Systems and Complexity Thinking in the General Practice Literature: An Integrative, Historical Narrative Review

Before 1995

In the years before 1995, phenomenologic experiences opened links to systems and complexity sciences. In the 1970s, general/family practice engaged in systems and complexity conversations, which led to a reexamination of clinical observations and experiences (refer to Table 2S at dx.doi.org/10.5061/dryad.kf251) and a recognition that systems and complexity theory could offer a more effective, context-based understanding of illness in individuals, families, and the community. During this period, the notion emerged that patients should be understood as belonging to interconnected, overlapping systems (open to environment) that interact and influence one another in nonlinear ways (adaptation, evolution, coevolution), with agents interacting through multicausal feedback loops (pattern of interaction, emergence, nonlinearity).

Family therapists were the first to realize that family as well as therapeutic relationships occur in a systems environment⁴ and could best be described through a cybernetic model.⁹ They noticed that interactions between variables hold the key to understanding observed patterns of health and illness.^6,9 To adequately understand a patient’s illness, practitioners must account for the effect of variables from overlapping systems such as family, suburb, or work.^6,9 Hence, clinical reasoning must follow nonlinear, feedback-driven, dynamics—reflected as the well-known circular process of “diagnosis as treatment” and “treatment as diagnosis.”⁹

Using this new approach, systems and complexity thinking challenged the prevailing scientific paradigm of disease causation, diagnostic reasoning, and therapeutic intervention for the first time. McWhinney^7,10 in particular recognized paradigmatic inconsistencies in the biomedical model and emphasized the need to move beyond reductionist scientific generalities of diseases to holistic understandings of particular individuals in their biosocial contexts. Recognizing that each patient’s presentation in the consultation is one particular instance of their health journey acknowledges that both illness and healing are dynamic adaptive states. McWhinney’s insights led to his proposal for shifting medical thinking towards a Geisteswissenschaft (humanities) of medicine that links the social with the natural sciences.

Furthermore, culture provides the framework for creating meaning out of the experience of health, and of feeling and being ill.¹¹ Illness and healing reside in a person’s complex interacting set of values, expectations, and images of oneself and others, and healing becomes a meaning/sense-making process in an individual context. Thus, context, long seen as the enemy of “natural” scientific research, became a rich field of study and knowledge generation.¹¹

These early investigations followed in Husserl’s phenomenologic tradition⁵⁹ and focused on both the philosophy and empiric science of health and disease, and on medical interventions and the role of the physician in health care and society.

1995–2000

The years 1995–2000 entailed a focus on the complex adaptive nature of general/family practice. Academic activity in general/family practice during this period exhibited a high degree of confidence in the capacity of systems and complexity science to advance our understanding of the nature of medical care (Table 3S at dx.doi.org/10.5061/dryad.kf251), especially the variability in patients’ illness presentation, behavior, and recuperation (nonlinear dynamics, emergence). A second line of activity applied systems and complexity thinking to the organization and function of health care systems, and highlighted the influence of attractors on the operation of, and learning in, health system organizations and their provision of clinical care (pattern of interaction, self-organization, adaptation, coevolution).

Family physicians increasingly looked to chaos theory to help guide their provision of care for individuals, treating their patients as multivariable, nonlinear, nonperiodic systems.¹⁴ McWhinney’s work¹⁵ was again pivotal. He outlined the characteristics of general/family practice in systems and complexity terms, emphasizing the following: (1) the discipline is defined by the relationship between physician and patient; (2) practitioners think about the nexus between the person and his/her illness; (3) practitioners embrace the unique multilevel and nonlinear properties of living systems—growth, regeneration, healing, learning, self-organization, and self-transcendence (what McWhinney termed the organismic metaphor); and (4) as a discipline, general/family practice transcends the dualistic separation of body and mind.

For McWhinney, systems thinking has 3 implications¹⁵: first, the etiology of illness and disease is understood in terms of causal network relationships rather than as a linear process; second, dichotomous (either/or) questions are meaningless; and third, through their properties of self-organization, learning, and transcendence, organisms behave in a mindful way, that is, each agent within a system has the capacity to act both independently and interconnectedly. Hence, disease labels may be less important than the context in which disease occurs.¹⁶ Understanding these interconnections explains the occurrence of illness without identifiable pathology—often termed somatization—and moves beyond the reductionism of mind-body dualism.

By looking at interconnections to better understand the function of general/family practices, researchers demonstrated the importance of attractors. It became clear that the agents in a practice are not truly independent, and that interactions among key agents provide a more powerful explanation for the behavior of a practice than individual knowledge and behaviors. Furthermore, change was more likely to occur where that change enhanced existing attractors or created new ones, or when practices became aware of their attractors and internal operational models. Lack of attention to these issues explains why guidelines, continuing medical education, and standardized interventions often failed. Rationality and knowledge do not matter to complex adaptive systems such as general/family practices unless they fit the practice system’s patterns and attractors.^18,20

The ongoing preference of patients and physicians for a personal relationship and for patient-centered care was recognized as the core attractor of general/family practice. Indeed, general/family practice emerged as the crucial discipline within the health care system that integrates the health care needs of patients across the life trajectory and in diverse care locations. Where general/family practice was unable to fulfill this role adequately, it was difficult to achieve high-quality patient care.²¹

2000–2005

The first 5 years of the new century saw a focus on the system dynamics of medical practice. During this period, researchers and academics increasingly applied systems and complexity science theories to broad general/family practice issues, including the place of this discipline in the community, professional leadership, and the therapeutic relationship (Table 4S at dx.doi.org/10.5061/dryad.kf251).

It is now recognized that general/family practice’s role in the health care system is to deliver care that meets the needs of both local communities and individuals. To achieve this goal, individual practices must be uniquely adapted to local conditions (ie, meet local expectations and community values), and understand their local competition and environment. To achieve such adaptation, each practice must develop unique local structures and functions. Its ability to do so, however, is constrained by external influences, such as the capacities of the health care system, financial limitations, regulation, and culture.²⁴

To better understand such adaptation, organization-focused research demonstrated that practice function arises from the nonlinear relationships and interactions among the practice’s agents (physicians, nurses, administrative staff). Observations also highlighted that well-functioning practices had transformational leaders able to use complexity insights to transform their organization, recognizing the interplay between its structure and self-organizing networks.²⁶ At the same time, clinical quality improvement strategies that adopted ideas from complexity science proved more successful than traditional linear approaches.²⁴

Likewise, exploration of the biopsychosocial model of care and the consultation showed that they are better understood as emergent, dependent on the individuals involved and the conditions at the time of presentation. Given this fact, models of care need to account for circular causalities, whereby feedback loops, either self-balancing (stabilizing) or self-reinforcing (amplifying), maintain observable patterns over time and have an impact on outcomes of care.^26,27

Systems and complexity tools thus provide the means to understand the patterns of patients’ illness dynamics and adaptive responses. Interrupting unhelpful dynamic patterns and facilitating the emergence of new ones will assist the healing process. Replacing objective detachment with reflective participation would result in greater warmth and caring in the therapeutic encounter.²⁸ Such deeper models will need to be complemented with approximations that describe the dynamics as a practical guide to clinical practice, however, always being aware that such approximations may have noticeable unintended consequences.²⁸ The overall dynamics of the therapeutic relationship appear to be reflected in the dynamics and the level of continuity of care, which in turn affect health outcomes.²⁷

After 2005

This last period involved application of complex system theories to health care, health care reform, and the future of medicine. It saw a transition from the largely theoretical engagement with complexity science to its pragmatic application—dealing with illness and disease, providing more effective consultations, leading change in practice organization, understanding primary care as an interconnected multidimensional system, and influencing health care reform (Table 5S at http://dx.doi.org/10.5061/dryad.kf251). We discuss these aspects in more detail below.

Leading Change in Practice Organizations

Complex adaptive systems thinking has pragmatic consequences for transforming general/family practice. First, systems usually return close to their previous state in response to disturbance. Looking at reform, the magnitude of change has little relationship to the size of the intervention, and unintended consequences occur frequently. Second, knowledge negotiated among members of a community of practice is more likely to be applied in response to challenges. Last, the education sector itself has come to recognize that adult education is best guided by concepts of continued self-organized learning.³⁶ Complex adaptive systems approaches are consequently more likely to achieve lasting quality improvement⁵⁵ and avoidance of “surprising unintended consequences.”³⁹

Understanding Primary Care as an Interconnected Multidimensional System

Since 2005, several authors have explored the multilevel, multidimensional nature of primary care.^49–53 Core to generalist practice are the view of the whole patient, the ability to deal with the uncertainties of people’s illnesses, and the changing context-dependent dynamics of illness over time. Good outcomes for a population are strongly linked to the level of primary care implementation, with its focus on individually appropriate and accessible care. Primary care functions in adaptive cycles of growth, accumulation, restructuring, and renewal within an ever-changing health system.

Influencing Health Care Reform Efforts

The health care vortex metaphor was proposed to demonstrate how personal health experiences could drive health care, health service delivery, and health care policy in an integrated fashion. The model highlights that achieving an integrated patient-centered health care system requires leaders able to simultaneously connect individual patient care, practice organization, and community needs with the system’s organizational and accountability demands, as well as policy makers able to allow the emergence of local solutions to common problems.^37,38