Purpose versus pragmatism: participant perspectives on the process of ethical review of animal research in Sweden

1 Introduction

In the EU, researchers seeking to use non-human animals (henceforth, animals) in their work must obtain prior approval from a competent authority tasked with assessing applications based on legal and ethical criteria (Chapter IV Section 3 Article 36 of Directive 2010/63/EU on the protection of animals used for scientific purposes (henceforth, Directive 2010/63/EU)). Only if this project evaluation, or ethical review, establishes that the project is legally required or justified from a scientific or educational point of view (Chapter IV Section 3 Article 38 p. 1a-b, 2a) and has adequately implemented the 3Rs (Replacement, Reduction, and Refinement) (Chapter IV Section 3 Article 38 p. 2b), and that the project’s anticipated benefits balance out the total harm (Chapter IV Section 3 Article 38 p. 2d), may animal use take place.

However, research indicates that the system of ethical approval of animal research remains fraught with challenges (Pietrzykowski, 2021; Azilagbetor et al., 2024; Hajosi and Grimm, 2024; Jörgensen et al., 2025). A recent study showed that submitted applications in Sweden frequently lack essential information, particularly concerning basic information about the animals, implementation of the 3Rs, descriptions of intended harm, and articulation of the project’s scientific purpose and expected benefits (Jörgensen et al., 2025). This impairs evaluation of whether the 3Rs have been adequately considered, as well as the performance of a meaningful harm-benefit analysis (HBA). The same study also found uncertainties in the AECs regarding legal obligations and inconsistent application of regulatory frameworks, indicating potential procedural shortcomings in need of attention.

In Sweden, scientific animal use is governed by three overlapping legal instruments that encompass Sweden’s implementation of Directive 2010/63/EU: the Swedish Animal Welfare Act (SFS 2018:1192), the Swedish Animal Welfare Ordinance (SFS 2019:66), and animal research-specific regulations issued by the Swedish Board of Agriculture (SBA). At the time this study was conducted, the latter were compiled in the Swedish Board of Agriculture’s Regulations and General Advice on Laboratory Animals (SJVFS 2019:9, 2020), commonly referred to as the L150.

The competent authority conducting the ethical review in Sweden consists of six regional Animal Ethics Committees (AECs) overseen by the SBA. Whilst many other countries have ethical committees linked to individual research institutions, the Swedish ones are distributed across the main university cities of the country, and each committee receives and reviews applications for all animal-based research in their geographical region, or according to special mandates. The AECs are entitled to approve, postpone or deny applications with legally binding decisions, that can be appealed by the applying researcher only. In addition to a chair and vice chair (positions occupied by lawyers deemed impartial by the SBA in accordance with Chapter 7 Section 14 of the Swedish Animal Welfare Ordinance) as well as a secretary, each regional committee is comprised of twelve seats. Six of the twelve positions are occupied by representatives from the scientific community (researchers, animal research technicians, animal research staff), nominated for selection by the SBA by universities and/or research institutions in the vicinity of the city where the committee is located (SBA Research Animal Unit, personal communication, 11 April 2025). The other six seats are held by laypersons in an effort to represent the interests of the general public. Less than half of the laypersons may be nominated by animal welfare organizations (Chapter 7 Section 14 of the Swedish Animal Welfare Ordinance), and the remaining majority are local politicians appointed through the municipal board (SBA Research Animal Unit, personal communication, 11 April 2025).

In Sweden, there is only one online application form available to researchers, irrespective of the kind of research or animal use planned. Once filled out and submitted, applications are first distributed to smaller working units within the AEC known as preparatory groups. These assess if any elaborations, clarifications or corrections are required before the applications can be further reviewed. If so, the applying researchers must be promptly notified and asked to supplement their application(s) (Chapter 7 Section 7 of the L150). Preparatory groups may also conduct their own research or seek input from external experts. An application is considered complete only when the committee has received sufficient information to conduct an ethical review of it (General Advice to Section 5 Chapter 7 of the L150). After undergoing preparatory work, the applications and all supplementary documents are circulated to the full committee, along with recommendations from the preparatory group to approve, postpone, or deny the proposed projects. The whole committee subsequently convenes in a plenary meeting to discuss and reach a final decision for each application.

For the ethical review process to function as intended, a number of requirements must be met. First, researchers must include relevant, accurate, complete and clearly presented information about the planned projects in their submitted applications (The European Commission Expert Working Group, 2013; Jörgensen et al., 2025). Second, AECs must be equipped with the time, resources and expertise necessary to interpret this information effectively (and to gather additional information if necessary) (The European Commission Expert Working Group, 2013). Third, members need to engage in thorough case-by-case deliberation for each project (taking all ethical considerations into account) (Chapter IV Section 2 Article 38 p.2 of Directive 2010/63/EU; The European Commission Expert Working Group, 2013), to finally make well-informed, justified and legally sound decisions (Section 32 of the Swedish Administrative Procedure Act; Lundmark and Säfsten, 2020; Azilagbetor et al., 2024). Naturally, for both parties to fulfil their obligations, each must know what is required of them by current relevant regulations in accordance with the rule of law (Fuller, 1969).

Jörgensen et al. (2025) identified issues in several of these areas of the Swedish ethical review based on written documentation from the review process. However, such data reveal only the documented outcomes, not the work performed behind the scenes, both of which must be understood to map challenges and to find reliable, long-term solutions. Against this background, the purpose of this study was to investigate, acknowledge, and better understand insider perspectives of the persons directly engaged in the ethical review process: researchers submitting applications for ethical review, and AEC members tasked with performing said review. More specifically, the study aimed to explore:

• Views and opinions amongst animal researchers and AEC members on the Swedish ethical approval process of animal research, and specific aspects therein.
• Personal experiences amongst animal researchers and AEC members of partaking in the Swedish ethical approval process of animal research.
• Suggestions for improvements of the Swedish ethical approval process of animal research by researchers and AEC members.

2 Materials and methods

2.1 Questionnaires

Members of the AECs were informed about the project and the questionnaire through an electronic newsletter distributed by the Swedish Board of Agriculture in November 2022. Researchers received the questionnaire without prior notification, as no similarly appropriate communication channel was available to reach them in advance.

2.1.1 Creation of questionnaires

Two electronic questionnaires, one for AEC members and one for animal researchers, (see Supplementary Materials 1, 2), were created using the software program Netigate. At the beginning of each questionnaire, mandatory background questions were included to map demographics. The ensuing main part of both questionnaires consisted primarily of the same mix of mandatory closed-ended questions and optional open-ended questions, organised into three sequential sections:

• Current regulations and roles within the ethical review of animal research
• The purpose and importance of ethical review in animal research
• How the ethical review of animal research is performed in practice

A final fourth section contained optional, open-ended questions, in which respondents were invited to share additional information. The AEC member questionnaire contained a total of 33 questions, whilst the animal researcher questionnaire contained 29 questions.

Both questionnaires were prefaced by a short informative text explaining that partaking in the survey was voluntary and that all responses would be handled anonymously and in line with the General Data Protection Regulation (GDPR) (REGULATION (EU) 2016/679, 2016). Written consent was obtained by all participants.

2.1.2 Distribution of questionnaires

Distribution of the questionnaires was done in the form of an email containing information about the project, a link to the project website, contact details to the research group and an embedded link to the questionnaire.

The AEC member questionnaire was distributed in April 2024 with three subsequent reminders and stayed open for 46 days. A list of committee members, both regular and substitute, in position at the time was obtained from the SBA and emails containing individual links were then sent to the email addresses of all listed. In total, 183 committee members received an email. Background data on committee affiliation, role and membership status were obtained from the same SBA records used to distribute the survey.

The researcher questionnaire was distributed in April 2024 with one reminder and stayed open for 38 days. Based on a list of registered Swedish Animal Welfare Bodies (AWBs) from 2023 provided by the SBA, a selection of AWB representatives were contacted and asked to forward the questionnaire to researchers within their organisation who had applied for ethical approval of animal research between 2020 and 2024. This time frame was chosen because the L150 entered into force in 2020 and remained in effect until November 2025, ensuring that respondents had experience working under the regulatory framework relevant to the study. AWBs connected to the largest and most renowned research institutions in Sweden were contacted first, followed by AWBs that could be approached through the professional connections of the research group. In total, 21 AWB representatives agreed to distribute the questionnaire.

The AEC member questionnaire was provided in Swedish, it being the official working language of AECs in Sweden, whilst the researcher questionnaire was offered in both Swedish and English to accommodate the large number of non-Swedish-speaking researchers working in Sweden.

Prior to the distribution of the questionnaires, ethical approval was sought from the Swedish Ethical Review Authority. It was concluded that no ethical approval was necessary (Decision ID: 2023-04401-01).

2.2 Data analysis

2.2.1 Analysis of questionnaire data

All email addresses and names were removed from the dataset previous to analysis so that responses could not be linked to individual persons. Instead, each respondent was automatically assigned an anonymous respondent-ID by Netigate.

Responses to closed-ended questions from both questionnaires were summarised descriptively¹, and statistical analyses were conducted on a subset of these responses. Replies to open-ended questions were analysed in two ways. First, the answers linked to closed-ended questions (e.g., asking the respondent to elaborate), and those linked to a delimited topic (e.g., asking committee members about their motivations for serving on an AEC), were analysed qualitatively (following the first two steps of reflexive thematic analysis, see below) and summarised descriptively. Second, answers to two open-ended questions inviting the respondent to freely share thoughts, ideas and experiences of the ethical approval process were analysed qualitatively using all six steps of reflexive thematic analysis (see below).

AEC member questionnaire responses were analysed both collectively and as subgroups defined by the respondents’ roles within their respective committees. As only one chairperson (2%, 1/55), one vice chairperson (2%, 1/55), and three secretaries (5%, 3/55) took part in the study, these roles (and their responses) have been excluded from the final data set due to low representation, and were not analysed either quantitatively or qualitatively. Hence, in the ensuing sections of this paper, AEC members are either scientific community representatives (54%, 27/50) or laypersons (46%, 23/50).

Due to a known overlap between the populations of committee member respondents (especially, but not exclusively, scientific community representatives) and animal researcher respondents, it was predicted that some respondents may receive both questionnaires. In total, four respondents accessed both questionnaires. As these persons may provide important dual perspectives, their responses were included in the final data set. This should be considered when reflecting on the data. In this paper, “scientific community representatives” (SCRs) refers to researchers, animal research technicians, or animal research staff who responded in their capacity as committee members to the AEC questionnaire, whilst “animal researcher respondents” refers to researchers who responded as researchers to the animal researcher questionnaire.

2.2.2 Statistical tests

A selection of closed-ended questions was analysed using Fisher’s exact test, a statistical test suitable for small sample sizes that examines whether a statistically significant association exists between categorical variables. To reduce the risk of chance findings, statistical analyses were only performed for questions and variables where descriptive statistics supported hypothesised differences or associations. To investigate differences between animal researcher and committee member respondents, responses to Question 20 of the committee member questionnaire and Question 15 of the animal researcher questionnaire were compared. Among committee member respondents, Questions 18 and 23 were examined for the effects of role (scientific community representative, public representative, or animal welfare organisation representative) and experience (<1 year vs. ≥1 year), whilst Question 29 was examined for the effect of role. Associations between committee member responses to Question 29 and two statements included in Question 23 were also assessed. The resulting p-values are reported in the Results section, and the questionnaire items included in each analysis are provided in Supplementary Materials 1 and 2. Statistical significance was set at p < 0.05.

2.2.3 Thematic analysis

For the thematic analysis, a six-step Reflexive Thematic Analysis process, as described by Braun and Clarke (Braun and Clarke, 2006, 2022), was performed. The first author was responsible for coding and thematising in all six steps whilst author four reviewed developing themes against data, during steps three to six. Several of the steps overlap and/or are revisited throughout the process, as described below.

First, a substantial amount of time was spent “getting to know” the data, placing particular emphasis on the open-ended questions and related responses. All replies to all open-ended questions from both questionnaires were meticulously read through several times and initial observations were recorded. Second, data coding for thematic analysis was performed using the online qualitative data analysis software tool Delve (https://delvetool.com). Responses to be coded were inserted into Delve as two separate documents: one containing committee member responses, and one for animal researcher responses. Coding was done one question at a time, multiple times. As recommended by Braun and Clarke, both the order of the questions and of individual responses were varied each round of coding, as the order in which a question and/or its responses are read in relation to others may impact interpretation, undesirably limiting the analysis. An inductive approach was applied to enable the discovery of unexpected codes that may not have become apparent through deductive coding tailored to specific research questions. Coding was both semantic and latent, resulting in 112 initial codes. These codes were summarised descriptively. Third, to develop the analysis on respondents’ perspectives, initial themes were generated by grouping and interpreting codes based on answers to the questions inviting respondents to share their thoughts freely. Some codes were reworked or renamed to provide more clarity and distinction, others were ultimately deleted, and a few were created anew. This process resulted in 97 final codes. Themes were identified by creating digital mind map-like structures using MindMup (https://mindmup.com). The themes were then reviewed and further developed in what constitutes the fourth and fifth steps of Braun and Clarke’s six-step approach. This was done analogously on paper, in MindMup, and through discussions with fellow colleagues. Themes recurred in questionnaires, between respondents, and in answers to both questions, leading us to conclude that we had reached data saturation (Saunders et al., 2018). In the final sixth step of the process, the themes were arranged into a “story” through which the outcome of the reflexive thematic analysis could be presented in detail. This resulted in a total of four main themes and three subthemes, bridged by a fifth theme to two underlying ideals (Supplementary Material 3). The underlying ideals, identified in this last step, can be understood as guiding the respondents’ perspectives: “Mind the animals!” and “Abide by the law!”. The thematic analysis is presented in section 3.6.1, starting with the underlying ideals and proceeding with the five themes and their subthemes.

3 Results

3.1 Demographics

3.1.1 Response rates

Although responses were mandatory for the majority of all questions, participants could opt out at any time. Consequently, response rates varied across different questions for both questionnaires, with a general decline observed toward the end of both questionnaires. Response rates reported in the Results section of this paper may vary between questions, as the order of questions therein has been rearranged for presentation purposes.

Among committee members invited to participate in the questionnaire (excluding chairpersons, vice chairpersons and secretaries), 28% (50/178) answered at least one question before discontinuing their participation, and 22% (40/178) completed the entire survey from start to finish. The total number of animal researchers who received a link to the questionnaire is unknown, but 102 accessed and at least partially filled out the questionnaire, with 54% (55/102) completing it in its entirety. Thirty-five percent (36/102) of animal researchers chose to access the English version of the questionnaire, whilst the remaining 65% (66/102) responded to the Swedish version.

3.1.2 Representation

For committee member respondents, representation was almost equally divided between scientific community representatives (collectively referred to as SCRs) and laypersons (Figure 1). Most laypersons were public representatives (PRs), and fewer were animal welfare organisation representatives (AWORs) (Figure 1). Due to variation in response rates across questions, representation was occasionally limited for certain respondent groups, particularly AWORs. This is addressed as a limitation of the study in the Limitations section.

The six regional AECs were fairly evenly represented amongst committee member respondents, with 16% (8/50) belonging to Gothenburg, 10% (5/50) to Linköping, 14% (7/50) to Malmö/Lund, 24% (12/50) to Stockholm, 24% (12/50) to Umeå and 12% (6/50) to Uppsala AEC. Two-thirds of committee member respondents were regular members (64%, 32/50) and one third were substitutes (36%, 18/50).

In an optional open-ended question, committee member respondents could share their reasons for becoming a member of an animal ethics committee. A total of 41 responses were provided. Many respondents shared the same motivations, such as personal interest, an ambition to “be of use” [our translation], possessing special competence(s) of use for the job, a wish to ensure animal welfare, and/or a belief that the ethical approval process should protect the interests of animals used in research. Many respondents also expressed interest in and recognised the need for animal ethics in research, as well as the overall importance of the ethical review process. Others reported becoming a committee member simply because they were asked to.

3.1.3 Work experience

Thirty-seven percent (18/49) of committee member respondents had less than one year of experience working on an ethical committee, whilst 29% (14/49) had 1–5 years, 20% (10/49) had 6–10 years, and the remaining 14% (7/49) had more than 10 years.

Among animal researcher respondents, only 3% (2/67) had worked in animal research in Sweden for less than one year. Ten percent (7/67) had done so for 1–5 years, 19% (13/67) for 6–10 years and the vast majority, 67% (45/67), for more than 10 years. Just less than half of the responding researchers (45%, 30/67) had previously worked in animal research in another EU country, and just over half (52%, 35/67) had done so outside the EU.

3.1.4 Education

When asked about what education they had received in their role as committee members, 68% (17/25) of SCRs, 57% (4/7) of AWORs and 56% (9/16) of PRs stated that they had been educated about the 3Rs. Similarly, 72% (18/25) of SCRs, 57% (4/7) of AWORs and 56% (9/16) of PRs had received education about ethics, and 72% (18/25) of SCRs, 57% (4/7) of AWORs and 50% (8/16) of PRs about animal research legislation. Almost one fourth (24%, 6/25) of SCRs, one third (29%, 2/7) of AWORs, and 6% (1/16) of PRs responded having received “other education” of some kind. When asked to elaborate, several respondents mentioned a short introductory course held by the SBA for new committee members, although one committee member respondent “… would not call it education due to its poor quality” [our translation], and another noted that many things, such as practical consequences of different degrees of severity, were not addressed. “Other education” was also described as knowledge gained from previously working as a researcher, from being FELASA-accredited to handle experimental animals, or from simply being “well educated in the field” [our translation]. Two respondents who reported having received “other education” contradicted their own responses by describing that they had, in fact, not received any education specifically linked to their role on the committee but had been granted access to material for self-studies or to relevant regulations. In total, 12% (3/25) of SCRs, 14% (1/7) of AWORs, and 25% (4/16) of PRs reported having received no education at all in their role as a member of an AEC.

Among animal researcher respondents, almost all stated that, in relation to using animals for scientific purposes, they had received education about the 3Rs (95%, 63/66), ethics (98%, 65/66) and animal research legislation (97%, 64/66). Fifteen percent (10/66) had received some form of other education, which, when asked to specify, included veterinary medicine, laboratory animal sciences training, alternative methods, experimental surgery, species-specific education and animal handling training. No animal researcher respondent selected the option “I have not received any education”.

3.2 Perceptions of preparedness

3.2.1 General conditions

The majority of SCRs (78%, 18/23) and PRs (92%, 11/12) believed that they, generally speaking, had the necessary general prerequisites to carry out their duties as animal ethics committee members in a satisfactory manner, citing previous knowledge, work-experience, and education as indispensable, and the mixed competence of the committees as helpful. However, a lower percentage of AWORs (57%, 4/7) believed that conditions were sufficient. The less satisfied committee member respondents cited frustration with long, complicated applications, time constraints and a lack of access to relevant databases and scientific literature for members for whom this was not available through other channels.

3.2.2 Knowledge and competence

All SCRs (100%, 23/23) agreed “completely or mostly” that their expertise/education was relevant and sufficient to satisfactorily perform all committee duties. In comparison, only 58% (7/12) of PRs and 43% (3/7) of AWORs thought as highly of their own competence, and role had a statistically significant effect on perceptions (p = 0.0013).

Committee members generally rated both their own knowledge of the 3Rs and the harm-benefit analysis, as well as that of their committees, as high (Table 1). However, AWORs were consistently less optimistic than PRs and SCRs, and role was found to be statistically significant for perceptions of one’s own knowledge about Replacement, as well as for the presumed knowledge of one’s committee concerning Replacement, Reduction, Refinement, and the harm-benefit analysis (Table 1).

Table 1. To what degree committee member respondents agree with statements about knowledge. Committee member responses (CMRs) can also be seen sorted according to committee role: scientific community representatives (SCRs), animal welfare organisation representatives (AWORs), and public representatives (PRs).

One AWOR wrote: “…Many members of the committee lack knowledge of the regulatory frameworks to abide by, as well as basic understanding of what the 3Rs and animal welfare actually are. Focus is placed solely on Reduce, and it is considered legitimate to subject research animals to far too much suffering simply to be able to claim that Reduce is being implemented. Refine and Replace are not even considered, as they lack knowledge about them.” [our translation]. Another AWOR wrote that “…Applicants can often explain the benefits, so additional support is rarely needed in that regard. But relevant assessments of the impact on the animals and of their suffering are always inadequate in the committees! … A lack of understanding and knowledge amongst the majority of committee members makes a high-quality ethical review difficult. …” [our translation].

Animal researcher respondents also generally rated their own knowledge as high. A majority agreed “completely or mostly” that they had sufficient knowledge of Replacement (89%, 50/56), Reduction (98%, 55/56) and Refinement (98%, 55/56) to always implement these into their own research projects, and of the harm-benefit analysis (89%, 50/56) to perform one themselves. Seventy-one percent (40/56) agreed “completely or mostly” and 27% (15/56) “somewhat” that they had sufficient knowledge of the harm-benefit analysis to understand what animal ethics committees must consider when forming decisions of ethical review. Only one respondent (2%, 1/56) did not agree with this at all.

3.2.3 Time and resources

Most committee member respondents believed they had both sufficient time and access to resources to adequately perform their duties, although, for thoughts on time, role was found to be statistically significant (p < 0.001), with AWORs experiencing time constraints as more of an issue than PRs or SCRs (Table 2).

Table 2. To what degree committee member respondents agree with statements about time and resources. Committee member responses (CMRs) can also be seen sorted according to committee role: scientific community representatives (SCRs), animal welfare organisation representatives (AWORs), and public representatives (PRs).

3.2.4 Committee member confidence

Presented with three statements regarding confidence in their responsibilities related to their role and the relevant legal frameworks, the majority of SCRs, PRs, and AWORs showed high levels of confidence in all three cases (Table 3).

Table 3. To what degree committee member respondents agree with statements about confidence. Committee member responses (CMRs) can also be seen sorted according to committee role: scientific community representatives (SCRs), animal welfare organisation representatives (AWORs), and public representatives (PRs).

3.3 Perceptions of tasks and responsibilities in the ethical review

3.3.1 The purpose(s) of the ethical review

Most committee member respondents believed it was important for the ethical review process to ensure animal welfare, legal compliance, the application of 3Rs by researchers, and ethical discussion by AECs (Figure 2). However, some variation was noted depending on role, and only AWORs unanimously ranked the fulfilment of all listed aspects as “of outmost importance” (Figure 2).

Animal researcher respondents ranked all four aspects as less important than committee member respondents collectively (Figure 2), although differences were not significant. Only amongst animal researcher respondents was the ranking “not important at all” selected: by one animal researcher respondent for “Complying with legal requirements regarding what animal research may and may not be carried out”, and another for “Ensuring that all research projects are carefully discussed from an ethical perspective” (Figure 2).

SCRs and PRs commonly believed that all four aspects were “very well” fulfilled by their committee, apart from “Ensuring application of the 3Rs”, for which equally many PRs responded “quite well” (Figure 3). No AWOR provided the response “very well” for any of the listed aspects, and were most critical of their committee’s ability to ensure ethical discussion of all research projects (Figure 3), with only one respondent (14%, 1/7) answering that this was done “quite well”, two (29%, 2/7) “not so well”, and the remaining majority (57%, 4/7) that discussing projects from an ethical perspective was “not at all” performed by their committee.

3.3.2 Representation of interests

Both committee members and animal researchers were asked whose interests they understood the review as theoretically intended to serve (Figure 4).

The median values for SCR responses indicate that, in theory, SCRs believed the interests of “animals used in research” should be given highest priority in the ethical review, whilst the interests of “the applying researcher” should be given lowest priority (Figure 4). AWORs showed the same median rankings of both interests, but with a large interquartile range for “animals used in research”, indicating greater variability in how this interest was ranked (Figure 4). Median responses for PRs assigned “animals used in research” second priority, and “the applying researcher” together with “patients” shared third, with a wider distribution of responses than observed amongst SCRs and AWORs for both interests (Figure 4). Median values for PRs did not assign any of the listed interests to either first or last priority.

Median values of animal researcher respondents indicate that they, like SCRs and AWORs, believed that first priority of the ethical review, in theory, should be to serve the interests “animals used in research”, and last priority should be to serve the interests of “the applying researcher” (Figure 4).

When asked whose interests they believed the ethical review process serves in practice, a different pattern emerged (Figure 5).

Median values indicate that, in practice as in theory, SCRs believed that the interests of “animals used in research” are prioritised first by the ethical review (Figure 5). Conversely, AWORs believed animal interests were in reality being prioritised last. PRs again showed wide response distribution when ranking this particular interest, with their median response values placing both “animals used in research” and “the applying researcher” third (Figure 5). Median values for rankings of “the applying researcher” show SCRs recognising this interest as being of fourth priority in practice, whilst AWORs unanimously perceived it as being first.

Animal researcher responses for both the interests of “animals used in research” and “the applying researcher” mirrored those of the SCRs: placing the former first in practice just as in theory, and the latter one step higher in practice than in theory (Figure 5).

All AWORs, and just over half (52%) of SCRs reported that they themselves primarily represent the interests of “animals used in research” when conducting committee work (Figure 6). PRs displayed less consensus, claiming to represent all interest holders to somewhat similar degrees, except for “the applying researcher”, whom no PRs considered themselves to primarily represent (Figure 6).

3.4 Perceptions of the importance of information in applications for ethical review

3.4.1 Necessary information in applications

When presented with different kinds of information relevant for understanding and assessing an animal research project, a clear majority of both committee member and animal researcher respondents agreed “completely or mostly” that all nine provided examples were important to include in applications for ethical review (Figure 7).

However, neither kind of information was unanimously reported by committee member respondents as always or often included before preparatory work (Figures 8, 9), and after preparatory work and potential supplementation, only information about “how animals would be housed, handled, and cared for during the project” was seen as always or often included (Figure 8). Whilst PRs and SCRs generally thought preparatory work increased the amount of all exemplified kinds of information in applications, AWORs did not believe it generated more information about “Reduction”, “humane endpoints”, or “the expected benefit of the project” (Figures 8, 9).

Despite having disclosed that important information was not always included in applications even after preparatory work (Figures 8, 9), the majority of both PRs (91%, 10/11) and SCRs (91%, 21/23) agreed “completely or mostly” that their committees gained sufficient understanding of planned projects to make well-informed ethical decisions. However, only just under one third (29%, 2/7) of AWORs shared this view, and again, role was found to significantly impact responses (p < 0.001).

Amongst animal researcher respondents, 86% (48/56) acknowledged that they had been asked to supplement at least one application, most commonly concerning planned procedures and their effects on the animals (Figure 10). Information about humane endpoints and the physical as well as mental suffering of the animals was also frequently requested (Figure 10). Fifteen percent (7/48) shared that they had been asked to supplement with “other information” than the examples provided, such as being asked to clarify what facility the research would be carried out in, provide graphical timelines of their experiments, specify what vehicles would be used for substance administration, account for genetic modifications and resulting phenotypes, and how animals would be trained for procedures. Some respondents had been required by the reviewing committee to lower humane endpoints and adjust procedures. One animal researcher respondent expressed frustration that the ethical committee could request changes to the application against the will of the researcher, describing “[a]n enormous focus on lowering humane endpoints despite repeated explanations from me as the responsible researcher.”, resulting in conflict between the principles of Refinement and Reduction: “In the final versions [of the application], the committee added an amendment (amongst many others about lowering humane endpoints) which means we have to set up double the number of animals in order to have enough animals for an experiment (since around half of the animals need to be euthanised because they reach the humane endpoint that the committee wrote into the application). Where are the 3Rs in this?” [our translation].

3.5 Perceptions of the ethical decision-making process

3.5.1 Committee discussion climate

Most SCRs and PRs agreed “completely or mostly” that committee members may generally speak and participate on equal terms, and that they themselves felt listened to and were able to express their views during committee work (Table 4). However, AWORs agreed less with both statements and role was found to be statistically significant for thoughts on the former (Table 4). One AWOR felt that”… Animal welfare organisation representatives’ opinions/arguments are always less valued than those of scientific community representatives. …” [our translation].

Table 4. To what degree committee member respondents agree with statements about discussion climate during committee work. Committee member responses (CMRs) can also be seen sorted according to committee role: scientific community representatives (SCRs), animal welfare organisation representatives (AWORs), and public representatives (PRs).

3.5.2 Ethical discussion and harm-benefit analysis

When asked if ethical decisions are always preceded by a thorough ethical discussion during plenary meetings, the majority of both SCRs (78%, 18/23) and PRs (64%, 7/11) agreed “completely or mostly” that this was the case. However, most AWORs did not agree at all (71%, 5/7), and role was found to be statistically significant (p < 0.001).

Role also influenced perceptions of the frequency of harm-benefit analyses, with SCRs and PRs perceiving it as a significantly more common occurrence than AWORs (Table 5). Moreover, statistically significant correlations were identified between respondents’ perceptions of how often their committees conducted harm-benefit analyses and the extent to which respondents believed committees engaged in thorough ethical discussions (p = 0.0012), as well as respondents’ assessments of their respective committees’ knowledge of what a harm-benefit analysis entails (p = 0.0011; Table 1).

Table 5. How often committee member respondents believe that their committee performs a harm-benefit analysis for projects undergoing ethical review. Committee member responses (CMRs) can also be seen sorted according to committee role: scientific community representatives (SCRs), animal welfare organisation representatives (AWORs), and public representatives (PRs).

However, only one in four SCRs (26%, 6/23), one in two AWORs (50%, 1/2), and no PRs (0%, 0/10) believed that harm-benefit analyses were in reality most commonly performed during plenary meetings. Instead, harm-benefit analyses were thought by a number of SCRs (65%, 15/23) and PRs (50%, 5/10) to be performed during preparatory meetings, or some other stage of the preparatory work (4%, 1/23 of SCRs, 40%, 4/10 of PRs). No AWOR believed either of these two options to be true. One respondent from each role argued that harm-benefit analyses were commonly performed “at another time”, the SCR (4%, 1/23) not knowing when this might be, the AWOR (50%, 1/2) claiming that no harm-benefit analysis was ever performed in their committee, and the PR (10%, 1/10) suggesting that harm-benefit analyses may be performed either by individual committee members when reading up on the applications, during preparations, or in plenary.

3.6 Open-ended questions

The following section will present the themes developed during the reflexive thematic analysis of two open-ended questions located at the very end of each of the questionnaires:

• - “If you were given the chance to change something about the ethical review of animal research, what would it be?”
• - “Is there anything you would like to add about your experiences or thoughts regarding the ethical review of animal research?”

Replying to these questions was optional, with 32 committee member respondents and 45 animal researcher respondents providing a response to the first question, and 18 committee member respondents and 27 animal researcher respondents to the second.

3.6.1 Analysis

3.6.1.1 Underlying ideals

3.6.1.1.1 Mind the animals!

In response to both questions, committee member and animal researcher respondents repeatedly voiced concern for the well-being of animals used in research and often expressed an ambition to prevent or reduce suffering when possible. In fact, this was so prevalent and commonly addressed with such emphasis that it was identified as one of two underlying ideals or values shared by the respondents. We call this first ideal: “Mind the animals!” (Figure 11). This was expressed by explicitly insisting that harm to animals must be minimised, that the ethical review should focus on animal suffering, or through frustration and disappointment that animal welfare was somehow being overlooked in the current system. Concern for animals was also indirectly expressed, such as by wishing that AWORs had more influence during the review process or that committee seats would be assigned to individuals with specific knowledge of animal welfare and alternative methods. Some respondents argued that the review process, contrary to its intended purpose, had direct negative consequences for the animals used: “It does not work. … As long as the system looks the way it does today, animals will continue to suffer unnecessarily and, above all, unlawfully.” [our translation].

3.6.1.1.2 Abide by the law!

In addition to concern for animal welfare, respondents repeatedly emphasised that the ethical review must comply with relevant rules and regulations. This was something which the review, according to respondents, did not always do. Identified as a second underlying ideal, we labelled this "Abide by the law!" (Figure 12). Respondents pointed out that legislative texts are not easily understood and expressed a desire for more concrete guidance on how the ethical review should be performed, such as a checklist for the review process, and detailed instructions for applicants. Respondents also argued for increased oversight of the current system, believing that there were insufficient consequences, or none at all, for flaws or misconduct. In the words of one committee member: “A complete overhaul is needed. Is it even possible to carry out a harm–benefit analysis? … More cases should be reported to the Parliamentary Ombudsman, the committees often make decisions that completely contradict applicable legislation without anything happening.” [our translation].

3.6.1.2 Themes and subthemes

3.6.1.2.1 The ethical review is not functioning as intended

Whilst some respondents explicitly expressed that the ethical review is important, its execution was also heavily critiqued. One researcher shared: “I do not find the ethical review process rewarding…” [our translation], and another candidly referred to the current system and its application form as “catastrophic” [our translation]. Ultimately, many respondents seemed to agree on what was identified as a theme of its own: that the ethical review is not functioning as intended. This theme serves as a conceptual link, or “bridge”, between the underlying ideals of respondents and a series of themes and subthemes that, by describing obstacles or issues emphasised by respondents, help explain how or why many respondents perceived the ethical review as flawed (Supplementary Material 3).

3.6.1.2.2 The ethical review process is ill-suited to the reality of how research is performed

One issue addressed by several respondents, the majority of whom were animal researchers, was that the ethical review process is ill-suited to the reality of how research is performed. For instance, due to its inherently uncertain nature, basic research was exemplified by several respondents as both difficult for the applying researcher to describe, and for AECs to assess. As expressed by one respondent: “… This [exploratory research] is of course difficult for the committee to review given that the exact path of the animal is unknown, however I believe the committee needs to be more open to this type of research and understanding of the difficulty in perfectly describing the path of the animal.”. Several respondents also addressed the fact that research projects are often conducted over long periods of time, with one animal researcher respondent explaining that “… [w]hen conducting research, it is not possible to know 1–4 years in advance exactly what type and combination of procedures mouse X will need to be subjected to. The research becomes poor if one needs to know the results before carrying out the experiment. …” [our translation]. In line with this, animal researcher respondents requested increased flexibility and an easier way to edit and rework already approved applications, strongly opposing both the financial costs and the time associated with submission. One animal researcher respondent predicted that if editing existing approvals would not be made easier and cheaper, “…more principal investigators will submit larger applications with more general descriptions in order to allow for experiments over five years, without having to submit amendment applications. It is difficult and expensive to submit amendment applications today, and this counteracts and makes the ethical review more difficult.” [our translation]. Animal researcher respondents expressed a desire for an ethical review that promotes good research and facilitates for researchers but described a reality in which it does the exact opposite, essentially functioning as an obstacle, providing the subtheme: the ethical review is negative for researchers. One animal researcher links the problematic application procedure to poor animal welfare, describing the application process as “…a very burdensome process for the researcher/principal investigator, because it always ends up focusing on details and the animals’ welfare seems completely forgotten.” [our translation]. Several respondents also expressed frustration that the application form was difficult to complete, further complicating the process.

3.6.1.2.3 It is difficult to achieve just the right amount of information in applications

This brings us to the next theme and issue: that it is difficult to achieve just the right amount of information in applications. Both committee member and animal researcher respondents described how applications must provide a good overview of planned projects and how enough information must be provided to enable a thorough ethical review. At the same time, however, submitted applications were thought to often be overly long, complex and detailed, and several respondents described how high application fees and a presumed wish to “cover all bases” when planning long-lasting research projects would push some researchers to include as much as possible in one application. Animal researchers reported struggling with knowing what was required of them, and described the application form as frustratingly repetitive, adding unavoidable but unnecessary length to applications: “…The current layout is confusing, and it feels like you’re repeating yourself. …we included things required by law to reduce animal suffering, but the committee did not want to read that, only what we do beyond the legal requirements, and it’s not easy to know that this is what was being asked for. You also think you need to add a lot of text since, for example, it says a maximum of 10,000 characters in the boxes to be completed, but the committee absolutely does not want a lot of text, only what is strictly necessary, so we had to spend a lot of time shortening it after our meeting where we were asked to supplement [the application]…” [our translation]. Comments on how it was often difficult to understand what the AEC was requesting, and the importance of AECs asking relevant questions and trusting the word of the applying researcher prompted the emergence of a subtheme: that good communication between applicants and committees is vital. Moreover, a number of animal researcher respondents shared how demands or conditions formulated by the AECs could be difficult to abide by in practice, further showcasing the importance of communication and understanding between both parties.

3.6.1.2.4 The AECs do not do a good job

By some respondents, AECs were accused of impeding the work of animal researchers and of ruining submitted applications through their review, sometimes even counteracting implementation of the 3Rs. Hence, that the AECs do not do a good job emerged as a theme. Within this theme, respondents described that AECs tend to get stuck on formalities within applications, and that they should not be discussing research planning or details of the proposed projects at all. Instead, several committee member respondents as well as animal researcher respondents expressed a wish that AECs should be discussing ethics more. Several respondents stated that applications were not vetted thoroughly enough and that ethical approvals were handed out arbitrarily. Respondents worried that AECs were not sufficiently well read up on reviewed applications and that they accept the words of applying researchers too readily. One animal researcher respondent wrote: “…I believe that most of the questions asked by the committee are relevant and important, however, answers from principal investigators are accepted too easily … even where there are still ambiguities” [our translation]. In order for AECs to adequately perform their assigned task, the importance of proper education and knowledge, especially in ethics, as well as sufficient time and resources was repeatedly emphasised, and that AECs lack the necessary resources and capabilities to perform their assigned task(s) was categorised a subtheme. Respondents specifically mentioned that AECs lack sufficient knowledge of: animals in general or of specific species; housing of animals; veterinary medicine; animal suffering, severity assessment and humane endpoints; the 3Rs; animal research regulations; and how animal research is actually conducted. Respondents also described how they believed that the different seats of the AECs should provide the committee and ethical review with different kinds of knowledge and expertise, but that this was not always the case. A number of respondents singled out laypersons as less knowledgeable. One respondent wrote: “In the committee I sit on, I experience a lack of knowledge, interest, and engagement amongst the members who are politicians (laypersons). In practice, it is the animal welfare organisation representative, along with the researchers, the veterinarian, and the chair, who actively participate and provide input. It concerns me that some members’ interest and engagement in approving animal research projects is so low, and I feel that the meetings are to some extent only ‘for show’ and do not fulfil their purpose. My experience is that all applications are approved, even if some are given specific conditions aimed at reducing the number of animals or their suffering.” [our translation].

3.6.1.2.5 The ethical review is biased and unpredictable

Respondents described how interest, knowledge and commitment vary significantly between different committee members, and how members are not always able to partake in discussions and decision-making on equal terms, prompting the notion that the ethical review is biased and unpredictable as a theme of its own. Respondents expressed a desire for a review where a wide range of opinions and stakeholders are represented, decisions are not overly influenced by the specific individuals or committees conducting the review, and all committees have the same prerequisites needed to form standardised judgements. One animal researcher wrote: “If one could wish freely, it would be for an ethics committee composed only of competent members who neither aim unilaterally to promote all kinds of research, nor do the opposite and try to sabotage all animal experiments. …” [our translation]. Some respondents expressed a fear of built-in bias in favour of research and researchers at the expense of animal interests, or of nepotism whereby “…local researchers have each other’s backs, and the views of the animal welfare representatives are not considered relevant. …” and “…(t)hose who are affected are the animals, the very ones actually supposed to be protected by the system.” [our translation]. Others felt that the use of animals in research was questioned too harshly, describing “…a misguided idea that there is a general interest amongst the public in these issues, when in reality it is about yielding to a small but vocal activist group. … since we are slowly giving in point by point, making it more costly, more time-consuming/complicated, and allowing more and more considerations to pass through, we risk having the fundamental decision [that it is permissible to use animals in experiments] called into question. … The solution is not to make further concessions but instead to state clearly that it is permitted, full stop. …” [our translation].

4 Discussion

4.1 Good intentions, poor performance?

This study set out to capture the perspectives and experiences of those directly engaged in the Swedish ethical review of animal research (AEC members and applying researchers) and to examine how these relate to the regulatory ideals the system is designed to uphold. Based on the data collected, quantitative replies in particular, many respondents seem generally satisfied with the functioning of the Swedish ethical review and their contributions to it, especially SCRs and PRs. However, respondents also reported experiencing that applications were not always complete, even at the final stages of review, and that key mandatory components, such as 3R assessment, ethical deliberation, and harm-benefit analysis, were not consistently performed by AECs. Notably, a non-negligible proportion of respondents, primarily AWORs, appear to experience a gap between how the ethical review should function in theory and how it operates in practice. Given the widespread concern for animal protection and regulatory compliance across all respondent groups, these contrasting views and reported shortcomings are interesting and merit further discussion.

4.2 A wide spectrum of diverging perceptions

Why do involved parties perceive the same process so differently? How can the indicated moat between theoretical ambition and practical performance in the Swedish ethical review of animal research be understood, and what are its potential consequences? In the following pages, we will discuss these questions and the findings that prompted them.

4.2.1 Epistemic inequality and asymmetric preparedness

By their own accounts, animal researcher respondents and SCRs had received more education than layperson committee members. SCRs also reported easier access to and greater satisfaction with resources than their layperson colleagues; rated their own knowledge higher, with 100% of SCRs believing it to be sufficient for committee work; and showed high levels of confidence in their understanding of the legal mandate of AECs and of their assigned role. Interestingly, PRs reported receiving the least amount of education, yet also rated their own knowledge as high, though they were less certain than SCRs that it was, in fact, sufficient. PRs also showed high levels of confidence in the roles and responsibilities associated with committee work. AWORs, on the other hand, saw themselves as less knowledgeable, were less confident, and most of them believed their knowledge was insufficient for committee work.

One reason behind the reportedly uneven distribution of knowledge, is likely the stark differences in how education is regulated for animal researchers versus AEC members. Apart from requiring that the AECs “shall consider expertise” in specific areas when performing the ethical review (Chapter IV Article 38 p. 3 of Directive 2010/63/EU, and Chapter 2 Section 15 of the L150), applicable regulations do not detail what knowledge or competence committee members should possess, let alone to what degree or how such knowledge should be acquired or overseen. The education and competence of animal research personnel (which would include any committee member working in this capacity outside the AEC) are substantially more regulated (Marinou and Dontas, 2023). Nonetheless, it is worth wondering if all AEC members should be expected to possess the same kind of knowledge, or if committee work is better served by complementary skills. However, the latter only works if all necessary areas of expertise, including insight into public views, are represented during important discussions and decisions. Otherwise, if there are knowledge gaps within the committee as a whole or important discussions and decisions are confined to smaller groups, as reported in our study, important perspectives and concerns may be missed.

The pattern observed amongst PRs could also reflect the so-called Dunning-Kruger Effect, whereby individuals with lower levels of knowledge or expertise in a given domain tend to overestimate their own competence, whilst those with higher levels of knowledge underestimate theirs (Dhingra, 2024). If this phenomenon is present amongst committee members, to whatever degree, it could potentially impact committee deliberations and decisions, as individuals’ perceptions of themselves are known to influence their work and the decisions they make (Adamecz et al., 2025). However, given the small sample size, and the intrinsic difficulty of accurately assessing one’s own abilities (Carter and Dunning, 2008), this interpretation should be treated with caution. That SCRs report high levels of education, as well as competence and confidence, could also be seen to challenge the presence of such an effect, or, at the very least, to raise the question of why this respondent group does not appear to be similarly affected. Differences in reported competence and confidence between roles could, however, at least in part, be explained by other factors. Respondents may, for example, have different expectations of the ethical review and their own role within it, or may place different weights on criteria such as “sufficient knowledge” or “adequate assessment”. Such factors and their potential implications will be discussed in more detail in upcoming sections. Why PRs and AWORs, despite describing similarly low levels of education, had such different views of their knowledge and confidence is perplexing. One theory is that AWORs’ intent on representing the interest of animals first may perceive a lack of education in 3R, ethics and legislation as an obstacle to doing so, prompting them to view knowledge gaps in these areas more negatively. This could also help explain why AWORs were less convinced that they had the necessary prerequisites to carry out their committee duties. PRs, on the other hand, who showed less consensus about who they primarily represent in their role, may not view limited knowledge of these topics correspondingly problematic if they do not deem them as relevant to safeguarding the interests of society, science or patients, or if they are more inclined to rely on others in their committee to know and address certain topics more in-depth.

Regardless of its underlying cause(s), the potential presence of epistemic inequality in AECs poses a number of risks. For example, assessing the 3Rs requires being able to determine which alternative methods can be used, what experimental design is most suitable for reducing the number of animals, and how animals can be housed or handled to minimise stress. These are all complex topics that require extensive expertise. Committee members with limited knowledge of the 3Rs may therefore struggle to do this and may consequently be less able to ask appropriate questions, propose alternatives, or challenge assumptions and/or statements in the applications. Influence and participation may be further skewed if AECs spend significant time discussing technical details, often requiring advanced terminology, a common phenomenon observed in this as well as other studies (Houde et al., 2003; Tjärnström et al., 2018). Unless all members understand the more technical aspects of a project, committees may also struggle to advance towards their main objectives: assessing the 3Rs, engaging in ethical deliberation, and conducting harm-benefit analysis. Members unfamiliar with ethics may likewise find it difficult to identify ethical issues or contribute meaningfully to deliberations. Similarly, deciding whether to approve or reject a proposal is likely harder if one is unfamiliar with the regulations that guide the decision, or if one lacks the legal competence to comprehend them. An individual’s contributions may also be automatically deemed less valuable by others, as described by some respondents in this study, because they are thought to lack the necessary knowledge to justify their participation, or due to prejudicial attitudes towards them, regardless of their actual competence (Schuppli and Fraser, 2007; Tjärnström et al., 2018; Galasiński et al., 2023; Stefanello, 2023). If happening in AECs, this is of course problematic, especially as lay persons do not necessarily lack relevant competence. Even if they reportedly receive less education specifically linked to their role on an ethical committee, their mandate is to represent the view of the general public or to advocate for animal welfare. Further, AWORs often have a background in veterinary medicine or animal science and are expected to bring this competence to the AEC table.

Both genuine epistemic divides and prejudicial attitudes can distort influence within AECs, marginalising non-technical viewpoints in favour of (certain) scientific disciplines (Schuppli and Fraser, 2007; Brister, 2012; Tjärnström et al., 2018; Miller and Pinto, 2022). In Swedish AECs, intentionally organised to allow for multiple interests to be represented, such uneven participation undermines the democratic function of layperson inclusion (Röcklinsberg, 2015).

4.2.2 Divergent perceptions of purpose(s) and performance

Most respondents agreed that the ethical review should ensure ethical deliberation, 3R assessment, legal compliance, and animal welfare, and should prioritise animal interests first and researcher interests last. However, when asked how well this is actually being done, different patterns emerged. Animal researcher respondents and SCRs perceived little gap between theory and reality. In their view, animal interests remained the top priority in practice, and SCRs believed that the review performed by their committees largely fulfilled its intended functions. AWORs, on the other hand, were critical of their committees fulfilling the listed functions of the review and believed that animal interests were deprioritised to last place, whilst researcher interests came first. PRs showed less internal agreement when ranking interests, but believed, like SCRs, that their committee generally did a good job of fulfilling the listed functions.

These varied accounts of the same institutional process could be explained by respondents having different expectations and views of the ethical review, its intended purpose(s) and how this/these should be fulfilled. Expectations and views that may, in turn, be related to respondents’ perceptions of their own role within the ethical review (Ideland, 2009). For example, someone who identifies as representing the interests of animals is likely to also believe that animal interests should be prioritised by the ethical review, and that animal welfare and 3R implementation are of utmost importance. Individuals may also have varying opinions on when and how fulfilment is obtained. What one person considers an ethical elaboration, another may not deem sufficient, and what one expects from the harm-benefit analysis, another may find unnecessary (Brønstad et al., 2016; Davies, 2018). Moreover, a member of a committee where ethical aspects are never or seldom discussed may not even recognise that something is missing. For this study, however, such a pattern cannot be neither confirmed nor denied as replies and respondents cannot be traced back to individual committees.

Regardless of its cause or causes, potential gaps between intended purpose(s) and practical performance of the ethical review threatens predictability, consistency, and transparency. As does disagreement amongst participants of the magnitude, and/or mere existence, of such gaps. After all, if the parties involved in the ethical review do not share the same view of its purpose, they cannot be expected to cooperate effectively towards achieving it, or to agree on when its overarching purpose or its necessary components have been achieved.

4.2.3 Perceived adequacy of information

Our study shows that applications are often perceived by committee member respondents as unnecessarily repetitive, comprehensive and complex, whilst confirming previous findings (Jörgensen et al., 2021, 2025) that applications may also lack important information, both before and after preparatory work. Still, interestingly, 91% of both SCRs and PRs believed their committees gained sufficient understanding of projects to go on to make well-informed ethical decisions, whereas only 29% of AWORs agreed.

One possible explanation for this discrepancy is that committee members likely hold different views on how the phrases “sufficient understanding” or “well-informed ethical decisions” should be interpreted in this context. This may be due to individuals applying slightly different definitions to the same concepts, or because they differ in their perceptions of what is essentially enough to be considered “sufficient” or “well-informed” in this situation. Regardless, disparate understandings of key terms could lead to friction between applying researchers and AECs, as well as amongst committee members. More fundamentally, without a shared understanding of key terminology, misunderstandings are inevitable, especially in the potential presence of epistemic inequality and/or the Dunning-Kruger Effect. In the long run, this introduces inconsistency and unpredictability to the ethical review process. Respondents may also have interpreted the question in the survey differently. When asked whether their committees gained sufficient understanding, some may have interpreted this as meaning an understanding based solely on information in the written applications, whereas others may have also considered information from other sources. After all, information may be obtained orally directly from the researcher during preparatory or plenary work (although, importantly, according to Section 27 of the Swedish Administrative Procedure Act (SFS 2017:900), oral information must be added in writing to the application documents), and external experts may be consulted. Committee members may also undertake independent background research on applications and their contents, something which may be more or less difficult depending on epistemic abilities, access to resources, familiarity with searching for and reading scientific literature etc.

If committees routinely approve applications that lack important information at the time of decision-making, as observed by Jörgensen et al. (2025), transparency is undermined. We also see a potential risk that such a pattern could lead to the threshold for what constitutes enough information or sufficient understanding lowering over time, gradually decreasing the quality of applications and ethical review. If important information about, for example, training of animals before procedures, Refinement strategies, pain relief, etc., is both left out in the application and not requested by the AEC, such a development could potentially endanger animal welfare, as animal care staff should be able to use the approved applications to guide their everyday work (Chapter 11 Section 1 of the L150, and Chapter 7 Section 8 of the current Swedish Board of Agriculture’s Regulations and General Advice on Laboratory Animal Use, SJVFS 2025:28)².

4.2.4 Exclusion of ethics and absent harm-benefit analyses

Although Directive 2010/63/EU requires a harm-benefit analysis “taking into account ethical considerations” to be performed as part of each ethical review process (Chapter IV Section 2 Article 38 p. 2), this is reportedly not always the case. Most SCRs and PRs believed that ethical discussions always precede decisions at plenary meetings, whereas most AWORs disagreed. At the same time, most SCRs and PRs believed that harm-benefit analyses were frequent, although commonly performed during preparatory stages rather than during the actual plenary meeting. AWORs, by contrast, perceived that harm-benefit analyses were rarely or never performed at all.

Numerous reasons could contribute to these findings. In the absence of a universally accepted and clearly defined methodology for how to perform the harm-benefit analysis (Brønstad et al., 2016; Grimm et al., 2019; Sena and Currie, 2019; Hajosi and Grimm, 2025), respondents are likely to be unsure of what is expected of them when completing this task. They may believe harm-benefit analysis and ethical discussion to be interchangeable concepts (Röcklinsberg et al., 2014), or, as discussed above, have different understandings of how to interpret (and apply) concepts of immediate relevance for the performance of an ethical review, such as “harm”, “benefit”, or “ethical considerations” (Brønstad et al., 2016; Eggel and Camenzind, 2020). Even well-defined principles like the 3Rs have been found to cause confusion (Nøhr et al., 2016; Lindsjö et al., 2021; Maestri, 2021; Jörgensen et al., 2025). Moreover, individuals have different affective responses to words, imagery and concepts that guide their decision-making (Slovic et al., 2007). Importantly, though, even if purely ethical discourse is seemingly rare, ethical concerns can be implicitly dealt with in technical or scientific discussions (Houde et al., 2003; Brønstad and Sandøe, 2020), for example, when addressing Refinement and Reduction. However, even these discussions may be discouraged if applications are too large and complicated or too brief and incomplete, especially when coupled with time pressure, a heavy workload, and epistemic challenges.

The apparent lack of ethics in a plenary context raises questions about procedural legitimacy. Conducting the harm-benefit analysis during preparatory stages means that not all committee members are included in this central evaluative task. In light of anecdotal evidence that 50% of committee members may not identify even severe ethical concerns unless applications are discussed in a larger group (Elliott and Hunter, 2008), this could prove detrimental. Given that, on average, 99% of all applications in Sweden are approved (SBA Research Animal Unit, personal communication, 10 April 2026), with similar figures observed across many other countries of the EU (Busquet, 2020), this poses significant risks to animal welfare, the scientific rigour of animal research, and to public trust in the review process. The lack of uniform guidelines for the harm-benefit analysis also jeopardises the transparency and harmonisation of ethical evaluation of animal research in the EU (Brønstad et al., 2016; Olsson et al., 2016; Hajosi and Grimm, 2024).

4.2.5 Competing interests and polarised perceptions

To uphold procedural justice and public trust in its independence and fairness, the ethical review requires the balancing, not only of harms and benefits, but also of competing interests (Pietrzykowski, 2021). As supported by the findings of this study, this can breed tension between involved parties when diverging perspectives are required to coexist and cooperate within the same system.

There is no consensus on how to assess the outcome of AEC work (Varga, 2013), confirmed by the fact that AWORs in our study present as generally, often significantly, more critical of the ethical review and its practical implementation than other committee member respondents. Some AWORs also described the feeling that their arguments were systematically devalued, that the opinions of SCRs’ carried disproportionate weight, and that local researchers “had each other’s backs” at the expense of animal interests. That AWORs feel underappreciated, overrun, or even belittled during committee work is not a new discovery (Röcklinsberg, 2015; Tjärnström et al., 2018). One potential underlying mechanism could, as has already been discussed, be their fundamental values. Values shape the understanding of concepts and of thresholds for when obligations are met (Bruskotter et al., 2019; Walker and McMillan, 2021), which, for committee members, likely influences both how they conduct their own work and judge the work of others, as well as their perceptions of the purpose and functioning of the ethical review as a whole. As shown, AWORs primarily see themselves as advocates for animal interests, and was the only respondent group to consider all four listed aspects (Figure 2) as “of utmost importance” for the review to fulfil. Hence, AWORs are likely to perceive any failure to protect animals, which is the purpose of Directive 2010/63/EU and its ethical review, as a devastating defeat.

A number of animal researcher respondents perceived the ethical review as inherently negative, which became evident primarily in responses to open-ended questions. Some even argued that the review ought to function in a facilitative capacity for researchers and their work. In view of this, it is worth recalling the meaning of the term review, defined by the Merriam Webster Dictionary as “an act or process of carefully looking at or examining the quality, condition, etc., of something or someone” (The Merriam-Webster Dictionary, 2026). By definition, a review entails scrutiny rather than accommodation. Given the fundamental point of departure for Directive 2010/63/EU, that animals shall not be caused unnecessary suffering, the burden rests with researchers to demonstrate that their proposed studies comply with applicable legal and ethical standards, and therefore, it is not reasonable to expect the ethical review process to be tailored to their needs or convenience. Whilst the process should not be unnecessarily burdensome, its primary function as a formal threshold and oversight mechanism must remain intact. The aim should therefore be to ensure that it is just, predictable, proportionate, aligned with its core purpose and transparent to the general public.

Despite the varying perspectives, experiences, and expectations registered in this study, respondents from all respondent groups clearly expressed concern for animal welfare and regulatory compliance, indicating a shared foundation that should not be overlooked. The indicated gap between theory and practice, in other words, is not primarily a gap of values. Rather, it mirrors an epistemological discrepancy and diverging perceptions of aims and intentions, manifesting itself as a gap of implementation. If this is adequately addressed, AECs might be able to bring practice closer to intent.

4.3 Limitations of the study

Several limitations should be considered when interpreting the findings of this study. The response rate amongst committee members was modest, and the total number of animal researchers who received the questionnaire is unknown, limiting generalisability. The two open-ended questions analysed thematically were placed late in the survey, which means that preceding questions may have influenced respondents’ open-ended responses in ways that cannot be quantified. Epistemic differences may also have had an impact. Observed correlations between respondent characteristics and answers do not automatically imply causation. The sample size was relatively small, and representation of particularly animal welfare organisation representatives (AWORs) was low. Although the statistical test used accounts for small sample sizes, findings of statistical significance associated with respondent role should therefore be interpreted in light of this limitation, as the inclusion of additional responses could affect the resulting p-values. Nevertheless, the distribution of committee member respondents closely reflects reality, and responses according to role mirror observations from previous research, supporting the general relevance of these results.

5 Conclusions

This study provides a comprehensive insider account of how both Swedish animal researchers and AEC members perceive and experience the ethical review, revealing that different participants of the same process may have very different perceptions thereof. Animal welfare organisation representatives (AWORs) stand out as the most critical across nearly all investigated dimensions, though the relatively low representation of this group should be considered when interpreting these findings. Possible, and often interrelated, explanations for this divide include: differing underlying values amongst respondents; divergent expectations of the aims and outcomes of the ethical review; contrasting views on the roles and responsibilities of the parties involved, as well as on when specific tasks and obligations are fulfilled; and differing understandings of key concepts.

To bridge the observed rifts in perceptions, a clearer alignment in the understanding of the purpose of ethical review, and in the roles and responsibilities of its participants, is needed. Only then, when the parties involved (particularly the committee members responsible for conducting the review), share such understandings, can they work effectively towards their fulfilment. However, shared understandings do not imply uniformity of views. On the contrary, a diversity of perspectives is essential to ensure adequate representation of relevant interests, as reflected in the mandatory inclusion of lay members in Swedish AECs. It is hence crucial to recognise the importance of including different viewpoints and ranges of interests in the ethical scrutiny of animal research, and that these contributions are accorded equal respect in the discussion.

With this in mind, questions arise as to whether the current system for the ethical review of animal research is capable of achieving this. Arguably, it may be placing unrealistic demands on committee members by expecting them to simultaneously examine and clarify technical, often complex, aspects of proposed projects, evaluate compliance with Replacement, Reduction and Refinement, and at the same time engage in ethical reasoning involving the difficult task of balancing of harms against benefits. Whilst each of these tasks is indispensable in its own right, combining them within a single review may exceed what can reasonably be asked of one body, and as a result, certain tasks and perspectives may be deprioritised in favour of others.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fanim.2026.1864730/full#supplementary-material

Supplementary material Supplementary material at the source

Footnotes

Edited by: Christine Janet Nicol, Royal Veterinary College (RVC), United Kingdom

Reviewed by: Sally Thompson Iritani, University of Washington, United States

Christian Rodriguez Perez, University of Basel, Switzerland

Footnotes

Percentages are rounded to the nearest whole number and may therefore not sum to exactly 100%.

In November 2025, the L150 was restructured into two separate documents: the Swedish Board of Agriculture’s Regulations and General Advice on Laboratory Animal Use, SJVFS 2025:28 (the L151), and the Swedish Board of Agriculture’s Regulations and General Advice on the Ethical Review of Animal Research, SJVFS 2025:29 (the L152).