Volume 57, Issue 11 p. 1036-1053
Open Access

‘Change talk’ among physicians in small group learning communities: An ethnographic study

Heather Armson

Corresponding Author

Heather Armson

Department of Family Medicine, Office of Continuing Medical Education and Professional Development, Cumming School of Medicine, University of Calgary, Calgary, Alberta, Canada

Foundation for Medical Practice Education, McMaster University, Hamilton, Ontario, Canada


Heather Armson, Sunridge Family Medicine Teaching Center, 2685-36th Street NE, Calgary T1Y 5S3, Alberta, Canada.

Email: [email protected]

Contribution: Conceptualization, Funding acquisition, Methodology, Validation, Writing - review & editing, Formal analysis, Supervision, ​Investigation, Visualization, Resources

Search for more papers by this author
Kathleen Moncrieff

Kathleen Moncrieff

Department of Family Medicine at Cumming School of Medicine, University of Calgary, Calgary, Alberta, Canada

Contribution: Writing - review & editing, Formal analysis, Validation, Visualization

Search for more papers by this author
Meghan Lofft

Meghan Lofft

Foundation for Medical Practice Education, McMaster University, Hamilton, Ontario, Canada

Contribution: Visualization, Writing - review & editing, Formal analysis, ​Investigation, Validation, Software, Resources, Data curation, Project administration

Search for more papers by this author
Stefanie Roder

Stefanie Roder

Foundation for Medical Practice Education, McMaster University, Hamilton, Ontario, Canada

Contribution: Writing - original draft, Methodology, Validation, Visualization, Writing - review & editing, Formal analysis, Project administration, Data curation, Resources, ​Investigation, Software

Search for more papers by this author
First published: 16 May 2023
Citations: 1
Funding information Funding for this project was provided by an internal research fund through the Foundation for Medical Practice Education, McMaster University in Hamilton, Ontario, Canada.



Physicians face uncertainties in complex clinical environments. Small group learning initiatives allow physicians to decipher new evidence and address challenges. This study aimed to understand how physicians in small learning groups discuss, interpret and assess new evidence-based information to make decisions for practice.


An ethnographic approach was used to collect data from observed discussions between practising family physicians (n = 15) that meet in small learning groups (n = 2). Physicians were members of a continuing professional development (CPD) programme that provides educational modules with clinical cases and evidence-based recommendations for best practice. Nine learning sessions were observed over 1 year. Field notes documenting the conversations were analysed using ethnographic observational dimensions and thematic content analysis. Observational data were supplemented with interviews (n = 9) and practice reflection documents (n = 7). A conceptual framework for ‘change talk’ was created.


Observations elucidated the following: Facilitators played a significant role in leading the discussion by focusing on practice gaps. As group members shared approaches to clinical cases, baseline knowledge and practice experiences were revealed. Members made sense of new information by asking questions and sharing knowledge. They determined what information was useful and whether it applied to their practice. They reviewed evidence, tested algorithms, benchmarked themselves to best practice and consolidated knowledge before committing to practice change(s). Themes from interviews emphasised that sharing of practice experiences played an integral part in decisions to implement new knowledge, helped validate guideline recommendations and provided strategies for feasible practice changes. Documented practice reflections regarding decisions for practice change(s) overlapped with field notes.


This study provides empirical data on how small groups of family physicians discuss evidence-based information and make decisions for clinical practice. A ‘change talk’ framework was created to illustrate the processes that occur when physicians interpret and assess new information to bridge gaps between current and best practices.



‘Family physicians work in a complex practice environment that requires them to balance patient-centered care with evidence-based practice’ (p. 117).1 Even with a strong evidence base, uncertainties remain regarding appropriate diagnosis and treatment decisions for optimal patient outcomes.2, 3 Medical research conducted outside primary care can be difficult to apply in family medicine, and lack of personal knowledge or practice experience can hinder effective knowledge translation and practice implementation.4, 5 Uncertainty motivates life-long learning as physicians continually look for information and try new solutions to problems in an ever-changing environment.6-8

The sheer volume of new research creates its own challenges.9 Searching for evidence related to clinical decisions is time consuming and requires specific skills.10, 11 Synthesised reviews of the literature (e.g. educational materials and clinical practice guidelines) provide physicians with information that is easier to implement in practice; however, uptake of synthesised reviews is low due to passive dissemination, lack of awareness, varying levels of evidence, conflicts in recommendations and point-of-care applicability.12-14

Clarifying the processes that lead from identifying the need to change based on new information to changing practices in the context of the clinical environment is complex.13, 15, 16 Many theoretical approaches exist describing how physicians learn and change.17, 18 Although these approaches have shaped educational interventions and continuing professional development (CPD) activities for practising physicians,9, 19 they have not elucidated the processes that describe how physicians move from new information to knowledge application in the form of changes in their practice.

CPD activities that encourage participatory learning experiences as opposed to passively disseminating knowledge are more likely to facilitate change.9, 19, 20 Case-based clinical conversations provide a structured forum where physicians can discuss their thinking behind practice change decisions.21 The conversation about how to enact change has been classified as ‘change talk’.22, 23 Getting to ‘change talk’ relies on adequate knowledge and practical experience to conceptualise approaches to cases, anticipate problems and make decisions in one's own practice context.21, 22 An example of ‘change talk’ in small group discussions is described in a recent scoping review24 in which ‘quality circles’ are described that help physicians scrutinise their practices in light of new evidence and deal with uncertainty, while enhancing their routine practices and confidence in their roles. Benefits of small group learning extend to the clinical environment, that is, encouraging physicians to order fewer and more appropriate tests, improving prescribing and enhancing critical incident reporting.24 Although this review included literature supporting behaviour change resulting from small group learning, no literature was found explaining how physicians decide to change practice.

Our study used a focused ethnographic approach to narrate the small group ‘change talk’ and identify the conversational elements that take place as physicians discuss new evidence-based recommendations in the context of their current practices. Ethnography ‘is concerned with the everyday experiences of individuals, organizations and society, with a commitment to understanding the cultural context in which these experiences and social interactions take place’ (p. 1455).25 Focused ethnography has been used previously in the medical education context.26 It can provide a lens into how physicians make sense of new evidence and what information is important for clinical practice change. This focused ethnography was done within the cultural context of a well-established Canadian CPD programme, to answer the following:
  1. How do physicians interpret new information?
  2. How do physicians assess applicability of new information to their practice?
  3. How do physicians discuss the integration of new knowledge into their clinical practices?
  4. What contributes to a decision to change practice?
  5. How do physicians address implementation issues?


The principal investigator (HA) observed family physicians' discussions in a natural setting of small group learning sessions. Observational data were supplemented with interviews of participating group members and practice reflection documents completed during the sessions.

2.1 Context

The study took place within the context of a well-established Practice-Based Small Group Learning (PBSGL) Programme (website). The PBSGL programme was designed by and for family physicians to make CPD interactive and relevant.27 At the time of this study, the PBSGL programme was in operation for over 25 years and supporting 873 small groups (approximately 6111 members) across Canada. Ongoing groups of 8–10 members meet monthly to discuss evidence-based educational modules (Data S1—sample module) on clinical topics guided by a trained peer facilitator.27 Facilitators utilise a specific instructional design called the ‘PBSGL process’ that has developed over many years through reflection, research and literature review.1, 27, 28 The PBSGL process includes identification of gaps between current and best practices, case-based discussions, sharing of practice experiences to enhance practice implementation, completion of a practice reflection tool (PRT) documenting decisions for practice change(s) using commitment-to-change (CTC) statements and subsequent follow-up on the success of planned practice changes (Data S2—PRT and follow-up PRT).1, 27, 28 Facilitators submit PRT and follow-up PRTs to the programme so that members can receive professional credits for the small group learning sessions.

It was deemed appropriate to do this focused ethnography within the cultural context of this CPD programme as all learning sessions presumably follow the same established PBSGL process thus providing an environmental constant when focusing observations on the small group discussions among family physicians. To obtain richness in data, the aim was to recruit two groups (one group that had been meeting for 1 year and another group that had met for more than 5 years) and observe several small group learning sessions over a period of 1 year.

2.2 Participant recruitment

Two groups were recruited from 20 PBSGL groups in Southern Alberta, Canada that were within driving distance of the observer (HA). Invitations were sent via email to group facilitators (July 2016). Reminder emails were sent 1 month apart (August and September 2016) to give facilitators time to present the proposed study to their groups. All members in the participating groups individually confirmed their study participation by signing a consent form. No incentives were given for study participation.

2.3 Reflexivity

The observer (HA) is a family physician, medical educator and education researcher who has been involved with the PBSGL programme in several capacities, including facilitating her own PBSGL groups. Fully understanding the PBSGL processs,1 HA was able to act as an observer of the group sessions without first having to immerse herself in the cultural context of PBSGL. Most participants knew HA as a colleague and accepted HA's presence as an observer taking notes. All research team members (HA, KM, SR, ML) participated in the analysis of field notes and are familiar with the programme. KM is also a family physician and medical educator that is involved in the PBSGL programme as a qualitative researcher and an active member of a PBSGL group. SR and ML are researchers with experience in qualitative research methodologies and knowledge translation.

All research team members have been trained in scientific methods and evidence-based medicine; thus, supporting the research philosophies of post-positivism, that is, using multiple methods to search for a logical truth in knowledge being aware that all methods are imperfect.29, 30 However, the research team members, in particular HA, are also familiar with the uncertainties of the clinical practice environment and the construction of knowledge through lived experiences and thus can identify with the research perspectives of social constructivism.29-31

2.4 Observations in natural setting

2.4.1 Data collection and processing

PBSGL sessions were not audio or video recorded to support a natural meeting environment and minimise any concern about contentious issues perceived by participants with discussion topics. The observer took detailed handwritten notes of the discussions, specifically focusing on the verbal interactions ‘who said it’ and ‘what was said’ (Figure 1). Fieldnotes also noted any references to the module, session activities (Table 1) and her own reflective thoughts that occurred while observing the discussions. These handwritten notes were typed after each session and shared with research team members for qualitative analysis. All field notes were imported into NVivo™12 software program33 for processing and analysis. Notes were reformatted by grouping each speaker's statements into paragraphs to easily visualise who made each statement. Conversations were also grouped into small group process categories for ease of coding (Table 2). Some coding was exported from NVivo™1233 into Microsoft Excel 201034 to help compare observations within and between sessions observed for two groups.

Details are in the caption following the image
Samples of field notes with assigned themes for two groups discussing the same module topic.
TABLE 1. Observational dimensions of small group learning sessions.32
Dimensions Description Observation notes
Space Natural setting

Group 1 meets at member's homes

Group 2 meets at a restaurant

Actors Description of participants, community of practice members

Group 1

  • 7 members (6 women, 1 man) meeting for 1 year
  • Some members from same clinic, some from different clinics; same facilitator for all observed sessions (n = 5)

Group 2

  • 8 members (5 women, 3 men) meeting for more than 15 years
  • All members from same clinic; mostly academic; alternate facilitator for some of the observed sessions (n = 4)

Activities Set of related actions by participants (enablers for discussion)
  • Prereading material, discussing module cases; referring to module information points, tables, figures, case commentaries, and practice tools provided within appendices (e.g. algorithms, summary tables and patient handouts)
  • Accessing external information via phone/computer online guidelines, videos, implementation tools (e.g. algorithms)
  • Completing PRT and follow-up PRT
Objects Materials to support discussion
  • Evidence based educational material—module (cases, information points, tables, figures, references to literature, appendices with practice implementation tools)
  • Phone/laptop—access to apps for implementation tools (e.g. algorithms, guidelines and videos)
  • Practice Reflection Tool (PRT)28—completed during session, document most useful information; gaps in practice; will change (barriers); considering change (enablers); confirmed practice; not convinced there is a need to change
  • Follow-up Reflection Tool—documents made change (enablers); unable to make change (barrier); unanticipated change; plan further changes
Acts Single actions by participants (enablers for discussion)
  • Facilitator asks group member to read case/information points in module
  • Group member reads a section in the module
  • Facilitator asks stimulus questions to provoke discussion among members
Events Activities that participants carry out (enablers or distraction from discussions)
  • PBSGL housekeeping activities: organise meetings (summer break); discuss whether to take on new members; reminder for membership fees.
  • Related to discussion of module topic: Group 1 played video, tried physical exam manoeuvres.
  • Unrelated to module topic: bring forward info from attending a conference; discussion of election results.
  • Interruptions: ordering/delivery of food at restaurant (Group 2); group member leaving.
Time Sequencing of events
  • General thoughts about module; comments on current module; any gaps—plan for PBSGL session
  • Reading and discussing cases in sequence (n = 2–3 cases per module)
  • Referring to information points related to cases (more Group 1)
  • Test algorithms; physical exam manoeuvres; watch videos; review tables and appendices
  • Review case commentaries (more Group 2)
  • Completing PRT (Group 1–5 PRTs; Group 2–2 PRTs)
  • Review of previously discussed module (at beginning of session for Group 2 for 3/4 session (shorter discussions) and at the end of session for Group 1 for 5/5 sessions (longer discussions)
Goals Things that people are trying to accomplish—domain, practice—focus on improve practice

Goals within PBSGL sessions (usually lead by facilitator):

  • Go through module; play video; practice manoeuvre; ‘let's go through case 1; let's talk about …’; review approaches; identify other issues to explore; review next steps; ‘let's finish case’

Goals re Practice Changes (unrelated to PRT documentation):

  • Order tests; develop handout; look up information; talk to cardiologist; consult pharmacy; more patient education
  • Reference to QI projects (Group 1: 1/5 sessions; Group 2: 3/4 sessions)

Planned changes documented on PRT (lead by facilitator)

Feelings Emotions felt and expressed—mediators for decision process to make change
  • Still feel scattered; I have forgotten; huge gap for me; confused; worried about; bad experience; nervous to start; was surprised; would be comfortable treating; not feel much more confident; sceptical; frustrated
  • Abbreviation: QI, quality improvement.
TABLE 2. Summary of themes/subthemes of conversation pieces during observed learning session categorised into process small group session processes.
Small group processes Themes Subthemes Field notes
Start of small group session
Review of a previously discussed module topic (follow-up of planned practice changes) Review session for Group 2 Practice change(s) made; unable to make change(s); outstanding issues from previous module; quality improvement (QI)
  • ‘The group agreed there were no outstanding issues’ (Group 2).
  • ‘Facilitator then reminded them that …. the opportunities identified by [member] for measuring via QI what they were doing’ (Group 2).
Administrative tasks (intermittently)

Group membership (Group 2)

Outlining plan for session (Group 1)

Registration fees; potential for new group member(s); organising time to spend on learning activities
  • ‘Facilitator then raised a new issue about group size …, she wondered about inviting younger doctors who were interested in joining’ (Group 2).
  • Facilitator: ‘I have set up the video to play important stuff in the cases’ (Group 1).
Conversation starters (facilitator stimulates start of discussion) Thoughts regarding the module topic Interest level; not read module/not watched video; testing tools; surprises
  • M1.4: ‘thought it was going to be boring’.
  • M1.1: ‘tried the tools and they helped’.
  • M1.5: ‘reframed her approach to …’
  • All reported they had [read the module] but M2.4 felt the module was without substance (Group 2).
  • M2.1: ‘noted that in both cases they automatically jumped to CT rather than re-Xray’.
Gap in practice No approach/different approach; lack of knowledge; comfort level; not aware; confused
  • M1.1: ‘I still feel scattered and do not have an approach’.
  • M1.2: ‘It is a weak area for me’.
  • M1.1: ‘I've always been confused when …’
  • M2.4: ‘Felt like a klutz as the names were all French for the maneuvers discussed’.
  • M2.6: ‘Underestimate severity of side effect and can miss autonomic things. Do not feel much more confident’.
Middle of small group session
Case discussions (2–3 cases) Approach to module case History, physical exam, investigation, treatment, management, referral, decision support, clinical context, mediators
  • M1.2: ‘Stop nasal decongestants [All agree]’.
  • M1.5: ‘Basic blood work and try to convince them to stop smoking’.
  • M2.3: ‘Ask where it hurts before examining; compare sitting, standing, walking’.
Knowledge clarification History, physical exam, assessment, screening, investigation, treatment, resources, evidence
  • M1.5: ‘How long to get spirometry?’
  • Facilitator: ‘Let's read info point 33 [tapering of steroids]. It's similar to how you taper domperidone for milk supply. A slow taper’ (Group 1).
  • Facilitator: ‘Would you do spectrometry? What is the consensus? [All said no and reviewed symptoms that would make them reconsider.]’ (Group 1)
  • M2.7: ‘No dyskinesis, what do they mean by cerebellar testing?’
Knowledge sharing History, assessment, treatment, resources, mediators
  • M1.2: ‘American Academy of Ortho handouts are great. They have both exercises and rehabilitation’.
  • M1.4: ‘If not compounded properly (diclofenac) it won't work, ask patient to bring it in and see if it is gritty – if so won't work’.
  • M2.5: ‘Saw [physician] at ASA and talked about maneuvers that were effective and constellation of syndromes. Presume there is one available for each joint?’
Knowledge application Activity (exam manoeuvre, using algorithm), clinical relevance
  • 10–15 minutes of trying exam maneuvers on each other with [facilitator] supervising (Group 1).
  • The group talked through the case using the algorithm (Group 1).
  • All looking at X-ray and reading information which is different between modules (Group 2).
Knowledge consolidation Clinical context, heuristic, history, practice application
  • Facilitator: ‘Let's summarize our decision. Would we treat for GERD? All: Not right away’ (Group 1).
  • Facilitator: ‘Examining above and below, look for compartment syndrome – good reminder to be thoughtful and systematic’ (Group 2).
Reflection on practice Approach to practice; current skill level/knowledge gap; information needs; resource needs; decision for implementation
  • M1.4: ‘Previously would use Ventolin first but now might try Atrovent’
  • M1.4: ‘I learned a lot of new things; M1.1: I wasn't thinking there would be so much; M1.4: things change so fast’.
  • M2.1: ‘Need refresher on names of tests’
Sharing of experiences (tacit knowledge) Patient stories/clinical cases; resources; implementation strategies
  • M1.4: ‘Other case was a high-speed motorcycle accident and had been seen by ortho and with ongoing pain was refereed to foot and ankle clinic and is not having surgery’.
  • M2.2: ‘stated that she tells pregnant woman food sources for iron and provides a handout’.
  • M2.2: ‘stated that in her pregnant patients she recommended 2000-4000 IU calcium and reminded that Mg should be taken with calcium’.
Benchmarking Against module information; in relation to what other group members do
  • M1.4: ‘Has anyone seen this?’
  • M2.2 [reads part 2 of case]: ‘We forgot to examine the Achilles tendon’.
  • M2.3: ‘The case commentary didn't say what it was’.
Appraisal of evidence/recommendations Low evidence/evidence based on expert opinion
  • M1.1: ‘looks it up and states that evidence is low’.
  • M1.5: ‘mentions that the case presentation does not follow the flow sheet’.
  • M2.4: ‘Other thing that struck me was the evidence. This was not evidence based on expert opinion’.
Side discussions unrelated to case discussion/module topic Sharing experiences Diagnosis challenges; treatment recommendations; patient challenges
  • M1.5: ‘One lady was indeterminant but wanted genetics. Genetics wanted her to have amnio’.
  • M2.5: ‘Know someone who from the physician side has gone through the MAID process. It was way harder than they thought’.
Sharing knowledge from conference
  • M1.1: ‘discusses new information that she received from a conference unrelated to the current discussion’.
Recent news
  • ‘Discussion of election results’ (Group 1).
  • M2.8: ‘stated that he wanted to discuss something altogether different; This led to a 5–10 minute discussion about political issues with no relationship to the prior discussion; No attempt was made by the facilitator or other group members to curtail this side discussion’.
End of small group session
Facilitator documents discussion outcomes (documentation of planned practice changes) Practice Reflection Tool (PRT)—decision(s) for change Most useful information discussed; gap(s) in practice; commitment-to-change (will change), barriers anticipated; consider change, enablers to change; confirm change, support practice; not convinced to change
  • M1.2: ‘Add on treatment is the biggest change in practice. Definitely did not do before’.
  • M1.2: ‘Thinking about specificity of tests – still a gap’.
  • M1.1: ‘Better physical exam. Maybe more topicals’.
  • M1.1: ‘[Incorporate the] use of apps [All agree]’.
  • Facilitator: ‘Better research, more evidence and consultation. More knowledge on …’ (Group 1)
  • ‘Discussion among group about difficulties in distinguishing diagnosis’ (Group 1).
  • M2.1 + M2.2 were reminded of the anatomy’.
Review of a previously discussed module topic (follow-up of planned practice changes) Review session for Group 1 Practice change(s) made; unable to make change, barrier to change; unanticipated change(s); further planned change(s)
  • M1.1: ‘No new patient [seen with this clinical condition]’.
  • M1.2: ‘Used app and figured out differences in guidelines and discussion of risk versus benefit’.
  • ‘Changes – recommend home BP monitoring; BP 135/85 target for automated testing; decrease acetaminophen in OA; increase topical NSAIDS’ (Group 1).
Administrative task Plan for next small group session Date; location; module topic
  • ‘Discussion of next meeting and whether current dates work for remainder of group. Next topic and location were decided’ (Group 1).
  • ‘Facilitator will email to set up next meeting’ (Group 2).
  • Note: Themes of field notes reflect data compiled from all small group learning sessions on various clinical topics.
  • Abbreviation: MX.X, member ID.

2.4.2 Data analysis

Approaches taken to analyse the field notes were specific to ethnographic observational dimensions32, 35 and thematic content analysis.36

Observational dimensions

Two team members (SR, ML) categorised the field notes using the nine ethnographic observational dimensions32, 35 (Table 1). The categorisation was reviewed, and discrepancies were discussed and resolved at research team meetings. Differences between sessions observed within and between the two groups were noted.

Thematic content analysis

Thematic content analysis used both conventional and directed approaches.36 Using the conventional approach, coding categories were derived directly from statements documented in the field notes. Using the directed approach, conversation pieces from the field notes were coded to research questions arising from prior research related to the PBSGL programme.1, 14, 27, 28, 37, 38

Conventional coding

Thematic data analysis started with two team members (SR, ML) reading field notes repeatedly to achieve immersion and gain full understanding of flow and content of the discussion. Understanding of field notes was clarified at team meetings with the observer who created the field notes. After initial impressions, each team member created thoughts and concepts for two sets of field notes in which both observed groups discussed the same module topic. Thoughts and concepts were compared and discussed at team meetings. A coding book was created (Table 2), which two research team members (SR, ML) used to code all field notes. Coding was compared, and any discrepancies were discussed and resolved with the observer. The coding themes were clustered into categories that provided a foundation for a conceptual framework for ‘change talk’.

Directed coding

A structured approach to thematic analysis was also used to focus on the study's research questions. The same research team members (SR, ML) reread the field notes and this time grouped conversation pieces to predetermined coding themes of the research questions: interpreting information, assessing application to practice, integration of knowledge, contributors to practice change decisions and addressing implementation issues. Again, coding was compared, and any discrepancies discussed and resolved.

2.5 Supplemental data collection and analysis

2.5.1 Individual semi-structured telephone interviews

All participants were invited to follow-up semi-structured individual telephone interviews between July and August 2019. The interviews focused on participants' understanding of the role of small groups in the process of implementing new knowledge into practice (Data S3—Interview guide). Interviews were conducted, audiotaped and transcribed verbatim by one team member (ML). Interview transcripts were independently coded in NVivo™1233 using the conventional approach to content analysis36 by two team members (SR, ML) to create an initial list of themes. Themes were discussed at team meetings, organised into a coding scheme and applied to all interview transcripts. Coding was reviewed by the principal investigator (HA) to confirm analysis and identify additional elements for the conceptual framework for ‘change talk’.

2.5.2 Practice reflection documents

PRT statements were used to compare field notes taken at the end of the session when the facilitator summarised the outcome of the group discussion. This was done by transferring field notes and PRT statements to Microsoft Excel 201034 spreadsheet. Comments about PRT completion from the field notes were compared with statements documented on PRT by group facilitator at the end of learning session (Data S4—Sample of field notes and PRT comparison).

2.5.3 Data triangulation

Observational, interview and PRT data were triangulated to capture concepts that address the study's research questions and identify additional concepts for the framework not captured in the observational data. Data triangulation29 involved a multi-step analysis process. Initially, field notes were compared with the statements documented on the PRTs by the group facilitator to determine the extent of overlap of the two data sources and help confirm the accuracy of at least some of the handwritten field notes. Next, after completion of the thematic analysis of the field notes followed by the interview transcripts, comparisons were made between all data sources, that is, looking for both similarities and differences in themes/concepts (Table 3). Similarities enhanced credibility of data collected, and differences provided additional concepts for the ‘change talk’ framework.

TABLE 3. Summary of themes from field notes, interviews and practice reflections to answer research questions.
Field notes Interviews Practice reflections
1. How do physicians interpret new information?

Identify gaps in knowledge: ‘Wouldn't have thought that …’/’Never heard of …’/’I did not know this’.

Clarify understanding: ‘Way I understand is …’/‘Do you know why …’

Assessing evidence: ‘Isn't evidence better for …’/‘summary -not great evidence’.

Express needs for more information: Discussing gaps in module or what ‘should have been included’

Sharing of information: ‘I've been reading about [topic], and there is a potential for false negative biopsies so take multiple samples’.

Identifying new information

‘As you go through the info points in the modules, sometimes that will come up where you'll have a piece of information that is new’ M1.4.

Validating recommendations

‘The notion of validation is that there might be a recommendation that is printed in the module, but on the other hand … we are able to discuss on whether or not it is an appropriate recommendation and whether or not we need to evolve our practice’ M2.7.

PRT—most useful information: ‘How to order a NIPT’./‘Discussion on local resources and how to access them’./‘Criteria and scoring algorithm’.

PRT—gaps in practice: ‘Lack of knowledge around NIPT’./‘Index of suspicion not high enough’.

PRT—confirmed practice: ‘Arranging amnio as a diagnostic test for a positive FTS’.

2. How do physicians assess applicability of new information to their clinical practice?

Asking questions: ‘Would it overwhelm the patient?’/‘I wonder if___ can be used instead of _’.

Testing out algorithm/practice tools: ‘let's go to the flowchart and look at inclusion and exclusion criteria’

Comfort level: ‘I would feel [un]comfortable doing this’/‘I hesitate adding on medications …’

Sharing patient stories/practice implementation experiences: ‘My patient had a good response with steroids … my patient was the same’

Voting for module topic

‘1st step by selecting or voting for the topic - already made choice that it is relevant to us’ M2.7.

Discuss feasibility of implementation

‘Sometimes the evidence is very strong and is challenging the way we practice, then we discuss on whether or not this potential change can be rolled out in our environment and becomes more operational or more practical’ M2.7.

PRT—will change: ‘use labs to guide diagnosis’/‘use of patient handouts …’/‘find out which resources patients are using’

PRT—not convinced: ‘first-generation antihistamines for non-allergic UACs (concerns re: tolerance and safety)’

3. How do physicians discuss integration of new knowledge into their clinical practice?

Identify new knowledge/approach: ‘New approach – METH- new suggestion- movement, elevate, traction and heat instead of ICE’

Sharing information: ‘discusses new information that she received from a conference’

Confirm each other's thinking: ‘look at Table 1 which speaks to what you're saying’

Talking through an algorithm: ‘Lot of information but don't know if it changes what I do – might think about NIPT’.

Share experience: ‘Sometimes people have experience from their own practice that you haven't seen that is helpful in terms of change’ M1.4.

Share ideas how to make change: ‘… then just kind of bouncing some of the different ideas about how to actually make the change off my fellow group members …’ M1.2.

Discuss reasons for change: ‘the discussion is really around variability in practice, and who chose to do what and why’ M1.1.

PRT—gaps in practice: ‘good resources to use for patients’/‘when is CT better than a follow-up x-ray at 7–10 days post injury’

PRT—will change: ‘higher index of suspicion for Lisfranc injuries’/‘use thrombosis Canada website tools’

4. What contributes to a decision to change practice?

Group consensus: All agreed. ‘[member] you convinced me’./‘[facilitator] let's summarize decision. Would we treat for ___? All – not right away’.

Implementation tools: ‘[member] now have PSA handouts in patient education drawer’

Awareness: ‘[member] more aware of Lisfranc’

Availability: ‘look at appendix 1 … includes things we do not have available’

Level of difficulty: [observer notes] ‘discussion among group about difficulties in distinguishing diagnosis and the use of ___ for prolonged treatment’

Have control in making change: ‘if it's just something completely in my control then we … like trying a different therapy or being more evidence-based about what labs you order … or using certain tools, … so I can do those things from day to day’ M1.1.

Evidence/shared experience: ‘it's the discussion around evidence and experience that makes me decide’ M2.2.

Disagreement with recommendations:

‘It is more common that the group … disagrees with the recommendation in the materials. You know whether it's based on different evidence that we had or sometimes it is just the provincial limitations of practice’ M2.4.

PRT—most useful information: ‘evidence for HTN treatment options’/‘colleagues' personal experiences and practical tools- tips’/‘instructions of prescribing topical nifedipine’

PRT—gap in practice:

‘not a high enough index of suspicion’/‘not a detailed enough ankle exam, including proximal fibula’

PRT—support practice

‘experience’/‘expert consensus’/‘evidence’

5. How do physicians address implementation issues?

Looking for resources: ‘It would be nice if we have a set of resources and know where to find it’./‘more evidence to inform change. Couldn't find anything on google scholar’

Looking up information: ‘Will check blue cross for special authorization’; ‘SOGC website for patients and on there are routine tests. I show patients and several look it up’.

Help remember information: ‘need to use mnemonics like in med school’

Share information/experiences: ‘they courier tests on Thursday’/‘one lady was indeterminant but wanted genetics. Genetics wanted her to have amnio’

Doing follow-up review to not forget:

‘When we have our practice reflection six months later … there's some changes that I completely forgot about. I think that I reflect on them and kind of say, “okay well why didn't I do this?”’ M1.2.

Focus on doing something that can be done:

‘Move on. There's always something else to do … If there are no resources, then the solution sort of becomes out of our control. We don't dwell on the issue, and we'll do something else’ M1.1.

PRT—barrier to change: ‘concern re: overmedication’/‘our memory’

PRT—enabler to change: ‘increased access to methacholine challenge’/‘consultation with specialists’

PRT—not convinced: ‘use of integrated screen (not available in our region)’/‘clinic policies may be needed’

  • Abbreviations: MX.X, member ID; PRT, Practice Reflection Tool.


Two groups were recruited for this study, Group 1 (7 members) meeting for 1 year and Group 2 (8 members) meeting for more than 15 years (Table 1—actors). Nine small group learning sessions (Group 1: n = 5 sessions; Group 2: n = 4 sessions) were observed between November 2016 and October 2017. Each small group session was 1.5–2.0 hours long and occurred between 5 and 9 pm. Each session had 3–7 members in attendance, including the facilitator. Interviews were conducted May–August 2019 with 9 of the 15 participants. Groups submitted 7 PRTs and 5 follow-up PRTs for 9 observed sessions.

A narrative approach39 was used to synthesise the findings of this ethnography to help with the conceptual understanding of the ‘change talk’ within PBSGL sessions. The narrative of the findings was divided into the three stepwise analytical approaches (see methods above), with focus first on field notes followed by integration of findings from supplemental data: (1) General observations taken from field notes to provide an overview of the structure of the PBSGL sessions; (2) Themes of conversation pieces taken from field notes to identify elements for the ‘change talk’; and (3) Outcome of the predetermined coding of field notes and triangulation with interviews and PRT data to illuminate additional elements for ‘change talk’ and answers to the research questions. The narrative ends with a summary of the elements of the ‘change talk’ used to create the ‘change talk’ framework.

3.1 General observations providing an overview of the structure of observed PBSGL sessions

The findings for the ethnographic observational dimensions are summarised in Table 1 with some key findings highlighted below.

The structure and flow of the group discussion for the most part was similar across all observed sessions and followed the PBSGL process.1, 27 Facilitators played a significant role leading the discussion. Some differences between Group 1 and 2 sessions were noted. Group 1 facilitator ensured each member contributed to the discussion. Group 2 sessions were less structured with the facilitator allowing discussion to flow freely. Both groups spent most of the session discussing approaches to module cases. Group 1 regularly referred to module's information points and case commentaries (Table 1—objects; Data S1—Sample module) to inform their approaches. Group 2 rarely referenced information points but shared information from their own clinical practices and then reviewed case commentaries to benchmark their own approaches. Group members used laptops and phones to look up guidelines and investigate implementation tools (e.g. algorithms) while working through cases. Group 1, during one meeting in a member's home, practised ankle and foot exam manoeuvres, which was not feasible for Group 2 meeting at a restaurant. More references to quality improvement (QI) projects/activities were noted for Group 2 compared with Group 1 (Table 1—goals). All sessions ended with PRT completion; however, Group 2 facilitator only submitted 2 PRTs for the 4 observed sessions for professional credits. The facilitator, in discussion with members, documented discussion outcomes using CTCs. Most sessions included a revisit of previously planned practice change(s) and documentation of implementation outcome(s) on a follow-up PRT. Group 1 reviewed previously discussed modules at the end of the session, and Group 2 did this at the beginning of the session (Table 1—time).

3.2 Themes of conversation pieces to identify elements of ‘change talk’

Themes developed directly from field notes (conventional coding) are shown in Figure 1. No differences were noted between themes of conversation elements captured for Group 1 session versus those for Group 2 sessions; thus, Table 2 summarises themes/subthemes gleaned from conversation pieces captured in field notes for all sessions. Themes were organised into the following small group process categories: administrative tasks, conversation starters, case discussions, side discussions, documentation and follow-up of planned practice changes. Themes within these small group process categories depicting elements for ‘change talk’ are summarised below (themes are italicised).

Sessions usually began with the facilitator asking about ‘thoughts on the module’ (M1.1) or ‘anything surprising in the module’ (M2.2). Members addressed these components in addition to identifying perceived gaps in their own practice related to the module topic. After brief reflections on the module, one member read aloud one case (modules have 2–3 cases). Group members then shared their approaches to the case, including questions to clarify the patient's history and proposed aspects of physical examinations.

As discussion around the case continued, members interpreted information provided in the module or suggestions given by group members. Some members shared knowledge, whereas others asked for clarification. Thoughts on investigation, treatment or management were generally triggered by stimulus questions (within module or from facilitator). Questions were answered through further knowledge sharing, including references to module information, sharing members' own clinical experiences and sharing resources (literature, clinical guidelines, opinions from colleagues, consultants, referrals). Knowledge clarification involved questions around assessments, type of treatments and resources. Members assessed evidence and challenged recommendations provided in the information section, tables, figures and appendices of the modules. They voiced opinions and shared resources. As members considered the feasibility of using suggested approaches in the context of their own practice, they asked what others were doing (benchmarking), which lead to sharing of practice experiences of what worked and did not work.

During some sessions, members used implementation tools (e.g. algorithms or links to online resources) provided in the module to practise managing the case (knowledge application). As tools were used, more information was shared. This included sharing of experiences with the tool or other useful resources. As members reflected on their own practice and benchmarked their own clinical practice, decisions were made with respect to whether any practice change(s) were warranted. This discussion flow was repeated for additional cases.

During the time of PRT completion, at the end of each session, discussions around decisions for practice change entailed further knowledge sharing and consolidation, sharing of resources and a focus on mediators to practice implementation.

Members reported the success of implementing their previously planned changes during the follow-up review either at the beginning (Group 2) or the end (Group 1) of the sessions. Again, sharing of knowledge and practice experiences, as well as brief discussions around implementation issues, was noted.

3.3 Predetermined coding themes to answer the research questions

Table 3 summarises the themes from field notes, interviews transcripts and documented practice reflections. Interpretations of the triangulated data provided additional concepts to the ‘change talk’ framework and were utilised to answer each research question as described below.
  1. Interpreting information—Small group members spent substantial time interpreting information in the modules. As already noted above, the conversations start with members sharing their thoughts about helpful module information and own clinical practice gaps, which were later documented on the PRT. During the case discussions, members not only interpreted module information but also information shared by other members. Indications of new information were ‘Never heard of …’ and/or ‘Didn't know about …’ and knowledge clarification/consolidation was ‘the way I understand is …’. In some instances, group members felt there was a gap within the module and provided ‘what should have been included …’. Members also assessed evidence and expressed a need for more information if they required further clarity.

  2. Assessing applicability to practice—Interview data indicated that participants began considering knowledge application to practice when selecting the module topic and reading through the module in preparation for the discussion: ‘Voting for the topic in the group, we've already made a choice that it is relevant to us’ (M2.7) and ‘Reading through a module I will reflect on the gaps that I have in my practice’ (M1.1). Field notes indicated that members assessed applicability to their practice during case discussion by stating ‘I wonder if ___ can be used instead of ___’ or shared their comfort level with a new approach by saying ‘I would feel [un]comfortable doing___’. Members also asked questions, tested out algorithm/practice tools and shared patient stories around implementation. Information documented on the PRT indicated what will be changed in practice and/or information members were not convinced about.

  3. Integration of knowledge—Field notes captured that, during case discussions, each member presented their own approach to stimulus questions provided with each case or asked by the facilitator. New information or evidence was identified and shared among group members. Members explored whether a particular approach would work in their practice: ‘Lot of information but don't know if it changes what I do – might think about NIPT [non-invasive prenatal testing]’ (M1.1). Interview data explored reasons for change: ‘The discussion is really around variability in practice, and who chooses to do what and why’ (M2.1). Interview participants indicated that sharing experiences and ideas on how to make changes was helpful: ‘just kind of bouncing some of the different ideas about how to actually make the change off my fellow group members’ (M1.2). Knowledge integration was evident in PRT statements that documented identified practice gaps: ‘good resources to use for patients’ (Group 2) and commitments-to-change: ‘[will have a] higher index of suspicion for Lisfranc injuries’ (Group 1) or ‘[will] use thrombosis Canada website tools’ (Group 1).

  4. Contributors to practice change decisions—Contributors to practice change decisions were multi-factorial and included level of evidence in recommendations, context of clinical environment, implementation feasibility and resource availability. An interview participant stated that group discussion was a very powerful contributor in deciding to change practice:

    The discussion around why we're changing is as important as the recommendations in the materials. We frequently disagree with the recommendations, and it won't be on major stuff, it'll be minor implementation things. But the fact that you're discussing it and debating it is powerful (M2.4).

    This was exemplified in the field notes: ‘You convinced me’ (M1.1) and ‘[Facilitator] Would we treat for ___? [All members] Not right away’ (Group 1). PRT statements indicated that evidence, practice tips and experiences contributed to practice change.

  5. Addressing implementation issues—Observations during the session helped inform how physicians addressed implementation issues. Often, members looked up information and resources on laptop computers and/or phones. Suggestions were provided to help remember information: ‘Need to use mnemonics like in med school’ (M1.1). Interview participants stated it was helpful to have follow-up reflections on planned practice changes and figuring out ‘Why didn't I do this?’ (M1.2) and that there were many changes that could be made in practice, and if barriers to implementing a specific change were present, one could always ‘move on. There's always something else to do … [especially] if there are no resources then the solution sort of becomes out of our control’ (M2.7). Enablers to change, as documented on the PRTs, were often related to access and availability of treatments and/or changes to clinical policies.

3.4 Conceptual framework

With the use of an analytical lens of how physicians interpret information and assess its application to practice, we created a conceptual framework (Figure 2) to present the various elements contributing to ‘change talk’ in small group discussions.

Details are in the caption following the image
Conceptual framework for change talk. [Color figure can be viewed at wileyonlinelibrary.com]

During small group sessions, physicians talked through clinical cases provided within educational material for the learning session. They discussed feasible approaches to hypothetical cases considering evidence-based information provided in modules and their own clinical knowledge and practice experiences. They tried to make sense of new information by asking questions and sharing knowledge. They determined what information was useful and whether it applied to their clinical practice. They reflected on their own experiences and benchmarked themselves to best practices reported in the module and practice experiences shared by group members. They looked at evidence and consolidated knowledge before making decisions for practice change. The CTC strategy (completion of the PRT) helped explicate desired practice changes (making CTC statements) at the end of a learning session and served as a reminder to review the outcome of the planned practice changes at a later session. Even in the absence of CTC statements, physicians would reflect further on what they had learned during the small group session and/or seek out additional information.


This study explores empirical observations of how family physicians within a small group learning setting navigate through evidence-based educational material and make decisions for clinical practice. A conceptual framework for ‘change talk’ was created to illustrate the components used during small group discussions when interpreting and assessing new information and making clinical decisions to bridge gaps between current and best practice(s).

The ‘change talk’ framework was created within the cultural context of the PBSGL programme. Within this long-standing CPD programme, family physicians come together in small group sessions to review and discuss case-based, evidence-based modules with the guidance of a peer facilitator.1, 27 In this ethnography, each small group learning session was observed to follow the program's established PBSGL process1, 27 that pushes group participants from gap identification to sharing of practice experiences in light of case-based discussions and finishes by making decisions for practice. The PBSGL programme promotes social interaction that fosters ‘change talk’ and group decision making, giving this ethnography an ideal environmental constant to study elements of the ‘change talk’.

The set-up for the ‘change talk’ gleaned from field notes and interviews begins with the group discussion prior to the meeting as to which module topic they want to discuss. Group consensus for the topic to be discussed would often be based on a challenging case encountered in clinical practice within the topic area, a recognition of knowledge gaps or awareness of a new relevant guideline. While preparing for the session, members would pre-read the module information and reflect on their current knowledge and practice experiences related to the topic. The personal baseline knowledge and practice experience including gaps in practice would then be shared with the group throughout the PBSGL session. Practice reflections were followed by extensive discussions of module cases where members shared, clarified, applied and consolidated knowledge (‘change talk’). The outcome of the ‘change talk’ included CTC statements documented by the facilitator on the PRT and/or seeking additional information as documented in field notes. Follow-up at another PBSGL session would provide opportunities to share and discuss additional knowledge and practice experiences gleaned from implementing any of the previous planned practice changes.

The progression through the components that contribute to practice change are frequently depicted as a linear process.40-42 However, ‘change talk’ leading to decisions for practice change in this study showed a discussion flow that was iterative, depicted by the bidirectional arrows within the ‘change talk’ framework. This discussion flow facilitated the movement from information, data that are untethered, to knowledge that is linked to the specific practice environments and clinical experiences of the group members.43, 44 Part of that process includes the continuous back and forth between questioning and sharing of experiences and resources that facilitate the application to the clinical context. This iterative questioning, sharing of experience, knowledge and resources within the context of personal clinical practice promoted construction of alternative perspectives, which is important in making the decision to change practice.45, 46 Planning for change is also iterative, often requiring additional practice reflections and information seeking beyond the small group sessions. CTC statements captured on the PRT are revisited and maybe revised in future sessions when implementation successes and/or challenges are discussed. This iterative process throughout every component of the conceptual model reflects the thinking, revisiting and refining that needs to occur for practice implementation.

Armson and MacVicar1 previously identified that family physicians in this CPD programme demonstrate characteristics of a Community of Practice (CoP).47, 48 CoP is defined as ‘a group of people who share a concern, a set of problems or a passion about a topic, and who deepen their knowledge and expertise in this area by interacting on an ongoing basis’ (p. 4).48 Systematic reviews49, 50 indicate that there is increased use of CoPs to promote generating and sharing of knowledge with the goal to improve clinical practice. This ethnography supports the importance of the small group as ‘a vessel for conversations to take place’ (p. 50)51 built upon questions situated in practice. The development of a set of enduring relationships as part of the programme generates more opportunities for new ideas and innovations.52 In both groups, the interactions reflected a climate of trust, and conversations build around real practice issues, both crucial components of a successful CoP.48, 53 A final characteristic of CoP demonstrated by the groups was the accountability embedded in the PBSGL process. Individuals are encouraged to bring their practice problems and experiences to the small group sessions,51 and reflection on practice54, 55 is an important component of the small group interaction. Finally, there is an expectation that members are integrating and applying new information into their practices through CTC and consistent follow-up on implementation of planned changes.52 The structure and function of CoPs in the medical literature vary considerably, making it difficult to assess how CoPs impact physician behaviour changes in clinical practice.49, 50 This study reinforces the characteristics of CoP that contribute to practice change and the conversational components that support the functions of the CoP.

Although the small groups were using a similar PBSGL process and the same conversational elements depicted in the ‘change talk’ framework, there were differences in the way each group used the educational material and timing of the follow-up review, as might be expected in a unique CoP. The reasons for these differences are difficult to elucidate in this ethnography as only two groups were observed. The differences may have been solely due to the different composition of group members (see Table 1—actors, i.e. members coming from different clinics vs same clinic or non-academic physicians versus academic physicians), or potentially, this study may have uncovered certain features of a CoP that has been learning and practising together for many years.49, 50 In general, groups that have been together for many years are thought to have a greater collective knowledge compared with groups that have only been together for a short period of time.56-58 Thus, with more collective knowledge, there is less need to look for new information, more reliance on the lived experience of the members and a tendency to benchmark against other group members rather than against the new information contained in the module. Further, the more experienced group was less likely to accept and apply new information and more likely to create new knowledge through engaging in QI projects.

This study identified contributors for decisions to change practice that were similar to mediators reported for self-directed learning (SDL).59, 60 SDL ‘is defined by Royal College of Physicians and Surgeons as “activities planned to address specific needs, enhance awareness of new evidence potentially relevant to practice or enhance the quality of multiple systems”’ (p. 1245).59 The commitment to SDL is an important foundation to the work that can be done within the CoP. Greater awareness and understanding of evidence underlying recommendations, useful implementation tools, group consensus and having control in making a practice change all contribute to the individual decision to change practice. This study illuminates the importance of social interaction supporting the ‘change talk’ with colleagues. Within the complexity and uncertainties of the clinical environment, physicians studying together in small groups can exchange knowledge, reflect on practice experiences and assure themselves that there always is something that can be done better and that it made sense to focus on those practice changes that can be implemented and not dwell on practice changes where resources are not available.

4.1 Limitations

The findings of this study are limited by the observation of only two small groups of physicians belonging to one established CPD learning programme. Although variability was observed within and between small group sessions, observations of small group discussions within other educational settings and/or with other health professionals might have uncovered other elements of ‘change talk’ not accounted for here.

Much of the data is based on handwritten field notes of discussions rather than recordings transcribed verbatim. This could have resulted in misinterpretations. Some of the side discussions were only noted as such (i.e. no details of the conversation were documented), and additional elements of the ‘change talk’ may have been missed.

The conceptual framework for ‘change talk’ needs to be verified in other educational settings using small groups as a strategy for learning and practice implementation. The framework suggests that social interaction promotes learning through collaborative knowledge sharing, clarification, application and consolidation and facilitates decision making by reflecting on practice, looking at the level of evidence, benchmarking to best practices and sharing experiences. Future research should take this study a step further by identifying elements of discussions focused on the follow-up of planned practice changes, implementation issues and maintenance of implemented changes. In addition, a more in-depth understanding of the interactions between small group members would help elucidate the social processes that foster implementation of new knowledge into practice.

4.2 Conclusion

This study provides detailed insights into how physicians discuss clinical topics with the purpose of understanding and implementing new knowledge into practice. Small group discussions with colleagues can help overcome individual uncertainties in making appropriate decisions regarding diagnosis and treatment for optimal patient care. The conceptual framework for ‘change talk’ identifies conversational elements that occur during small group learning (knowledge sharing, clarification, application and consolidation) and leads to decisions for practice change. Medical educators need to be aware of these conversational elements to promote educational strategies using interactive approaches to learning that include understanding evidence for best practice, practice reflection, benchmarking and sharing of tacit knowledge to promote appropriate decisions to maintain practice competence.


HA as first author and principal investigator of this study made substantial contributions to the conception, design of the study, the acquisition (observation and field notes), analysis and interpretation of the data and development of the new ‘change talk’ framework and actively revising and editing the manuscript. KM as co-author of this study made significant contributions to the analysis and interpretation of the data and in the revisions of the ‘change talk’ framework and actively involved in editing the manuscript. ML as co-author of this study made significant contributions to the acquisition (interviews), analysis and interpretation of the data and was actively involved in creating the ‘change talk’ framework and actively involved in editing the manuscript. SR as co-author of this study made significant contributions to the design of the study, the acquisition (PRT and follow-up PRTs), analysis and interpretation of the data and was actively involved in creating the ‘change talk’ framework as well as writing the initial drafts of the manuscript and creating tables and figures for the manuscripts. All authors read and approved the final manuscript and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.


The authors thank the two PBSG groups and their members for participating in this study. They also give special thanks to Juli Finlay for her practical guidance in the development of this ethnographic study and Linda Mayhew for critical feedback on the manuscript. This work was presented in parts at the Society for Academic Continuing Medical Education (SACME) Virtual Meeting 26 February 2021; the Virtual Richard K. Reznick Wilson Centre Research Month, 20 October 2021, and the Association for Medical Education in Europe (AMEE) hybrid conference 27–31 August 2022.


    All authors are affiliated with the non-profit Foundation for Medical Practice Education (FMPE; www.fmpe.org) that developed and manages the PBSGL program. HA is FMPE executive director and research director, KM is the FMPE assistant research director, ML is research coordinator with module development and SR is the coordinator of the research programme at FMPE.


    This study was approved by the University of Calgary Conjoint Health Research Ethics Board. Ethics ID: REB16-0379.


    The data that support the findings of this study are available from the corresponding author upon reasonable request.