Revisiting Precede–Proceed: A leading model for ecological and ethical health promotion

Five defining contributions to health promotion

Precede–Proceed provides an eight-phase framework for practitioners to determine, develop, implement and evaluate health promotion programmes, including the application of health promotion theories systematically within such programmes (see Figure 1) (Gielen et al., 2008; Green and Kreuter, 2005). The model was first published as an evaluation framework (Green, 1974), then as Precede (Green et al., 1980) and finally as the full framework in 1991 (Green and Kreuter, 1991). The different Precede phases (for Predisposing, Reinforcing and Enabling Constructs in Educational Diagnosis and Evaluation) aim to map educational ‘diagnosis’ and planning. Proceed (Policy, Regulatory and Organisational Constructs in Educational and Environmental Development) on the other hand guides socio-ecological assessment and planning.

OPEN IN VIEWER

Figure 1.

The Precede–Proceed Model for Health Programme Planning and Evaluation.

The model’s authors define the model’s hallmark approaches as being ‘(1) flexibility and scalability, (2) evidence-based process and evaluability, (3) its commitment to the principle of participation, and (4) its provision of a process for appropriate adaptation of evidence-based “best-practices”’ (Green and Kreuter, 2005: 18). In this section, I will expand on these hallmarks to outline some of the defining theoretical, practical and ethical underpinnings of the Precede–Proceed health promotion model that differentiate it from other models.

As its authors note, the Precede–Proceed model is a framework that invites and guides health promoters to choose the theories they feel will best guide them in each phase of their work. A chapter by Gielen et al. (2008) summarises the model this way:

Precede-Proceed can be thought of as a road map and behaviour change theories as the specific directions to a destination. The road map presents all the possible avenues, whereas the theory suggests certain avenues to follow. Unlike the theories described in previous chapters, the main purpose of the Precede-Proceed Model is not to predict or explain the relationship among factors thought to be associated with an outcome of interest. Rather, its main purpose is to provide a structure for applying theories and concepts systematically for planning and evaluating health behaviour change programmes. (p. 408)

One early critique called the model ‘a-theoretical’ (McLeroy et al., 1993: 307). However, as the model’s authors and others claim in a reply (Green et al., 1994), the model has robust theoretical and practical underpinnings, though also is designed to help practitioners plan effective programmes using theory and not to predict or explain the relationships among factors.

Certainly, the model’s theoretical approach is highly eclectic, including explicitly inviting users to apply individualistic behaviour change theories in Precede phases. Arguably, the framework unites a collection of theories and, importantly and somewhat uniquely, ethical principles. The model could be cast as a theoretical and ethical framework that guides practitioners through each layer of a socio-ecological model of health, with the predictive value being that a more effective, relevant, comprehensive and democratically determined outcome will result from its use.

Green and Kreuter (1999, 2005) have subtitled the last two editions of their Precede–Proceed book, ‘an educational and ecological approach’. As this implies, the model draws on two radically different theoretical traditions – one cognitive and individual and the other structural and communal – to support comprehensive health programme planning that accounts for multiple layers of determinants of health.

Five important and defining approaches of Precede–Proceed are that it is:

Precede–Proceed leads health promoters through the layers of a socio-ecological model, from individual characteristics to broad socio-political conditions. The framework has been a leader in encouraging practitioners to view health and individual health behaviours within the overall ecosystem of health determinants and to improve those conditions (Green and Kreuter, 1999; Green et al., 1996; Kemm, 2015: 61; Richard et al., 2011).

The model’s population, rather than individual, lens on health also relates to its socio-ecological approach. Green and Kreuter (2005) urge planners to go ‘beyond the clinical, one-on-one aspect of acute health care’ to consider the relationship between groups of people and their environments (p. 2). While parts of the model can be applied to individuals, this is not its intent, in contrast with intrapersonal theories of health behaviour such as the Health Belief Model, the Transtheoretical Model, and Theory of Planned Behaviour. Assessments in the Precede phases are designed to be done at the population level. Yet, drawing on Bandura’s (1978) social cognitive theory, Green and Kreuter (2005) do not subsume personal agency with their population-level perspective, noting that ‘people learn continuously from their environmental and social surroundings and can develop, individually or collectively, the knowledge and skills to modify them’ (p. 30). This emphasis appears in several phases in the use of more agentic, individual theories such as cognitivism and exchange theories. Green and Kreuter (2005) write,

Without education and a commitment to the open exchange of idea, biases, and assumptions, the process of planning runs the risk of becoming a manipulative, social engineering enterprise. Without the policy support for social change, on the other hand, education efforts, shown to be effective on an individual basis, often prove to be too weak to yield a population-wide benefit. (p. 6)

This passage points to a third defining feature of the model: an emphasis on democracy and the participation of the ‘target population’ in every phase of assessment, prioritisation, planning, implementation and evaluation with emphasis on personal agency in ‘democratic social and behavioural change’ (Green and Kreuter, 2005: 6). Here, the authors draw from participatory action research methods and concepts from Paulo Freire’s (1970) transformation learning theory, including conscientisation and cultural invasion versus synthesis.

A fourth feature is its focus on beginning at the end, with the end being a higher quality of life rather than behaviour change or even health. Starting with the first edition of the Precede model, the authors define health as an instrumental value; ‘people cherish health because it serves other ends’ (Green et al., 1980: 18). This view, later enshrined in the Ottawa Charter for Health Promo-tion (1986), which describes health as ‘resource for everyday life, not the objective of living’, is widely expanded upon in later editions.

The model is also notable for being strongly grounded. In response to critiques and comments from practitioners applying the model in the field, the authors have gradually but substantially changed their model since Green’s inception 40 years ago of what became Precede (Green, 1974; Green and Kreuter, 1999). Even the early version of the Precede model was ‘rigorously evaluated [in] “real world” clinical trials’ (Green et al., 1980: 11). As the authors explain in a footnote,

the theoretical grounding and evolution of the model has been influenced as much from its various applications and the theories brought to bear in those applications as in the original theories and research that led to the formulation of the model. (Green and Kreuter, 2005: 180–181)

Examining the Precede–Proceed phases

In the most recent edition of Health Programme Planning, the Precede–Proceed model is said to have eight stages (see Figure 1) (Green and Kreuter, 2005). Precede has four assessment and planning stages to guide the health promotion partners in selecting what problem to address, examining its underlying causes, and planning an intervention. Proceed on the other hand includes four implementation and evaluation phases.

Phase 1, the social assessment and situation analysis, identifies and assesses potential areas for health action. By engaging the community in this diagnosis, using ‘both objective and subjective’ (Green and Kreuter, 2005: 31) information from multiple sources, the goal is to identify the population’s priorities in improving their lives. The authors suggest that community member priorities are as relevant as the ‘actual’ (quotes theirs) medical or health needs assessed in the next phase (Green and Kreuter, 2005: 38, 40).

Phase 2, the epidemiological assessment, identifies and prioritises health issues and sets change objectives. This phase draws heavily from the empirical traditions of the biomedical sciences, but also mandates ecological approaches for identifying structural barriers to health by assessing the physical, social, political and economic determinants of health.

Phase 3, the educational and ecological assessment, urges examination of the broader causal factors behind the social and health issues prioritised in the earlier stages. The authors draw heavily from several intrapersonal health promotion theories in how to assess these causal factors, particularly the Health Belief Model, Social Cognitive Theory, Theory of Reasoned Action, and the initial stages of the Transtheoretical Model. Practitioners have also successfully used motivational interviewing (Monteiro et al., 2011; Riegel and Carlson, 2002; Taplin et al., 2000).

The authors group the Phase 3 causal factors into three categories: predisposing, reinforcing and enabling. Predisposing factors, which include individual knowledge and attitudes, draw most heavily from the intrapersonal health promotion theories above. Enabling factors are the resources and skills required to make desired behavioural and environmental changes; for example, the availability, accessibility and affordability of fresh produce enables or inhibits eating a healthful diet. Enabling factors often overlap environmental factors identified in the previous stage. The authors draw on concepts such as social capital, community capacity and collective efficacy and they suggest community organising and social action strategies to identify problems, set goals and effect structural change. Reinforcing factors are those that follow a behaviour that ‘determine whether the actor receives positive (or negative) feedback and is supported socially afterward’ (Green and Kreuter, 2005: 167). They again draw from intrapersonal health promotion and other expectancy value theories in this part of their model. All three categories of factors should be prioritised and then addressed in the intervention, which is planned in the next stage.

In Phase 4, the health promotion coalition designs the action plan for meeting the objectives set in the first three phases, selecting interventions that are most likely to be successful in achieving each objective and that are within the capacity of the team. For making these choices the model’s authors suggest some practical tools (e.g. making Gannt charts, using ‘MATCH – Multilevel Approach to Community Health’ or ‘PATCH – Planned Approach to Community Health’ models). The authors also recommend conducting an internal policy assessment within the planning organisation to ensure policies align with the intended intervention plan and to gauge political forces likely to impact implementation; as they put it, ‘some barriers will be essentially attitudinal or political or reflect power relationships that you cannot politely make a matter of public record in your formal plan, but you ignore them at the peril of your programme’ (Green and Kreuter, 2005: 225). To aid in assessing political forces, the authors briefly discuss a wide range of organisational theories and approaches – including exchange, conflict, utilitarian and critical theories – that may aid in assessing the power relationships at play.

Proceed includes phases 5–8, mainly providing a participatory health promotion spin on standard project planning, management, implementation and evaluation frameworks. Proceed’s goal is to make the programme ‘available, accessible, acceptable and accountable’ (Green and Kreuter, 2005: 245), including being practically feasible for all partners and appropriate to the population’s needs, aspirations and circumstances. Phase 5 is implementation. Here, the authors provide guidance by example through three chapters of case studies. Phases 6–8 encompass the process, impact and outcome evaluations. Although these three evaluation phases are presented discretely as final phases, the model suggests conducting process evaluation activities in every phase.

Brief history of the model in health promotion

As a field, public health examines the interaction of environment, agents and hosts. In this triad, a focus on host ‘lifestyles’ was identified through chronic disease research in the 1960s, particularly in the USA (e.g. Kannel et al., 1961; US Surgeon General’s Advisory Committee on Smoking and Health, 1964). This focus on individual behaviour and its determinants drove the expansion of health education into health promotion.

Health education traditionally involved informing individuals about how their behaviours influence health. Health promotion is marked by preventing disease through the study and change of both behaviour and environmental influences on behaviour; its emergence as a part of public health is often dated to the 1974 LaLonde Report (Bell, 2003; Buchanan, 2000). This Canadian policy document announced that ‘the Government of Canada now intends to give to human biology, the environment and lifestyle as much attention as it has to the financing of the health care organisation’ (LaLonde, 1974: 6). Explicitly building on this report (though omitting its emphasis on equity), the office of the US Surgeon General officially marked a related shift with the first Healthy People report (Department of Health and Human Services [DHHS], 1979). In parallel, the quest for understanding the loci of responsibility for and solutions to health was aided by the adaptation of social and human ecological theory (Bronfenbrenner, 1977) to a socio-ecological model of determinants in health in the late 1980s (McLeroy et al., 1988). ¹

Health promotion officially came into its own in 1986 with the defining document of the field, the Ottawa Charter for Health Promotion. This normative charter, which remains the touchstone and cornerstone for the field, defines health promotion as ‘the process of enabling people to increase control over, and to improve, their health’. Several papers trace shifts in dominant discourses of the field as a whole from a focus on individual behaviour change to a socio-ecological approach for empowerment and equity to, today, a population approach located within capitalist economics (Porter, 2007; Robertson, 1998; Robertson and Minkler, 1994). ²

The 40-year history of the Precede–Proceed model unfolded within this history of health promotion. Green began brewing the Precede model in the late 1960s. He was influenced by the socio-ecological approach of Rogers (1960), who wrote that

the most pressing human health problems and the most probable solutions for them are more and more involved in the necessity of knowing more about the relationship between man [sic] and his total environment. Given a basic, genetic start it appears that what happens to man is thenceforth largely determined by these external forces. (p. vii)

Green later teamed with Kreuter and others to publish the book Health Education Planning: a Diagnostic Approach (1980), in which the Precede term was coined. The model itself was nearly identical to that in Green’s 1974 paper, except that it begins with the more positively oriented ‘quality of life’ rather than a ‘social problem’ diagnosis. Both publications focused on cost-benefit analysis, but it is the planning model that has since received the spotlight. This ‘pioneering’ (Rissel and Bracht, 1999: 61) book helped define the scope of what health education means and how to conduct it, in much of the USA at least.

The original Precede model focuses nearly exclusively on how to help people make voluntary health behaviour changes through democratic educational methods. This educational piece remains intact in later models, including participatory methods that the authors call ‘mutual planning and diagnosis’ (Green et al., 1980: 5). This was progressive for health education at that time and possibly helped pave the way for the emergence of community-based participatory research in the late 1990s (Israel et al., 1998) as public health’s version of Freirian (1970) participatory action research.

The early Precede model is missing the environmental and socio-ecological analysis and action developed in later versions. Although briefly referenced under enabling factors, these outer layers of today’s socio-ecological health model are mainly lumped together as ‘economic, genetic and environmental factors … acknowledged here because of the power they have, however indirect, to influence health’ (Green et al., 1980: 13). This reflects and perpetuates the limited scope of the health education field overall at that time. As the authors of the model note, ‘except on their own time health professionals usually are not expected (and sometimes are not allowed) to intervene in non-health matters’ (Green et al., 1980: 37). However, Green and Kreuter foreshadow the model’s future expansion from health education to health promotion, and from the educational to the socio-ecological:

… health education is sometimes accused of ‘blaming the victim’, because it appears to place all the responsibility for protection of health on the individuals whose health is threatened. Recognising the non-behavioural causes of health problems acknowledges that there are other threats to health beside the behaviour of the victim. (Green et al., 1980: 54)

In a strategies chapter, they also address the organising and social action methods required to impact these ‘non-behavioural causes’.

A major contribution of this original model was challenging the limitations of an exclusively KAP/KAB approach (Knowledge, Attitudes and Practices/Behaviours) to health behaviour change and KAB’s underlying premise that if only people had the right information they would change their behaviour. The authors instead emphasise voluntary, democratic behaviour change and draw on Freire to contest the ‘empty vessel’ approach, in which some ‘health educators behave as if all they have to do to ensure the success of their programme is to pour health information into the empty minds of an eagerly awaiting target population’ (Green et al., 1980: 6). They begin to challenge ‘victim blaming’ – noting that ‘the system may be at fault, rather than the patient’ (Green et al., 1980: 77). The breadth, depth and force of these arguments grew in the book’s next edition (Green and Kreuter, 1991).

Over a decade later, in keeping with and perhaps also partly leading broader developments in the field, Green and Kreuter (1991) explicitly expanded the model’s scope from health education to health promotion, titling the second edition of their book, Health Promotion Planning: an Educational and Environmental Approach. This edition looks very similar to the current (fourth) one, adding environmental assessments and a policy and power analysis. It also adds the Proceed portion of the model to help users negotiate the more political terrain this expansion encompasses. Among health promotion models and theories, Precede–Proceed led the way in socio-ecological health assessments and planning; only relatively recently have socio-ecological models for health been adopted in public health as a guiding frame for understanding and improving health (Lang and Rayner, 2012), including becoming the anchoring concept for US public health goals (US DHHS, 2010).

The third edition (Green and Kreuter, 1999), titled Health Promotion Planning: an Educational and Ecological Approach, expanded the environmental approach to an even broader ecological one because, as the authors noted, health promotion with a population focus

… demanded more than merely taking forces outside the person into account in planning programmes. It demanded an intersectoral, interdisciplinary, and interorganisational strategy for integrating the forces operating at several levels and in various spheres to support people in their efforts to gain greater control over the determinants of their health. (p. xxvi)

The revised model also gave increased emphasis to community participation, coalition formation, capacity building and sustainability, drawing on work published in the interim, such as asset mapping (Kretzmann and McKnight, 1993). It shifted even further away from a biomedical approach, marked by calling the Precede phases ‘assessments’ rather than ‘diagnoses’.

The most recent, fourth edition of the book is entitled Health Programme Planning: an Educational and Ecological Approach (Green and Kreuter, 2005). Shifts include the addition of genetic factors, suggestions for application shortcuts, and the folding of the assessment of health, behaviour, environment and genetics into one phase, thus reducing the model from nine to eight phases. This last consolidation emphasises viewing these factors as dialectic and reciprocally determined, rather than dualistic and discrete. The authors also expand the intended application of the model. As the title signals, they suggest its usefulness not just for health promotion but also for public and population health planning. Finally, to make the model ever more grounded, every successive edition has addressed feedback and critiques from users and has incorporated new best practices and processes in its application.

Limitations of health promotion models

Health promotion models and theories generally, the Precede–Proceed model specifically, and this paper about the model, all have their limitations. The influence of Precede–Proceed on the field is examined in this paper, as opposed to focusing on evaluations of the predictive value of the model. However, this section briefly examines the health promotion model evaluation literature.

In an introductory chapter to the Glanz et al. (2008b) health behaviour text mentioned above, the editors note that ‘theories that gain recognition in a discipline shape the field, help define the scope of practice, and influence the training and socialisation of its professionals’ (p. 31). In the same passage, the editors observe that ‘today, no single theory or conceptual framework dominates research or practice in health promotion and education. Instead, one can choose from a multitude of theories’ (Glanz et al., 2008b). One possible reason that no single model or theory dominates practice is that none have proven to be especially effective for promoting health. Some statistical reviews suggest that the use of health promotion theories and models to plan programmes has little or even no association with improved outcomes (e.g. Munro et al., 2007; Webb et al., 2010). Some qualitative reviews suggest that using health promotion theory to plan programmes can lead to better health outcomes (Glanz and Bishop, 2010), although cited examples (e.g. Ammerman et al., 2002; Noar et al., 2007) tend to provide evidence for particular behaviour change techniques (e.g. tailoring and goal-setting) rather than whole theories or models. For the Precede–Proceed model, the strongest evidence for effectiveness concerns predisposing, enabling and reinforcing factors being predictive of – or, at least, highly correlated with – health behaviours (e.g. Aboumatar et al., 2012; Chang et al., 2005; Polcyn et al., 1991; Tejeda et al., 2009).

Commonly cited practical limitations of the Precede–Proceed approach include the fact that its comprehensiveness and participation imperative make it time and cost prohibitive to apply in full in the real world, while also not providing detailed guidance for each step (Gielen et al., 2008: 417; MacDonald and Mullett, 2009: 165; Sharma and Romas, 2012: 48). In response, the authors suggest using the model in part, and in tandem with more specific theories.

For those who consider the ethical foundations of the model an asset, a critique could be made that the authors sometimes use rather technical language to describe (or perhaps to disguise) concepts that were ground-breaking for a model that became mainstream in health. They suggest prioritising health equity using cloaked language about statistical indicators of health problems. They also use terms standard in health promotion such as ‘target population’ and ‘patient’ that discursively cast people, communities and citizens in passive roles. And they do so even when making statements, cited above, that were radical in health fields for 1980 such as ‘the system may be at fault, rather than the patient’, that ‘both objective and subjective’ information should count, and that a programme being ‘available, accessible, acceptable and accountable’ is as important as it being measurably effective.

Limitations of this paper

A main limitation of this review is that the links suggested between the ethical and practical implications of Green and Kreuter’s framework with trends in health promotion and public health are associational, not necessarially causal. Evidence for causality is limited to, at best, temporal precedence. Other weaknesses are in the limited scope of the paper, including only a truncated review of histories of the field and, moreover, of the overall socio-political contexts of its development. Also, the assessment of the very extensive literatures containing weak evidence for the effectiveness of this (or any other) model in health promotion is partial. Finally, the lens employed to provide this overview of the model, its development, and its implications for the field is wide and, as with all perspectives, situated. Others might note different features standing out as worthy of attention, even if they were to look through the same lens from the same location.

Why the Precede–Proceed model matters

Public health involves ‘ensuring the conditions for people to be healthy’ (Institute of Medicine, 1988) and, within this, health promotion is ‘the process of enabling people to increase control over, and to improve, their health’ (Ottawa Charter, 1986). However, for a field and a discipline that should be about enabling control and improving health conditions, health promotion strategies tend towards highly individualistic behaviour change approaches. ³

The Precede–Proceed model embodies public health and health promotion goals and guides practitioners in reaching them. It allows health promoters to use popular individual health behaviour change theories as part of larger strategy for socio-ecological change that improve health conditions and enables people to control their own health in ways most meaningful to them. The framework provides a guide to improving health through change in multiple layers of an ecological model.

The model’s authors have offered not only a practical framework for using multiple theories in planning health programmes. They have been providing moral leadership in the health promotion field in two ways. First, the authors reject the persistent behaviour change myths of health promotion – that behaviour change is an end in itself; that information dissemination alone will lead to behaviour change; and that behaviour change is mainly a function of rational, individual decision making. Following this model, improved quality of life should be the goal of health work, with behaviour change and enhanced health simply as means to that end. Effective interventions should include action on structural factors that influence health. Second, long before the advent of community-based participatory research in health, the model’s authors had been advocating since 1974 that ‘target populations’ must be intimately involved in identifying, assessing and resolving their own community health issues. Couching their model in Cartesian, techno-rational terms may have made this process seem less radical, but also more acceptable in mainstream health promotion – serving as a Trojan horse, perhaps, for health promotion tackling social determinants of health more than health behaviour, and for respecting the experience, expertise and priorities of community members in determining their own health. If none of these tenets are news today, this may be in some part due to the leadership offered by this approach. The fact that socio-ecological and participatory approaches are hardly ubiquitously applied means that the Precede–Proceed model still has much to offer in guiding health promotion planning and practice.