A critical review of the literature is underway. This type of review provides an in-depth analysis of existing publications on a specific topic. Its purpose is to synthesize current knowledge; identify trends, gaps, and future research directions; and position new research within the context of prior work. Rather than merely summarizing the literature, it critically analyzes and evaluates previous studies, adopting a reasoned and reflective stance on their findings [18]. Given that the concept of “humanism” is polysemic, this approach will make it possible to identify the main concepts associated with humanism in health care and to list their theoretical and operational definitions, where applicable [19,20].

A literature search was conducted in Academic Search Complete, CINAHL Plus, Education Source, ERIC, MEDLINE, and SocINDEX (Textbox 1). The review includes literature related to health sciences professionals, educators, and students. Inclusion criteria were restricted to peer-reviewed papers published between 2015 and 2025 in either French or English. The corpus included empirical studies (qualitative, quantitative, and mixed methods), systematic reviews, conceptual and theoretical studies, and critical essays. Sources were included if they explicitly addressed humanism and its conceptual, pedagogical, or evaluative dimensions in the health sciences. Anecdotal texts, nonscholarly publications, and works focused solely on general professional ethics without explicit reference to humanism were excluded.

A 2-level content analysis has been initiated to systematically interpret textual data pertaining to the concept of humanism in health care [21]. Manifest content analysis focused on extracting explicit information, particularly the identification and synthesis of formal definitions of humanism reported in the academic literature. In parallel, latent content analysis aims to uncover implicit conceptual structures and attributes embedded within these theoretical formulations, including both convergent and divergent perspectives. To enhance the analytical depth of the latent analysis, the research team adopted a structured framework based on interrogative pronouns (who, what, when, where, why, how, and about what), as proposed by Levasseur et al [22] in their scoping review of vulnerability definitions. This methodological strategy facilitates the identification of underlying assumptions, contextual determinants, and relational dynamics that are often overlooked in surface-level analyses. Applied to the study of humanism, this approach allows a nuanced and contextually grounded exploration of its theoretical dimensions, thereby contributing to the refinement and operationalization of the concept within health care research and practice.

The screening process followed a structured and transparent approach aligned with the objectives of a critical review. Following literature searches conducted within the EBSCOhost databases, titles and abstracts were screened by 2 research assistants using predefined inclusion and exclusion criteria. This process was supported by an academic librarian, who contributed to refining the search strategy and clarifying eligibility criteria. As the search strategy evolved, a member of the research team validated all adjustments to ensure consistency and methodological coherence. All potentially relevant records were retained for full-text review, which was conducted by research professionals. Throughout both screening stages, cases of uncertainty or conceptual ambiguity were discussed with a senior researcher, who provided oversight and ensured alignment with the study’s conceptual framework. The screening process was managed using Rayyan systematic review software, which facilitated record organization, eligibility tracking, and transparent documentation of inclusion decisions. Consistent with the epistemological orientation of a critical review, formal interreviewer agreement statistics were not calculated. Rather, methodological rigor was ensured through iterative discussions, expert oversight, and reflexive calibration at key decision points, particularly when interpretive judgment was required.

This step constitutes the experiential component of phase 1. Its objective is to identify the core theoretical concepts that define humanism, along with their operational definitions, using an inductive approach grounded in lived experiences. Specifically, the concept of humanism in health care is explored through the distinctive and complementary perspectives of 3 categories of key informants: clinician-educators, learners, and members of the public, who have directly experienced humanism in health care settings. This approach enables the conceptualization of humanism to be informed by real-world experiences, thereby enriching and contextualizing its theoretical foundations.

Three categories of key informants were invited to participate in focus groups (Textbox 2) using a convenience nonprobability sampling strategy [23].

Participant recruitment was guided by both general eligibility criteria applicable across all key informant categories and specific inclusion parameters tailored to each subgroup. The universal eligibility criteria for participation in phase 1 focus groups were as follows: participants were required to be aged ≥18 years, demonstrate sufficient proficiency in French to actively participate in focus group discussions, have access to a computer equipped with a functional camera and microphone, and be able to independently use the Teams (version 26134.1702.4747.7366; Microsoft Corp) app.

The recruitment process followed a consistent structure across the 3 key informant categories, although the intermediaries involved in facilitating outreach differed, as outlined in Textbox 3. The research team initiated recruitment by emailing selected intermediaries, inviting them to share a recruitment poster via their social media platforms (eg, Facebook [Meta Platforms Inc] and Instagram [Meta Platforms Inc]) or through direct email to their respective networks (clinician-educators, learners, or members of the public). The poster presented a brief overview of the study and included both a hyperlink and a QR code directing interested individuals to a short eligibility and interest questionnaire. Respondents who completed the questionnaire were then contacted by a member of the research team, who provided further information about the study and addressed any questions. Following this initial contact, eligible participants received an email containing a link to the electronic information and consent form, along with a sociodemographic questionnaire. They were also invited to indicate their availability for scheduled focus group sessions. Within 2 weeks of receiving the completed forms, participants received a formal invitation with the Microsoft Teams link for their assigned focus group.

Two complementary data collection strategies were implemented. Focus groups served as the primary data collection method for this stage of the study and were conducted remotely and synchronously via the Microsoft Teams platform. Videoconferencing offers enhanced flexibility and accessibility, removing barriers related to geographic location and travel time [24,25]. Each focus group lasted between 80 and 100 minutes and included key informants from 3 categories: clinician-educators (n=7), learners (n=6), and members of the public (n=5). All focus groups were audiovisually recorded and cofacilitated by 2 members of the research team using a structured discussion guide developed specifically for each informant category. These guides included open-ended questions organized around the 7 interrogative pronouns (who, what, about what, why, when, where, and how), as proposed by Levasseur et al [22], to capture the unique experiential dimensions of humanism from each group. Participants were invited to recall a significant experience related to humanism from their perspective as clinician-educators, learners, or members of the public. A variety of facilitation techniques—including reformulation, clarification questions, and triadic and circular questioning—were used to minimize formulation bias and promote in-depth exploration. The order and sequence of questions were adapted to the natural flow of each discussion. All recorded sessions were transcribed by a research assistant for subsequent thematic analysis.

The second strategy involved the use of a reflective journal, which consists of reflective and critical introspection aimed at documenting the subjectivity of the research team, including perceptions, assumptions, values, decisions, and interpretations. These data served a complementary function by enhancing the transparency of the research process and facilitating the establishment of links between the data collected and the analytical interpretations [26].

The transcripts from the focus groups were subjected to a thematic coanalysis conducted by the research team. A mixed coding strategy, combining both deductive and inductive approaches, was applied in accordance with the methodology proposed by Miles et al [27]. To ensure that the research team’s interpretation accurately reflects the content of the focus group discussions, a summary report was drafted and emailed to participants, providing them with an opportunity to offer feedback on this collective output. This triangulation of perspectives (theoretical and experiential) will facilitate the identification of statements for an initial formulation of the conceptual (ie, abstract statements describing what humanism is) and operational definition of humanism (ie, concrete statements describing how humanism manifests in health care) to be submitted to a panel of experts. Although the number of statements cannot be predetermined, they will offer semantically rich and nuanced formulations.

Each e-Delphi round will span approximately 4 to 6 weeks. At the launch of each e-Delphi round, the survey home page will include a reminder of the method’s objectives, its sequential stages, and the current round’s position within the process. Before accessing the questionnaire, participants will be informed that their responses will be anonymous.

In the first round of the e-Delphi process, participants will be asked to rate the importance of each pre-established statement using a 4-point Likert scale (1=“Not a priority,” 2=“Low priority,” 3=“Rather a priority,” and 4=“High priority”). Each statement will be accompanied by a brief conceptual rationale to support informed evaluation. Panelists will also be invited to suggest modifications to proposed statements, formulate additional statements, and provide qualitative input through free-text responses. Participants will have 2 weeks to complete the survey and may revise their responses at any time until the survey closes. The estimated completion time is 30 to 45 minutes.

Upon survey closure, descriptive statistics will be used to summarize panelists’ characteristics (eg, professional role, discipline, and years of experience) and Delphi response data during each round. Specifically, for each statement, we will analyze measures of central tendency (median) and dispersion (IQR), as well as the distribution of ratings across response categories. These methods are appropriate given the ordinal nature of Delphi rating scales and are widely recommended for consensus-building studies. Therefore, descriptive statistics will be used to assess the level of agreement; identify areas of convergence or divergence among experts; and guide decisions regarding the retention, modification, or elimination of statements across Delphi rounds. Qualitative comments provided by panelists will complement the quantitative summaries to support the interpretation and refinement of statements. Consensus will be defined a priori as ≥70% of participants rating a statement as either “Rather a priority” or “High priority.” Qualitative data from free-text responses will be thematically analyzed and narratively summarized to enrich the interpretation of the quantitative results by providing contextual depth and insight into expert perspectives. Any newly suggested statements or modifications will be discussed among the research team to determine whether they are relevant for inclusion in the following round.

Following completion of round 1, anonymized narrative summaries will be shared with all panel members to foster continued engagement, a sense of ownership, and collaborative partnership between participants and the research team. Panelists will receive their individual responses from the first round and the group-level median and IQR for each item.

Participants will be invited to reevaluate and rerate each statement using the same 4-point Likert scale. Additionally, they will be asked to assess any newly proposed or reformulated statements that emerged from round 1.

Descriptive statistical analyses will be conducted to assess consensus using the same criteria described previously: a statement will be considered to have reached consensus if ≥70% of participants rate it as either “Rather a priority” or “High priority.” Qualitative data from open-ended responses will be analyzed and summarized narratively. The research team will review all new suggestions to determine their relevance for inclusion in subsequent rounds.