Is postgraduate leadership education a match for the wicked problems of health systems leadership? A critical systematic review

The drive and mandate for leadership education for physicians continues to grow and there are calls for systematic and critical reviews of medical leadership education. This work (1) provides systematic synthesis of the educational approaches and outcomes in current postgraduate leadership education and (2) offers analytical perspectives on some of the discordance between medical leadership education and the practice demands of health systems leadership. Ultimately, it draws on contemporary theory and evidence on leadership development to present implications for practice in medical leadership education.

Concerns about effective leadership are gaining prominence in discourses of health system operations and transformation. And, although definitional consensus is elusive [1], it is broadly acknowledged that leadership of healthcare systems extends beyond the clinical leadership [2] activities of directing and collaborating on the treatment of patients at the point of care. Rather, health systems leadership also entails active participation in the stewardship and transformation of healthcare service delivery, resources, and policy in the public interest [1, 3, 4]. The growing emphasis on the importance of health systems leadership coincides with the extraordinary challenges facing global health systems that must contain costs while meeting public expectations for high quality care at a time of rapid technological advances, demographic shifts, and rising income inequality [1, 5]. It is in this environment that there is growing contention for physician engagement in health systems leadership given their influence over healthcare quality and resource utilization [6‐9]. Indeed, there is modest, albeit mixed, empirical evidence to support a positive link between having physicians leaders in senior, structural roles, and indicators of health system effectiveness such as quality of care [10]. At the same time, evidence on physician leadership in practice suggests that health systems leadership work is fraught with challenges and paradoxical tensions. There are persistent concerns with general reluctance among physicians to engage in health systems leadership activities [11‐13]. And, despite being motivated to engage in leadership in order to contribute to health system improvement, physicians leaders report feeling unable to effect system change [14, 15] and perceive health systems leadership work as costly (personally, socially, and materially) [16‐18]. Tensions between physicians’ professional group (clinical) identity and leadership work demands [13, 19] are commonly reported as are relational tensions across professional and hierarchical boundaries [20, 21]. Critically, physician leaders often report feeling unprepared for their leadership roles [19, 22], and scholars have questioned whether traditional postgraduate training provides physicians with adequate education for leadership roles [7, 23].

The desire for greater physician participation in health systems leadership—and contention that better leadership education is needed—has put pressure on educators to develop leadership education programs across the continuum of medical training and practice. However, questions remain about whether leadership programs are adequately preparing upcoming physicians for the challenges they will encounter in practice. Notably, there have been critiques about investments in leadership training in healthcare, and questions about the efficacy of leadership education. These include questions as to whether leadership education in medicine can develop collective agency for reform as opposed to maintaining the status quo or entrenching medical power [24]. It also includes questions about the potentially heavily circumscribed impact of leadership development given the limited discretionary authority that leaders can have within their institutions’ policy, regulatory, and cultural constraints [25]. Furthermore, a review [26] of physician leadership development programs raised questions about the educational strategies that were utilized by extant programs. Across programs, the authors found evidence of a particularly narrow focus on individual-level (vs system) outcomes, as well as limited use of advanced leadership education tools such as interactive learning and feedback [26]. The aforementioned concerns and review findings point to a potential preparation-practice gap, whereby leadership curricula in medicine may not be designed to meet the demands of health system workplaces. Thus, it is imperative that scholars and educators update our understanding of how contemporary leadership education in medicine is functioning and whether it is evolving to better address the demands of physicians as health system leaders.

In this work, we systematically and critically examine the nature and outcomes of leadership education in medicine in light of emerging trends and evidence in education and leadership practice. Over the past decade, discussions on leadership in healthcare have started to embrace contemporary conceptualizations of leadership that define leadership as mutually influencing, power-sharing, and collective [27, 28]. Theoretical conceptualizations in this area include terms such as ‘shared,’ ‘collective,’ ‘complex adaptive’ and ‘relational’ leadership among others [28]. Notwithstanding conceptual distinctions among these approaches, they all have at their core a reframing of leadership as exercised within groups along relational lines that are not necessarily defined by formal roles or traditional hierarchies [28]. Further, they attend to the socio-political and material complexity that has led many health system challenges to be characterized as ‘wicked problems’—problems that are dynamic, and multi-faceted, with conflicting demands that are recalcitrant to linear resolution [29]. These contemporary conceptualizations of leadership may necessitate different pedagogical approaches to leadership development. Thus, it is important to examine whether these newer leadership discourses are influencing the landscape of leadership education in medicine, as well as study how they may be influencing program outcomes.

A critical systematic review, such as this one, provides the opportunity to discuss patterns, trends, and outcomes of leadership education in relation to a concrete synthesis of evidence about extant programs. Here, we focus on synthesizing evidence on the characteristics and efficacy of postgraduate leadership development programs. Ultimately, we aim to inform critical discussion on the current landscape of leadership education in medicine in relation to contemporary evidence on leadership development in addition to exploring potential preparation-practice gaps for physician leaders.

This study was part of a larger program of inquiry that included a state-of-the-art review on research about physician health systems leadership and a formative evaluation study on the development of new leadership development programs.

Review findings are presented in tabular and narrative form. The 31 identified studies represented five countries across three continents. Overviews of program features and content are presented in Tab. 2 and Tab. 3 (in the online Supplementary Material); comprehensive details about individual programs are presented in Tab. 4, which can be found in the online Supplementary Material. The average MERSQI score across two raters for the 18 quantitative studies/study components was 8.4 (range 6 to 12.5). Kappa co-efficient for inter-rater reliability was 0.94 (almost perfect agreement between raters).

Our review found modest evidence that extant postgraduate leadership education can generate positive, individual-level outcomes (e.g., skill gains) and that trained residents can contribute effectively to health service improvements. However, like earlier work reviewing physician leadership development more broadly, [26] we found that the scope of leadership education programs in postgraduate medicine remains relatively narrow, and largely focused on individual-level development. Similarly, our findings also highlight the limited range in leadership development strategies that are employed in postgraduate programs. Whereas our findings raise some concerns that the academic preparation-practice gaps alluded to by Frich et al. [26] may be persistent, our review identifies new, modest trends toward the inclusion of broader perspectives on leadership. Ergo, in the ensuing discussion we draw attention to key observations and discuss the relation of preparation-practice gaps to existing evidence about leadership education and health systems leadership.

First, there remains a prevailing programmatic emphasis on individual leader competence versus leadership development¹ that is focused on capacity building for a collective (e.g., teams, institution)—with attention to collective outcomes [57]. Whereas individual leader competence is undoubtedly necessary, leadership scholars argue that individual-level competence is insufficient for effective health systems leadership where the challenges faced are often too intractable for traditional individualist/heroic approaches to leadership [29, 58]. As noted earlier, Rittel and Webber’s [59] typology of critical, tame, and wicked problems is a useful one in thinking about leadership [60]. Critical problems in medicine (e.g., an acute heart attack) demand swift action leaving little time for uncertainty or procedure; tame problems (e.g., elective heart surgery) may be complicated puzzles but they are manageable with relatively linear tools and are likely to have been resolved before [60]. Wicked problems, however, are intertwined with deeply complex, social and cultural issues; they are situated across institutions and may have no clear ‘stopping points’ at which the problem could be said to be solved (e.g., developing equitable arrangements for providing government funded services to ageing populations at a time of increasing medical ability to maintain life) [29, 60]. Individual leader development may suffice to prepare future leaders to navigate critical and tame problems that manifest in acute clinical microsystems but be inadequate in preparing physicians—individually and as a collective—to engage in the leadership activity required to address the wicked problems of contemporary health systems. Addressing wicked problems, Grint argues, requires a reframing of leadership as ‘influencing a collective to take joint responsibility for collective problems’ [60]. Contemporary paradigms of leadership that espouse shared, process-oriented, mutually influential views of leadership places more value on developing collective capacity for leadership than traditional models that focus on skill enhancement [61]. However, only a few programs reported being explicitly informed by these contemporary views of leadership.

We must note here that physician participation in both traditional, individualist/heroic leadership practices (e.g., transformational/transactional leadership) [62] and contemporary boundary-spanning, collaborative leadership practices has been linked to the successful stewarding of health system reforms [14, 63, 64]. Thus, the argument for integrating more contemporary, collectivist views of leadership in leadership education is not wedded to a binary perspective of heroic vs post-heroic leadership [65]. Rather it is an argument that educating residents to understand and enact these types of approaches better prepares them for the demands of leadership. In practice, managing health system problems (such as implementing reform) demands hybridized or interwoven approaches where heroism and collectivism can coexist as demanded by shifting contexts and circumstance [65, 66]. That notwithstanding, leadership development in medicine is heavily weighted in the individualist realm. Thus, the inculcation of newer, collective conceptions of leadership would require leadership development approaches that work to shift mindsets about the primacy of individual leaders in addressing health system problems [67]. Tackling wicked problems demands forms of influence that cannot be conferred simply by occupying senior roles. Others cannot be forced to follow you, they must have the volition to help [60] and knowledge workers in modern health systems are disinclined to enact top-down visions [27]. Leading in such environments involves facilitating contexts in which others are also willing and able to lead to address shared problems [58] within the scope of their expertise. It thus behoves medical educators to be clear and balanced about the models of leadership they are promoting and ensure that curricula are strategically aligned with the underlying program aims.

Second, our findings show that although almost half of all residency leadership programs covered quality improvement and patient safety, fewer than 15% covered content domains explicitly related to system-level leadership (e.g., policy, equity, systems thinking). Given that most physicians operate within clinical microsystems—‘small, functional units at the front-lines of delivering care to patients’ [68]—it is important that emerging leaders are equipped to manage clinical microsystems. However, as patients move across multiple clinical microsystems in their life journeys it is important that those in leadership roles understand the macro-level issues that influence the success of inter-dependent clinical microsystems. Extant research tells us that leaders in general operate across vast knowledge and practice domains—often requiring a breadth of both professional-industry specific and general-administrative knowledge and skill [69]. It is argued that true mastery in leadership means the ability to cross and bridge differing domains—becoming an expert in a mega-domain with many subsets [69]. With this view, health systems leadership is a significantly broader domain of practice than clinical leadership and the skills acquired during clinical leadership training may not be transferrable to health systems leadership environments without concerted effort to develop those skills for that context. Evidence on developing adaptive expertise in clinical reasoning shows that contextual variation (seeing concepts in multiple contexts) is critical to building the cognitive store of exemplars necessary for expertise, as well as for preparing students for the transfer of learning to practice [70]. As our understanding of expertise in other domains is also relevant to leadership education [69], leadership programs should strive to provide learners with opportunities to learn across multiple contexts.

Third, like Frich et al. [26] our findings suggest that key leadership development tools such as experiential learning, mentoring, coaching, and feedback remain underutilized in postgraduate medical education. This runs counter to evidence on leadership development which demonstrates that challenging, practical experiences are powerful learning tools that function optimally when complemented by feedback on learning progress [71]. Although learning from experience is not guaranteed, residents within leadership programs may be particularly suited to benefit from such placements because they are likely to be in learning mode [72] during such placements (i.e., they are positioned to frame and pursue their placement experience with an orientation toward growth and learning) [72]. Promisingly, approximately 60% of longitudinal programs did report using experiential placements. Shorter-term programs should strive to develop alternative strategies that provide workplace-based exposure to leadership in practice (e.g., shadowing, rounds).

Further, leadership education remains exclusively uni-professional, and generally targeted toward single specialities. This persistent uni-professionalism raises concerns about whether budding physician leaders are prepared enough for the multi-professional and inter-sectoral relationship building that will be demanded of them in practice. The expansion of boundary-crossing social networks and social capital has been identified as an important feature of successful leadership development [57, 73]. Successful leadership has been linked to the quality of the social network in which a leader is embedded [73]. It may be particularly important for enhancing collective leadership capacity given the distributed nature of health systems whereby leadership activity must be spread across organizations/units in order to adequately respond to challenges that appear at different units of governance [74]. Indeed, research on physicians leadership in practice suggests that the absence of relationships among leaders across specialities and across clinical/non-clinical boundaries [14] impedes leaders’ efforts to enact health service improvements. On the other hand, a recent critical examination of inter-professional education research and practice [75] suggests that uni-professional educational environments may be more efficacious for teaching collaborative skills than inter-professional educational settings—if coupled with workplace-based interventions. However, gains in collaborative skills may not translate into social capital across a given system without dedicated opportunities for relationship building. It is important for educators to design curricula strategies that support the development of social capital across professional boundaries.

Promisingly, our findings present modest evidence of an emergent new wave of residency leadership interventions where leadership education is formally integrated into a broader residency curriculum, giving learners ongoing exposure to both clinical skills training and leadership development experiences. Integration of leadership education into clinical practice training may also encourage physicians to view leadership work as part of their regular professional practice, thus minimizing perceived commonly experienced tensions between clinical and leadership identities [13]. Many of these ‘new wave’ programs [32, 33, 40, 50, 56, 76] were also more likely to have curricula content on macro level health systems issues, and explicitly strive to enhance physician engagement in health systems leadership. Higher quality evaluative studies are needed to ascertain the impact of these types of programs on both individual-level and collective outcomes.

In summary, our review highlights preparation-practice gaps in leadership development and integrates knowledge on evidence-informed educational strategies that may be useful in addressing these gaps. Furthermore, we contend that alternative pedagogies may be needed to instil the value of contemporary, collectivist approaches to leadership. Specifically, it may be useful to incorporate approaches that go beyond predominant paradigms of education [77] (that are generally focused on skill acquisition and assimilation). For example, in leadership programs where the goal is to develop leaders’ change agency for system reform, approaches such as transformative learning [78, 79] would be appropriate as they encourage learners to critically reflect on and address factors that contribute to the maintenance of the status quo [80].

As noted earlier, previous scholarship has identified similar concerns about potential preparation-practice gaps in extant leadership education. The persistence of these gaps raises questions about the difficulties inherent in transforming established educational cultures. These challenges may have to do with resource constraints as well as inadequate knowledge mobilization and faculty development. Deeply rooted cultural change resistance may also be at play; senior educators/leaders may inadvertently reproduce antiquated frames of reference about leadership that may no longer be appropriate for today’s health system challenges [13]. Furthermore, contemporary or evidence-informed approaches to leadership education and practice may be at odds with existing ways of organizing across health systems that are still largely structured hierarchically. Deeper structural and policy changes may be required to better address health systems wicked problems, and to promote needed cultural and evidence-informed shifts in leadership development in medicine. Encouragingly, our findings on newer programs show subtle but promising movements in the structure and content of leadership programming for residents.