjustments rather than block research. Rejected trials showed more RFIs related to clinical trial rationale and methodological robustness, underlining the importance of a well-justified design [8]. Conditionally approved trials typically had more Part II RFIs, particularly concerning the consent form, information form, and recruitment. These remaining conditions in the final decision suggest that initial RFIs were not always fully resolved. Unaddressed issues are a basis for approval conditions, although they are not considered sufficient to prevent the trial from starting. Regardless of the outcome, many RFIs focused on TLC issues, primarily in the informed consent forms (Fig. 1), a trend that has also been confirmed by earlier research [15, 16]. While clarity in participant documents is critical, the volume of TLC RFIs suggests that some MRECs are effectively acting as copyeditors, raising concerns about efficient use of their time. The demand for faster reviews reinforces the need for better-prepared submissions.

Trials accepted after a single round of RFIs had fewer issues related to data protection and ICF structure, possibly indicating that clear, compliant, and well-organized materials contribute to faster approvals. Based on our findings, we conclude that the overall distribution of RFIs highlights the need to improve dossier preparation, particularly in areas such as study design, participant information, and regulatory alignment. Strengthening these elements could help reduce the number of conditional approvals and outright rejections. This improvement can be further supported by involving ethics advisors early in the research design process [31].

The role of MRECs in part I evaluations: ethics creep or essential oversight?

In the pilot period, Part I considerations primarily focused on statistical and methodological issues. In contrast, during the CTIS period, the ‘Other’ category became the most frequent, encompassing a broader range of concerns not directly covered in the assessment report, including ethnicity as a study variable, decentralized trial elements, and contraception requirements. While statistical and methodological issues remained an important part, the emergence of new challenges might reflect the evolving landscape of clinical research.

One notable shift was the increasing use of decentralized trial components, accelerated by the COVID-19 pandemic. As access to healthcare facilities became restricted, remote monitoring and home-based procedures expanded. While these approaches improve access and participation, they also raise new ethical, legal, and technical challenges, particularly regarding patient safety, data integrity, and compliance [32].

Despite the shift in focus under the CTR, between 2022 and 2024, 106 out of 411 Part I RFIs in the CTIS dossiers still addressed statistical and methodological concerns. These included issues related to randomization, blinding, sample size, interim analyses, and overall study design. This pattern is consistent with earlier research, which also found that a significant proportion of RFIs targeted the scientific and methodological foundations of clinical trials [8, 9, 14, 33, 34]. These considerations can be essential to ensure methodological rigor and to avoid enrolling participants in research lacking scientific validity. While sometimes referred to as “ethics creep”, the notion that ethics committees overstep into scientific or regulatory domains, such reviews, when addressing fundamental flaws, are ethically justified [8, 13, 33]. A poorly designed study not only wastes resources but also exposes participants to risk without producing meaningful knowledge [8]. The severity of the RFIs varied widely: From requests to rephrase unclear formulations for better clarity, to concerns about potential bias, or even the complete absence of statistical descriptions. This range reflects a broader trend seen in other categories as well, spanning from essential errors or omissions to relatively minor details. Ensuring scientific validity is a core ethical responsibility, as emphasized in the Nuremberg Code [35] and reaffirmed in the Declaration of Helsinki [36]. MRECs thus play a vital role in safeguarding both the integrity and ethical justification of clinical research. We feel the question of “ethics creep” may be more relevant when considering legal analyses done by MRECs versus ethical ones, which is further discussed in the section on legal review and compliance issues.

Another recurring issue during the CTIS period was the collection of ethnicity or race as a study variable, reflecting broader societal discussions on diversity, inclusion, and discrimination. Several guidance documents have been developed [37] to address the importance of collecting such data while ensuring that it is gathered ethically, used responsibly, and framed appropriately [38]. In clinical trials for pharmaceutical investigational products, these variables are often justified for scientific and regulatory purposes. However, ethical concerns arise when data collection risks reinforcing discrimination, bias, or stereotypes [39]. MRECs must carefully evaluate whether such data collection aligns with the principles of beneficence and justice, ensuring that it provides value both to the groups described and to society as a whole, while also complying with GDPR and national legal requirements. Crucially, the collection of ethnicity or race data should be explicitly justified by a clearly defined research question or a hypothesis that the trial seeks to test.

A comparison between RMS and MSC evaluations reveals a notable disparity in the number of RFIs issued for Part (Fig. 3). When the Belgian MREC acted as RMS, evaluations generated a median of 2 RFIs; as MSC, the median was 0. This likely reflects the RMS’s lead role in coordinating the EU-wide review process. The finding raises an important question: Is a full multi-state review of Part I necessary? Given the limited input from MSCs, streamlining the process could reduce redundancy and improve efficiency. Releasing MSC reviewers from duplicate Part I assessments may allow them to focus more meaningfully when acting as RMS. Nonetheless, all Part I documents remain essential for Part II evaluations, as MREC members must assess their consistency with the ICF.

The expanding role of mrecs: ethical guardians or compliance gatekeepers?

During both the pilot and CTIS periods, Part II RFIs consistently focused on the quality of written information in ICFs, as noted in earlier research [13–15, 23, 34]. Most legal issues, such as compliance related to biological samples or the use of personal data, were addressed in Part II RFIs. Although the AR includes specific checkpoints for data protection in Part I (Table S3) our research showed that concerns about data protection were often raised only in Part II, particularly in relation to the written information provided to participants (Table S4). These RFIs frequently went beyond the ethical obligation to inform participants, as required by the principle of autonomy, and reflected broader legal evaluations of the trials data processing practices. During the pilot phase, data protection emerged as a key issue, especially following the introduction of the GDPR in 2018 [40]. In the CTIS period, data protection remained a central concern, highlighting the possible ongoing difficulties both sponsors and ethics committees face in interpreting and applying these rules in health research. We expect this trend to continue as clinical trials increasingly depend on large and complex datasets. It raises important questions about current legal frameworks and whether the expertise available within MRECs is sufficient to address the evolving legal and ethical challenges posed by the use of big data in research [41].

During the CTIS period, new issues emerged in the ‘Other’ category, particularly those related to decentralized trial elements and electronic consent. These topics also raised concerns about data protection. While initially driven by the COVID-19 pandemic [42], decentralization has now become a standard feature in trial design. This shift has made ICFs more complex, especially in cases involving subcontractors managing tools like e-diaries and questionnaires. Many RFIs focused on GDPR Article 13 disclosures [40], highlighting the growing need for transparency in participant information amidst the increasingly complex exchange of personal data between sites, sponsors, and third-party service providers.

Adding to the issue of legal review by MRECs, and confirming earlier research [8, 13, 20, 21], substantial variability in feedback between MRECS persists. This ranges from relatively harmless contradictory suggestions, such as whether to express blood volume in teaspoons or milliliters, to conflicting interpretations of legal obligations or guidelines by different MRECs (Table S1). In addition, some of these legal or compliance-related RFIs are, in certain cases, based on incorrect legal interpretations. For example, one RFI questioned the retention of participant data for 25 years, which the MREC considered excessive, even though this retention period is legally required under the CTR. In another case, a seven-year retention period for financial data managed by a subcontractor responsible for participant payments was challenged, despite being fully compliant with Belgian financial legislation.

This raises a broader concern: is the ethics review process shifting toward compliance auditing? Legal input is important, but when legal professionals are absent, non-experts may apply legal standards inconsistently or overly strictly. As observed in this study and others [9, 20], ethics committees are increasingly tasked with compliance checks, which may overshadow their core role in ethical reflection. Could this, in fact, be the real ‘ethics creep’? We suggest categorizing the assessment into two types: compliance-related and ethics-based. As Tusino et al. note, centralization risks reducing ethics review to a procedural exercise [5]. Our findings support this, suggesting that excessive focus on regulatory compliance may undermine the primary mission of ethics committees: to protect participants, not merely ensure formal legal adherence. It is understandable that MREC reviewers may not be familiar with All specific legal retention requirements. Nevertheless, certain standards, such as the 25-year retention period under the CTR, should be widely known among MREC members. Involving legal experts in a preliminary review phase could help clarify what is legally required, reduce unnecessary RFIs, and support a more efficient and focused ethics review process.

A revised assessment strategy that separates ethical and compliance evaluations could help. In Belgium, the NCP already performs initial document validation. Introducing a structured precheck process focused on legal and regulatory compliance could help address many of these issues early in the review process. As noted in other research [13], this would allow MRECs to focus more effectively on their core mandate, which is the ethical evaluation of clinical trials. Government actors, better placed to interpret legal standards, could thus ensure consistent application while freeing MRECs to focus on participant protection and ethical scrutiny.

Standardization practices in ethics review: balancing structure and contemplation

Is meticulous attention to detail, such as referencing page numbers for minor issues, truly within the scope of MREC ethics review [22, 43]? While overly granular feedback may appear excessive, unclear or poorly structured ICFs can undermine participant understanding and, ultimately, patient autonomy. The introduction of a standardized ICF template in 2019 by the Belgian government [44] has possibly already improved submission quality (Fig. 2), reducing the number of MREC comments and enabling more focused reviews, particularly when deviations from mandatory text are clearly flagged. However, the template remains optional. Making its use mandatory, with defined consequences for non-compliance, could further enhance consistency and dossier quality. In Belgium, a policy shift in this direction is expected from August 2025, though its practical impact and uptake by sponsors remain to be evaluated. That said, we question the ethical value of excessively long informed consent forms, often extending to 40 pages and filled with legal disclaimers. Such documents may comply with formal requirements but fail to support meaningful autonomy or truly informed decision-making. The ICF template of the Belgian government and BAREC itself increasingly reflects the tension between compliance and ethics, raising questions about whether legal thoroughness serves, or hinders, the ethical principle of informed consent.

In relation to the usefulness of an ICF template, we support the view also reflected in the literature that a certain degree of standardization can enhance the efficiency and consistency of ethics reviews [7, 9, 16, 20, 21]. This is particularly important given the earlier mentioned contradictions between MRECs in interpreting similar issues. Standardization may also help ensure that the review process remains focused on ethical evaluation instead of legal evaluations. However, standardization alone is not sufficient. While it may seem at odds with flexibility, we emphasize that evaluating elements beyond strict regulatory compliance requires room for ethical reflection. The key is striking the right balance: promoting transparency for submitters [7, 9], through tools such as a standardized ICF template, where unchanged sections are pre-approved by all Belgian MRECs, while still preserving space for meaningful, unstructured ethical dialogue within committees. This requires that MREC members are trained in ethical reflection rather than simply memorizing legal provisions, though core legal knowledge remains essential. Such training can be supported through ongoing ethics education [9, 22, 34], engagement in peer discussions at national and international ethics conferences [21], and transparent information-sharing practices [21, 22]. In this study, the AR was used to categorize RFIs (Table S2-S4), yet many were still placed in the “Other” category. This suggests that the current AR does not fully capture the range of concerns MRECs encounter during ethics review. Revising the assessment report into a more comprehensive tool focused on ethical reflection, rather than maintaining its current compliance-driven tone, could enhance its relevance for ethical evaluation, as Morton has also suggested through a reflective equilibrium approach [45]. It is therefore essential that such a tool does not become a rigid checklist. Instead, we propose the development of an “ethics contemplation guidance framework” that offers structure while preserving space for meaningful ethical deliberation, as the literature has suggested may require a shift in mindset in some cases [12].

Transparency in ethics review: enhancing communication between MRECs and submitters

Previous research suggests that improving transparency and communication in ethics review can help shift the perception of MRECs from bureaucratic bodies to trusted evaluators [7, 8, 14, 17, 21, 22]. In our research we noticed that the quality of RFIs remains inconsistent. While the professionalism of RFI letters has improved compared to the pilot period, when feedback was provided through written letters, the centralisation of RFIs via CTIS has not fully resolved ongoing concerns about clarity and consistency. Similar to previous research, RFIs remain vague or poorly formulated [8, 16], causing confusion and avoidable delays [13]. For instance, a comment like “The word racial or race has to be replaced in all ICF documents” offers no guidance on appropriate alternatives. Similarly, “We just like to know what is the title and the role of the physician who signed for the institution” lacks clarity on relevance or necessity. Such ambiguity raises questions about whether quality control is consistently applied before issuing RFIs.

Well-structured RFIs are essential to a constructive and efficient review. Previous studies emphasize that an effective RFI includes three core elements: (1) a clear rationale for the question, (2) a well-formulated question, and (3) an indication of the desired solution. For example: “1. The standard of care (SOC) treatment is X, 2. but here Y is used. 3. Please justify this deviation in the protocol. Solutions should be evidence-based and, when possible, reference relevant guidelines.” Clear RFIs allow submitters to respond effectively, ideally within a single review round, whereas ambiguous or unfocused feedback increases the likelihood of conditions and prolongs the review process [7, 8].

Structural constraints in Belgium also impact transparency. MREC independence, while important, currently requires all communication to pass, in written, through multiple government intermediaries, limiting the opportunity for direct clarification. This can contribute to inefficiencies, delays, and perceptions of bureaucratic detachment [22]. Allowing direct oral communication with submitters, especially to clarify unclear points, could streamline the process and reduce unnecessary RFIs [7, 17]. Oral communication could significantly improve transparency and efficiency in the ethics review process. Brief discussions or structured Q&A sessions can prevent misunderstandings and lead to more precise, timely responses. Recognizing these limitations, Belgian authorities recently advised including principal investigators (PIs) in MREC meetings to provide direct clarifications [30]. While this solution may work well in academic trials, it is less practical for commercial EU trials, which make up approximately 60% of submissions. Local PIs often lack detailed knowledge of the study protocol and must relay MREC questions to the sponsor, increasing the risk of miscommunication and delays. To address this, we recommend allowing direct communication between MRECs and sponsor representatives, without mandatory filtering by government bodies. It is important to emphasize that MRECs in Belgium operated effectively under such a model for over two decades prior to the implementation of the CTR, without, to our knowledge, any reports of undue sponsor influence or ethical compromise. Restoring this practice would enable more accurate, efficient exchanges while maintaining the integrity and independence of the ethics review process.

The limitations of guidelines and the role of policy advisors in MREC feedback

Clear guidelines could contribute to greater harmonization in ethics reviews. However, as previously noted, such guidelines are not always consistently followed and may even be contradicted in practice. This limits their effectiveness and underscores the need for both clearer interpretation and shared understanding among MRECs (Table S1). In Belgium, the BAREC represents all MRECS involved in clinical research evaluations. BAREC regularly organizes meetings, forms ad hoc working groups, and engages with federal healthcare authorities on ethics issues [46]. Although BAREC has no legal authority and is composed of volunteer MREC members and policy advisors, it has issued several guidance documents to address recurring RFIs, including topics such as ‘Compensation for Contraception’ [47], ‘Guidance on Compensation for Clinical Research Participants’ [48], ‘Statement on Decentralized Trials’ [49], ‘ICF and Trial at a Glance’ [50], and ‘Insurance Coverage’ [51]. Despite these efforts, our study conducted early in the guideline implementation phase does not show an elimination of RFIs on these topics. In addition to a possible lack of awareness among submitters, a further explanation can lie in how these guidelines are applied and enforced during the review process.

Interim presentations of this study at BAREC symposia in 2022, 2023 and 2024 highlighted the central role of policy advisors in the operations of Belgian MRECs [26, 27, 29]. These staff members are employed by the host institutions and act as the operational backbone of the committees. In contrast to MREC members, typically volunteers with clinical or patient advocacy roles, policy advisors bring continuity, regulatory expertise, and in-depth procedural knowledge. They structure RFIs, ensure coherence and quality in committee feedback, and contribute to national and European policy discussions. However, as they are not legally appointed MREC members, they cannot formally participate in ethical deliberations or correct RFIs, even when these contradict official guidelines. This structural limitation was illustrated by an RFI issued in 2023 stating: “Note that in Belgium, electronic consent is not at all standard and will not be accepted,” despite the existence of a government-approved guideline permitting e-ICFs under specific conditions [52]. That such a comment was issued, despite the policy advisors’ awareness of the guideline, underscores the need to better align authority with expertise.

Strengths and limitations

This study contributes to the ongoing discussion about the evolving role of ethics committees in balancing compliance demands with ethical integrity in clinical trials. While the analysis highlights important trends in RFI content, volume, and review practices, several limitations must be acknowledged. The purposive sampling strategy and underrepresentation of certain years limit generalizability, and dossier complexity (e.g., multiple ICFs) was not controlled for. Additionally, MREC identity was anonymized, potentially skewing representation, and some categorization relied on subjective judgment in the absence of standardized assessment reports. The study was limited to Belgium, the broader applicability of the findings may be constrained. The classification was conducted by a single researcher. Nevertheless, interim presentations of the findings to the 15 MRECs and relevant government bodies confirmed that the trends observed were recognizable and not disputed [26, 27, 29]. The use of recent, nationwide data covering both the pilot and CTR periods offers rare empirical insight into current ethics review practices. The findings suggest actionable opportunities to reduce administrative burdens, enhance dossier preparation, improve RFI structuring, and clarify the boundary between ethics and compliance. The formal recognition of policy advisors and the implementation of structural reforms, such as assigning the legal compliance review to government bodies through a more thorough precheck and reserving the ethics review by MRECs for genuinely ethical questions, could significantly improve the review process. Additional reforms may include revising the AR into an ethics contemplation checklist, eliminating multiple Part I revisions by different Member States in favor of mutual trust in the assessment of the reporting Member State, reconsidering the composition of MRECs to better address challenges posed by big data research, and allowing direct communication between submitters and MRECs for ethics-related questions. Furthermore, training opportunities through international conferences for MREC members and policy advisors, along with improved international information sharing, could further support a more consistent, efficient, and ethically focused review system. Although context-specific, the trends and challenges identified in this study are likely relevant for MRECS in other EU Member States operating under the CTR. Shared issues such as growing legal complexity, inconsistent RFI quality, and limited review timeframes make these insights broadly applicable to ongoing efforts toward harmonized, high-quality ethics evaluation across Europe.

Conclusion

This study identifies key opportunities to improve ethics review under the CTR. Strengthening pre-submission support through structured checks and clearer guidance could reduce avoidable RFIs, especially in areas like statistics, informed consent, and compliance. Revising the assessment structure to distinguish ethical from legal evaluations would help MRECs focus on substantive ethical concerns rather than administrative editing. While sometimes labeled as “ethics creep”, MREC scrutiny of methodological and scientific validity remains an ethically necessary function. We also recommend restoring limited, direct communication between MRECs and sponsors. Controlled direct contact would improve response accuracy and reduce avoidable delays.

Though this study is limited to Belgium, the patterns observed have broader relevance. The future of ethics review lies in striking the right balance: efficiency and harmonization must be pursued without losing the depth, flexibility, and critical reflection that define ethical oversight.

While greater consistency is necessary, it must not come at the expense of flexibility in ethical reflection. Ethics review is not a mechanical checklist but a deliberative process requiring thoughtful ethical scrutiny. The primary goal should not be simply to minimize the number of RFIs, but to ensure that all relevant ethical concerns are thoroughly addressed.

Abbreviations

AR

Assessment Report

BAREC

Belgian Association for Research Ethics Committees

CTR

Clinical Trials Regulation

CTIS

Clinical Trial Information System

DP

Data Protection

EU

European Union

GCP

Good Clinical Practice

GDPR

General Data Protection Regulation

ICF

Informed Consent Form

MDR

Medical Device Regulation

MREC

Medical Research Ethics Committee

MSC

Member State Concerned

NCP

National Contact Point

PI

Principal Investigator

RFI

Request for Information

RMS

Reporting Member State

SOC

Standard of Care

SPSS

Statistical Package for the Social Sciences

TLC

Typographical, Linguistic, and Coherence

Audrey Van Scharen

holds Master’s degrees in Law and Political Science from the Vrije Universiteit Brussel (VUB). She is Manager of Ethics at the University Hospital Brussels (UZ Brussel), where she also serves as Vice President of the Medical Ethics Committee of the Vrije Universiteit Brussel and Chair of the Health Ethics Committee of UZ Brussel.

She chairs two working groups within the Belgian Association of Research Ethics Committees (BAREC), focusing on decentralized trials and medical devices. As an affiliated professor at the VUB Faculty of Medicine and Pharmacy, she teaches courses in bioethics, research ethics, and ethics in health management and policy.

Drawing on her practical experience in ethics evaluations and the application of regulatory frameworks such as the CTR, GDPR, MDR, and IVDR, she is currently pursuing a PhD at the Faculty of Medicine and Pharmacy of the Vrije Universiteit Brussel. Her research focuses on the sense and non-sense of the roles attributed to medical research ethics committees, under the supervision of Prof. dr. Pieter Cornu, Prof. dr. Maarten Moens, and Prof. dr. Michel Deneyer.

Authors’ contributions

A.V.S. conceived the study, conducted the research, performed the analysis, and wrote the full manuscript. P.C. provided support with statistical analysis and interpretation. All authors reviewed and approved the final manuscript.

Funding

na.

Data availability

The data that support the findings of this study are available from the Belgian Federal Agency for Medicines and Health Products (FAMHP). Restrictions apply to the availability of these data, which were used under license for the current study and are therefore not publicly available. Data are, however, available from the authors upon reasonable request and with permission of the FAMHP. We have updated the Data Availability statement in the system accordingly.

Declarations

Ethics approval and consent to participate

NA

Consent for publication

NA

Competing interests

The authors declare no competing interests.