Unravelling the concept of integrated public health policy: Concept mapping with Dutch experts from science, policy, and practice.

G Model HEAP-3351; No. of Pages 11

ARTICLE IN PRESS Health Policy xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Health Policy journal homepage: www.elsevier.com/locate/healthpol

Unravelling the concept of integrated public health policy: Concept mapping with Dutch experts from science, policy, and practice Luuk Tubbing ∗ , Janneke Harting, Karien Stronks Department of Public Health, Academic Medical Center, University of Amsterdam, PO Box 22660, 1100 DD Amsterdam, The Netherlands

a r t i c l e

i n f o

Article history: Received 25 June 2014 Received in revised form 26 December 2014 Accepted 28 December 2014 Keywords: Integrated public health policy Health in all policies Healthy public policy Intersectoral action for health Concept mapping

a b s t r a c t Background: While expectations of integrated public health policy (IPHP) promoting public health are high, assessment is hampered by the concept’s ambiguity. This paper aims to contribute to conceptual clarification of IPHP as first step in further measurement development. Methods: In an online concept mapping procedure, we invited 237 Dutch experts, 62 of whom generated statements on characteristics of IPHP. Next, 100 experts were invited, 24 of whom sorted the statements into piles according to their perceived similarity and rated the statements on relevance and measurability. Data was analyzed using concept mapping software. Results: The concept map consisted of 97 statements, grouped into 11 clusters and five themes. Core themes were ‘integration’, concerning ‘policy coherence’ and ‘organizing connections’, and ‘health’, concerning ‘positioning health’ and ‘addressing determinants’. Peripheral themes were ‘generic aspects’, ‘capacities’, and ‘goals and setting’, which respectively addressed general notions of integrated policy making, conditions for IPHP, and the variety in manifestations of IPHP. Measurability ratings were low compared to relevance. Conclusion: The concept map gives an overview of interrelated themes, distinguishes core from peripheral dimensions, and provides pointers for theories of the policy process. While low measurability ratings indicate measurement difficulties, the core themes provide pointers for systematic insight into IPHP through measurement. © 2015 Elsevier Ireland Ltd. All rights reserved.

1. Introduction The purpose of integrated public health policy (IPHP) is to positively impact population health through contributions from various policy sectors and organizations. The rationale behind IPHP is grounded in the epidemiological understanding that health is influenced by different types

∗ Corresponding author. Tel.: +31 020 5664892. E-mail addresses: [email protected], [email protected] (L. Tubbing).

of determinants Lalonde [16]: genetic makeup, lifestyle, social and physical environmental factors, and the quality and accessibility of prevention and health care. Many of these determinants are mainly influenced by policies of non-health sectors. IPHP can be characterized as coherently addressing the different types of determinants of health through purposeful contributions from multiple policy sectors and organizations [13,16,19]. As health is an outcome of complex and dynamic relationships among various actors and determinants [13], IPHP is a complex undertaking. It at least requires a variety of knowledge and capacities, the agency of multiple actors on multiple

http://dx.doi.org/10.1016/j.healthpol.2014.12.020 0168-8510/© 2015 Elsevier Ireland Ltd. All rights reserved.

Please cite this article in press as: Tubbing L, et al. Unravelling the concept of integrated public health policy: Concept mapping with Dutch experts from science, policy, and practice. Health Policy (2015), http://dx.doi.org/10.1016/j.healthpol.2014.12.020

Definition main concept

Aspects mentioned

Intersectoral action for health

‘A recognized relationship between part or parts of the health sector with part or parts of another sector which has been formed to take action on an issue to achieve health outcomes (or intermediate health outcomes) in a way that is more effective, efficient or sustainable than could be achieved by the health sector acting alone.’

Milio/2001/glossary

Places healthy public policy within other policy constructs and invites testing in use and discussion

Healthy public policy

Freiler et al./2013/glossary

Presents concepts and terms useful in understanding the implementation of health in all policies as a cross-sectoral policy

Health in all policies

Kickbusch and Gleicher/2011/report

Tracks governance innovations that have been introduced to address priority determinants of health and summarizes them as strategic approaches to smart governance for health

Governance for health

‘Healthy public policies improve the conditions under which people live: secure, safe, adequate, and sustainable livelihoods, lifestyles, and environments, including housing, education, nutrition, information exchange, child care, transportation, an necessary community and personal social and health services. Policy adequacy may be measured by its impact on population health.’ ‘Health in All Policies [. . .] is [. . .] a governmental strategy to improve population health by coordinating action across health and non-health sectors.’ ‘A single case of HiAP reflects a multisectoral initiative towards healthy policymaking involving the national or state/provincial level of government where sectors collaborate (often through processes of cooperation, coordination or integration) to develop policies and programmes that include population health initiatives for preventing the manifestation of inequities [in the determinants of health].’ ‘Governance for health is defined as the attempts of governments and other actors to steer communities, whole countries or even groups of countries in the pursuit of health as integral to well-being through both whole-of-government and whole-of-society approaches.’ ’Governance for health promotes joint action of health and non-health sectors, of public and private actors and of citizens for a common interest. It requires a synergistic set of policies, many of which reside in sectors other than health as well as sectors outside of government, which must be supported by structures and mechanisms that enable collaboration’

Discusses determinants of health, intersectoral action at work, partnerships and strategies, policy and program implications Presents elements of successful intersectoral action for health: social mobilization and empowerment; leadership, champion and/or catalyst; analysis/priority-setting; mutually beneficial relationships; integrated action at micro and macro level; human, technical and financial resources; variation in institutional long-term policy; institutionalization of health impact/gain assessment; training, tools and capacity development; coordination and integration mechanisms, partnering. Also explains the following concepts: policy, public policy, organization policy, policy goals, healthy public policies, policy making process, policy stakeholders/players, actors, policy environment, policy instrument, political strategy, strategic information

Discusses: policy implementation, intersectoral action, intersectoral engagement, agenda setting, raising awareness, win-win approach, capacity building, institutional capacity, expert capacity, prior experience

Determinants of health, governance for health and well-being, whole-of-society/government approach, interdependence, complexity, co-production Presents intersectoral action for health, healthy public policy and health in all policies as three successive waves in governance for health and places them on a continuum of less-to-more policy integration

ARTICLE IN PRESS

Main concept

Provides consensus on the definition of intersectoral action for health

G Model

Contribution

WHO/1997/Report

L. Tubbing et al. / Health Policy xxx (2015) xxx–xxx

Author/year/publication type

HEAP-3351; No. of Pages 11

2


Table 1 Selected core publications on integrated public health policy.


ARTICLE IN PRESS

Aspects mentioned Definition main concept

‘Health in All Policies is the policy practice of including, integrating or internalizing health in other policies that shape or influence the social determinants of health. These determinants include transport, housing, tax and agricultural policies, to name a select few. Health in All Policies is more concerned with the ‘big issues’ and less with individual programmes or projects. Depending on the institutional context of a country, these policies may be found at the national, regional, local level or dispersed in multilevel governance systems. Health in All Policies is a policy practice adopted by leaders and policy-makers to integrate consideration of health, well-being and equity during the development, implementation and evaluation of policies.’ Health in all policies McQueen et al./2012/report

Main concept Contribution

Provides an analytical framework for intersectoral governance consisting of intersectoral governance structures and actions

Author/year/publication type

Table 1 (Continued)

Discusses two dimensions of governance as to achieving HiAP: 1) the structures that bring actors together and 2) the actions flowing from their mutual engagement and deliberations Intersectoral governance structures include: interdepartmental committees and units, joint budgeting, delegated financing, public engagement, stakeholder engagement, industry engagement Intersectoral governance actions include: evidence support, setting goals and targets, coordination, advocacy, monitoring and evaluation, policy guidance, financial support, providing legal mandate, implementation and management.


3

levels [1,20], and consideration of different interests and normative assumptions [3]. As a result of this multiplicity and complexity, a variety of theoretical approaches is apparent [6,13]. Besides, IPHP practices are often of experimental nature, while actual manifestations of IPHP are highly heterogeneous and mostly hard to identify [10,22]). As clear operational criteria and defining characteristics to assess IPHP are lacking [10], evidence to support IPHP remains ‘scarce, preliminary and anecdotal [17,26].’ So far, instruments tend to assess certain aspects of the approach, like either the epidemiological ground or intersectoral efforts and network configurations to develop and implement IPHP (e.g. [11,24]), rather than to cover the approach in its entirety. It can additionally be questioned whether instruments like these sufficiently cover the policy coordination and policy integration that can be regarded as core characteristics of IPHP [6]. This means that further conceptual clarification of IPHP could be a first step in the identification of the operational criteria and defining characteristics needed to monitor and evaluate IPHP in its extensiveness and thus to further strengthen the evidence-base of the approach. Previous publications on the conceptualization of IPHP provide already useful insights. A summary of selected core publications is presented in Table 1. A first observation is that within the scope of IPHP different main concepts have been identified, such as intersectoral action for health [29], healthy public policy [21], and health in all policies [9,13]. These concepts, of which the definitions partly overlap, can be regarded as three successive waves of governance for health, an even broader integrated policy approach, and positioned on a continuum of less-to-more policy integration [13,14]. A second observation is the multitude of policy aspects that is presented and elaborated on within the scope of IPHP. Although this may be expected given the multiplicity and complexity of the IPHP concept, few attempts seem to have been made to thematically organize the different aspects, for instance in relation to policy content and policy process factors, policy input and policy output factors, or the different phases of the policy process. A third observation is that previous publications tend to elaborate on specific parts of the IPHP concept, such as on the development process (e.g. [19]) or the implementation process [9], thereby apparently neglecting other parts that cannot be properly understood separately. While the previous works might together cover IPHP conceptually in its entirety, we conclude that further clarification is needed if we want the concept to guide policy development, implementation and evaluation. Based on these observations, in order to contribute to conceptual clarification, we decided to go back to the drawing board through concept mapping [12]. This method extends beyond literature reviews as it provides an empirical basis by relating different aspects of a concept. 2. Methods 2.1. Concept mapping Concept mapping offers a standardized participative methodology for gathering and organizing ideas of a



ARTICLE IN PRESS L. Tubbing et al. / Health Policy xxx (2015) xxx–xxx

4

diverse group of experts and rapidly form a common framework about complex, multidimensional concepts [28]. Concept mapping is recommended as a particularly suitable method to clarify ill-defined problems of human interactions in ‘soft’ systems, where health is coproduced and much of the most important policy analysis and management occurs [27]. In a concept mapping procedure participants generate and structure statements, which are then represented in a two-dimensional concept map through statistical analysis and interpretation of the researchers and the participants. Through the participation of a diversity of experts a wide array of theories is incorporated, while the sorting and rating of statements provides an organizing mechanism that exposes interconnections and core areas [12].

2.2. Preparation We made a list of possible ‘focus prompts’, intended to generate the type of input desired from the participants, one of which was pilot tested (n = 5) and found suitable: ‘A specific characteristic of local integrated public health policy is. . .’ 237 potential participants across different work fields (science, policy, practice) in The Netherlands were identified, largely from the health sector and for a smaller portion from other sectors, e.g. policy and organization researchers and social policy makers or advisors. The participant pool was further fairly diversified based on geography, gender, amount of experience and position.

2.4. Structuring of statements A subgroup of 100 participants, including the 62 respondents and a selection of 38 non-respondents from the statement generation phase, was invited to perform two structuring tasks: sorting and rating of statements. For the sorting task, each participant was instructed to sort the statements into piles ‘in a way that makes sense to you’, not to sort all the statements into one pile, not to sort every statement as its own pile (although some statements may be grouped by themselves), and not to sort statements into more than one pile [12]. The participants were asked to rate the statements based on relevance and measurability. Seven-point Likert-type scales were applied, where ‘1 = not at all relevant/very difficult to measure’ and ‘7 = very much relevant/very easy to measure’. In order to collect information for response profiling and subgroup comparison, the participants were asked to indicate their gender (male, female), age category (younger than 35, between 35 and 50, older than 50), primary field of work (science, policy, practice), primary area of expertise (directly health-related, not directly health related), and years active in their current area of expertise (less than 5 years, between 5 and 10 years, more than 10 years). A total of 24 unique participants completed one or more of the structuring tasks, yielding a response rate of 23% for the clustering task and 22% for the rating task (Table 2). All participants were from the health sector, while four public health practitioners took part in the clustering assignment and two in the rating assignment. 2.5. Data analysis

2.3. Generation of statements Through an e-mail with a link to a collective brainstorming page, participants were invited to anonymously submit statements that complement the focus prompt. They were given two weeks in which they could contribute as much statements as they wanted whenever they wanted. To increase response rate two reminders were sent at equal intervals, of which the first was planned in advance and the second optional in case of low response. Also, it was emphasized that participants could make a valuable contribution within just 5 min and it was appreciated if they contributed more. A group of 62 respondents generated 177 statements, yielding a response rate of 26%. The fields of science, policy and practice were all fairly represented among them (Table 1), eight of whom were from outside the health sector. The statements were edited by the researchers: statements that contained more than one idea were split, duplicate ideas removed, largely overlapping ideas merged, and language mistakes corrected. In addition, 10 ideas from the fields of public health, governance and policy, politics, and organization science were added. These ideas were perceived as both missing and essential by the researchers and an external health promotion expert. This resulted in a final statement set of 97 unique exemplars. This number is considered adequate in order to keep the set manageable for further participant assignments and data analysis [12].

Using the concept mapping software, a binary symmetric similarity matrix was computed for each participant. The program then calculated an aggregated matrix by counting the individual matrices. A high value in this matrix indicates that many participants sorted that pair of statements together in a pile and implies that the statements are perceived as conceptually similar in some way. Through multidimensional scaling analysis, each statement was located on a two-dimensional map. Statements that appear on the map closer to each other are generally grouped together by the sorters more frequently, while statements that appear more distant from each other are generally grouped together less frequently. To assess the degree to which the configuration of the map matches the data from the similarity matrix, the stress value is calculated ranging from 0 to 1, where a lower value indicates a better overall fit [15]. Any value between 0.24 and 0.39 can be considered appropriate [23]. The point map undoubtedly met this demand (0.293). The concept mapping software partitions the statements into clusters by hierarchical cluster analysis and suggests various cluster label alternatives. Using a checklist, a panel of five experts from science, policy, and practice individually assessed the different cluster solutions provided by the software with either ‘agree’, ‘disagree’, or ‘undecided’. In this procedure, the number of clusters per solution decreased step-by-step from 20 to 4, which can be considered a conceptually appropriate range on the


G Model

ARTICLE IN PRESS

HEAP-3351; No. of Pages 11


5

Table 2 Participants: response rate, total and by domain. Sortingb

Ratingb

Primary field of worka

Brainstorming Invited (n)

Response (n)

Response rate (%)

Invited (n)

Response (n)

Response rate (%)

Invited (n)

Response (n)

Response rate (%)

Science Policy Practice Total

81 67 89 237

21 17 24 62

25.9 25.4 30.0 26.2

35 26 37 100

11 8 4 23

31.4 30.8 10.8 23.0

35 26 37 100

11 9 2 22

31.4 34.6 5.4 22.0

a In the brainstorming assignment, the primary field of work of participants was based on publicly available data, while in the clustering and rating assignments this is based on self-reporting. b 21 Participants completed both clustering and rating, 2 completed only clustering, 1 completed only rating.

continuum of abstract-specific [12]. The experts additionally indicated which cluster labels where most appropriate from their point of view. No consensus on a cluster solution and labels was reached among the experts. The researchers decided on the final cluster solution and cluster labels, taking into account the suggestions of the software and the input from the expert panel. Occasionally, the researchers also used bridging values, which reflect the frequency with which a statement is sorted together which another statement, to support the final cluster solution. The final cluster solution determined all further analyses. First, in order to provide a better grasp of shared conceptual meaning, we partitioned the clusters into themes by selecting a more aggregated cluster solution using the concept mapping software. Second, for similar conceptual reasons, we used the distance between clusters provided by the software to identify dimensional axes in our final concept map. Third, pattern match analyses were applied to explore remarkable patterns in the data, comparing the mean relevance and mean measurability ratings of clusters as well as differences between subgroups of participants. 3. Results The 97 statements (Table 3) were grouped into 11 clusters, which were in turn partitioned into five themes (Fig. 1). 3.1. Relevance The eastern part of the concept map yielded substantially higher mean relevance ratings than the western part. The themes ‘integration’ (5.33) and ‘health’ (5.26) received the highest mean relevance ratings, separating them from the other three themes that can be seen as more peripheral considering their lower mean relevance ratings: ‘generic aspects’ (4.94), ‘capacities’ (4.83), and ‘goals and setting’ (3.96). The theme ‘integration’ was all about connection and synthesis (e.g. of policies, organizations, budgets, sectors, and domains). Its highest rated cluster ‘policy coherence’ (5.53) was composed of statements about the coherence of policies of multiple sectors in terms of approaches (e.g. integrated approach to all areas of life), arrangements (e.g. integration of policy notes), and outcomes (e.g. mutual reinforcement). Its other cluster ‘organizing connections’

(5.17) was involved with the process of connecting sectors in terms of domains, people, resources, knowledge, and responsibilities. The ‘health’ theme was about the substantive multiplicity of underlying determinants and the role of health issues within the policy content and process. Its highest rated cluster ‘positioning health’ (5.62) pointed to different roles that health issues can play within the policy process, starting with an integrated analysis of health issues while emphasizing that health is intertwined with other issues and is one of the many values in public policy. Its other cluster ‘addressing determinants’ (4.94) referred to the content of IPHP in terms of the determinants involved, with a focus on health gains. The theme ‘generic aspects’ is named after the only cluster it contains. It comprised rather general statements, referring to concepts such as collectivity and common goals. Located at the center of the concept map, this cluster can be seen as a connector that bridges the several surrounding themes. The ‘capacities’ theme entailed conditions for establishing IPHP. The cluster ‘network and actors’ (5.05) was about the networking of different actors that can play a role in IPHP. In the cluster ‘conditions for collaboration’ (4.88), the statements described several conditions and ‘things to do’ in a multi-stage collaboration process. The ‘boundary work’ (4.79) cluster emphasized boundary crossing activities and actors inherent to IPHP. The cluster ‘learning process’ (4.55) accommodated statements on challenges arising from uncertainty and complexity and their associated required competencies. The theme ‘goals and setting’ referred to the diversity of possible forms of IPHP that can be identified on the local level. The cluster ‘scope’ (4.03) consisted of statements on the comprehensiveness and multiplicity of IPHP practices, which can range from small, short-term, and explorative to ambitious, long-term, and focused. The cluster ‘citizen and neighborhood’ (3.89), contained statements related to the target population and their role. Finally, two axes were identified. First, a horizontal axis, representing different policy phases, from design at the right-hand side (the ‘integration’ theme) to implementation at the left-hand side (the ‘goals and setting’ theme). Second, a vertical axis, from input at the lower side (the ‘capacity theme’ referring to ‘the how question’) to output at the upper side (the ‘health’ theme referring to ‘the what question’).




6

Table 3 Mean relevance en measurability ratings of themes, clusters and statements. Hierarchically grouped and ranked by mean relevance. Statement number THEME ‘INTEGRATION’ Cluster ‘Policy Coherence’ 36. 29. 39. 81. 45. 18. 16. 75. 12. Cluster ‘Organizing Connections’ 35. 91. 86. 22. 85. 96. 70. 43. 71. 13. 87. THEME ‘HEALTH’ Cluster ‘Positioning Health’ 79. 82. 58. 21. 68. 59. 93. 44. 50. 48. 24. 80. 27. 57. Cluster ‘Addressing Determinants’ 63. 15. 53. 90. 19. 95. 25. 20. THEME ‘GENERIC ASPECTS’

Statement

‘That policies across multiple sectors reinforce each other’ ‘That it is also initiated from other policy sectors’ ‘Achieve each other’s goals from different policy areas’ ‘The integration in policy notes of different policy sectors’ ‘An integrated approach to all areas of life’ ‘That it is part of a more comprehensive integrated program’ ‘The complex coherence with other policies’ ‘That it counteracts to departmentalization’ ‘That the budgets of more municipal departments are utilized’ 5.17 ‘The involvement of sectors that affect a health problem’ ‘The joint responsibility for health’ ‘The search for common interests’ ‘The use of knowledge from research, policy, and practice’ ‘The joint process of policy development’ ‘It is a win-win situation for all parties’ ‘That it connects policy and budgets’ ‘That related sectors blend into each other’ ‘The connection between domains, people, and resources’ ‘A fair distribution of contributions and benefits’ ‘The connection between cure and care’ 5.26 5.62 ‘An integrated analysis of health issues’ ‘That health is not just the domain of the health sector’ ‘That it is wider than health policy’ ‘That health is a permanent value in policy development’ ‘That work is being done also on other subjects than health’ ‘An integrated analysis of social issues’ ‘Making explicit the value of health for other sectors’ ‘That professionals integrate health aspects into their work’ ‘The weighing of health interests against other interests’ ‘That health benefits count as income for other sectors’ ‘That politics governs for health’ ‘A coherent and thought-out plan’ ‘That health is seen as a means’ ‘That it is often not labelled as health policy’ 4.64 ‘That it focuses on a healthy environment’ ‘That it focuses on health gains’ ‘That it focuses on social determinants’ ‘That it addresses complex health issues’ ‘The mix of information, facilities and regulation’ ‘That health benefits are not the primary objective’ ‘That it saves health care (costs)’ ‘That it focuses on a healthy lifestyle’ 4.94

Mean relevance rating

Mean measurability rating

5.33 5.53 6.36

4.03 4.12 3.36

5.91

4.68

5.86

3.95

5.59

5.68

5.32 5.27

3.41 4.36

5.23 5.18 5.05

2.91 3.23 5.50

3.95 6.27

4.55

5.91 5.77 5.73

3.36 4.55 4.14

5.59 5.36 5.05 4.64 4.55

4.32 4.00 4.32 3.27 3.45

4.05

3.09

3.95 4.22 4.10 6.32 6.14

4.45

6.05 6.00

4.50 3.86

5.91

4.91

5.86 5.73

4.55 4.09

5.64

3.68

5.41

3.45

5.32

3.77

5.27 5.27 5.23 4.50 4.42 5.68 5.18 5.14 4.86 4.59

3.95 4.86 2.73 3.36

4.45

4.09

3.64 3.59 3.93

3.14 5.41

4.91 4.73

4.64 4.00 4.91 4.14 5.05




7

Table 3 (Continued) Statement number

Statement



Cluster ‘Generic Aspects’ 28. 2. 64. 77.

4.94 ‘Room for local customization’ ‘Pursuing a common goal’ ‘That it is structural’ ‘Breaking through existing patterns, classifications, and ownership’ ‘The necessary support from national policy’ ‘That it concerns the formation, execution, monitoring, and evaluation of policy’ ‘That it provides positive energy’ 4.83 5.05 ‘Powerful network formation around a local problem’ ‘The connection between the circuits of administration, governance, and professionals’ ‘That stakeholders share their knowledge’ ‘The collaboration between residents, relevant organizations, and the municipality’ ‘The collaboration between public and private parties’ ‘The collaboration with health insurers’ 4.88 ‘The early involvement of relevant actors’ ‘Jointly determining priorities’ ‘That parties make formal agreements with each other’ ‘That also the possibilities of the partners determine the focus’ ‘Making explicit the individual goals of all involved’ ‘A preliminary stakeholder analysis’ ‘Giving room to the other parties’ ‘The contribution of unique resources by each party’ ‘That the reward focuses on collaboration’ ‘That normative assumptions fit with the rationales of the stakeholders’ 4.79 ‘The presence of a connecting director’ ‘The embedding in politics, governance, and administration’ ‘Looking beyond the boundaries of the subject matter’ ‘Immerse oneself in other people’s languages’ ‘The deployment of an entrepreneur with a broad perspective’ ‘The necessity of boundary work’ ‘That the municipality directs the policy’ ‘That you need boundary workers’ ‘That partners of the municipality direct the process’ ‘That nobody pursues their own benefit’ ‘That the involvement is not always voluntary’ ‘That no one feels responsible’ 4.55 ‘That it requires different competencies from professionals’ ‘That it requires political competencies’ ‘That it is a learning process’ ‘That the process is difficult to map’ ‘The influence of media on processes and outcomes’ 3.96 4.03 ‘That it has short-, medium-, and long term goals’ ‘That it can be input, output, as well as outcome of policy’

3.93 5.55 5.41 5.41 5.18

3.86 4.68 5.05 2.86

4.86 4.32

3.77 4.55

3.86 3.93 4.33 5.86

2.73

5.73

3.27

5.14 5.14

4.09 4.59

4.41

5.50

4.05 4.15 5.86 5.59 5.36

4.82

5.23

3.18

5.14

4.64

4.82 4.59 4.50

6.14 2.68 3.82

4.45 3.23

3.36 2.36

3.78 6.23 5.82

5.55 3.68

5.64

3.23

5.55 5.05

3.09 4.82

4.73 4.45 4.45 3.95

2.68 5.18 3.32 4.36

3.95 3.86 3.82 3.37 5.77

2.45 3.50 3.50

5.68 4.41 3.91 2.95

3.64 2.91 3.45 2.91

3.95 4.25 5.14

5.36

4.27

3.55

62. 67. 33. THEME ‘CAPACITIES’ Cluster ‘Network and Actors’ 14. 4. 84. 47. 60. 5. Cluster ‘Conditions for Collaboration’ 78. 46. 65. 92. 10. 83. 26. 54. 88. 66. Cluster ‘Boundary Work’ 7. 42. 49. 8. 73. 52. 9. 31. 97. 51. 23. 37. Cluster ‘Learning Process’ 61. 40. 6. 76. 94. THEME ‘GOALS AND SETTING’ Cluster ‘Scope’ 69. 55.

3.73

4.59 4.95 5.82

3.95




8 Table 3 (Continued) Statement number

Statement



74.

‘The achievement of a small visible success on the short term’ ‘The use of specific instruments, such as a health impact assessment’ ‘That its impact is difficult to measure’ ‘That there is not immediately a concrete goal or project’ ‘That a term without health fits better’ 3.89 ‘That it focuses on vulnerable citizens’ ‘That it enables people to organize their lives according to their desires and abilities’ ‘That it is a recognizable theme for residents’ ‘That the perspective of the customer is central’ ‘That it departs from local knowledge of residents’ ‘That it focuses on the neighborhood level’ ‘That it supports professionals in disadvantaged neighborhoods’ ‘The battle over who the citizen is and what he/she wants’

4.05

5.36

3.91

5.77

3.82 3.73

3.27 3.82

3.27 3.69 4.59 4.50

2.64

4.18 4.05 3.86

4.68 3.45 3.14

3.73 3.73

5.27 3.91

2.50

1.86

38. 34. 89. 32. Cluster ‘Citizen and Neighborhood’ 1. 30. 41. 56. 11. 3. 72. 17.

3.2. Measurability Mean measurability ratings of the themes were in relative proximity and were quite low compared to relevance. The theme ‘health’ came out on top (4.22), followed by ‘integration’ (4.03), ‘goals and setting’ (3.95), ‘capacities’ (3.94), and ‘generic aspects’ (3.93). Among the clusters, ‘addressing determinants’ (4.42) received the top mean measurability rating, followed by ‘network and actors’ (4.33), and ‘scope’ (4.25). The clusters with the lowest mean measurability ratings were ‘boundary work’ (3.78), ‘citizen and neighborhood’ (3.39), and ‘learning process’ (3.37). 3.3. Pattern match analyses A pattern match comparing mean relevance and measurability ratings of the clusters showed that most clusters

4.64 2.59

were found to be considerably more relevant than measurable (Fig. 2). Pattern match analyses on participant demographics showed that, on a consistent basis, participants from the policy domain produced higher mean relevance ratings than scientists did, while this seemed to be the other way around with respect to the measurability ratings (data not shown). Comparison on other participant demographics did not result in any notable patterns. 4. Discussion We conducted a concept mapping procedure on IPHP at the local level among experts from science, policy, and practice. Core themes of the concept map were ‘integration’ and ‘health’, the first being about ‘policy coherence’ and ‘organizing connections’, and the latter about ‘positioning

Fig. 1. Concept map.




9

Fig. 2. Pattern match of mean relevance and measurability ratings of clusters.

health’ and ‘addressing determinants’. More peripheral themes were ‘generic aspects’, ‘capacities’, and ‘goals and setting’, which respectively addressed general notions of integrated policy making, conditions for IPHP and the variety in manifestations of IPHP. Mean measurability ratings of most themes were low compared to relevance. 4.1. Implications While organized differently, the content of the themes and clusters that resulted from the concept mapping procedure largely correspond to the various issues discussed by recent theoretical works on IPHP (see Table 1). A first added value of the concept map is that it constitutes a framework that thematically organizes a major part of earlier separately reported and presented aspects into themes, clusters and statements, and shows the extend of interconnectedness between them. By providing such a layered and interrelated conceptualization of local IPHP, the concept map appreciates the complexity of the phenomenon it aimed to clarify. As such, while this complexity suggests that IPHP can probably not be adequately measured by a single all-encompassing instrument, the concept map shows which aspects can be measured separately and be brought together in order to provide a more complete image than is possible with the current instruments. A second added value is the distinction the concept map makes between core and peripheral themes. This difference indicates that clusters covered by core themes, namely ‘positioning health’, ‘organizing connections’, ‘policy coherence’, and ‘addressing determinants’, may be regarded as aspects that are typical for IPHP [13,22] while

clusters covered by peripheral themes, such as ‘network and actors’, ‘conditions for collaboration’, ‘learning process’ and ‘scope’, may be considered as generic policy aspects [21,25]. The first category is critical in assessing the extent to which public policy is integrated from a health perspective, the latter in understanding the required structures and actions to effectively develop and implement IPHP. In future research, it could be valuable to compare interpretations of distinctions between core and peripheral themes between experts with different characteristics with large enough samples of sub groups. A third added value is that the concept map constitutes a stepping stone towards developing theories of the policy process. The development of such s theory is described by [4] as to ‘formulate propositions on the conditions under which certain policy phenomena (e.g. preferences for certain interventions, decisions on implementation issues, allocation of resources, inclusion or exclusion of certain stakeholders, etc.) are observed and impact on policy outcomes [5,7].’ The way the concept map positions core and peripheral themes amidst a frame consisting of a development-implementation and an input-output axis, suggests that to achieve improved population health outcomes, IPHP at least requires (1) both a specific policy content (e.g. instruments addressing determinants of health) and particular policy processes (e.g. policy integration from a health perspective); (2) policy integration throughout the entire process from policy development (e.g. organizing connections) to policy implementation (e.g. reaching citizens); (3) the continuous but tailored input of a variety of capacities (e.g. provided by actors in boundary spanning networks). Propositions like these need to remain



10


tentative, as concept maps in general do not allow for causal inferences (REF), but may nevertheless serve the development of more specific theories of the policy process while putting IPHP in local policy practices. A final added value is that the statements of the concept map can be easily operationalized as they are formulated in a rather specified manner. As such, the concept map serves the purpose of conceptual clarification as a first step in the identification of the operational criteria and defining characteristics needed to adequately assess IPHP and further strengthen the evidence-base. Regardless of these added values, the rather low mean measurability ratings imply that operationalizing IPHP may be a difficult undertaking. While some aspects of IPHP seem quite tangible and fairly easy to quantify or identify (e.g. ‘addressing determinants’), others are more intangible and hard to identify or extract as they are highly dependent on context and lose their meaning when reduced to items on a checklist. Measurability ratings give cause to assume that the concept of IPHP can be mapped partially quantitatively and partially qualitatively. For example, the core themes provide a foundation to develop measurements in the form of a checklist to quantitatively assess the degree of IPHP, especially on a local level. In contrast, the theme ‘capacities’ might be hard to quantify according to low mean measurability ratings, yet provides pointers to, for instance, develop a qualitative instrument such as a reflexive evaluation method to identify and guide the quality of joint processes, for example through stakeholder workshops. Despite its correspondence with recent literature, it should be acknowledged that the concept map hardly explicitly touches upon the topics of politics, budgeting, and law—subjects that are inherently part of IPHP [2,18]. This finding confirms the lack in health promotion research and practice to respectively describe and apply the concepts that play an important role in policy theories [4]. For instance, politics in the narrow sense, i.e. ‘the process through which desired outcomes are achieved in the production, distribution and use of scarce resources in all the areas of social existence [2],’ remained underexposed in the concept map as there was no theme or cluster on politics and the words ‘politics’ or ‘political’ appeared only three times in the statement set. This is illustrative for recent criticism declaring that the role of politics in health promotion research is underdeveloped [8]. The same lack applies to budgeting aspects, such as joint budgeting or funding across sectors, and judicial aspects of policy making, such as legal mandates for intersectoral collaboration [19]. The relative underrepresentation of these elementary policy aspects can perhaps be explained by the concept map’s focus on the local level. Subjects as politics, budgeting and law may likely be more present on higher governmental levels, where frames for local policy making are predetermined. A second possible explanation is that politics, budgeting and law can be perceived as not worth mentioning because it is seen as naturally interwoven in many policy processes and structures, as is indeed implied by some of the statements in the concept map (e.g. ‘that the budgets of more municipal departments are utilized’). A final explanation may be that the respondents were mainly

from a public health background and perhaps not involved in politics, budgeting and law on a regular basis. 4.2. Limitations Our concept map should be interpreted with some caution as it has some specific methodological limitations. First, the overall response was moderate and reasons for non-response remained largely unknown. Still, the respondents in the statement generation phase were fairly representative for the target population, decreasing the probability of selective non-response [12]. This is further confirmed by the high degree of resemblance between individual aspects of our concept map and prior IPHP literature. Second, the response in the structuring phase was rather low. A quantitative pooled study analysis of 69 individual concept mapping studies suggests that between 20 and 30 sorters ensures the consistency of fit in the concept map [23], i.e. data reaches a point of saturation above this number. The same study found a moderate correlation between the number of raters and inter-rater reliability, with average responses of 82 and 66 for the first and second rating question respectively. Thus, while such our response is considered sufficient for sorting, it is presumably not sufficient for reliable rating data. Third, the response in the structuring phase was substantially less varied than in the statement generation phase: all respondents were from within the health sector, of whom most were researchers and few were health practitioners. Further, the number of health practitioners decreased disproportionately between the sorting and rating assignment (from 4 to 2). On the one hand this might have influenced the positioning and, to a larger extend, the rating of statements. On the other, a wide variety of concepts was already included in the data due to the wide variety of participants in de statement generation phase. Also because of the mixed composition of the expert panel that verified the cluster solution, we believe the final cluster solution to represent sufficient input from science, practice and policy. Fourth, although inevitable, it should be noted that the judgment of the researchers played some role in deciding on the final cluster solution and cluster labels. Finally, as the participants were all Dutch, the generalizability of the concept map to other countries could be limited. However, as the focus prompt did not direct the context, the statements, probably with one or two exceptions, seem to be generalizable to at least other Western democracies. 4.3. Conclusions The concept map gives an overview of interrelated themes, distinguishes core from peripheral dimensions, and provides reference points for developing a theory of the policy process. While low measurability ratings indicate that measurement development is certainly not a done deal, the core themes provide a first step to gain systematic insight into IPHP through measurement. Ultimately, we propose that the concept map is not an ‘end state’ and should be seen as a ‘living organism’ that can evolve as knowledge is advancing.




Conflict of interest statement None declared.

[12] [13]

Acknowledgements The authors thank the participants of the concept mapping procedure for their valuable contributions and the members of the Instrument for Integrated action (i4i) Consortium for bringing forward their ideas about our research. This research was funded by The Netherlands Organization for Health Research and Development (ZONMW, Grant no. E271534).

[14]

[15] [16] [17]

References [18] [1] Arredondo A, Orozco E. Application of the ecohealth model to translate knowledge into action in the health sciences. Environmental Health Perspectives 2012;120:104–5. [2] Bambra C, Fox D, Scott-Samuel A. A politics of health glossary. Journal of Epidemiology and Community Health 2007;61:571–4. [3] Bekker M. The politics of healthy policies: redesigning health impact assessment to integrate health in public policy. Rotterdam: Erasmus University; 2007. [4] Breton E, De Leeuw E. Theories of the policy process in health promotion research: a review. Health Promotion International 2010;26:82–90. [5] Breton E, Richard L, Gagnon F, Jacques M, Bergeron P. Health promotion research and practice require sound policy analysis models: the case of Quebec’s Tobacco Act. Social Science & Medicine 2008;67:1679–89. [6] Christensen T, Lægreid P. The whole-of-government approach: regulation, performance, and public-sector reform. Bergen: Stein Rokkan Centre for Social Studies; 2006. [7] De Leeuw E. Policies for health: the effectiveness of their development, adoption, and implementation. In: McQueen D, Jones C, editors. Global perspectives on health promotion effectiveness. 2007. [8] De Leeuw E, Clavier C. Healthy public in all policies. Health Promotion International 2011;26:ii237–44. [9] Freiler A, Muntaner C, Shankardass K, Mah CL, Molnar A, Renahy E, et al. Glossary for the implementation of Health in All Policies (HiAP). Journal of Epidemiology and Community Health 2013;67:1068–72. [10] Hendriks A-M, Habraken J, Jansen MWJ, Gubbels JS, De Vries NK, van Oers H, et al. ‘Are we there yet?’—operationalizing the concept of integrated public health policies. Health Policy 2014;114:174–82. [11] Hoeijmakers M. Local health policy development processes: health promotion and network perspectives on local health policy-making

[19]

[20] [21] [22]

[23]

[24] [25]

[26]

[27]

[28]

[29]

11

in The Netherlands. Maastricht: University Maastricht; 2005. Kane M, Trochim WM. Concept mapping for planning and evaluation. Thousand Oaks, CA: Sage Publications; 2007. Kickbusch I, Gleicher D. Governance for health in the 21st century. Bakoe, Azerbaijan: World Health Organization, Regional Office for Europe; 2011. Koivusalo M. The state of Health in All policies (HiAP) in the European Union: potential and pitfalls. Journal of Epidemiology & Community Health 2010;64:500–3. Kruskal JB, Wish M. Multidimensional scaling. Beverly Hills, CA: Sage; 1978. Lalonde M. A new perspective on the health of Canadians. Ottawa, Canada: Information Canada; 1974. Lin V, Jones C, Synnot A, Wismar M. Synthesizing the evidence: how governance structures can trigger governance actions to support health in all policies. In: McQueen D, Wismar M, Lin V, Jones C, Davies M, editors. Intersectoral governance for health in all policies. Copenhagen, Denmark: World Health Organization; 2012. Malhotra K, Heiman HJ. Public health policy is political. American Journal of Public Health 2012;102:e1. McQueen DV, Wismar M, Lin V, Jones C, Davies M. Intersectoral governance for health in all policies: structures, actions and experiences. Copenhagen, Denmark: World Health Organization; 2012. Milio N. Public policy as the cornerstone for a new public health: local and global beginnings. Family & Community Health 1988;11:57–71. Milio N. Glossary: Healthy public policy. Journal of Epidemiology and Community Health 2001;55:622–3. Peters D, Harting J, Oers vH, Schuit A, Vries dN, Stronks K. Manifestations of integrated public health policy in Dutch municipalities. Health Promotion International 2014. Rosas SR, Kane M. Quality and rigor of the concept mapping methodology: a pooled study analysis. Evaluation and Program Planning 2012;35:236–45. Scott-Samuel A. Health impact assessment. British Medical Journal 1996;313:183. Smith KE, Katikireddi SV. A glossary of theories for understanding policymaking. Journal of Epidemiology and Community Health 2013;67:198–202. Ståhl T, Wismar M, Ollila E, Lahtinen E, Leppo K. Health in all policies: prospects and potentials. Helsinki: Finnish Ministry of Social Affairs and Health; 2006. Trochim WM, Cabrera D. The complexity of concept mapping for policy analysis. Emergence: Complexity and Organization 2005;7:11–22. Trochim WM, Cabrera DA, Milstein B, Gallagher RS, Leischow SJ. Practical challenges of systems thinking and modeling in public health. American Journal of Public Health 2006;96:538–47. WHO. Intersectoral action for health: a cornerstone for health-for-all in the in twenty-first century. In: Halifax, Nova Scotia, Canada: Report of the International Conference 20–23 April 1997. WHO; 1997.


Manifestations of integrated public health policy in Dutch municipalities.

Values in Health Policy - A Concept Analysis.

Epidemiologic science and public health policy.

Advancing One Health Policy and Implementation Through the Concept of One Medicine One Science.

Basic science and public policy.

Climate change and public health policy: translating the science.

Framing the ecosystem concept through a longitudinal study of developments in science and policy.

Law and policy on the concept of bullying at school.

Public policy and health manpower.

Letter: Private practice and public policy.

Public health week: marketing the concept of public health.

Policy, Research and Residents' Perspectives on Built Environments Implicated in Heart Disease: A Concept Mapping Approach.

Swedish public health policy: Impact on regional and local public health practice and priorities.

The Oregon Public Health Policy Institute: Building Competencies for Public Health Practice.

The "child size medicines" concept: policy provisions in Uganda.

Health Policy and Management: in praise of political science. Comment on "On Health Policy and Management (HPAM): mind the theory-policy-practice gap".

Subclinical hypothyroidism and pregnancy: the intersection of science, the art of medicine, and public health policy.

Public health policy for the 1990s.

Prioritizing Sleep Health: Public Health Policy Recommendations.

Public participation in the process of local public health policy, using policy network analysis.

The paradox of health policy: revealing the true colours of this 'chameleon concept' Comment on "The politics and analytics of health policy".

Community mental health: The evolution of a concept in social policy.

Public health and international drug policy.

A model for training public health workers in health policy: the Nebraska Health Policy Academy.