Volume 18, Issue 7 p. 603-609
ORIGINAL RESEARCH
Open Access

Domains of professional fulfillment for pediatric hospital medicine: A concept mapping study

Sarah Webber MD

Corresponding Author

Sarah Webber MD

Department of Pediatrics, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin, USA

Correspondence Sarah Webber, MD, Department of Pediatrics, University of Wisconsin School of Medicine and Public Health, H4/472, Mailcode 4108, 600 Highland Ave, Madison, WI 53792-4108, USA.

Email: [email protected]

Search for more papers by this author
Heidi Kloster MD

Heidi Kloster MD

Department of Pediatrics, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin, USA

Search for more papers by this author
Kristin A. Shadman MD

Kristin A. Shadman MD

Department of Pediatrics, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin, USA

Search for more papers by this author
Michelle Kelly MD, PhD

Michelle Kelly MD, PhD

Department of Pediatrics, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin, USA

Search for more papers by this author
Daniel Sklansky MD

Daniel Sklansky MD

Department of Pediatrics, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin, USA

Search for more papers by this author
Ryan J. Coller MD, MPH

Ryan J. Coller MD, MPH

Department of Pediatrics, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin, USA

Search for more papers by this author
First published: 09 May 2023

Graphical Abstract

Abstract

Background

We know little about how pediatric hospital medicine (PHM) physicians conceptualize their professional fulfillment (PF). The objective of this study was to determine how PHM physicians conceptualize PF.

Objective

The objective of this study was to determine how PHM physicians conceptualize PF.

Methods

We performed a single-site group concept mapping (GCM) study to create a stakeholder-informed model of PHM PF. We followed established GCM steps. For brainstorming, PHM physicians responded to a prompt to generate ideas describing the concept of PHM PF. Next, PHM physicians sorted the ideas based on conceptual relatedness and ranked them on importance. Responses were analyzed to create point cluster maps where each idea represented one point, and point proximity illustrated how often ideas were sorted together. Using an iterative and consensus-driven approach, we selected a cluster map best representing the ideas. Mean rating scores for all the items in each cluster were calculated.

Results

Sixteen PHM physicians identified 90 unique ideas related to PHM PF. The final cluster map described nine domains for PHM PF: (1) work personal-fit, (2) people-centered climate, (3) divisional cohesion and collaboration, (4) supportive and growth-oriented environment, (5) feeling valued and respected, (6) confidence, contribution, and credibility, (7) meaningful teaching and mentoring, (8) meaningful clinical work, and (9) structures to facilitate effective patient care. The domains with the highest and lowest importance ratings were divisional cohesion and collaboration and meaningful teaching and mentoring.

Conclusion

Domains of PF for PHM physicians extend beyond existing PF models, particularly the importance of teaching and mentoring.

INTRODUCTION

Physician well-being is a national priority and critical to a sustainable healthcare workforce.1, 2 An important driver of physician well-being is professional fulfillment (PF), defined as the summary of intrinsic and extrinsic rewards of medical practice.3, 4 PF is multidimensional and includes work-related happiness, worthwhileness, satisfaction, meaning, values-action alignment, and control in dealing with difficult problems at work.3

To promote PF, healthcare organizations must understand the determinants of PF for physicians. The commonly used Stanford Professional FulfillmentTM model focuses on three domains: a culture of wellness, practice efficiency, and personal resilience.5 Though important, focusing only on these three domains may neglect other important areas, particularly in academic medicine. For example, academic physicians also participate in education, mentorship, and research. These activities, clinical care, and their interactions are likely important components affecting PF.6-9 To our knowledge, there are no other PF models which incorporate elements of academic medicine.10

Moreover, strategies to address PF likely differ among physicians across specialities and work settings. For example, scribe programs are increasingly used to improve PF through enhancing practice efficiency in outpatient clinics.11, 12 However, the evidence and cost-effectiveness of such interventions may be difficult to translate to inpatient settings. For these reasons, understanding PF in different medical specialties and practice settings is important to generating effective solutions.

Pediatric hospital medicine (PHM) is the newest board-certified pediatric subspecialty and its practice scope and structure are distinct from outpatient practice models. For example, PHM physicians often have clinical service weeks alternating with academic and administrative time, are caring for patients and families during medical crises without prior established relationships, and often have several trainees at different levels on their team at one time. The state of PHM PF is not well understood, though limited studies have reported high job satisfaction overall.13 Both academic practice and good mentorship may correlate with increased PF among PHM physicians, while high weekend, overnight, and in-house clinical effort may threaten workforce sustainability.9, 13, 14 Interventions to promote PHM PF may be most effective when tailored to these concepts. The objective of this study was to describe how PHM physicians at our institution conceptualize PF.

METHODS

Setting and study design

This study was conducted with academic PHM faculty at a Midwest Quaternary Children's Hospital. We designed this project to better understand the concept of PF among our PHM faculty. Data were collected from July to December 2022. We used the group concept mapping (GCM) approach.15 GCM methodology was chosen because it is designed to enable stakeholders to generate an informative visual conceptualization of a complex and poorly understood topic. GCM has been used to create models for abstract topics within other healthcare research, for example creating a model describing quality long-term PICU care and articulating stakeholder perspectives of quality health outcomes for medically complex children.16, 17 Two of the authors are experienced with GCM and facilitating group processes (S. W., R. J. C.).17, 18 GCM uses four steps to leverage both qualitative and quantitative data from stakeholders to develop a conceptual framework for a complex topic. The four steps include (1) brainstorming, (2) sorting and rating, (3) representation, and (4) interpretation. These steps are outlined in Figure 1 and described in detail below.

Details are in the caption following the image
Group concept mapping study steps. PHM, pediatric hospital medicine.

This project was designated program improvement by the University of Madison-Wisconsin Institutional Review Board Program Evaluation Self-Certification Tool and did not require full IRB review. Concept mapping data collection and analysis were conducted using The Concept System® Groupwisdom™ software: Concept Systems, Inc. Copyright 2004–2022; all rights reserved.

Participants and recruitment

All full-time and part-time PHM faculty were invited to voluntarily participate in this multistep process during regularly scheduled division meetings. Meeting time is often used for collaboration on projects and contributing ideas to group development. Time to participate in GCM was provided during division meetings, however brainstorming, sorting, and rate were voluntary online activities without any ability for the division chief to know who participated. The only person with direct access to the data was S. W., the leader of the project and content expert in physician well-being.

At the time of this project, there were 16 PHM faculty; 11 attended the hospital wards only, 3 practiced complex care (inpatient and outpatient), and 2 practiced pediatric sedation (1 of whom also practiced inpatient pediatrics). All are academic faculty with responsibilities that include nonclinical scholarship and education of trainees. Each step of the process was led by S. W. who acted as a facilitator. This is consistent with GCM methodology; as the process can be guided by an internal member of the group.19

Step 1: Brainstorming

The goal of brainstorming was to create a stakeholder-generated list of ideas that serve as the foundation for understanding the construct of the study focus. Within Groupwisdom™, PHM faculty were instructed to provide as many possible answers in response to the prompt, As an academic pediatric hospitalist, happiness, meaningfulness, self-worth, self-efficacy, or satisfaction at work includes…. We deliberately included multiple concepts from the literature in the prompt to ensure that the resulting conceptualization of PF would be as comprehensive as possible. When adding their ideas to the list, participants could view the ideas submitted by others. After data collection was complete, the brainstormed list was reviewed by two PHM faculty (S. W. and R. C.) who removed duplicates and combined similar ideas into a consolidated list of unique ideas. This was performed first independently, then coming together for consensus.

Step 2: Sort and rate

The purpose of this step was to understand how the brainstormed ideas were related conceptually, and how important they were to PHM faculty.15 The consolidated list from Brainstorming was uploaded to Groupwisdom™ where participants are able to virtually and independently sort and rank each idea. During sorting, PHM faculty were instructed to sort each idea from the list into groups based on conceptual relatedness (not order of importance). They were instructed to create at least two groups of ideas and to provide a label for each group that summarized the ideas within that group. During rating, PHM faculty were then instructed to rate each idea from the consolidated list on a 4-point scale (Drawing from your own experience and point of view, please rate each statement on the degree of importance from not important to very important). S. W. reviewed all responses to ensure completeness (at least 75% of ideas sorted and rated) before inclusion in the analysis.

Step 3: Representation

The goal of this step was to analyze sorted data into a visual point map using multidimensional scaling (MDS) and hierarchical clustering.15 Groupwisdom™ applied MDS to sorted data to generate a single point map, where each point represented one idea from the brainstorming list. On this map, the distance between points reflects how frequently statements were grouped together, with closer points reflecting ideas that were sorted together frequently. Next, hierarchical clustering created cluster maps, where boundaries were created around groups of points that were conceptually related. Multiple maps were created within Groupwisdom™, including maps with fewer clusters (more points per cluster) and more clusters (fewer points per cluster). Next, a core set of investigators reviewed the cluster map options and identified the map that best represented the ideas. In this case, we invited all PHM faculty to review these maps, starting with the seven-cluster solution and sequentially increasing clusters to identify the solution that best represented the data. A smaller group (S. W., R. J. C., H. K., K. A. S.) reviewed the final cluster map and created labels for each cluster that represented the main ideas. Finally, we calculated the mean importance rating of items within each cluster.

Step 4: Interpretation

During interpretation, the concept map, labels, and rating data were brought back to all PHM faculty for group review and revision. During a PHM faculty meeting, the cluster labels were presented one by one. Labels were revised during the meeting after which we sought additional anonymous, independent feedback from faculty using an open-ended survey before finalizing the cluster labels.

RESULTS

All 16 PHM faculty (100%) participated in brainstorming, 14 in sort and rank (88%), and 14 in interpretation (88%). Most faculty (75%) were between 36 and 45 years of age. Our division includes 12 women and 4 men. One physician identified as Asian while the remainder identified as White. In total, 101 ideas were brainstormed, and 90 unique ideas remained after consolidation. We did not measure the demographics of participants within each step.

Domains of PHM PF

The final cluster map identified nine domains for PHM PF (Figure 2). Domain descriptions and example ideas are listed in Table 1. Two clusters (Meaningful clinical work and Meaningful teaching and mentoring) describe how meaning is derived from clinical and educational work. Several clusters illustrate how the healthcare system influences PF, including operational and interpersonal aspects: structures to facilitate effective patient care, a supportive and growth-oriented environment, and person-centered climate. Internal representations of PF were represented by two clusters, confidence, contribution, and credibility and a second cluster, feeling valued and respected. One cluster illustrated team-based interpersonal aspects of work (divisional cohesion and collaboration) and one described the intersection of work and personal life (work-life fit).

Details are in the caption following the image
Concept map of pediatric hospital medicine (PHM) faculty professional fulfillment (PF). Each point represents one idea generated during brainstorming. Point proximity represents the frequency that ideas were sorted together. Borders around points represent ideas that likely conceptually relate, forming cluster domains. We identified eight domains conceptualizing PHM PF.
Table 1. Representative ideas from each domain of professional fulfillment identified by PHM faculty.
Domain name Description of domain Example idea
Divisional cohesion and collaboration Belonging to a division where cohesion and collaboration promote shared accomplishments

“Time to collaborate with division members, resulting in synergy of efforts and ideas”

“Working with colleagues who I admire and respect deeply”

“Shared responsibility within our group”

Supportive and growth-oriented environment A work environment that is inclusive in valuing contributions, respecting people, and allowing for mistakes

Recognition for work that may not result in publications, regional reputation, or teaching accolades”

“An environment where it is ok to make mistakes”

“Being able to work through interpersonal problems/conflicts”

Confidence, contribution, and credibility Desire to be recognized and respected as an expert in ones' field and able to contribute that expertise

“Feeling I add value”

“Fixing things and solving problems”

“Feeling respected by consultants”

Meaningful teaching and mentoring Having mindset, time, and skills to help others grow

“Watching learners flourish and grow”

“Having enough confidence to mentor others”

Structures to facilitate effective patient care Feeling that time, tools, workload, and compensation are just and promote work fulfillment

“Consolidated times to focus on clinical service”

“Having the tools I need to help patients”

“A patient load that allows time to think about patients”

Meaningful clinical work Clinical work that facilitates meaningful connection with patients and families, as well as personal growth and learning

“Easing the experience of children and families who are suffering”

“Learning from patients and clinical situations”

“Variety in clinical work”

People-centered climate Climate that supports, resources and centers relationships and engagement

“Reduction of non-meaningful tasks, where possible”

“Time for fun at work”

“Being a part of a team”

Work-personal fit Work culture and boundaries that facilitate compatibility between work and personal life

“Ability to set boundaries”

“Time for mental rest”

“Adequate exercise”

“Adequate sleep”

Feeling valued and respected Feeling work and contributions are valued and recognized by colleagues and institution

“Institutional practices that demonstrate trust in me”

“Feeling valued by and connected to the inpatient nurses”

“Being respected within the department, [children's hospital], and larger [hospital system].”

  • Abbreviation: PHM, pediatric hospital medicine.

Cluster ranking

Cluster ratings are depicted in Table 2. The cluster rated most important was Divisional cohesion and collaboration (mean 3.44, SD 0.31), while the cluster with the lowest importance rating was Meaningful teaching and mentoring (mean 2.95, SD 0.16).

Table 2. Domains of professional fulfillment ranked by mean importance.
Domain name Mean importance rating, range 0–4 (SD)
Divisional cohesion and collaboration 3.44 (0.31)
Supportive and growth-oriented environment 3.35 (0.22)
Meaningful clinical work 3.27 (0.36)
People-centered climate 3.26 (0.38)
Confidence, contribution, and credibility 3.21 (0.28)
Work-personal fit 3.20 (0.41)
Feeling valued and respected 3.18 (0.23)
Structures to facilitate efficient patient care 3.10 (0.41)
Meaningful teaching and mentoring 3.00 (0.16)

DISCUSSION

This study identified how academic PHM faculty conceptualize PF within nine domains. Although some elements of this concept map reflect previously described models of physician PF,3, 5 several differences emerged: the addition of meaningful teaching and mentoring, a focus on person-centered and growth-orientated environments, and a reframing of resilience concepts. The domains in this model provide deeper granularity about the nature of PHM work that physicians value, and the contributing circumstances.

Importantly, our findings advance existing conceptualizations of physician PF5, 20, 21 by adding, affirming, and reframing concepts. To our knowledge, none of the established physician PF models include meaningful mentoring and teaching as a domain contributing to PF. Mentoring and teaching are core elements of academic medicine and PHM. Other studies have described their importance; academic surgeons reported teaching as the best aspect of their job and academic pediatricians described the tension between academic responsibilities like teaching and mentoring and clinical roles as barriers to well-being.3, 7 Adding nonclinical meaningful work like teaching and mentoring to models of academic physician PF may be needed.

The importance of social aspects of work like clinical teamwork, communication, leadership support, fairness, and inclusiveness was affirmed in our model.5, 10, 21 This is not surprising; social connection and belonging are core human motives.22 Other studies have identified elements of social support important to physicians: trust between colleagues, friendships with staff, and time for community building.21, 23 The need to belong occurs across the spectrum of interpersonal work-life including between individual people, within groups and teams, and among people and their organizations.22 This is illustrated in our domains, which describe the need for belonging and social support within each level of the work environment: divisional, cross-specialty, and larger organizational climate and structures.24, 25 Interventions aimed at social connection and belonging can occur at each of these levels.

Finally, our model reframed the concept of resilience within the PF construct. Resilience is one of the three elements of PF in the Stanford model but was not a domain in our model.5 One explanation for this difference may be that PHM physicians in this study did not view personal resilience as a major factor in PF. There has been a growing emphasis on systems over individual physician resilience-building approaches, and this may correlate to physician perspectives about the role of resilience in PF.26 PF experts and PHM physicians may also perceive resilience differently. For instance, sleep, exercise, time for mental rest, having a supportive family, and the ability to take off work when sick were ideas within our study. These ideas clustered within the work and personal fit domain. This might indicate that PHM physicians in this study were attuned to the factors that accommodate or interfere with resilience within work and personal life, rather than resilience as a state or intrinsic quality. The idea that social support, reflection, and physical health contribute to resilience is supported in other fields.27, 28 Future work should further explore physician perceptions of resilience and its' relationship to PF and consider how personal resilience is represented within existing PF models.

In addition to conceptual advancements, these domains offer a starting point for implementing interventions within PHM to promote PF. For example, division leaders can create systems for recognizing accomplishments or contributions that don't result in publications or traditional scholarly work, and individuals can establish boundaries that prevent nonclinical work interference with patient care and teaching. Coaching, which has been a method of improving imposter symptoms, may be useful in promoting confidence and self-compassion.29 Divisions could also use these domains to identify areas for specific improvement based on local feasibility and predicted degree of impact.

Finally, it is also important to consider how our findings may relate to current PF measures. Our domains suggest that current measures may need to expand and evolve to better represent the variation in practice and needs across physician specialties and settings. The validated Professional Fulfillment Index is a feasible measure that examines feelings of happiness, worthwhileness, satisfaction, feeling in control, meaning, and sense of contribution.3 Our model suggests this might be more inclusive of the physician PF experience if expanded beyond personal contributions and feelings of success and include experiences of collaboration, sense of belonging, and other established culture and climate measures.15, 30 The experiences of teaching, mentoring, and research may also be added to measures for academic physicians.

There are several limitations that should be considered. First, data were collected from PHM physicians at a single site, which limits generalizability. However, concept mapping intends to generate a “broad sampling of ideas rather than a representative sampling of persons,”31 so while the results may not be specifically generalizable, they offer a visual map to understand stakeholder conceptualization of a complex topic. GCM results are dependent on the focus prompt and other questions may have brought different issues to light. This process was led by physicians within the group, though the GCM process can be guided by an internal member of the group. Our division includes physicians who are majority White and women, so we are likely missing ideas from identities not represented in our division. Different investigators or stakeholder groups may have produced different cluster labels. Future studies should continue investigating specialty-specific conceptualizations of PF, and drivers that are amenable to intervention. Investigations should include a diverse spectrum of PHM physicians, including race, ethnicity, age, years of practice, and scope of practice. Perspectives of nonphysician PHM providers, including nurse practitioners and physician assistants, should also be explored.

CONCLUSION

Stakeholder-driven investigations of PF may move us further toward a meaningful understanding of both drivers and outcomes. The nine domains identified here reinforce the known importance of the system and environments on physician PF and provide more detail about unique components of social support like meaningful collaboration. Models of PF for academic physicians should consider mentoring and teaching to better encompass the important features of PF in these settings.

ACKNOWLEDGMENTS

The authors would like to thank the Pediatric Hospital Medicine division for actively participating in this project.

    CONFLICT OF INTEREST STATEMENT

    The authors declare no conflict of interest.