Interventions Designed to Improve the Learning Environment in the Health Professions: A Scoping Review

This article was migrated. The article was marked as recommended. Purpose: To identify and describe interventions designed to affect the learning environment (LE) in health professions education, summarize factors that influence the LE, and determine gaps that require additional research. The LE can be thought of as a dynamic and complex construct co-created by people in a particular setting. A positive LE represents a welcoming climate for learning, which enhances satisfaction, well-being, academic performance and collaboration, while a negative LE restricts participation and learning, leading to emotional exhaustion, depersonalization and burnout. Method: A six-step scoping review methodology was followed to identify and report on literature that describes interventions affecting the LE in the health professions education: 1. Identify the research question, 2. Identify relevant studies, 3. Select studies to be included, 4. Chart the data, 5. Collate, summarize and report results, and 6. Consult with stakeholders. Results: 2,201 unique citations were identified and reviewed using titles and abstracts. 240 full-text articles were retained for detailed review, resulting in the inclusion of 68 articles. Study results are reported in relation to essential components of the LE: personal, social, organizational, physical and virtual spaces. Results of four different types to the studies of the LE are described: specific interventionsimpacting the LE, comparisonsof perceptions of the LE by two or more different groups, associations with other variable such as well-being with the LE, and descriptivestudies of the LE. Major influences included accreditation regulations, curricular interventions, faculty/staff development grading practices, instructional interventions, placements, physical and virtual spaces, and support services; and are reported along with specific interventions. Conclusion: These results reflect the complexity of the LE and the need for conceptual clarity. Since the quality of the evidence was not evaluated, the identified influences should be viewed as potential opportunities to improve the LE.


Introduction and Purpose
Educational learning environments (LE) dramatically affect the way participants think and feel, engage and work.Positive LEs support learning and are welcoming, collaborative, (Chinthammitr and Chierakul, 2014;Thomson et al., 2014;Tackett et al., 2017) and respectful while negative or "chilly" LEs (Janz and Pyke, 2000) are destructive and restrict participation and learning.LEs describe the dynamic, co-constructed perceptions, experiences and behaviors of participants in the physical and virtual spaces within which learning occurs.But more importantly, it also refers to the tone of the educational climate or culture, and the routine way people interact.LEs affect a wide variety of factors important to learners and providers alike: burnout, depersonalization and emotional exhaustion; satisfaction and wellbeing; identity formation; performance and collaboration (Darcy A. Reed et al., 2011;Thomson et al., 2014;Castillo-Angeles et al., 2017;Tackett et al., 2017).While interventions designed to improve LEs for health professionals have targeted many of these factors, which interventions have been studied?Given the diversity of ways LEs have been defined, how can these interventions be identified and categorized?If we could find such interventions, we could better target efforts to improve the learning environment for all.The purpose of this scoping review is to identify and classify interventions designed to improve the environment for learning in the health professions.
By interventions, we mean the introduction of a planned new activity (e.g., near peer coaches) or organizational change (e.g., curriculum, training site, duty hours) that is anticipated to have an impact on the learning environment.Our primary purpose in this review is to identify interventions that could improve the LE, but we also recognize that it is important to understand the factors that influence the LE, whether included in formal interventions or not.Thus, we cast our net more broadly than just a focus on interventions per se.
The learning environment (LE), which appears frequently in the health professions education literature, is a complex theoretical construct that lacks a unified definition (Genn, 2001b(Genn, , 2001a;;Roff and McAleer, Sean Sue Roff, 2001).The conceptual ambiguity surrounding this term has arisen, in part, from the varying disciplines and associated theoretical lenses used to investigate this phenomenon (i.e., anthropology, education, psychology, and sociology).The LE can describe personal experiences and perceptions (psychology and education), social interactions (sociology and education), organizational culture and practice (anthropology and sociology), physical facilities and online spaces (sociology and education) within which learning occurs.It can be associated with formal and informal learning experiences that occur in classroom, online, simulation and clinical settings.
The LE is often used interchangeably with such terms as atmosphere, educational environment, learning climateandorganizational culture.The LE has been defined as "a set of features that gives each circumstance and institution a personality, a spirit, a culture and describes what it is like to be a learner within that organization" (p.553) (Holt and Roff, 2004).However, just what these features are is inconsistent from one situation to another and from one study to another.The LE can be thought of as a complex psycho-social-physical construct co-created by individuals, groups, and organizations in a particular setting, and shaped by contextual climate and culture (Palmgren, 2016).
There is little disagreement that the LE is important, linked to various educational outcomes (Genn, 2001a(Genn, , 2001b)), and the focus of a number of accreditation regulations (e.g., LCME, ACGME, GMC).While the perceived importance of the LE has led to numerous efforts to measure it (Colbert-Getzet al., 2014), there is still a lack of clearly identified, evidencebased interventions or conditions that positively impact the environment for learning in the health professions.
Conceptual framework for the learning environment Although many authors do not provide an explicit theoretical perspective on the LE in their studies, we believe that the LE can be best understood and studied through the lens of sociocultural learning theories that include situated cognition, situated learning, ecological psychology, and workplace learning.The LE is conceived by different people in different ways, is dynamic and emergent, and is co-constructed through interactions and activity.Within the situated learning framework, learning involves acculturation into a new knowledge community or community of practice through active participation -initially as a legitimate peripheral participant and emerging into a full participant (Lave and Wenger, 1991).Ecological psychology and workplace learning emphasize that social interaction is facilitated through affordances in the learning/working environment (tools, scaffolded relationships, tasks, language, concepts) and the active engagement of learners (through their agency, engagement and emerging autonomy) (Billett, 2001).Situated cognition theorizes that learning is social and involves an interaction between persons and environment -thus linking learning, situations and culture.Specifically, knowledge is embedded in the activity, context and culture in which it is learned (Brown et al., 1989).
Each of these theories emphasizes the importance of interactions and collaborations with others "as the means for students' learning/participation, both through learning knowledge and skills from others, and through becoming familiar with the norms, cultural beliefs and attitudes existing in the communities to which they (the learners) are being introduced" (p.739) (Schönrock-Adema et al., 2012).However, the LE construct extends beyond typical sociocultural frameworks to include intra-individual psychological characteristics (learning preferences and history), as well as institutional culture, organizational structures and physical and virtual spaces in which students learn.It should be emphasized that the LE is not "owned" by any particular theoretical perspective.Neither is the LE often a central concern, which leaves the construct in something of a theoretical limbo.

Components of the learning environment
Lacking a canonical theory of the LE from the health professions education literature, we sought to synthesize multiple conceptual frameworks (Moos, 1974(Moos, , 1980;;Genn, 2001b;Schönrock-Adema et al., 2012;Gruppen et al., 2015;Gruppen and Stansfield, 2016;Gruppen, Rytting and Marti, 2017) and identified four overlapping and interactive core components (Figure 1): 1. Personal Component.The individual learner interacts with the LE through activity, develops perceptions of the LE, and engages in personal growth through clarity about goals, selection of relevant and meaningful learning; and in the process develops professional identity and increasing autonomy.
2. Social Component.Learners engage with others and navigate multiple relationships, which shapes their perceptions of and experiences with the LE.These relationships include: peer-to-peer (competition, cooperation, shared values and learner culture), learner-to-faculty/staff (trust, feedback, communication, instructional strategies, mentoring), and learner-to-patient (responsibility, acceptance and trust).All of these social relationships influence what and how students learn.
3. Organizational Component.Organizations provide structure, guidance and support for learning, including curriculum resources and artifacts, geographic placements, accreditation rules as well as organizational practices, culture and policies (orderly environment, rule clarity, duty hours, regulatory environment, teacher control, curriculum, placements, technology infrastructure).One example of this is the Clinical Learning Environment Review (CLER) implemented by ACGME.The underlying premise of the CLER program is that the educational program and patient care will be improved if constructive actions are taken regarding patient safety, health care quality, care transitions, supervision, fatigue management, and professionalism (Accreditation Council on Graduate Medical Education, no date; Weiss, Wagner and Nasca, 2012).We also include placements in the community (geographical settings/locations) in this organizational component as well.
4. Physical and Virtual Component.Learning and practice take place within physical spaces of educational and practice settings.Similarly, informational infrastructures and resources (e.g., online resources, electronic health records) also provide a virtual "space" in which learning is fostered or obstructed.
These components serve as an organizing framework for the diverse and often implicit definitions of LE for this review, but they do not constitute a complete theory of the learning environment in the health professions education.Such a theory will require considerable debate and discussion within the community.Nor are our categorization of individual studies definitive; most studies include elements from more than one component.

Studies of the learning environment
We conducted a scoping review of the literature to identify and characterize interventions that appear to affect the LE in order to better prepare health professionals for delivering quality patient care and engaging in a fulfilling practice.O'Malley, 2005) for scoping reviews.This framework includes six steps, which we used to organize our methods (Steps 1-3) and results (Steps 5-6).
Step 1: Identify the research question Based on several conference calls, we collectively discussed and agreed upon the purpose and rationale for this review, which informed the formulation of our research questions.In our discussions, we considered the population, types of relevant interventions, and impact on the LE.
Step 2: Identify relevant studies We assembled a research team with expertise in health professions education, clinical medicine, and information science.All team members had interest and experience in health professional LEs as well as experience in conducting literature reviews in health professions education.
LM, a health professions education researcher trained in information science, collaborated with a medical librarian to search and manage results from PubMed, Embase, Scopus, CINAHL and ERIC.With input from the team, search strategies were crafted using Boolean operators to combine controlled vocabulary terms (e.g., medical subject headings) and key words for all relevant concepts (search details available in Appendix 1).Our searches, were carried out beginning in August 2017 and were finalized 11 October 2017.The searches focused on journal articles written in English.No date limits were set and both quantitative and qualitative studies were included.
Step 3: Select studies to be included in the review The research team collaboratively determined inclusion criteria based on our research questions.For inclusion, articles needed to describe a study of an educational intervention or associated factor that measured outcomes related to the learning environment and that targeted health professions trainees and/or practitioners.Therefore, we excluded articles only focused on measuring the LE and/or that did not include a clearly identified intervention on the LE.
Our initial study selection, based on titles and abstracts, was an iterative process conducted over regular phone meetings.
To ensure concordance on the inclusion criteria, we participated in several rounds of selecting studies as a group.In total, each reviewer examined approximately 500 titles and abstracts.When moving to independent selection, we continued group discussions for any studies for which inclusion was uncertain.If consensus was unmet based on the title and abstract, the full-text was reviewed and consensus was achieved.
Step 4: Chart the data We collectively created a data charting form, which was adapted from a data extraction tool utilized by the Best Evidence Medical Education Collaboration for knowledge syntheses in health professions education (Issenberget al., 2005) and tailored to our research questions.Before implementation, we tested the form on four citations as a group to ensure agreement.Upon agreement, we each independently charted data for approximately 50 articles with one of the authors (DI) reviewing an additional 40.Following data charting, we held weekly phone calls to pose questions and ensure consistency in how we extracted study information.
Step 5: Collate, summarize and report results Our database search identified 2,201 unique citations; 68 met the inclusion criteria.See results.
Step 6: Undertake consultations with stakeholders This will be completed at a conference convened by the Josiah Macy, Jr. Foundation in April 2018 to identify policy recommendations for improving the LE for the health professions.

Results
Our search retrieved 2,662 articles (PubMed=1,491; CINAHL=77, ERIC=132, Scopus=244, Embase=718); with duplicates removed there were 2,201 unique citations.Based on examination of all titles and abstracts, 240 full-text articles were selected for review.Following full-text review, 68 articles were retained for inclusion (Figure 2).In the set of studies, there were 18 nations represented and six professions studied (medicine n=54; nursing n=11; dentistry n=1, pharmacy n=1, veterinary n=1, chiropractic n=1).Preclinical and clinical students were the primary population (n=45), but several studies also included residents (n=12) and/or faculty members (n=4).In some cases, studies included more than one population, setting and/or profession.
We identified four approaches to the study of interventions in the LE.First, there are studies designed to assess the impact of a specific intervention or series of interventions on the LE.These include studies of changes in duty hours, curricula, placements, and faculty development and their impact on LE.These we term interventional studies.Second, investigators compared two different groups' assessments of the LE related to instructional formats, curriculum models, geographical placements, and grading practices.We titled these comparison group studies.Third, one or more variables of interest, such as resilience, burnout, mistreatment, achievement, and well-being, were associated with perceptions of LE.We called these association studies.Fourth, descriptive studies using qualitativemethods illuminate participant perspectives and identify themes associated with interventions in the LE, such as establishing a welcoming environment and teaching culture, continuity of participants, and availability of learning/practice space.We termed these as descriptive studies.Each of these four approaches offer important insights into interventions impacting the LE.
The results of the review are organized around these four approaches to studying LE interventions.

Interventional Studies
Sixteen studies described specific interventions to improve the LE (Table 1).One study aimed at the personal component, three studies addressed the social component, and 12 examined the organizational component; none targeted the physical/ virtual component.In the personal component, time spent by students on direct patient contact is positively related to their perceptions of the quality of the LE.In the social component, a formative assessment tool supported student's clinical learning and improved perceptions of LE, and supervision by the same preceptor created a more supportive relationship.

Comparison Group Studies
Comparison group studies were the most common of the four approaches and also quite diverse in how the comparison conditions were defined.Some were naturally occurring differences in the LE (e.g., in two different clinical sites) whereas others were side-effects of events or changes (e.g., institution of team-based learning).These 29 studies were sorted by personal component (1 study), social component (3 studies), organizational component (23 studies), and the physical/ virtual component (2 studies).See Table 2. Within the personal component, nursing students with and without prior experience with elder care perceived the nursing home LE similarly.In the social component, distance learning compared with local live learning were perceived similarly, yet learners tended to prefer traditional classroom environments.Blended learning, the combining of online and in-person learning, was preferred to traditional instruction.
In the organizational component, geographical placements were compared (rural/remote preferred to metropolitan referral centers) and curriculum models contrasted (integrated and problem-based preferred to traditional discipline curriculum).Also, school features, the presence of learning communities and pass/fail grading practices effects on LE were explored.In terms of their effects on the LE, highly-rated departments had legitimacy, good clerkship arrangements, and a focus on personal development and engagement of learners; schools with learning communities had more positive student perceptions of LE than schools without learning communities; and students in schools with grades had higher stress, emotional exhaustion and depersonalization than students in pass-fail schools.
In the physical/virtual component, medical students had higher overall satisfaction than residents with Veterans Administration hospital training, although students' satisfaction declined over time while residents improved.The LE for obstetrics and gynecology residents in community hospitals was perceived to be better than at tertiary care/referral hospitals.

Association Studies
We found 14 studies that reported associations of another important variable (such as burnout, career choice, department academic support) with the LE.These studies included 7 in the personal component, 2 in the social component, 5 in the organizational component, and none in the physical/virtual component (Table 3).In the personal component, resident performance on their certifying exams was positively associated with perceptions of the LE.Similarly, nursing student effort and grade point averages were also positively related to perceptions of LE.Student well-being was positively associated with having a community of peers, good quality of life and less emotional exhaustion and depersonalization.Students with higher resilience levels had better quality of life and better perceptions of the LE.Resident worries about future endurance/capacity predicted exhaustion and lower ratings of the LE.
In the social component, department educational leadership skills were not related to ratings of the LE.In the organizational component, when clerkships were sorted into provision of high and low supervision of students, students perceived that low supervision clerkship sites offered too few opportunities to examine patients independently, insufficient supervision/no feedback, staff lacked motivation to teach and held negative attitudes towards students, the site had too many students, and there was a lack of organization.Residents perceiving adequate support to succeed had less burnout, better resilience, better job satisfaction, better organizational support, and were more likely to have high performance on the in-service exam.Compliance with common program requirements in residency training was associated with better resident perceptions of the LE.Themes from Descriptive Studies While the vast majority of studies of the LE were quantitative and used standardized measures of the LE, a few descriptive studies used qualitative research methods to explore learners' perceptions of the LE.We found nine descriptive studies that addressed all four components (Table 4).Student perceptions of a constructive LE were associated with resilience, a focus on personal growth, feeling that they were learning in a meaningful place and becoming part of a community, and that they trusted the system to support them.In the social component, students described constructive LEs as being welcoming with scaffolding relationships and a strong teaching culture.Preceptors were perceived to enjoy teaching and provided appropriate instruction, feedback and role modeling.A poor social environment was characterized by mistreatment, neglect and negative attitudes toward learners, unclear expectations, insufficient supervision and too few opportunities to examine patients independently.In the organizational component, the teaching arrangements were well organized, and there was continuity of participants.Smaller and more rural clinical sites were perceived to be better as was a PBL curriculum.Destructive organizational attributes included lack of clear expectations for learners, failure to integrate students into teams, too many students, and lack of organization.In the physical/virtual component, availability of adequate space for students to interview patients was identified.

Discussion
The vast majority of studies included in this scoping review reported on interventions and influences that had a positive impact on the LE in 18 different countries representing medicine, nursing, dentistry, pharmacy, and veterinary medicine.
All four types of studies (interventional, group comparisons, associations with another key variable, and descriptive) described influences on one or more components of the LE.The majority of studies were focused on the organizational component, followed by the social component and the personal component.Very few studies examined the impact of the physical or virtual space component.
Our scoping review sought to answer three research questions, the first of which was: What interventions affect the LE in the health professions?A synthesis of the reported interventions aimed at influencing the LE are reported in Table 5.There were seven classes of influences on the LE (accreditation regulations, curricular interventions, faculty/staff development, grading practices, instructional interventions, placements, physical and virtual spaces, and support services) and 20 specific targets for possible interventions.Since the strength of the interventions displayed in Table 5 were not assesses, the list should be viewed as potential opportunities for improving the LE.perceptual dimensions of a particular setting.Interventions or factors positively associated with the personal component of LE included: time focused on direct patient care, having a community of peers, a good quality of life and high levels of resilience, learning in a "meaningful" place, and trust in a regulated system to support them.Factors with negative associations were poor quality of life leading to more emotional exhaustion, depersonalization, and worries about future endurance and capacity.These factors are less about interventions and more about the psychological characteristics of the learners and their perceptions of the environment.
Sociocultural learning theories associated with situated learning, situated cognition, ecological psychology, workplace learning explain these findings (Brownet al., 1989;Lave and Wenger, 1991;Billett, 2001).A supportive learning community encourages participation and scaffolds learning in the context of the setting.Motivation theory, which emphasizes autonomy, purpose/goals, mastery and relatedness also connect with these recommendations (Deci, Koestner and Ryan, 1999;Pintrich, 2003).Learners are intrinsically motivated to learn, develop autonomy, pursue a goal and purpose larger than themselves, and work collaboratively with others, especially if they are supported in the process.

Social Component of LE
Studies exploring the social component of learning reinforced the importance of interpersonal relationships in fostering a constructive LE.These relationships include teacher and learner (e.g.face-to-face or blended instruction and longitudinal clinical mentoring), learner to learner (e.g.peer instruction and support), as well as faculty to faculty (e.g.leadership performance).Studies did not address the learner and patient relationship.These studies also underpinned the importance of longitudinal relationships as well as the value of setting and revisiting expectations about performance and relationships.The descriptive studies highlighted the role of a strong teaching culture, strong role model skills and values, multiple levels of learners working together (e.g.near peer teaching) as well as the need to avoid mistreatment, unclear expectations, and insufficient supervision without feedback.Teamwork and its relationship to LE were not explicitly

Physical and Virtual Spaces
No studies addressed in the studies included in our review.These findings are consistent with situated learning (communities of practice and legitimate peripheral participation), situated cognition, and deliberate practice theory, as noted above.

Organizational Component of LE
The organizational component of the LE model was most frequently studied through comparative studies of contrasting LEs.Frequently, these contrasting environments were "natural experiments" rather than carefully designed studies specifically of the impact on the LE.Many of these were comparisons of alternative curricular models (e.g., problembased learning, team-based learning) or specific curricular interventions (e.g., augmenting feedback, faculty development, team-work skills) or larger setting of school comparisons (rural vs urban, alternative clinical settings within a larger academic institution).The uncontrolled and non-randomized nature of these studies limits the confidence one can place in the results, but the evidence is generally positive in indicating that some environments are perceived as better than others.These include: Courses or innovations to augment feedback, increase respect and well-being, and reduce mistreatment Faculty development programs focused on aspects of the LE rather than specific teaching skills Structural features like duty hour implementation, grading systems, supervisory models, and dedicated educational units Rural settings, smaller clinical placements, learning communities, and elective rotations, which may be surrogates for having more attention given to learners.
Given the diversity in study outcomes, disciplines, countries, and focus, it is not surprising that the results are often mixed.There is not a critical mass of studies on any given variable to provide convincing conclusions.Understanding the dynamics of how organizational features relate to the LE clearly builds on the theories of sociocultural and interpersonal interactions cited in the sections on the personal and social components of our model.However, the organizational component also leads to considerations of institutional and organizational culture that are seldom cited in LE studies.Organizational change (Kotter, 1995;Bolman and Deal, 2013), leadership models (Avolio, Walumbwa and Weber, 2009), and systems science (Miller, 1978)are a few of the conceptual domains that may be relevant and beneficial for better understanding how the LE functions at higher level human systems.

Physical and Virtual Space Component of LE
The physical/virtual space component of the LE encompasses the physical spaces of educational and practice settings in which learning and practice occur, and the virtual or online learning spaces.We identified three studies, two of which were comparison studies (Cannonet al., 2008;Diwadkar and Jelovsek, 2010)and one a descriptive study (Seltzet al., 2016), all of which were conducted in the US.Within these studies, physical components of the LE are peripheral rather than the main focus of the study.For example, in a survey of 125 Veterans' Affairs hospitals, physical space is one of four investigated subdomains that are associated the LE (Cannonet al., 2008).This study notes that for residents and medical students the maintenance and cleanliness of hospital facilities impacts the LE.
The lack of identified studies and limited coverage suggests a gap in the health professions literature and opportunities for future research.Health professions education researchers might refer to other fields, such as environmental psychology and higher education, as they have long studied the physical/virtual components of the LE and recognize the impact of space on learning (Oblinger and Lippincott, 2006).Furthermore, a need for knowledge about physical/virtual components of the LE will become more pronounced as health professions education institutions implement blended learning (Mehtaet al., 2013;Prober and Khan, 2013).Using blended learning approaches, faculty intentionally plan their teaching to engage trainees online and in-person to optimize the affordances of both modalities.While blended learning moves some of the learning out of the physical space and into the ether, it underscores the need for those opportunities in the physical learning space to directly support small group learning.In addition, as interprofessional education and practice increase, new spaces for conferences and huddles in the workplace will be needed.Ambulatory clinic space is also required for medical student practice, especially in the early stages of learning when they are inefficient.
We note that the physical and virtual space component received the least attention of the four components in our organizational framework, especially given the amount of time, energy and financial resources devoted to fundraising campaigns targeting expanded and improved physical spaces and online courses (Association of American Medical Colleges, no date).This lack of coverage may in part reflect the absence of sociocultural theoretical stances, where the location and its interaction with participants is a key element.Indeed, we suspect that clarity on definitional and theoretical stance would lead to more (needed) investigations of this component.

Recommendations
We have several recommendations that arise from this review: There is a significant need for theoretical development to provide a more comprehensive framework for both defining the learning environment and studying its impact on various educational outcomes.The need for better definitional and theoretical clarity became evident early in our review process.This lack of clarity led to challenges in constructing our literature search, as well as in synthesizing our findings.We believe that enhancing the definitional and theoretical clarity of the LE is a critical next step to improve our understanding of interventions, the components to target, and addressing practice gaps.
Similarly, the over-reliance on learner self-reported perceptions as a measure of the learning environment need to be supplemented by assessment methods that better address other viewpoints and the characteristics of the LE at the group and institutional levels.Reviews of assessment instruments are available and note the lack of consistent theoretical frameworks (Schönrock-Ademaet al., 2012;Colbert-Getz et al., 2014).
There are a number of gaps that warrant research attention: exploring the patient's impact on the LE, investigating how interprofessional and intra-professional teams influence the LE as well as the design and testing of interventions that are inclusive of multiple components from our model would be worthy of future investigations.Similarly, potential interventions to improve the LE should carefully consider creating a community of peers, ensuring support especially in times of transition and stress, emphasizing meaning in the work, and supporting personal resilience and autonomy.Physical and virtual spaces as settings for learning are also under-represented in the literature.
Educational scholars and practitioners must recognize that the contextual, background nature of the LE makes it a construct that may or may not be explicitly identified in individual studies.For example, our search returned only two articles (Lachanceet al., 2014;Schumacher et al., 2014)on resident duty hours as an element of the LE.There are, obviously, many more articles that examine the impact of duty hour changes on educational outcomes, but these are seldom labeled as "learning environment" and were thus missed in our search.Care must be taken to search more broadly in a given LE intervention to include articles that do NOT mention "learning environment".

Limitations
A particular challenge of conducting a comprehensive literature search for a construct like the LE, is that it has no uniform definition and is often a background phenomenon rather than an explicit component of a study.This challenge meant crafting a search strategy that was focused on the inclusion of the term "LE" and several synonyms.Despite our best efforts, we may have failed to retrieve all relevant articles on the LE because we did not use the right terms (LE or its synonyms).Additionally, we restricted our search to English language journal articles and thus may have excluded relevant research in non-English languages.Since the review was focused on interventions that impact the learning environment, studies that described the LE or validated a LE instrument were excluded.Some of these may have provided further insights into interventions.

Conclusions
The context in which people learn clearly has an impact on the learning process and its outcomes.This context includes numerous factors at the personal, social, and organizational levels.It also includes physical and virtual spaces.Because of this scope, discussing all of these factors under the term LE would appear to be a gross over-simplification.We argue that research in this area can only progress if investigators and practitioners become clear and precise about what they mean by LE.Clarity and precision will be facilitated by the development of more detailed theoretical models and congruent assessment tools.For example, the model we have developed from this review would suggest that authors should address the "personal learning environment" as distinct from the "social learning environment," the "organizational learning environment" or the "physical and virtual learning environments".Such distinctions are necessary to advance future research on the LE by focusing on a subset of components, variables and/or interventions rather than the enormity of all possible contextual influences.Similarly, because the specific LE in a given study is defined by the educational purpose, actions, and outcomes, further theoretical development of the LE concept must incorporate these foreground educational issues in order to understand the dynamics of the LE "background."

Take Home Messages
There is a significant need for theoretical development to defining the dynamics of the learning environment and studying its impact on various educational outcomes.
There is an over-reliance on learner self-reported perceptions as a measure of the learning environment.Other assessment methods are needed to better address other viewpoints and characteristics of the LE at the group and institutional levels.
Additional research attention is needed in such areas as exploring the patient's impact on the LE, investigating how interprofessional and intra-professional teams influence the LE, creating a community of peers, ensuring support especially in times of transition and stress, emphasizing meaning in the work, and supporting personal resilience and autonomy.Physical and virtual spaces as settings for learning are also under-represented in the literature.
The contextual, background nature of the LE makes it a construct that may or may not be explicitly identified in individual studies.For example, there are many articles that examine the impact of duty hour changes on educational outcomes, but these are seldom labeled as "learning environment" and were thus missed in our search.Care must be taken to search broadly in a given LE intervention to include articles that do NOT mention "learning environment".

Declarations
The author has declared that there are no conflicts of interest.

Ethics Statement
This is not human subjects research, only an analysis of the published literature.

External Funding
This article has not had any External Funding

Acknowledgments
This paper was originally prepared for and presentation at the Josiah Macy Jr Foundation Consensus Conference on the Learning Environment in the Health Professions, April 2018.

Figure 1 .
Figure 1.Four interactive components of the learning environment: personal, social, organizational, physical and virtual.

Figure 2 .
Figure 2. Review and selection of articles on learning environment interventions in health professions education.

Table 1 .
Themes from 16 studies of interventions to improve the learning environment in the health professions.
Nishioka (2014) (Nishioka et al., 2014) USA Nursing Implemented a dedicated education units (DEU) for students (+) Students perceived clinical learning experiences and mentoring

Table 2 .
Themes from 29 comparison studies of the learning environment in the health professions.

Table 2 .
Continued These two questions are addressed in relation to each of the four components of the LE.Personal Component of LEThe personal component of our LE model describes how individual learners interact with the LE, develop perceptions of the LE, engage in personal growth and develop professional identity.It describes the psychological, experiential and

Table 3 .
Themes from 14 association studies of the learning environment in the health professions.

Table 5 .
Selected targets for possible interventions to improve learning environments derived from 68 reviewed studies in the health professions.

Table 5 .
Continued Larry Gruppen is Professor, Department of Learning Health Systems at the University of Michigan where he directs an innovative, competency-based Master of Health Professions Education program.His scholarly interests center around clinical reasoning, assessment, faculty development, and research methods.David Irby is Professor Emeritus, Department of Medicine and Senior Scholar, Center for Faculty Educators, UCSF school of Medicine.His scholarly interests center around faculty development and clinical teaching.Steven Durning is Professor, Department of Medicine and Pathology and Director, Graduate Programs in Health Professions Education.His scholarly interests center around clinical reasoning, assessment, educational theory, and research methods.Lauren Maggio is associate professor, Department of Medicine, and the Associate Director of Distributed Learning and Technology in the Graduate Programs in Health Professions Education at the Uniformed Services University of the Health Sciences.Her research interests focus on scholarly communication, knowledge syntheses, and health information use.