National health behavior guidelines offer evidence-informed recommendations for how to obtain health benefits and mitigate health risks.1 Guidelines exist for a variety of health behaviors and are useful for goal setting, developing public health messages, and monitoring individual behavior.2 Although guidelines are necessary for public health action, they are not often mobilized into practice or policy.3 Indeed, public awareness and knowledge of many national health behavior guidelines is low, and behavioral targets remain unmet.3,4 This knowledge-to-practice gap exists worldwide.5–11
One approach that holds promise for addressing this knowledge-to-practice gap is the development of partnerships with intermediary organizations (ie, named for the “mediating” role they play in moving evidence into practice12) to contribute to guideline development and dissemination (ie, the purposive distribution of information to target audiences13). Indeed, as the dissemination of national health behavior guidelines is typically underresourced,3 working collaboratively with intermediary organizations at the outset of guideline development may be an efficient, low-cost strategy to increase the reach of guidelines to target audiences, and improve the adoption of guidelines across professional networks.12 Moreover, intersectoral organization networks are often well-positioned to address complex public health concerns, such as the adoption of national health behavior guidelines.14 Despite the potential for interorganization networks to be harnessed for knowledge mobilization (KMb), little is known about the degree to which intermediary organizations work collaboratively to influence public health outcomes, and there is limited guidance for how the features (ie, social structure and organization characteristics or attributes) of interorganization networks can be used to support KMb activities (ie, synthesis, dissemination, exchange, and application15) more broadly.
Contemporary conceptualizations of KMb describe the “gap” between knowledge and practice as an “evidence ecosystem, living, interacting, and evolving with intersecting, and, overlapping boundaries and roles.”14 From this perspective, successful KMb is reliant on intersectoral relationships (ie, networks) and a mutual understanding of the benefits of the knowledge that is being shared.14 Social network analysis (SNA) is an empirical approach that may be used to explore how the relationships among individuals and/or organizations influence a given outcome.16 Applied in a KMb context, SNA offers valuable information about how social networks, and the characteristics of actors within them, influence KMb.17 Practically, this information may be used to inform KMb activities within a network, such as the design of network-altering interventions (eg, to address network gaps or holes), to improve important KMb outcomes. Accordingly, the overarching purpose of this study was to apply SNA to explore how the features of interorganization networks influence the dissemination of national health behavior guidelines. Specifically, the objectives of this study were to (1) examine the connectedness of organizations and/or subgroups within a national health behavior guideline network and (2) identify organization characteristics or attributes (eg, sector and size) associated with influential network positions (eg, highly central or visible positions within a network). Such an approach will deepen the understanding of existing connections among intermediary organizations that can facilitate or impede the dissemination of national health behavior guidelines and add to the body of literature exploring how SNA may be used to bolster KMb activities more broadly.
Context
In October 2020, Canada’s Physical Activity Guidelines were replaced with the Canadian 24-Hour Movement Guidelines for Adults aged 18–64 years and Adults aged 65 years or older (24HMG)18. The 24HMG include recommendations for all intensities of physical activity, sedentary behavior, and sleep with a focus on the compositional impact of movement behaviors over a 24-hour period.
In recognition of the diverse expertise required to develop and integrate guidelines into practice, 2 groups—a Consensus Panel and a KMb Advisory Committee—were established at the outset of guideline development.19 The Consensus Panel provided oversight and strategic direction to the development of the 24HMG, whereas the KMb Advisory Committee spearheaded the KMb efforts for the 24HMG (eg, formative research, strategy selection, intervention design, and evaluation). The Consensus Panel was comprised of content experts (ie, physical activity, sedentary behavior, and sleep researchers), representatives from each of the funding partners, target audiences (ie, health professionals and members from the general public), international consultants, and methodology consultants.18 The KMb Advisory Committee consisted of content experts (ie, KMb researchers); representatives from each of the funding partners; intermediary organizations that promote physical activity, sedentary behavior, or sleep; intermediary organizations with ties to various guideline target audiences (eg, postsecondary students, health professionals, and adults 65 y or older); target audiences; a KMb methodologist; and student trainees.19 Collectively, the Consensus Panel and KMb Advisory Committee comprised the “immediate network” through which the 24HMG were disseminated. Intermediary organizations who were professionally connected to members of the “immediate network” are described as the “extended network.” Together, the “immediate network” and the “extended network” form the “24HMG network.”
Methods
Study Design
This work is both pragmatic and exploratory in nature. Philosophically, this study is grounded in critical realism, which combines a realist ontology with a constructivist/interpretivist epistemology.20 Realist ontologies are premised on the existence of a single reality that is external to the individual while acknowledging that this reality may never be fully known or understood.21 Constructivist/interpretivist epistemologies assume that participant experiences cannot be separated from the research process, and meanings derived from research are coconstructed through transactions between the participant and the research team.20 Accordingly, it is important to acknowledge that members of the research team (Tomasone, Latimer-Cheung, and Faulkner) held leadership positions within the immediate network, which may have influenced study participation and the interpretation of study results.
A partial network design (ie, the exploration of connections among an initial set of respondents and their nominated partners22) was used to examine the relations among organizations in Canada that have the potential to disseminate the 24HMG. A snowball sampling approach was used to identify actors in the network.22 For feasibility, this study relied on a realist approach (ie, actors self-identify the network boundaries by reporting their connections with other actors)23 to set the network boundary wherein organizations within 2 sampling waves of the immediate network were included in data collection and analysis.24 Further, due to the influence that each additional nominee placed on data collection, responding participants were requested to nominate up to 20 organizations within their organization’s network with the potential to disseminate the 24HMG.25
Procedure
Following ethics approval, eligible (ie, national) members of the immediate network were invited to participate in this study in October 2019 (ie, 1 y before the release of the 24HMG). Members were invited by email to participate in an online survey (Supplementary Material [available online]) on behalf of their organization. Participants were asked to provide demographic information about themselves (eg, history with the organization) as well as attributes of the organization that they were representing (eg, location, level, sector, and size). Next, participants were asked to list other organizations within their organization’s professional network that have disseminated or have the potential to disseminate national-level guidelines pertaining to physical activity, sedentary behavior, and/or sleep in Canada. For each organization identified, participants were requested to provide contact information for an individual working within the nominated organization and the movement behaviors (eg, physical activity, sedentary behavior, or sleep) promoted by each organization. Last, participants were requested to identify whether their connection to each nominated organization was established prior to, or as a result of, their involvement in the 24HMG initiative. Note that in some cases, members of the immediate network were affiliated with more than one organization. These individuals were requested to complete the survey multiple times (ie, once for each organization) or to nominate a second individual within their organization who could complete the survey.
Organizations nominated by participants in the immediate network (ie, extended network) were contacted to participate in a second online survey in January 2020. This survey contained similar items to the survey distributed to the immediate network. However, members of the extended network were not asked to indicate whether their connections with other organizations were formed through their involvement with the 24HMG, as they were not directly involved in the 24HMG initiative (ie, the Consensus Panel or KMb Advisory Committee). To maximize study participation, follow-up reminders were sent to nonresponding participants in the immediate and extended networks 2 weeks following the first recruitment email. In cases where a contact was no longer employed with the nominated organization, the first author attempted to identify and contact an alternate employee within the nominated organization. Informed consent was obtained from all participants included in this study.
Analysis
Frequencies and descriptive statistics were calculated for participant demographics and organization attributes using IBM SPSS Statistics, version 26.0.26 In cases where actors in the immediate network represented the same organization, their responses were collapsed, as all ties were reported at an organizational level. Organizations nominated to be included in the 24HMG network that were located outside of Canada were excluded from analysis.
To examine network structure and connectedness, survey data were exported into Excel,27 and information pertaining to organizational ties was converted into adjacency matrices, wherein a “one” represented a tie between 2 organizations and a “zero” represented the absence of a tie. Organizations nominated by actors in the immediate or extended network were included in data analysis even if they did not participate in data collection. Matrices were generated for the 24HMG (ie, all organizations in the immediate and extended networks) and responding networks (ie, responding organizations in the immediate and extended networks) and separate physical activity, sedentary, and sleep behavior subnetworks (ie, all organizations in the immediate and extended networks that were described as promoting physical activity, sedentary, or sleep behavior, respectively). The network analysis was performed using UCINET (version 6).28
Network-level measures (ie, analyses that occur at the structural level that provide information about the network as a collective unit or whole, such as the size and density of a network24) were calculated to examine the connectedness of organizations within each network (ie, objective 1). Whereas, node-level measures (ie, analyses that occur at the node-level that provide information about the position of actors within a network, such as their centrality or closeness to other actors24) were calculated to assess network position (ie, objective 2) for organizations within the responding network only, as centrality measures are influenced by the number of responders within a network. Both network- (ie, whole network centralization and density, fragmentation, reciprocity, and isolates) and node-level (ie, in-degree, out-degree, betweenness, in-closeness, and out-closeness centralities) measures were included in analysis. As published thresholds for SNA measures are not available, the value obtained for each network- and node-level measure was interpreted relative to the network generated. NETDRAW software was used to generate network maps to visualize network structures. Table 1 provides a summary of terms and measures used in SNA and their implications for national health behavior guideline KMb.
SNA Terms, Definitions, Interpretations, and Their Implications for KMb and National Health Behavior Guideline Dissemination
| SNA term | Definition | Score interpretation | Implication for KMb | Implication for guideline dissemination |
|---|---|---|---|---|
| Network | An interconnected group of actors (eg, people and organizations). | None. | Provides the social context within which KMb occurs. | Provides the social context within which guideline dissemination occurs. |
| Actor | A point (node) in a network map that represents an individual, organization, or entity connected to other actors (through ties). | None. | Represents the actors involved in KMb processes. | Represents the actors involved in the process of guideline dissemination. |
| Tie | The relations or connections among actors in the network. | None. | Represents the interactions, collaborations, or relationships involved in KMb. | Represents the relationships between actors involved in the process of guideline dissemination. |
| Whole network centralization | The extent to which interconnections are unequal across the network (ie, concentrated around one or more central individuals). | The actual sum of differences in centrality in a network with n actors divided by the maximum value of the sum of differences in centrality in a network with n actors. | Thought to enhance ease of knowledge sharing and to promote standard practices of existing protocols. | Thought to enhance the ease of guideline dissemination within the dissemination network. |
| Whole network density | An index of the proportion of existing ties relative to all possible ties in a network. | Total number of ties divided by the total number of possible ties in a network. | Proxy for efficiency of information flow, solidarity, or cohesiveness within a network. | Proxy for dissemination network cohesiveness (a dense network reflects connectivity between actors in the guideline dissemination network). |
| Fragmentation,a,24 | An inverse measure of the amount of connectedness or redundancy in a network. | The proportion of mutually reachable actors as each actor is removed from the network. | Proxy for the efficiency of information flow within a network. | Proxy for the efficiency of guideline dissemination within a network (eg, a higher score indicates less efficiency). |
| Reciprocity,a,24 | Assesses the proportion of mutual ties within the network. | The total number of mutual ties divided by the total number of possible mutual ties in a network. | Proxy for the strength of relationships between network actors. | Proxy for the strength of relationships between actors within the guideline dissemination network or miscommunication among actors. |
| Degree centralityb | In-degree centrality: Number of individuals who send (identify) ties to an actor. Out-degree centrality: Number of direct ties an actor sends (identifies) to others. | The sum of ties sent to an actor by others in the network. The sum of ties an actor sends to others in the network. | A directional measure that can be considered an index of importance, power, or influence. A directional measure that may quantify access to resources through colleagues, exposure to evidence, and others’ practices. | An indicator of importance, power, or influence in the guideline dissemination network. Actors with a high in-degree centrality may be considered champions or opinion leaders in the guideline dissemination network. An indicator of reach within the guideline dissemination network. Actors with a high out-degree centrality may be strong guideline disseminators. |
| Betweenness centralityb | The extent to which an individual is tied/connected to others who are not connected themselves. | Number of times an actor connects pairs of other actors. | A measure that may act as a proxy for control of KMb processes; high values reflect a favorable position for information flow. | A measure that may act as an indicator of a favorable position for guideline dissemination. Actors with a high betweenness centrality may be leveraged to reach organizations that are not connected to others in the guideline dissemination network. |
| Closeness centralityb | In-closeness centrality: Proportion of actors that can be reached in one or more steps. Out-closeness centrality: Proportion of actors that can reach others in one or more steps. | The shortest path between all actors is calculated. A score is assigned to each actor based on its sum of shortest paths. | A directional measure that may act as a proxy for the degree of access to information or efficiency in communicating with the network (relative reach). | A directional measure that may act as a proxy measure for the efficiency of guideline dissemination within the network. |
| Components/isolates | Portions of the network that contain actors connected to one another but disconnected from actors of other subgroups. | Number of actors in a network that are not connected to any others. | Subgroups and isolates can be targeted to increase connectedness, share information, or mobilize action. | Subgroups or isolates can be targeted to increase guideline dissemination network connectedness. |
Last, the association of organization attributes and network position (ie, objective 2) was explored using Point-Biserial and Spearman rank correlation coefficients for nominal, and, ordinal variables with more than 5 levels, respectively.30
Results
Seventeen participants representing 14 organizations in the immediate network participated in the first online survey (response rate of 63.64%). Of note, 4 connections were reported as being formed through participation in the 24HMG initiative. Participants from the immediate network nominated 71 organizations to be included in the extended network. Of the 71 organizations in the extended network, 20 participants completed the second online survey (response rate of 28.17%). Combined, 20 participants representing 20 organizations in the extended network nominated 96 organizations that were described as having the potential to disseminate movement behavior guidelines in Canada. Across the 2 sampling waves, a total of 189 unique organizations were identified as belonging to the 24HMG network. Three organizations were excluded as they were located outside of Canada, resulting in a total of 186 unique organizations. See Figure 1 for an overview of participant and organization flow during data collection.
—Participant and organization flow diagram. 24HMG indicates 24-Hour Movement Guidelines for Adults; n, participants; k, organizations.
Citation: Journal of Physical Activity and Health 22, 4; 10.1123/jpah.2024-0337
Responding participants reported working for an average of 10 (7.76) years within their respective organizations. Participants were employed in a range of positions including directors, policy advisors, clinical research managers, professors, program officers, and managers of marketing and communications. Most participants described their organizations as not-for-profit organizations (70.6%) that promote physical activity (88.2%) and are located in the province of Ontario (61.8%). See Table 2 for a summary of organization attributes for responding organizations (ie, responding network).
Attributes of Responding Organizations (k = 34)
| Organization attribute | Frequency (%) |
|---|---|
| Province | |
| Alberta | 4 (11.7) |
| British Columbia | 4 (11.7) |
| Ontario | 21 (61.8) |
| Prince Edward Island | 2 (5.9) |
| Quebec | 1 (2.9) |
| Saskatchewan | 1 (2.9) |
| Yukon | 1 (2.9) |
| Level | |
| National | 14 (41.2) |
| Provincial/territorial | 15 (44.1) |
| Local | 5 (14.7) |
| Sector | |
| Government | 2 (5.9) |
| Not-for-profit | 24 (70.6) |
| Education | 6 (17.6) |
| Other | 2 (5.9) |
| Number of full-time employees | |
| 0 | 2 (5.9) |
| <10 | 10 (29.4) |
| 10–19 | 5 (14.7) |
| 20–29 | 5 (14.7) |
| 30–39 | 0 (0) |
| 40+ | 9 (26.5) |
| Not collecteda | 3 (8.8) |
| Number of part-time employees | |
| 0 | 2 (5.9) |
| <10 | 18 (52.9) |
| 10–19 | 3 (8.8) |
| 20–29 | 0 (0) |
| 30–39 | 0 (0) |
| 40+ | 8 (23.5) |
| Not collecteda | 3 (8.8) |
| Volunteers | |
| 0 | 12 (35.3) |
| <10 | 2 (5.9) |
| 10–19 | 4 (11.8) |
| 20–29 | 2 (5.9) |
| 30–39 | 1 (2.9) |
| 40+ | 10 (29.4) |
| Not collecteda | 3 (8.8) |
| History of movement behavior promotion | |
| <5 y | 2 (5.9) |
| 5–10 y | 6 (17.6) |
| 11–15 y | 4 (11.8) |
| 16–20 y | 6 (17.6) |
| 20+ | 16 (47.1) |
| Primary movement behavior(s) promoted | |
| Physical activity | 30 (88.2) |
| Sedentary behavior | 12 (35.3) |
| Sleep | 5 (14.7) |
| Resources allocated to primary movement behavior | |
| <20% | 12 (35.3) |
| 20%–39% | 0 (0) |
| 40%–59% | 4 (11.8) |
| 60%–79% | 2 (5.9) |
| 80%–100% | 13 (38.2) |
| Not collecteda | 3 (8.8) |
| Secondary movement behavior(s) promoted | |
| Physical activity | 2 (5.9) |
| Sedentary behavior | 9 (26.5) |
| Sleep | 7 (20.6) |
| Other | 18 (52.9) |
| Previous movement behavior guideline dissemination | |
| Physical activity | 28 (82.4) |
| Sedentary behavior | 15 (44.1) |
| Sleep | 11 (32.4) |
| None | 4 (11.8) |
Abbreviation: k, organization. Note: In cases where k > 34, participants selected multiple responses for their organization.
aContent experts were unable to answer item on behalf of their organization.
Objective 1: Network Structures (All Organizations)
Figure 2 depicts the network structure of each of the 5 networks explored in this study. Overall, a lack of cohesion was observed across each network. The centralization (12.9%–38.6%) and density (1%–5%) scores suggest a lack of connectedness and possible constrained information sharing between organizations (see Table 3). The fragmentation scores for 4 of the 5 networks ranged from 73% to 95%, suggesting that many of the organizations belonging to each network cannot be easily reached (ie, information cannot easily be transferred among organizations). Interestingly, the sedentary behavior network had a fragmentation score of 7%, suggesting that the proportion of organizations that could either directly or indirectly reach each other was relatively high. Last, the reciprocity value for the responding network was 21%, indicating that many of the ties reported between organizations were not mutual (ie, they were initiated by 1 organization). As such, there is a higher potential for miscommunication or ineffective communication among organizations belonging to the responding network.
—Sociograms depicting the network structure of the 5 dissemination networks. Note. The sociograms display the nodes that represent organizations (designated using letters) in each guideline dissemination network. Lines connecting nodes indicate a connection or tie between organizations. Nodes (ie, organizations) without connections to the network are isolates (upper left corner). Distances between nodes (ie, line length) are not meaningful; however, the size of each node is proportional to centrality (ie, larger node depicts a higher centrality score). Organization attributes for the 24HMG Network and the Responding Network are color coded and best viewed in a digital format. 24HMG indicates 24-Hour Movement Guidelines for Adults. (Color figure online).
Citation: Journal of Physical Activity and Health 22, 4; 10.1123/jpah.2024-0337
Network Measures Across 5 Dissemination Networks
| Network measure | 24HMG network | Responding network | Physical activity network | Sedentary behavior network | Sleep network |
|---|---|---|---|---|---|
| Number of nodes | 186 | 34 | 115 | 90 | 28 |
| Number of ties | 228 | 57 | 158 | 117 | 21 |
| Whole network centralization | 13% | 39% | 20% | 24% | 22% |
| Whole network density | 1% | 5% | 1% | 2% | 3% |
| Fragmentation | 95% | 73% | 91% | 7% | 95% |
| Reciprocity | — | 21% | — | — | — |
| Isolates | 11 | 3 | 5 | 10 | 12 |
Abbreviation: 24HMG, 24-Hour Movement Guidelines for Adults. Note: Reciprocity was calculated for the responding network only as the value is influenced by the number of responders in a network.
Objective 2: Network Position and Organization Attributes (Responding Organizations)
Degree Centrality
In-degree centrality (ie, the number of times an actor is identified as having a tie to other actors in a network) scores for organizations in the responding network (k = 34) ranged from 0 to 9. On average, organizations were identified as having 2 ties to other organizations (SD = 2.11). Three organizations had high in-degree centrality scores (5–9) relative to other organizations in the network. These organizations may be considered champions or opinion leaders within the responding network. Out-degree centrality (ie, the number of ties an actor identifies as having to other actors in the network) scores ranged from 0 to 9. Participants reported, on average, that their organization was professionally connected to 2 other organizations in the responding network (SD = 2.32). Six organizations had high out-degree centrality scores (5–9) relative to other organizations in the network, indicating that these organizations have a high reach within the network and may be strong disseminators of information.
Betweenness Centrality
The range of betweenness centrality scores was wide (0–151). Nine organizations had a betweenness centrality score of >1, suggesting that these organizations may be potential gatekeepers of information and may be leveraged to reach other organizations that are not well-connected in the responding network.
Closeness Centrality
Little variability for in- (0.17–0.26) and out-closeness centrality scores (0.17–0.29) was observed. Organizations with higher in-closeness centrality scores can quickly be reached within the responding network. Organizations with higher out-closeness centrality scores may be efficient disseminators of information within the network. All centrality measures for the responding network are summarized in Table 4.
Centrality Measures for Responding Network
| Centrality measure | Mean score (SD) | Minimum score | Maximum score |
|---|---|---|---|
| In-degree | 1.68 (2.11) | 0 | 9 |
| Out-degree | 1.68 (2.32) | 0 | 9 |
| Betweenness | 11.5 (32.93) | 0 | 150.8 |
| In-closeness | 0.21 (0.03) | 0.17 | 0.26 |
| Out-closeness | 0.21 (0.05) | 0.17 | 0.29 |
Last, several statistically significant relationships were identified between organization attributes and network position (ie, centrality measures). These findings are summarized in Table 5. First, a negative relationship was observed between organizations located in the province of British Columbia and in-closeness centrality, rpb(32) = −.35, P = .04. Next, a negative relationship was observed between organizations belonging to the education sector and in-closeness centrality, rpb(32) = −.37, P = .03, whereas a positive relationship was observed with out-closeness centrality, rpb(32) = .34, P = .05. Regarding organization size, a positive correlation was observed between organizations with part-time employees and out-degree centrality, rs(29) = .38, P = .03, and out-closeness centrality, rs(29) = .42, P = .02. Further, a positive correlation was found between the percentage of resources allocated to movement behavior promotion and betweenness centrality, rs(29) = .40, P = .03. Last, a positive relationship was observed between organizations that had previously disseminated sedentary behavior guidelines and in-closeness centrality, rpb(32) = .42, P = .01. No other statistically significant relationships were noted.
Correlations of Organizational Attributes With Centrality Measures for the Responding Network
| Attribute | In-degree centrality | Out-degree centrality | Betweenness | In-closeness | Out-closeness |
|---|---|---|---|---|---|
| Full timea | −.21 | .19 | −.06 | −.20 | .21 |
| Part timea | −.21 | .38* | .11 | −.27 | .42* |
| Resources allocated to primary movement behaviora | .30 | .08 | .40* | .26 | −.01 |
| Volunteersa | −.16 | .00 | −.08 | −.11 | .03 |
| History of movement behavior promotiona | .18 | −.08 | .06 | .22 | −.12 |
| Provinceb | |||||
| Alberta | .06 | .01 | .10 | .13 | −.12 |
| British Columbia | −.08 | .09 | −.13 | −.35* | .12 |
| Ontario | .14 | .05 | .12 | .15 | .10 |
| Prince Edward Island | −.20 | −.07 | −.09 | −.29 | −.03 |
| Quebec | −.06 | −.13 | −.06 | .16 | −.16 |
| Saskatchewan | −.06 | .03 | −.06 | .11 | .09 |
| Yukon | .03 | −.13 | −.06 | .13 | −.16 |
| Levelb | |||||
| Local | −.10 | −.01 | −.09 | −.16 | −.18 |
| Provincial/territorial | −.26 | −.08 | −.15 | −.18 | .09 |
| National | .33 | .09 | .22 | .29 | .04 |
| Sectorb | |||||
| Not-for-profit | .21 | .11 | .23 | .19 | −.12 |
| Government | .10 | −.13 | −.08 | .26 | −.08 |
| Education | −.30 | .07 | −.16 | −.37* | .34* |
| Behavior promotedb | |||||
| Physical activity | .08 | .19 | .13 | −.07 | .19 |
| Sedentary behavior | .09 | .16 | .08 | .15 | .18 |
| Sleep | −.10 | .13 | −.08 | .08 | .16 |
| Previous disseminationb | |||||
| Physical activity guidelines | .19 | .14 | .16 | −.05 | .10 |
| Sedentary behavior guidelines | .17 | .02 | .09 | .42* | .06 |
| Sleep guidelines | −.10 | −.18 | −.20 | .15 | −.19 |
| None | −.16 | −.03 | −.13 | −.03 | .05 |
aSpearman rank correlation. bPoint-Biserial correlation.
*Correlation is significant at the .05 level (2-tailed).
Discussion
The overarching purpose of this study was to examine network properties that may influence the dissemination of national health behavior guidelines across interorganization networks. Several network measures examined in this study demonstrated the potential for inefficient and constrained information sharing among each network. Overall, findings highlight that the consideration of the relational factors (eg, network structure or connectedness) that influence guideline dissemination is warranted to support the meaningful and sustained KMb of future national health behavior guidelines.
The centralization and density scores for each dissemination network ranged from 12.9% to 38.6% and 1% to 5%, respectively. For reference, the density scores reported in a recent scoping review by Glegg et al17 ranged from 0.5% to 31%. These data suggest that knowledge cannot be shared easily or efficiently among organizations belonging to or connected to the 24HMG initiative. Importantly, these findings were consistent across each network, suggesting that the connections between organizations do not vary greatly across movement behavior domains (ie, physical activity, sedentary, and sleep behavior). While this result may be a function of low response rates, it may also suggest that network-altering interventions31 may be needed if intermediary organization networks are to play a larger role in the dissemination of future movement behavior guidelines. Future guideline initiatives may benefit from using community-building strategies (eg, search conferences32 and learning collaboratives31) to foster new connections and partnerships among organizations within the 24HMG network and improve the capacity for guideline dissemination among intermediary organizations.31 Taken together, it is unsurprising that previous guideline dissemination efforts in Canada have had a limited impact on practice and policy to date, resulting in poor awareness, knowledge, and adoption of the guidelines among target audiences.5 Within KMb science, there is a growing recognition of the value of intersectoral relationships (eg, social networks) for mobilizing knowledge across contexts.14 Practically, these findings provide support for the consideration of the social mechanisms that underpin KMb outcomes.
At an organizational level, this study identified several attributes that were significantly associated with network positions (ie, location, sector, size, resource allocation, and previous dissemination of sedentary behavior guidelines). First, a moderate negative relationship was observed between in-closeness centrality and organizations located in the province of British Columbia, implying that organizations located in British Columbia cannot be efficiently reached within the existing network. Although national health behavior guideline dissemination is typically performed on a national scale, it is possible that the overrepresentation of organizations in Ontario in the immediate network (ie, many national movement behavior-promoting organizations are based in Ontario) and the free recall approach used in this study led to a negative correlation observed between these variables. Further, it is important to note that although a negative relationship was identified between in-closeness centrality and organizations located in the province of British Columbia, there were several provinces in Canada that were not represented by organizations in the 24HMG network (ie, Manitoba, New Brunswick, Newfoundland and Labrador, Nova Scotia, Northwest Territories, and Nunavut). Next, a moderate negative relationship was observed between organizations belonging to the education sector and in-closeness centrality, whereas a moderate positive relationship was observed with out-closeness centrality. This finding suggests that organizations in the education sector (ie, universities) may not be efficiently reached by other organizations within the existing network; however, they demonstrate a high potential to disseminate information efficiently within the network. Although responding participants may not view academic institutions as strong disseminators of information, guideline initiatives may benefit from leveraging the networks of academic institutions to disseminate future health behavior guidelines. Interestingly, a moderate positive relationship was observed between organizations with part-time employees and out-degree and out-closeness centralities. This finding may be attributed to the size of organizations that have a high number of individuals employed part time. For example, it is possible that larger organizations have a higher number of part-time employees, and thus a larger professional network to disseminate information more efficiently. Next, there was a moderate positive relationship between the percentage of resources allocated to movement behavior promotion and betweenness centrality. This finding suggests there may be merit to engaging organizations that allocate a high percentage of resources to movement behavior promotion to be involved in future movement behavior guideline KMb initiatives, as these organizations may act as “gatekeepers,” and have the potential to reach other organizations that are not well-connected to the network. Last, a moderate positive relationship was observed between organizations that had previously disseminated sedentary behavior guidelines and in-closeness centrality. Participants belonging to organizations that promote healthy sedentary behavior in Canada are also considered to be highly influential content experts in the field of physical activity and sedentary behavior globally. Accordingly, it is possible that these individuals and their organizations were more easily recalled by responding participants.
Strengths and Limitations
This study has numerous strengths. Practically, these findings provide insight into the network features that may have contributed to the blunted impact of previous health behavior guidelines and highlight gaps that need to be addressed if intermediary organizations are to play a larger role in facilitating successful guideline dissemination. Methodologically, this study highlights the utility of using SNA as a precursor to guide dissemination activities within a national interorganization network to improve dissemination outcomes and inform the allocation of resources. For example, these findings may be used by international guideline KMb teams to leverage organizations holding favorable network positions (ie, organizations with high in-degree, out-degree, and betweenness centrality) as champions, or knowledge brokers, to increase the reach or adoption of movement behavior guidelines across health-promoting organizations, in their respective countries. Such an approach aligns with current commentaries in KMb science regarding complexity that encourage KMb researchers to understand and work with the “system” within which KMb occurs rather than attempting to change it or push against it.14
Despite our valuable findings, some limitations must be acknowledged. First, we found it difficult to estimate authentic ties reported by organizations that participated in this study. Although our research team requested for emails to be provided by survey participants as a means of ensuring that reported connections between organizations were valid, there were some instances where no emails were included for nominated organizations. In these cases, the first author attempted to find an appropriate email for the individual listed, and when no email could be found, a general inquiries email for the organization was used. Next, the free recall approach used in this study may have omitted potentially relevant organizations that could contribute to guideline dissemination, as an average of 2 ties were reported per organization. Further, due to concerns related to feasibility, we asked participants to nominate up to 20 organizations within their professional network, which may have resulted in an underestimate of network size. Although the purpose of this study was not to generate an expansive network of health-promoting organizations in Canada with the potential to disseminate national movement behavior guidelines, it is possible that using a roster method (eg, providing participants with a list of organizations) for data collection might allow for more organizational ties within the network to be reported.22 Next, our research team found it challenging to discriminate between interindividual and interorganization networks. Although this study made efforts to engage individuals in communication positions for their respective organizations, it is possible that we collected data that is more reflective of individual connections rather than organizational connections. For example, as members of the immediate network were often champions or content experts within their fields of research, it is possible that if they left their organization, their influence and their connections in the interorganization network would follow them. It is important for future researchers engaging in this type of work to be aware of the fluidity inherent in the study of interorganization networks and to be strategic with their use of language during recruitment and data collection. Moreover, although this study provides insight into the social structure of the networks examined in this study, we did not assess the value or strength of ties between organizations. Social Support Theory suggests that information is disseminated more effectively between networks of actors with strong social ties (eg, those that collaborate frequently).33 As such, it is possible that tie strength may moderate KMb outcomes; this may be an important direction for future research in this area. Last, it is important to acknowledge that the small sample size and response rate (ie, 34 participating organizations out of 93 invited) of this study may limit the generalizability of these findings.
Conclusions
Despite the resources and time dedicated to the development and KMb of national health behavior guidelines, they continue to have a limited impact on practice and policy. These findings provide insight into the network features that may have contributed to the blunted impact of previous national health behavior guidelines while also demonstrating the utility of SNA for understanding and facilitating KMb across interorganization networks. With this information, researchers are more effectively positioned to design KMb interventions that lead to more meaningful and sustained guideline dissemination outcomes.
Acknowledgments
The authors would like to thank the members of the 24HMG network for their collaboration and contributions to this study and Dr. Alex Lithopoulos for his contributions to data analysis. Funding Source: This study was made possible through funding from the Public Health Agency of Canada, the Canadian Society for Exercise Physiology, and Queen’s University. The views of the funding agencies had no influence on the content or conduct of this study. Guy Faulkner was supported by a CIHR-PHAC Chair in Applied Public Health. Kaitlyn D. Kauffeldt was supported by a Social Sciences and Humanities Research Council Doctoral Fellowship.
References
- 1.↑
Gagliardi AR. Translating knowledge to practice: optimizing the use of guidelines. Epidemiol Psychiatr Sci. 2012;21(3):231–236. doi:
- 2.↑
Brawley L, Latimer A. Physical activity guides for Canadians: messaging strategies, realistic expectations for change, and evaluation. Appl Physiol Nutr Metab. 2007;32(suppl 2):S170–S184.
- 3.↑
Tomasone JR, Kauffeldt KD, Morgan T, et al. Dissemination and implementation of national physical activity, sedentary behaviour, and/or sleep guidelines among community-dwelling adults aged 18 years and older: a systematic scoping review and suggestions for future reporting and research. Appl Physiol Nutr Metab. 2020;45:S258–S283. doi:
- 4.↑
Bull FC, Al-Ansari SS, Biddle S, et al. World Health Organization 2020 guidelines on physical activity and sedentary behaviour. Br J Sports Med. 2020;54(24):1451–1462. doi:
- 5.↑
Dale L, LeBlanc A, Orr K, et al. Canadian physical activity guidelines for adults: are Canadians aware? Appl Physiol Nutr Metab. 2016;41(9):1008–1011. doi:
- 6.
Knox M, Taylor IM, Biddle SJ, Sherar LB. Awareness of moderate-to-vigorous physical activity: can information on guidelines prevent overestimation? BMC Public Health. 2015;15(1):392. doi:
- 7.
Kay MC, Carroll DD, Carlson SA, Fulton JE. Awareness and knowledge of the 2008 physical activity guidelines for Americans. J Phys Act Health. 2014;11(4):693–698.
- 8.
National Health Survey. Chronic Obstructive Pulmonary Disease. Australian Bureau of Statistics; 2023. Accessed March 21, 2023. https://www.abs.gov.au/statistics/health/health-conditions-and-risks/chronic-obstructive-pulmonary-disease/latest-release
- 9.
Guthold R, Stevens GA, Riley LM, Bull FC. Worldwide trends in insufficient physical activity from 2001 to 2016: a pooled analysis of 358 population-based surveys with 1.9 million participants. The Lancet Global Health. 2018;6(10):e1077–1086. doi:
- 10.
Health Survey for England 2018 [NS]. NHS England. 2019. Accessed March 21, 2023. https://digital.nhs.uk/data-and-information/publications/statistical/health-survey-for-england/2018
- 11.↑
Ministry of Health NZ. Annual update of key results 2018/19: New Zealand health survey. 2019. Accessed March 21, 2023. https://www.health.govt.nz/publication/annual-update-key-results-2018-19-new-zealand-health-survey
- 12.↑
Latimer-Cheung AE, Copeland JL, Fowles J, Zehr L, Duggan M, Tremblay MS. The Canadian 24-hour movement guidelines for children and youth: implications for practitioners, professionals, and organizations. Appl Physiol Nutr Metab. 2016;41(6):S328–S335. doi:
- 13.↑
Rabin B, Brownson R, Haire-Joshu D, Kreuter M, Weaver N. A glossary for dissemination and implementation research in health. J Public Health Manag Pract. 2008;14(2):117–123. doi:
- 14.↑
Kitson A, O’Shea R, Brook A, et al. Using complexity and network concepts to inform healthcare knowledge translation. Int J Health Policy Manag. 2018;7(3):231–241. doi:
- 15.↑
Knowledge Translation at CIHR. CIHR. 2016. Accessed March 21, 2023. https://cihr-irsc.gc.ca/e/29418.html
- 17.↑
Glegg S, Jenkins E, Kothari A. How the study of networks informs knowledge translation and implementation: a scoping review. Implement Sci. 2019;14(1):879. doi:
- 18.↑
Ross R, Chaput J, Giangregorio L, et al. Canadian 24-hour movement guidelines for adults aged 18-64 years and adults aged 65 years or older: an integration of physical activity, sedentary behaiour, and sleep. Appl Physiol Nutr Metab. 2020;45:S57–S102. doi:
- 19.↑
Tomasone JR, Flood SM, Latimer-Cheung AE, et al. Knowledge translation of the Canadian 24-hour movement guidelines for adults aged 18-64 years and adults aged 65 years or older: a collaborative movement guideline knowledge translation process. Appl Physiol Nutr Metab. 2020;45(10):S103–S124. doi:
- 20.↑
Ryba TV, Wiltshire G, North J, Ronkainen N J. Developing mixed methods research in sport and exercise psychology: potential contributions of a critical realist perspective. Int Rev Sport Exerc Psychol. 2022;20(1):147–167. doi:
- 21.↑
Poucher ZA, Tamminen KA, Caron JG, Sweet SN. Thinking through and designing qualitative research studies: a focused mapping review of 30 years of qualitative research in sport psychology. Int Rev Sport Exerc Psychol. 2019;10:9858. doi:
- 22.↑
Morris M. Network Epidemiology: A Handbook for Survey Design and Data Collection. Oxford University Press; 2004:9–18.
- 23.↑
Pazitka V, Wojcik D. The network boundary specification problem in the global and world city research: investigation of the reliability of empirical results from sampled networks. J Geogr Syst. 2021;23:97–114. doi:
- 24.↑
Hawe P, Webster C, Shiell A. A glossary of terms for navigating the field of social network analysis. J Epidemiol Community Health. 2004;58(12):971–975. doi:
- 27.↑
Microsoft Corporation. Microsoft excel [Internet]. 2018. Accessed September, 2020. https://office.microsoft.com/excel
- 28.↑
Borgatti SP, Everett MG, Freeman LC. UCINET for Windows: Software for Social Network Analysis. Analytic Technologies, Inc; 2002.
- 31.↑
Bunger AC, Yousefi-Nooraie R, Warren K, Cao Q, Dadgostar P, Bustos TE. Developing a typology of network alteration strategies for implementation: a scoping review and iterative synthesis. Implement Sci. 2023;18:10. doi:
- 32.↑
Emery M, Purser R. The Search Conference: A Powerful Method for Planning Organizational Change and Community Action. Wiley; 1996:382.
- 33.↑
Slaughter J. Social Support: Theory, Research, and Intervention – by Alan Vaux. Praeger Publishers; 1988.
