Article Text

Download PDFPDF

A concept mapping approach to identifying the barriers to implementing an evidence-based sports injury prevention programme
Free
  1. Alex Donaldson1,2,3,
  2. Aisling Callaghan4,
  3. Mario Bizzini5,
  4. Andrew Jowett4,
  5. Patrick Keyzer2,
  6. Matthew Nicholson1
  1. 1 Centre for Sport and Social Impact, La Trobe University, Melbourne, Australia
  2. 2 Law School, La Trobe University, Melbourne, Australia
  3. 3 Australian Collaboration for Research into Injuries in Sport and their Prevention (ACRISP), Ballarat, Australia
  4. 4 Football Federation Victoria, Melbourne, Australia
  5. 5 Schulthess Clinic, Zurich, Switzerland
  1. Correspondence to Dr Alex Donaldson, Centre for Sport & Social Impact, La Trobe University, Bundoora 3086, Australia; a.donaldson{at}latrobe.edu.au

Abstract

Background and aim Understanding the barriers to programme use is important to facilitate implementation of injury prevention programmes in real-word settings. This study investigated the barriers to coaches of adolescent female soccer teams, in Victoria, Australia, implementing the evidence-based FIFA 11+ injury prevention programme.

Methods Concept mapping with data collected from 19 soccer coaches and administrators.

Results Brainstorming generated 65 statements as barriers to 11+ implementation. After the statements were synthesised and edited, participants sorted 59 statements into groups (mean, 6.2 groups; range, 3–10 groups). Multidimensional scaling and hierarchical cluster analysis identified a six-cluster solution: Lack of 11+ knowledge among coaches (15 statements), Lack of player enjoyment and engagement (14), Lack of link to football-related goals (11), Lack of facilities and resources (8), Lack of leadership (6) and Lack of time at training (5). Statements in the ‘Lack of 11+ knowledge among coaches’ cluster received the highest mean importance (3.67 out of 5) and feasibility for the Football Federation to address (3.20) rating. Statements in the ‘Lack of facilities and resources’ cluster received the lowest mean importance rating (2.23), while statements in the ‘Lack of time at training’ cluster received the lowest mean feasibility rating (2.19).

Conclusions A multistrategy, ecological approach to implementing the 11+—with specific attention paid to improving coach knowledge about the 11+ and how to implement it, linking the 11+ to the primary goal of soccer training, and organisational leadership—is required to improve the uptake of the 11+ among the targeted coaches.

  • barriers to programme implementation
  • sports injuries
  • adolescent female soccer players
  • sports coaches
  • concept mapping.

Statistics from Altmetric.com

Introduction

Implementation and injury prevention

Over a decade ago, the editor of Injury Prevention asked, “what needs to be done to ensure that solid research findings are put into practice?”,1 and Brussoni and colleagues lamented the lack of context-specific injury prevention programme implementation data.2 In 2010, McClure and colleagues suggested adopting an ecoepidemiological approach to injury prevention, incorporating a focus on the injury context and the upstream influences on programme implementation.3 Despite these and other similar calls,4 little guidance is available for researchers and practitioners, and few studies have investigated implementation issues specific to injury prevention.5

In the sports injury prevention field, Finch developed the Translating Research into Injury Prevention Practice model in 2006, with a stage focused specifically on understanding context-related implementation issues.6 However, only 1% of sports injury prevention publications focus on programme implementation and effectiveness.7 Since then, despite numerous calls to pay more attention to translating and implementing sports injury prevention research evidence into real-world practice,8 9 reporting on these issues remains inadequate.10 Nonetheless, a recent investigation reported that International Sports Federations consider implementation planning research to be important.11

Understanding what influences the use of evidence-based programmes in real-world settings is important in maximising the public health impact of such programmes.5 6 12 13 Although the barriers and facilitators to the use of sports injury prevention programmes have been described,14–17 it has frequently been as a secondary component of efficacy studies, often identifying barriers to participating in research rather than to programme use per se.17

Lower limb injuries and their prevention in female soccer

Worldwide, 30 million women play soccer, and over half of all female participants are under 17 years of age.18 Participating in soccer, like all physical activity, confers many health benefits on those who participate regularly.19 However, injuries do occur in soccer,20 and lower limb injuries are particularly common among female soccer players.21 Fédération Internationale de Football Association (FIFA), the international governing body for soccer, developed the 11+ warm up programme to prevent lower limb injuries in soccer players, and it is particularly effective with adolescent female soccer players.22 The 11+ has been widely promoted and programme resources and implementation guidelines are available.23 In addition, several national football federations integrate the 11+ into their coach education programmes, and about 5000 coaches from more than 40 countries have been instructed on how to deliver the 11+.23 Nonetheless, implementing the 11+ in community soccer clubs is an ongoing challenge.22

To facilitate the implementation of externally developed evidence-based programmes, it is vital to understand how programme implementers interpret and perceive such programmes, within the context of their real-world experience.24 This study aimed to explore the barriers to coaches of adolescent female soccer teams in Victoria, Australia, implementing the 11+. The information generated will be used to enhance the uptake and sustained high-fidelity use of the 11+ among the targeted coaches.

Methods

Concept mapping

This study used concept mapping (CM),25 a mixed-methods approach to generating and organising the ideas of a group of individuals with insight into a topic of interest. Previous implementation26 27 and injury-related28 studies have used CM, including to identify factors that influence the use of evidence-based practices.29 30

The key steps in CM are described by Kane and Trochim.25 How these steps were used in this study by the 11+ implementation project group (composed of the lead author (AD) and two Football Federation Victoria (FFV) representatives), is outlined below. The Concept Systems Global MAX31 web platform was used to brainstorm ideas, structure statements, analyse data and generate concept maps.

Step 1: preparation

Focus

The focus prompt used to generate ideas (Step 2) was: “A barrier to coaches of adolescent (12–18 year old) female football teams in Victoria implementing the FIFA 11+ injury prevention programme at training would be…”. The questions used to structure the statements (Step 3) were: “On a scale from 0 (least important) to 5 (most important), how important do you think this barrier is to coaches of adolescent female football teams implementing the FIFA 11+ programme at training with their players?”; and “On a scale of 0 (least feasible) to 5 (most feasible), how feasible do you think it is for FFV to address this barrier?”.

Participants

The FFV representatives identified 20 key individuals involved in soccer in Victoria who they believed were well informed about the project focus.

Scheduling and logistics

A FFV representative emailed potential participants inviting them to participate in the CM exercise. Those who agreed to participate were then emailed a link to the ‘generating the ideas’ process (Step 2) by the lead researcher (author AD). Two reminder emails were sent before this step was closed. All participants, even if they had not participated in Step 2, were emailed an invitation to participate in the Statement Structuring process (Step 3). Three reminder emails were sent before this step was closed. The CM data collection was conducted over 5 weeks in early 2017, immediately prior to the start of the 2017 community soccer season. All emails related to this project were accompanied by a Participant Information Sheet explaining the reasons for the study, who was funding the study and what was expected of study participants. Participants were required to provide informed consent to participate in the project the first time they registered on the Concept Systems Global MAX web platform.

Step 2: generating the ideas

Brainstorming

Participants were asked to write a brief (ie, one thought) statement to complete the focus prompt (see above). They could add multiple statements, review statements made by others and search the collected statements to see if their idea was already there.

Ideas analysis

The project steering group met to synthesise and edit the brainstormed statements to ensure all statements were relevant to the project focus, obtain a unique set of ideas with each idea represented once, and to improve the clarity and comprehension of all statements. This process was undertaken by (1) identifying and deleting statements that were not related to the focus prompt; (2) identifying and splitting compound statements that contained two or more distinct ideas; (3) grouping statements around key ideas, identifying duplicate statements that represented the same ideas and selecting the statement that was thought to best capture the essence of the idea; and (4) identifying and editing statements as required to ensure all members of the group understood the essential meaning of the statement. Discussion and editing continued until the three members of the project steering group reached consensus.

Step 3: structuring the statements

Demographics

Participants provided information about their gender, age, current role in soccer, years in current role and total years involved in soccer, before undertaking their first CM activity.

Statement sorting

Participants sorted the synthesised statements into piles in a way that made sense to them. They were instructed to group the statements according to how similar in meaning they were to one another and to name each group based on its theme or contents. They could put statements into piles on their own if they thought the statements were unrelated to all other statements. They were asked to put every statement somewhere and to not create piles such as ‘miscellaneous’ or ‘other’. They were informed that the number of piles people create varies, and that 5 to 15 piles usually works well to organise the number of statements they were asked to sort.

Statement rating

Participants rated each statement on ‘importance’ and ‘feasibility’. They were asked to use the full rating range (0–5) and to rate each barrier relative to the other barriers in the list, rather than as an absolute rating.

Step 4: concept mapping analysis

Multidimensional scaling analysis was used to locate each sorted statement as a separate point on a two-dimensional ‘point map’. Hierarchical cluster analysis was then used to partition the statements on the point map into clusters of related statements (‘cluster maps’). For a detailed description of the multidimensional scaling, including the stress index calculation, and hierarchical cluster analysis used in the Concept Systems Global MAX31 web platform, see Kane and Trochim (Concept Systems Incorporated pp. 87–100).

There is no universally correct number of clusters and no mathematical formula to select the most appropriate number of clusters. Therefore, the project group followed the guidance from Kane and Trochim (Concept Systems Incorporated pp. 101–103)31 by examining the cluster maps for a 10-cluster solution through to a 5-cluster solution, paying particular attention to which statements were merged together as the number of clusters was reduced. The aim was to find the cluster level that retained the most useful detail between clusters, while merging those clusters that seemed to logically belong together, based on the group’s knowledge of the community soccer context. Once the most appropriate cluster level was identified, if a statement on the map qualitatively seemed to belong in an adjacent cluster, it was reassigned from the original cluster to the neighbouring cluster to which it seemed a better conceptual fit.27

Importance and feasibility rating descriptive statistics were calculated and used to generate ‘Go-Zones’25 in which each statement’s mean rating was plotted on a graph. The resulting scatterplot was divided into four quadrants using the mean of each rating.

Results

Participants

Nineteen of the 20 invited participants contributed CM data—17 brainstormed ideas, 16 sorted statements, 16 rated importance and 14 rated feasibility. The characteristics of the 17 participants who provided both CM and demographic data are summarised in table 1.

Table 1

Characteristics of participants (n=17)

Ideas generated

Seventeen participants contributed 65 statements during the generating ideas step. After statement synthesising and editing, 59 statements were available for participants to sort and rate (table 2).

Table 2

Statements generated during the concept mapping brainstorming process including cluster in which the statement fits, mean importance and feasibility ratings, and Go-Zone quadrant for each statement

Statement sorting

Sixteen participants sorted the 59 statements in to groups (mean number of groups, 6.2; range, 3–10 groups). One participant left two statements unsorted, and two participants did not name the groups they created.

Statement rating

Fifteen participants rated all 59 statements for importance, while one rated 57 and one rated 56 statements for importance. Fourteen participants rated all 59 statements for feasibility. Statement 11 (‘Lack of adequate training of coaches and coach educators in injury prevention’) received the highest mean importance rating (4.12 out of 5). Statement 36 (‘Not trained in providing the exercises’) received the highest mean feasibility rating (4.21). Statement 34 (‘At least 50% of female players are there to socialise’) received the lowest mean importance (1.41) and feasibility (1.29) rating.

The statements in the ‘Lack of 11+ knowledge among coaches’ cluster received the highest mean importance (3.67) and feasibility (3.20) rating. The statements in the ‘Lack of facilities and resources’ cluster received the lowest mean importance rating (2.23), while the statements in the ‘Lack of time at training’ cluster received the lowest mean feasibility rating (2.19). See table 2 for the mean importance and feasibility ratings for the 59 statements and six clusters.

Concept maps

Cluster map

Figure 1 presents the cluster map that emerged following the multidimensional scaling analysis and hierarchical cluster analysis. The distance between the points illustrates the degree of similarity between statements (ie, the statements sorted together by more participants appear closer to each other on the map). For example, statement 18 (‘Discrimination in terms of resources and focus on female players in some clubs/areas’) and statement 26 (‘Tightness of training space allocated’) were considered closely related by participants, while statement 5 (‘The girls lose attention and don’t focus on the exercises, after the initial few’) and statement 27 (‘Coaching courses dont teach how to incorporate the 11+ into training session’) were not considered closely related by the participants. The stress index—an indication of the degree to which the two-dimensional point map represents the grouping data—was 0.296, close to the average stress value across a broad range of CM projects.25

Figure 1

Six-cluster map of barriers to coaches of female adolescent soccer teams implementing the 11+ (dashed lines indicate unaltered clusters before statement reassignment).

The project team agreed that a six-cluster map retained the most useful detail while merging those clusters that seemed to logically belong together (see figure 1). The six clusters around which the statements were grouped were Lack of 11+ knowledge among coaches (15 statements), Lack of player enjoyment and engagement (14 statements), Lack of link to football-related goals (11 statements), Lack of facilities and resources (eight statements), Lack of leadership (six statements) and Lack of time at training (five statements). See table 2 for a full list of the statements within each cluster, including details of the six statements that were reassigned to neighbouring clusters to which there was a better conceptual fit.

Go-Zones

Figure 2 is the Go-Zone for all 59 statements. It is divided into quadrants above and below the mean of each rating, showing a  ‘Go-Zone’ quadrant of statements in the top right, which are above average on both ratings. Go-Zones for the statements within each cluster are provided in  online supplementary file 1. To aid interpretation of the Go-Zone, see  table 2  for the details of each statement including its mean importance and feasibility ratings.

Figure 2

Go-Zone of barriers to coaches of female adolescent soccer teams implementing the 11+.

Discussion

Most previous studies of barriers and facilitators to implementing injury prevention interventions used either surveys—where participants rated their level of agreement with, or the perceived importance of, a researcher-defined set of barriers and facilitators—12 15or qualitative methods—where researchers thematically analysed and coded the data.5 32 This study, the first to use CM to identify the barriers to implementing an injury prevention intervention, addressed these methodological limitations by having study participants both identify and thematically group the implementation barriers.

The six clusters of implementation barriers identified in this study span the range of intervention levels identified in the sports setting matrix for understanding the implementation context for community sport33—from the engagement of individual participants to the leadership of the national governing body for the sport. In addition, at least one statement in every cluster was located in the top right-hand Go-Zone quadrant. These findings support previous calls9 15 34 for multistrategy ecological or systems approaches, incorporating active (eg, education of coaches) and passive (eg, policy development and organisational change) measures, to overcome the barriers to implementing injury prevention interventions in community sport.

While acknowledging the need for a systems response as outlined above, it is clear that coach education and linking the 11+ more closely to game-related skills and actions should be priorities for facilitating 11+ implementation. Both these clusters included a relatively large number of statements (15 and 11, respectively), and a relatively large proportion of statements in the top right-hand quadrant of the Go-Zone (80% and 45%, respectively). Previous research supports the finding that the lack of education of coaches14 16 and the non–sport-specific nature of the exercises in injury prevention programmes35 36 are key barriers to coaches using such programmes. It has also been suggested that successful widespread implementation of the 11+ would be enhanced if the programme was integrated into existing coach education and accreditation programmes, and supported by a ‘train the trainer’ delivery model.22 To ensure that future sports injury prevention interventions are more closely related to game skills and actions, and more compatible with the needs of the end-users and the implementation context, an intervention development process that incorporates more than just research evidence is recommended.37 A novel approach worth exploring to address these barriers would be for experts in developing neuromuscular injury prevention programmes (like the 11+) to collaborate with game-based training38 experts to develop interventions that achieve the dual outcomes of injury prevention and sport-specific skill development.

Although the ‘Lack of leadership’ cluster that emerged in this study only contained six statements, four of these were located in the top right-hand quadrant, and all six were located in the top half of the Go-Zone plot. This finding, supported by implementation science principles9 and previous research,22 39 suggests that organisational leadership at the national, state and local club levels, and role modelling by elite players, could be effective and relatively easy to implement strategies to enhance the uptake of the 11+ by community soccer coaches.

Limited time at training, difficulties generating commitment or interest from coaches and players, and the need for extra resources are frequently cited barriers to coaches adopting and implementing exercise-based sports injury prevention programmes.14–17 Although these concepts did emerge as barriers in this study, they were all rated as relatively less important and less feasible for the FFV to address than the other barriers identified in this study. This suggests that, relative to the other barriers identified in this study, they may not be the significant barriers to implementing exercise-based sports injury prevention programmes that previous research has indicated. Nonetheless, to tackle these time-related, engagement-related and resource-related challenges, those developing injury prevention programmes should ensure that programmes are fun16 and compatible with the constraints experienced in real-world implementation contexts.37 Including practical training in ‘how’ to implement injury prevention interventions efficiently and engagingly, alongside information about ‘what’ to implement, in any coach education could also help address some of these barriers.16

There are a number of limitations to CM as a research method that should be noted. CM shares methodological limitations commonly associated with other qualitative research methods including issues around the reliability, validity and generalisability of the findings due to non-random sampling, small sample size and over-reliance on the skills of the researcher.40 In this study, the subjective judgement of the members of the project team was used to synthesise and edit the brainstormed statements, decide on the number of clusters for the cluster map and reassign some statements to neighbouring clusters. Therefore, even though the project team followed the detailed guidance of Kane and Trochim,25 a similar study involving the same participants conducted by a different project team may have produced different results. Given that CM involves the same participants in a series of often time-consuming activities that build on each other, it can be difficult to recruit and retain study participants. However, this was not the case in this study as 19 of the 20 invited people agreed to participate in the study, and 13 of the 19 completed all four steps.

Conclusion

Understanding why targeted end-users do and do not use evidence-based injury prevention programmes is a key step in implementation planning and, ultimately, ensuring interventions are effective in real-world settings. Using CM, this study has confirmed that, relative to the other barriers identified in this study, a lack of coach knowledge about the 11+ and how to implement it, and the lack of a strong link between the content of the 11+ and the primary goal of soccer training, are two key barriers to coaches implementing the 11+. It has also highlighted the relative importance of organisational leadership, at the local, state and national level, as a key driver of the implementation of injury prevention interventions in community sport.9 However, perhaps the most significant contribution that this study makes is that it provides a concrete example of how important a multistrategy, ecological approach is to planning the implementation of evidence-based injury prevention interventions.

What is already known on this subject

  • Understanding the barriers and facilitators to the widespread, high-fidelity use of evidence-based injury prevention programmes is required to ensure they can have a significant public health impact.

  • Lower limb injuries are a significant problem in soccer, particularly among young female participants.

  • The 11+ is an effective injury prevention programme, but there are challenges in ensuring it is well implemented in grass roots soccer.

What this study adds

  • This is the first study to use concept mapping to identify barriers to the implementation of an injury prevention programme.

  • The barriers to coaches of female adolescent soccer teams implementing the 11+ cluster around the themes of Lack of link to football-related goals, Lack of time at training, Lack of 11+ knowledge among coaches, Lack of leadership, Lack of facilities and resources, and Lack of player enjoyment and engagement.

  • An ecological, systems approach, with a particular focus on coach education and linking the 11+ more closely to game-related skills and actions, is required to address the barriers to coaches implementing the 11+.

Acknowledgments

The Australian Collaboration for Research into Injury in Sport and its Prevention (ACRISP) is one of the International Research Centres for the Prevention of Injury and Protection of Athlete Health supported by the IOC. The authors would like to thank Maddy Goldsmith and Kate Jacewicz from the Football Federation Victoria for their contribution to synthesising the statements and deciding on the number of clusters in this study. We also thank the soccer coaches and administrators for their support of, and participation in, this research.

References

Footnotes

  • Contributors All authors contributed to the writing and critical review of this manuscript. AD, MB and AJ developed the idea for the study. AD managed the concept mapping process. AD and AC analysed and interpreted the concept mapping data. Final approval of the contents of the manuscript was obtained from all authors. All authors take responsibility for the integrity of the work from conception to publication.

  • Funding This study was funded by grants from the Medibank Better Health Foundation and the La Trobe University Sport, Exercise and Rehabilitation Research Focus Area.

  • Competing interests None declared.

  • Ethics approval La Trobe University Human Research Ethics Committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.