A Generalizable Process for Developing Sports

ARTICLE IN PRESS JSHS317_proof ■ 4 August 2016 ■ 1/8

HOSTED BY Available online at www.sciencedirect.com

Available online at www.sciencedirect.com ScienceDirect ScienceDirect

Journal of Sport and Health Science xx (2016) 1–8 Production and hosting by Elsevier www.jshs.org.cn

1bs_bs_query Q1 Original Article

2bs_bs_query

3bs_bs_query Scientiﬁc evidence is just the starting point: A generalizable process for

4bs_bs_query

5bs_bs_query developing sports injury prevention interventions

6bs_bs_query a, b c a,d a,e 7bs_bs_query Alex Donaldson *, David G. Lloyd , Belinda J. Gabbe , Jill Cook , Warren Young , a a 8bs_bs_query Peta White , Caroline F. Finch a 9bs_bs_query Australian Centre for Research into Injury in Sport and its Prevention (ACRISP), Federation University Australia, Ballarat, Victoria 3353, Australia b 10bs_bs_query Centre for Musculoskeletal Research, Grifﬁth Health Institute, Grifﬁth University, Gold Coast, Queensland 9726, Australia c 11 bs_bs_query Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Victoria 3181, Australia d 12bs_bs_query La Trobe Sport and Exercise Medicine Research Centre, La Trobe University, Melbourne, Victoria 3083, Australia e 13bs_bs_query School of Health Sciences, Federation University Australia, Ballarat, Victoria 3353, Australia

14bs_bs_query Received 12 February 2016; revised 13 May 2016; accepted 23 May 2016

15bs_bs_query Available online

16bs_bs_query

17bs_bs_query Abstract

18bs_bs_query

19bs_bs_query Background: The 2 most cited sports injury prevention research frameworks incorporate intervention development, yet little guidance is available

20bs_bs_query in the sports science literature on how to undertake this complex process. This paper presents a generalizable process for developing implementable

21bs_bs_query sports injury prevention interventions, including a case study applying the process to develop a lower limb injury prevention exercise training

22bs_bs_query program (FootyFirst) for community Australian football.

23bs_bs_query Methods: The intervention development process is underpinned by 2 complementary premises: (1) that evidence-based practice integrates the best

24bs_bs_query available scientiﬁc evidence with practitioner expertise and end-user values and (2) that research evidence alone is insufﬁcient to develop

25bs_bs_query implementable interventions.

26bs_bs_query Results: The generalizable 6-step intervention development process involves (1) compiling research evidence, clinical experience, and knowledge

27bs_bs_query of the implementation context; (2) consulting with experts; (3) engaging with end users; (4) testing the intervention; (5) using theory; and (6)

28bs_bs_query obtaining feedback from early implementers. Following each step, intervention content and presentation should be revised to ensure that the ﬁnal

29bs_bs_query intervention includes evidence-informed content that is likely to be adopted, properly implemented, and sustained over time by the targeted

30bs_bs_query intervention deliverers. For FootyFirst, this process involved establishing a multidisciplinary intervention development group, conducting 2

31bs_bs_query targeted literature reviews, undertaking an online expert consensus process, conducting focus groups with program end users, testing the program

32bs_bs_query multiple times in different contexts, and obtaining feedback from early implementers of the program.

33bs_bs_query Conclusions: This systematic yet pragmatic and iterative intervention development process is potentially applicable to any injury prevention topic

34bs_bs_query across all sports settings and levels. It will guide researchers wishing to undertake intervention development.

36bs_bs_query license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

37bs_bs_query

38bs_bs_query Keywords: Australian football; Intervention development; Lower limb injuries; Sports injury prevention; Implementation; Research-to-practice; Translation

39bs_bs_query

40bs_bs_query 1. Introduction Both the Translating Research into Injury Prevention Practice 53bs_bs_query 4 7 41bs_bs_query framework and the Sequence of Prevention model require prac- 54bs_bs_query

42bs_bs_query Evidence-based sports injury prevention interventions are 1-3 titioners and researchers to identify potential injury prevention 55bs_bs_query 43bs_bs_query not well implemented in real-world settings, often because solutions and develop appropriate prevention measures guided by 56bs_bs_query

44bs_bs_query the interventions are not directly relevant to speciﬁc implemen- 4,5 high-quality epidemiologic and etiologic studies. Most research 57bs_bs_query 45bs_bs_query tation contexts. Interventions should be informed by research 8,9 remains in the early stages of these models and frameworks, and 58bs_bs_query

46bs_bs_query evidence and be widely adopted, properly implemented, and 4,6 this limits the potential for injuries to be prevented. In practice, 59bs_bs_query 47bs_bs_query sustained over time. preventive measures are often based on anecdotal experience or 60bs_bs_query 48bs_bs_query 4 current practice, and the scientiﬁc literature rarely provides 61bs_bs_query

49bs_bs_query

insights into the complex process of intervention development in 62bs_bs_query 50bs_bs_query Peer review under responsibility of Shanghai University of Sport. 10 real-world settings. Although systematic reviews and meta- 63bs_bs_query 51bs_bs_query * Corresponding author.

52bs_bs_query E-mail address: [email protected] (A. Donaldson). analyses can identify promising interventions, their conclusions 64bs_bs_query

http://dx.doi.org/10.1016/j.jshs.2016.08.003 2095-2546/© 2016 Production and hosting by Elsevier B.V. on behalf of Shanghai University of Sport. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Please cite this article in press as: Alex Donaldson, et al., Scientiﬁc evidence is just the starting point: A generalizable process for developing sports injury prevention interventions, Journal of Sport and Health Science (2016), doi: 10.1016/j.jshs.2016.08.003 ARTICLE IN PRESS JSHS317_proof ■ 4 August 2016 ■ 2/8

2 A. Donaldson et al.

1 65bs_bs_query Table 1

66bs_bs_query Summary of the primary methods used to develop FootyFirst.

67bs_bs_query Evidence-based practice Aims Contribution to the development of FootyFirst

68bs_bs_query element

69bs_bs_query Compilation and quality To ensure that the best available research • Compilation of published and previously unpublished community AF injury 12 70bs_bs_query assessment of research evidence relevant to the speciﬁc problem data (Fig. 1, step 1)

71bs_bs_query evidence was identiﬁed and applied in the • Review of the scientiﬁc literature to systematically evaluate the evidence about

72bs_bs_query development of the intervention the beneﬁts of lower limb injury prevention protocols aimed at reducing the 20 73bs_bs_query most common, severe lower limb injuries in community AF (Fig. 1, step 1).

74bs_bs_query Incorporation of clinical To ﬁll in the gaps where there was limited • Application of health promotion, implementation science, physiotherapy,

75bs_bs_query expertise and practitioner information in the literature or where no biomechanical, and sports science clinical and research expertise to develop

76bs_bs_query knowledge and views successful intervention for a speciﬁc injury the exercise training program (Fig. 1, step 1)

77bs_bs_query was identified • Delphi consultation to achieve expert consensus on the specific content of the 21 78bs_bs_query To fit the available evidence to the specific exercise training program (step 2)

79bs_bs_query circumstances, populations, and needs 34 80bs_bs_query Consideration of end user To ensure that the intervention is • Focus groups, following standard focus group methods, with community AF

81bs_bs_query preference, capacity, and appropriate for, and reﬂects the capacity of, senior coaches, strength and conditioning/ﬁtness/high-performance coaches,

82bs_bs_query values the implementation context players, sports trainers, and administrators (Fig. 1, step 3)

83bs_bs_query • Testing of the exercise training program with delivery agent representatives

84bs_bs_query and conducting a “train-the-trainer” session (Fig. 1, step 4)

85bs_bs_query • Evaluation of the program against the attributes of innovations from the 25 86bs_bs_query diffusion of innovations theory (Fig. 1, step 5)

87bs_bs_query • Feedback from early implementers on the content and presentation of the

88bs_bs_query program (Fig. 1, step 6) = 89bs_bs_query Abbreviation: AF Australian football.

90bs_bs_query

91bs_bs_query are rarely directly applicable to speciﬁc real-world settings, and considers the implementation context, the end user’s perspectives, 126bs_bs_query 11 92bs_bs_query translation into effective practice is challenging. and long-term sustainability, injury prevention programs are 127bs_bs_query 4,5 93bs_bs_query Australian football (AF) is a popular sport at the community unlikely to be widely used and will therefore have limited impact. 128bs_bs_query

94bs_bs_query level in Australia. It is a dynamic sport that incorporates running, Three methods underpinned the application of the interven- 129bs_bs_query

95bs_bs_query rapid acceleration and deceleration, changing direction, jumping tion development process to FootyFirst: (1) literature search to 130bs_bs_query

96bs_bs_query and landing, full body contact including tackling and bumping, identify published research evidence, (2) use of clinical exper- 131bs_bs_query

97bs_bs_query and kicking and marking (catching) a ball. As in many other tise and expert opinion via a Delphi process, and (3) focus 132bs_bs_query

98bs_bs_query sports, preventing lower limb injuries (LLIs) is a priority in groups to identify end user preferences, capacities, and values 133bs_bs_query 12 99bs_bs_query community AF. Although several evidence-based LLI preven- (Table 1). The speciﬁc methods used to establish LLIs as a 134bs_bs_query 13-15 12 100bs_bs_query tion programs exist, how they were developed is largely unre- priority, compile and assess the quality of exercise protocols 135bs_bs_query 20 101bs_bs_query ported, and only 1 targeting selected LLIs is speciﬁc to aimed at reducing LLIs in similar sports, and achieve expert 136bs_bs_query 16 21 102bs_bs_query community AF. For example, the only published information consensus on the contents of FootyFirst are described else- 137bs_bs_query

103bs_bs_query available on the development of the well-known and widely used where. Federation University Australia (E13-004) Human 138bs_bs_query

104bs_bs_query Fédération Internationale de Football Association (FIFA) “FIFA Research Ethics Committee approved the study protocol. 139bs_bs_query

+ 140bs_bs_query 105bs_bs_query 11 ” program states that it was developed by an expert group,

17 3. Results 141bs_bs_query 106bs_bs_query and tested by 1 club, before it was implemented in trials. 142bs_bs_query

107bs_bs_query This paper presents a generalizable process for developing The intervention development process can be encapsulated 143bs_bs_query

108bs_bs_query evidence-informed sports injury prevention interventions that in 6 steps (Fig. 1). The application of these steps and the 144bs_bs_query

109bs_bs_query need to be widely and sustainably implemented in real-world outcomes of each step when developing FootyFirst are summa- 145bs_bs_query

110 bs_bs_query settings. An example application of the process is provided rized in Fig. 2 and Table 2. 146bs_bs_query

111 bs_bs_query based on the development of an exercise training program As recommended in Translating Research into Injury Pre- 147bs_bs_query 4 112 bs_bs_query (called FootyFirst) to prevent LLIs in community AF. This vention Practice stage 3, a multidisciplinary FootyFirst Devel- 148bs_bs_query

113 bs_bs_query paper serves as a guide to researchers wishing to progress their opment Group (FFDG) was established, consisting of 2 sports 149bs_bs_query

114 bs_bs_query research through the intervention development process. physiotherapists (authors JC and BJG), 1 biomechanist (author 150bs_bs_query

115 bs_bs_query DGL), 1 sports scientist (author WY), and their research teams. 151bs_bs_query

116 bs_bs_query 2. Methods Alongside their clinical and research experience, the FFDG had 152bs_bs_query 117 bs_bs_query

considerable exercise and rehabilitation experience in commu- 153bs_bs_query 118 bs_bs_query Two complementary ideas underpin the process described in this

nity and elite sport as well as involvement in previous commu- 154bs_bs_query 119 bs_bs_query paper: (1) evidence-based practice integrates the best available 18 16 nity AF LLI prevention research. 155bs_bs_query 120bs_bs_query scientiﬁc evidence with practitioner expertise and end-user values,

156bs_bs_query 121bs_bs_query and (2) research evidence alone is insufﬁcient to develop 2 3.1. Step 1: Use the research evidence and clinical 157bs_bs_query 122bs_bs_query implementable interventions. This process addresses the criticism experience 158bs_bs_query

123bs_bs_query that evidence-based practice devalues practitioner expertise, 159bs_bs_query

124bs_bs_query ignores community values, and promotes a “1-size-ﬁts-all” This initial step is necessary to maximize the likelihood that 160bs_bs_query 19 125bs_bs_query approach. It also acknowledges that unless intervention design the developed intervention will “work” by ensuring ﬁrm 161bs_bs_query

Developing sports injury prevention interventions 3

Sequence of prevention of sports Translating Research into Injury injuries model7 Six-step intervention Prevention Practice (TRIPP) framework4 development process

Step 1 1. Injury surveillance 1. Establishing the extent of Use the research the sports injury problem evidence and clinical experience 2. Establish etiology and mechanism of injuries Step 2 2. Establishing aetiology and Consult the experts mechanism of injuries

Step 3 3. Develop preventive Engage the end users measure 3. Introduce preventive measures Step 4 Test the intervention 4. Ideal conditions/scientific evaluation

4. Assess their effectiveness Step 5 Evaluate against theory by repeating step 1 5. Describe intervention context Step 6 Obtain feedback from early implementers 6. Evaluate effectiveness in implementation context

162bs_bs_query Fig. 1. Six-step intervention development process.

163bs_bs_query

164bs_bs_query grounding in the available epidemiologic and etiological evi- cally assess the content of the ﬁrst draft of FootyFirst (June to 195bs_bs_query 21 165bs_bs_query dence. To begin the development of FootyFirst, the FFDG sys- July 2011). Three rounds of revisions and consultations went 196bs_bs_query

166bs_bs_query tematically evaluated the strength and quality of the evidence into creating the second draft of FootyFirst, which the FFDG was 197bs_bs_query

167bs_bs_query on the benefits of exercise protocols with the potential to reduce then confident contained the “right” exercises and progressions. 198bs_bs_query 20 168bs_bs_query common community AF LLIs (March to May 2011). In However, whether it “fitted” the community AF context remained 199bs_bs_query

169bs_bs_query summary, the collated research evidence suggested that an LLI unknown. 200bs_bs_query

201bs_bs_query 170bs_bs_query prevention program should include balance and control and

20 3.3. Step 3: Engage end users 202bs_bs_query 171bs_bs_query eccentric hamstring, plyometric, and strength exercises. 203bs_bs_query

172bs_bs_query However, speciﬁc details of effective training routines were Development of any intervention requires that the proposed 204bs_bs_query

173bs_bs_query often unavailable, many studies focused on elite athletes, and strategies and program components are relevant and acceptable 205bs_bs_query

174bs_bs_query there was limited evidence for the prevention of some LLIs. to potential program deliverers (e.g., strength and conditioning 206bs_bs_query

175bs_bs_query The FFDG used the synthesized research evidence, their staff, coaches, etc.) and participants (e.g., athletes, players, 207bs_bs_query 5,6 176bs_bs_query collective clinical and research experience, knowledge of exist- etc.). Two focus groups—1 with 3 participants from clubs in 208bs_bs_query

177bs_bs_query ing LLI prevention protocols and exercise training principles a metropolitan region and the other with 12 participants from 209bs_bs_query

178bs_bs_query (e.g., progressive overloading, speciﬁcity, and regularity), and clubs in a regional league—were conducted (July and August 210bs_bs_query

179bs_bs_query prior experience with community AF to develop the ﬁrst draft 2011) to engage potential end users of FootyFirst in determin- 211bs_bs_query

180bs_bs_query of FootyFirst. For example, the research evidence on exercises ing whether the program was suitable for implementation by 212bs_bs_query 20 181bs_bs_query to prevent hamstring injuries was quite conclusive, requiring community AF coaches. Participants were recruited with the 213bs_bs_query

182bs_bs_query minor modiﬁcation to suit community AF. Conversely, research assistance of administrators from the relevant local governing 214bs_bs_query

183bs_bs_query evidence was sparse for hip and groin injury prevention, requir- league and included community AF coaches, strength and 215bs_bs_query

184bs_bs_query ing greater reliance on etiologic studies and clinical experience. conditioning/ﬁtness/high-performance staff, sports trainers, 216bs_bs_query

185bs_bs_query administrators, and players. Participants reviewed the expert- 217bs_bs_query 186bs_bs_query 3.2. Step 2: Consult the experts

agreed version of FootyFirst before discussing its content and 218bs_bs_query 187bs_bs_query

188bs_bs_query This step ensures that the developed intervention is speciﬁc to presentation, the methods that would be required to support 219bs_bs_query

189bs_bs_query the sport and injury mechanisms of interest. In the case of community AF clubs and coaches to implement FootyFirst, and 220bs_bs_query

190bs_bs_query FootyFirst, the extent to which the information gathered in step 1 the capacity of end users to deliver or complete FootyFirst. 221bs_bs_query

191bs_bs_query could be translated into the AF context was unknown. Therefore, Focus-group ﬁndings suggested that there was general 222bs_bs_query

192bs_bs_query 55 purposively selected LLI prevention experts working in high- support for replacing current warm-up programs with FootyFirst. 223bs_bs_query

193bs_bs_query performance AF clubs and other environments were invited to Minor revisions to the way in which FootyFirst was presented 224bs_bs_query

194bs_bs_query participate in an online consensus-generating approach to criti- were suggested, and ideas to facilitate its implementation were 225bs_bs_query

4 A. Donaldson et al.

1 Timeline Step 1: Use the research evidence and clinical experience March 2011 Review of the scientific literature Application of clinical experience of research team Application of exercise training principles Integration of knowledge of implementation context

1st draft of FootyFirst*

Step 2: Consult the experts Expert consensus on content of FootyFirst using a Delphi process

June 2011 2st draft of FootyFirst*

Step 3: Engage end users Step 4: Test the intervention Consultations with community coaches, Testing FootyFirst with community administrators, etc. using focus group coaches and “train-the-trainers session”

December 2011 Step 5: Evaluate against theory FootyFirst FootyFirst content and presentation launch to 22 evaluated against diffusion of Innovations community attributes Australian Football clubs

3rd draft of FootyFirst

January 2012 Step 6: Obtain feedback from early implementers Feedback from coaches and clubs who had used FootyFirst

4th draft of FootyFirst

February 2012 Expert graphic designer and editor produce program resources

May 2012 FootyFirst Final Programme#

FootyFirst Coaches Manual FootyFirst Posters FootyFirst Compact Disc

226bs_bs_query Fig. 2. Application of the 6-step intervention development process to FootyFirst. *Details of these drafts available in Appendix (online). #Available at

227bs_bs_query https://Footyﬁrstaustralia.wordpress.com/footyﬁrst-program.

Developing sports injury prevention interventions 5

228bs_bs_query Table 2

229bs_bs_query The process and outcomes of developing FootyFirst, from the initial exercise selection to the ﬁnal protocol and resource production.

230bs_bs_query FootyFirst program version Key program development processes Summary of process outcomes and protocol revisions

231bs_bs_query 1st Draft • Literature review (Fig. 1, step 1) 5-level, progressive exercise

232bs_bs_query • Clinical research, health promotion, and program developed consisting of

233bs_bs_query implementation science experience of brief warm-up and exercises to

234bs_bs_query project team (Fig. 1, step 1) reduce the risk of groin, hip,

235bs_bs_query • Training principles (e.g., progressive hamstring, knee, and ankle injuries

236bs_bs_query overloading, speciﬁcity, and regularity) among community AF players

237bs_bs_query (Fig. 1, step 1)

238bs_bs_query • Knowledge of implementation context

239bs_bs_query (Fig. 1, step 1)

240bs_bs_query 2nd Draft • Expert consultation (Fig. 1, step 2) First draft revised as follows:

241bs_bs_query • Instructions for some exercises revised

242bs_bs_query • Two exercises removed and 2 exercises added to program

243bs_bs_query • The term “core” (muscles) removed and all related exercises referred to as hip

244bs_bs_query strength exercises

245bs_bs_query • Order of some exercise changed

246bs_bs_query 3rd Draft • End user engagement (Fig. 1, step 3) Second draft revised as follows:

247bs_bs_query • Exercise program tested for feasibility and • Instructions for some exercises revised

248bs_bs_query acceptability (Fig. 1, step 4) • Some exercises modiﬁed to be more focused on AF and more like usual

249bs_bs_query • Program content and presentation evaluated community AF training, to increase relevance and acceptability 25 250bs_bs_query against diffusion of innovations attributes • “Frequently asked questions” section added to instruction manual

251bs_bs_query (Fig. 1, step 5) • Number of repetitions of some exercises reduced

252bs_bs_query • Key points for good technique (i.e., high ﬁdelity) added to instruction manual

253bs_bs_query • Instructions and photographs on how to identify common faults for most

254bs_bs_query exercises added to instruction manual

255bs_bs_query • Justiﬁcation for each exercise added to instruction manual

256bs_bs_query • Exercise added to program

257bs_bs_query 4th Draft • Feedback from early program implementers Third draft revised as follows:

258bs_bs_query (Fig. 1, step 6) • Instructions for some exercises revised

259bs_bs_query • Endorsement from the AFL, a high-performance AFL coach, and the AFL

260bs_bs_query Medical Ofﬁcers Association added

261bs_bs_query FootyFirst ﬁnal • Expert graphic designers and editors Fourth draft revised as follows:

262bs_bs_query program engaged to produce program resources • Wording of instructions for some exercises revised to improve clarity for the

263bs_bs_query level of the intended audience

264bs_bs_query • Layout and presentation ﬁnalized

265bs_bs_query • Logo designed and associated color scheme for resources agreed upon

266bs_bs_query • Coach’s instruction manual, posters, and compact disk produced = = 267bs_bs_query Abbreviation: AF Australian football; AFL Australian Football League.

268bs_bs_query

269bs_bs_query provided. Discussants reported similarities between the warm-up • capacity of community-AF players to perform some of the 291bs_bs_query

270bs_bs_query component of FootyFirst and their existing pretraining activities. entry-level exercises; 292bs_bs_query

271bs_bs_query However, most of the strength and conditioning and neuromus- • concern that some exercises might increase the risk of injury 293bs_bs_query

272bs_bs_query cular training exercises were considered less compatible with and/or muscle soreness; and 294bs_bs_query

273bs_bs_query existing community AF training. Participants indicated that most • the challenge of implementing a progressive exercise train- 295bs_bs_query

274bs_bs_query clubs had experienced or qualiﬁed people to deliver FootyFirst, ing programme with players within a single team/club with 296bs_bs_query

275bs_bs_query although help with implementing the program into training ses- varying ﬁtness levels and training commitment. 297bs_bs_query

298bs_bs_query 276bs_bs_query sions was needed. They also advocated for including a rationale

Consequently, the structure of some FootyFirst exercises was 299bs_bs_query

277bs_bs_query for each exercise to help coaches promote FootyFirst to their

revised to align more closely with typical community AF train- 300bs_bs_query

278bs_bs_query players. Finally, although they acknowledged and appreciated

ing activities while maintaining their ﬁdelity from a clinical and 301bs_bs_query

279bs_bs_query that FootyFirst was underpinned by scientiﬁc evidence, focus

biomechanical perspective. The AFL logo, a foreword or intro- 302bs_bs_query

280bs_bs_query group participants suggested that FootyFirst should also be

duction from a high-performance AF coach, and a “frequently 303bs_bs_query

281bs_bs_query endorsed by respected, high-performance AF coaches and the

asked questions” section were also added to the FootyFirst 304bs_bs_query

282bs_bs_query peak body, the Australian Football League (AFL).

instruction manual. 305bs_bs_query

283bs_bs_query Participants identiﬁed other implementation-related issues 306bs_bs_query

bs_bs_query 284bs_bs_query that needed to be considered when developing FootyFirst, 3.4. Step 4: Test the feasibility and acceptability of the 307

bs_bs_query 285bs_bs_query including the following: intervention 308

286bs_bs_query 309bs_bs_query

287bs_bs_query • time available to complete the intervention (typically Before assessing intervention efﬁcacy, the feasibility and 310bs_bs_query

288bs_bs_query 15-20 min); acceptability of an intervention should be tested to be conﬁdent 311bs_bs_query

289bs_bs_query • equipment available to implement the intervention in a that it can be delivered as intended and that participants can 312bs_bs_query

290bs_bs_query typical club; successfully complete all required tasks. In the case of 313bs_bs_query

6 A. Donaldson et al.

1314bs_bs_query FootyFirst, the postexpert consultation version of the program Following this discussion, the injury prevention and 371bs_bs_query

315bs_bs_query (second draft in Fig. 2 and Table 2) was tested on several performance-related advantages of FootyFirst over current 372bs_bs_query

316bs_bs_query occasions by the FFDG. First, 6 exercise science and physio- warm-up activities were further emphasized in the instruction 373bs_bs_query

317bs_bs_query therapy students, who were all community-level athletes, were manual. FootyFirst’s compatibility with the existing values and 374bs_bs_query

318bs_bs_query videotaped and photographed performing the exercises cor- practices of community AF clubs was also highlighted by 375bs_bs_query < 319bs_bs_query rectly and incorrectly (August 2011). This led to further reﬁne- explicitly stating that FootyFirst would take 20 min to com- 376bs_bs_query

320bs_bs_query ment of some exercise instructions and the addition of plete and could replace, rather than be added to, existing 377bs_bs_query

321bs_bs_query photographs and descriptions of common technique faults in warm-up activities. More extensive use of footballs in some 378bs_bs_query

322bs_bs_query the program resources. exercises was also instituted to enhance compatibility with 379bs_bs_query

323bs_bs_query The intervention was then tested again. A coach was given typical AF game components. FootyFirst was also promoted as 380bs_bs_query

324bs_bs_query the revised FootyFirst instruction manual and observed while a way to align community AF training more closely with train- 381bs_bs_query

325bs_bs_query delivering the intervention to 2 community AF players (October ing at high-performance AF clubs by adding endorsements 382bs_bs_query

326bs_bs_query 2011). The aim was to determine whether a typical community from a respected high-performance AF coach and the AFL 383bs_bs_query

327bs_bs_query AF coach who had preread the instruction manual could deliver Medical Ofﬁcers’ Association. The complexity of the program 384bs_bs_query

328bs_bs_query it as intended. The time taken to complete the program was was considered by ensuring that all exercises were explained 385bs_bs_query

329bs_bs_query reviewed to ensure compatibility with the time allocated to clearly and straightforwardly and by emphasizing that 386bs_bs_query

330bs_bs_query complete pretraining activities at a typical community AF train- FootyFirst did not require any expensive or sophisticated 387bs_bs_query

331bs_bs_query ing session. This test highlighted the need to emphasize proper equipment. 388bs_bs_query

332bs_bs_query technique in the instructions, and it showed that additional rest 389bs_bs_query

bs_bs_query 333bs_bs_query periods were required between repetitions for some exercises. 3.6. Step 6: Obtain feedback from early implementers 390

391bs_bs_query 334bs_bs_query Importantly, it also identiﬁed that some exercises needed modi-

Despite all the effort invested in steps 1-5, it is not guaran- 392bs_bs_query

335bs_bs_query ﬁcations (e.g., reduction in number of sets or repetitions) to

teed that the process will develop a perfect intervention, espe- 393bs_bs_query

336bs_bs_query reduce the time taken to complete them. Further minor word

cially at the ﬁrst attempt. Therefore, it is useful to ask end users 394bs_bs_query

337bs_bs_query changes were also made to improve clarity and assist interven-

to use the intervention in their settings and obtain feedback 395bs_bs_query

338bs_bs_query tion delivery.

from them about the content and presentation before the inter- 396bs_bs_query

339bs_bs_query High levels of trainer competency and self-efﬁcacy are 22,23 vention is formally evaluated. The post-theory-evaluated 397bs_bs_query 340bs_bs_query acknowledged drivers of implementation success. There-

version of FootyFirst (third draft in Fig. 2 and Table 2)was 398bs_bs_query

341bs_bs_query fore, a key strategy for enhancing FootyFirst implementation

launched and disseminated to 22 community AF clubs afﬁliated 399bs_bs_query

342bs_bs_query within targeted community AF clubs was training coaches how 24 with 1 AF regional league in Victoria, Australia (December 400bs_bs_query 343bs_bs_query to deliver the intervention to players. A “train-the-trainer”

2011). This provided an opportunity to obtain feedback from 401bs_bs_query

344bs_bs_query session was held to explain FootyFirst to a respected, high-

early implementers of the program prior to the football season. 402bs_bs_query

345bs_bs_query proﬁle, high-performance coach afﬁliated with a community

Eligible clubs and coaches were asked to suggest revisions to 403bs_bs_query

346bs_bs_query AF club in the targeted region. This coach was observed as he

the program before ﬁnal resource production (February 2012). 404bs_bs_query

347bs_bs_query delivered the intervention to 2 community AF players and was

No further changes to the program content were suggested, 405bs_bs_query

348bs_bs_query provided with feedback (January 2012). Following the train-

although minor changes were made to the language and pre- 406bs_bs_query

349bs_bs_query the-trainer session, it became apparent that the progression for

sentation where inconsistencies were identiﬁed. A professional 407bs_bs_query

350bs_bs_query 1 particular exercise might be too difﬁcult for some community

editor/graphic designer was then engaged to produce the ﬁnal 408bs_bs_query

351bs_bs_query AF players, and a new exercise was added to ensure adequate

FootyFirst resources, which were distributed to the 22 targeted 409bs_bs_query

352bs_bs_query preparation for the transition between levels. Minor revisions

clubs (March 2012). Video footage of the FootyFirst exercises 410bs_bs_query

353bs_bs_query were also made to the instructions and photographs for some

was produced, in partnership with the AFL, and made available 411bs_bs_query

354bs_bs_query exercises.

to participating clubs in DVD and online formats prior to the 412bs_bs_query

355bs_bs_query 29 2013 community AF season. 413bs_bs_query

356bs_bs_query 3.5. Step 5: Evaluate against theory 414bs_bs_query

357bs_bs_query 4. Discussion 415bs_bs_query

358bs_bs_query When developing an intervention, 1-way to enhance the 416bs_bs_query

359bs_bs_query likelihood of its adoption, implementation, and maintenance by Although developing preventive measures is integral to 417bs_bs_query

360bs_bs_query the target audience in the real-world setting is to evaluate it the 2 most widely used sports injury prevention research 418bs_bs_query 4,7 361bs_bs_query against a relevant theory. To further ﬁt FootyFirst into the com- frameworks, to our knowledge, this is the ﬁrst detailed 419bs_bs_query

362bs_bs_query munity AF context, the research team evaluated the program description of a process to achieve this. This paper provides a 420bs_bs_query

363bs_bs_query content and presentation against the relative advantage, com- step-by-step generalizable guide on how to develop evidence- 421bs_bs_query

364bs_bs_query patibility, and complexity attributes of innovations as outlined informed and context-speciﬁc interventions. The guide con- 422bs_bs_query 25 365bs_bs_query in the diffusion of innovations theory during a structured ﬁrms that intervention development is a complex, iterative 423bs_bs_query 30 366bs_bs_query round-table discussion (December 2011). The diffusion of process requiring a balancing of evidence and experience. 424bs_bs_query

367bs_bs_query innovations theory was selected because it is speciﬁcally appli- Physical training is the second most studied sports injury 425bs_bs_query 26 9 368bs_bs_query cable to the introduction of new ideas into communities and is prevention intervention, yet when step 1 of the intervention 426bs_bs_query

369bs_bs_query one of the few theories that has been previously applied in development process was applied to FootyFirst, it was identiﬁed 427bs_bs_query 24,27,28 370bs_bs_query sports injury prevention research. that the scientiﬁc evidence underpinning physical training as an 428bs_bs_query

Developing sports injury prevention interventions 7

20 429bs_bs_query injury prevention intervention is limited. Randomized con- Step 4 in the intervention development process focuses on 486bs_bs_query

430bs_bs_query trolled trials to reduce injury rates may be ineffective because testing the intervention. In our experiences with FootyFirst, this 487bs_bs_query

431bs_bs_query the principles underpinning the trialed prevention strategies was challenging owing to the short break between the end of 1 488bs_bs_query

432bs_bs_query were not based on the evidence of injury etiology and mecha- community AF season and the start of preseason training for the 489bs_bs_query

433bs_bs_query nisms, or because the implementation context was not next season (often as little as 6 weeks). FootyFirst was not 490bs_bs_query

434bs_bs_query adequately considered during the intervention development tested with a full community AF team, and coaches were not 491bs_bs_query 4,31,32 4,9 435bs_bs_query process. These issues are common in sports medicine, observed delivering it to their teams before the program was 492bs_bs_query 33 436bs_bs_query and in public health more broadly, and both are addressed by made publically available. Only small-scale tests of the program 493bs_bs_query

437bs_bs_query the proposed intervention development process (in steps 1 and were possible, which did not reﬂect its intended real-world use. 494bs_bs_query

438bs_bs_query 3 speciﬁcally). Furthermore, the program required participants to progress 495bs_bs_query

439bs_bs_query Step 2 provides a practitioner-expert perspective on the from 1 level to another over a relatively short period of time and 496bs_bs_query

440bs_bs_query intervention under development. This is particularly valuable a number of training sessions. Even after the evidence under- 497bs_bs_query

441bs_bs_query when full scientiﬁc evidence about what might work is either pinning the content of a program is conﬁrmed, intervention 498bs_bs_query

442bs_bs_query absent from, or limited in, the scientiﬁc literature and interven- development is a long and time-consuming process, ideally 499bs_bs_query

443bs_bs_query tion developers need to rely on critical evaluation of the inter- involving long lead times with early testing of the proposed 500bs_bs_query

444bs_bs_query vention by others in the ﬁeld. In the example of FootyFirst, the interventions within the target setting. 501bs_bs_query

445bs_bs_query expert consensus opinion generally reﬂected the available sci- Following the development of an intervention, it is important 502bs_bs_query

446bs_bs_query entiﬁc evidence. The exercises that were more strongly sup- to develop an evidence-informed and context-speciﬁc imple- 503bs_bs_query

447bs_bs_query ported by published evidence remained largely unchanged. The mentation plan to facilitate uptake of the intervention by end 504bs_bs_query 24,28 448bs_bs_query round in which consensus was reached on each exercise was users. The effectiveness of the intervention and implemen- 505bs_bs_query 20,21 449bs_bs_query commensurate with the level of evidence available. A side tation plan should then be evaluated under both ideal and real- 506bs_bs_query 4 450bs_bs_query beneﬁt of this expert consultation was that it assisted in attain- world conditions. 507bs_bs_query

508bs_bs_query 451bs_bs_query ing endorsement of FootyFirst from relevant experts working in 5. Conclusions 509bs_bs_query

452bs_bs_query high-performance AF. 510bs_bs_query

453bs_bs_query The intervention development process and its application to Sports scientists, sports medicine clinicians, and sports 511bs_bs_query

454bs_bs_query the development of FootyFirst highlighted that the intervention injury prevention researchers need to be competent in develop- 512bs_bs_query

455bs_bs_query content experts often lacked expertise in effectively communi- ing interventions that are informed by the best available 513bs_bs_query

456bs_bs_query cating those interventions to the target audience. The FFDG research evidence, suitable for the context in which they are to 514bs_bs_query

457bs_bs_query focused initially on ensuring that scientiﬁc evidence and exer- be implemented, and compatible with the needs, capacity, and 515bs_bs_query

458bs_bs_query cise training and biomechanical principles underpinned the values of potential end users. The 6-step intervention develop- 516bs_bs_query

459bs_bs_query program content, rather than precisely describing the exercises. ment process outlined here provides, for the ﬁrst time, a prac- 517bs_bs_query

460bs_bs_query Interpretation difﬁculties experienced while testing the inter- tical, feasible, and generalizable process for doing this. It is 518bs_bs_query

461bs_bs_query vention (step 4 speciﬁcally) were not anticipated by the FFDG, applicable to intervention development across a range of topics 519bs_bs_query

462bs_bs_query nor were they identiﬁed during either the expert (step 2) or end and sports settings and levels. 520bs_bs_query

521bs_bs_query 463bs_bs_query user (step 3) consultations. Participants in the end user consul- Acknowledgments 522bs_bs_query

464bs_bs_query tations were asked if they understood the instructions and 523bs_bs_query

465bs_bs_query whether they believed that community AF players could com- Thanks to Alasdair Dempsey, Clare Minahan, Jane Grayson, 524bs_bs_query

466bs_bs_query plete the exercises. They were not asked if they understood Jace Kelly, and Nadine Andrew for their contributions to devel- 525bs_bs_query

467bs_bs_query exactly how to do the exercises. The challenges experienced oping FootyFirst. We also thank all the experts and community 526bs_bs_query

468bs_bs_query while testing FootyFirst highlight the distinction between gen- Australian Football representatives who participated in consul- 527bs_bs_query

469bs_bs_query erally understanding an intervention and being able to deliver or tations and testing of FootyFirst. 528bs_bs_query

470bs_bs_query instruct someone else to participate in it with high ﬁdelity. This study was funded by an National Health and Medical 529bs_bs_query

471bs_bs_query Although some program planning frameworks recommend Research Council (NHMRC) Partnership Project Grant (ID: 530bs_bs_query

472bs_bs_query that potential program participants are involved right from the 565907) which included additional support (both cash and 531bs_bs_query 6 473bs_bs_query needs assessment/problem identiﬁcation stage, in the sports in-kind) from the following project partner agencies: the Aus- 532bs_bs_query

474bs_bs_query medicine and science context, end users are not typically tralian Football League; Victorian Health Promotion Founda- 533bs_bs_query

475bs_bs_query directly engaged in the intervention development process out- tion; New South Wales Sporting Injuries Committee; JLT 534bs_bs_query

476bs_bs_query lined in this paper until step 3. However, this should not pre- Sport, a division of Jardine Lloyd Thompson Australia Pty Ltd; 535bs_bs_query

477bs_bs_query clude intervention developers from conﬁrming earlier that the Sport and Recreation Victoria, Department of Transport, Plan- 536bs_bs_query

478bs_bs_query injury issue being addressed is important and ensuring that the ning and Local Infrastructure; and Sports Medicine Australia— 537bs_bs_query

479bs_bs_query proposed method of intervention is acceptable from the end National and Victorian Branches. Caroline Finch was supported 538bs_bs_query

480bs_bs_query user’s perspective. This may be particularly relevant when by an NHMRC Principal Research Fellowship (APP1058737). 539bs_bs_query

481bs_bs_query addressing injury issues or proposing intervention methods that Belinda Gabbe was supported by an NHMRC Career Develop- 540bs_bs_query

482bs_bs_query are less well documented than for LLIs in team ball sports. It ment Fellowship (APP1048731). Jill Cook was supported by a 541bs_bs_query

483bs_bs_query may also be advantageous to gain feedback from the end users NHMRC Practitioner fellowship (APP1058493). Alex Donald- 542bs_bs_query

484bs_bs_query when critically evaluating a proposed intervention against a son and Peta White held Research Fellowships funded through 543bs_bs_query

485bs_bs_query theory or model (step 5). the major NHMRC Partnership Project Grant. 544bs_bs_query

8 A. Donaldson et al.

13. Hübscher M, Refshauge KM. Neuromuscular training strategies for bs_bs_query 1545bs_bs_query Authors’ Contributions 610

bs_bs_query 546bs_bs_query preventing lower limb injuries: what’s new and what are the practical 611

547bs_bs_query CFF, DGL, BJG, JC and AD conceived of and planned the implications of what we already know? Br J Sports Med 2013;47:939-40. 612bs_bs_query

14. Hübscher M, Zech A, Pfeifer K, Hänsel F, Vogt L, Banzer W. 613bs_bs_query 548bs_bs_query study. BJG coordinated the literature review with support from

Neuromuscular training for sports injury prevention: a systematic review. 614bs_bs_query 549bs_bs_query DGL and JC. DGL, BJG, JC and WY developed the ﬁrst draft

Med Sci Sports Exerc 2010;42:413-21. 615bs_bs_query

550bs_bs_query of FootyFirst and contributed to all stages of the programme’s 15. Emery CA, Roy TO, Whittaker JL, Nettel-Aguirre A, van Mechelen W. 616bs_bs_query

551bs_bs_query revision. AD coordinated and conducted the expert consulta- Neuromuscular training injury prevention strategies in youth sport: 617bs_bs_query

552bs_bs_query tion, end-user engagement, feedback from early implementers a systematic review and meta-analysis. Br J Sports Med 2015;49: 618bs_bs_query

865-70. 619bs_bs_query 553bs_bs_query of FootyFirst and production of ﬁnal FootyFirst programme

16. Finch CF, Twomey DM, Fortington LV, Doyle TL, Elliott BC, Akram M, 620bs_bs_query 554bs_bs_query material. AD coordinated and all authors contributed to the

et al. Preventing Australian football injuries with a targeted neuromuscular 621bs_bs_query

555bs_bs_query evaluation against theory. DGL and WY coordinated the testing control exercise programme: comparative injury rates from a training 622bs_bs_query

556bs_bs_query of FootyFirst with support from BJG and JC. PW coordinated intervention delivered in a clustered randomised controlled trial. Inj Prev 623bs_bs_query

557bs_bs_query the early stages of the development of FootyFirst and gathered 2015. doi:10.1136/injuryprev-2015-041667 624bs_bs_query

17. Soligard T, Myklebust G, Steffen K, Holme I, Silvers H, Bizzini M, et al. 625bs_bs_query 558bs_bs_query the background information for the manuscript. AD conceived

Comprehensive warm-up programme to prevent injuries in young female 626bs_bs_query 559bs_bs_query the ‘6 step process for developing sports injury prevention

footballers: cluster randomised controlled trial. BMJ 2008;337:a2469. 627bs_bs_query

560bs_bs_query interventions’ with feedback from all authors. All authors con- doi:10.1136/bmj.a2469 628bs_bs_query

561bs_bs_query tributed to early drafts of the manuscript and have read and 18. Haynes RB, Devereaux PJ, Guyatt GH. Clinical expertise in the era 629bs_bs_query

562bs_bs_query approved the ﬁnal version of the manuscript, and agree with the of evidence-based medicine and patient choice. Evid Based Med 2002;7: 630bs_bs_query

36-8. 631bs_bs_query 563bs_bs_query order of presentation of the authors.

19. Straus SE, McAlister FA. Evidence-based medicine: a commentary on 632bs_bs_query 564bs_bs_query

common criticisms. CMAJ 2000;163:837-41. 633bs_bs_query 565bs_bs_query Competing interests

20. Andrew N, Gabbe BJ, Cook J, Lloyd DG, Donnelly CJ, Nash C, et al. 634bs_bs_query 566bs_bs_query

bs_bs_query 567bs_bs_query None of the authors declare competing ﬁnancial interests. Could targeted exercise programmes prevent lower limb injury in 635

community Australian football? Sports Med 2013;43:751-63. bs_bs_query 568bs_bs_query 636

21. Donaldson A, Cook J, Gabbe B, Lloyd DG, Young W, Finch CF. Bridging 637bs_bs_query 569bs_bs_query Appendix: Supplementary material

bs_bs_query 570bs_bs_query the gap between content and context: establishing expert consensus on the 638

571bs_bs_query Supplementary data to this article can be found online at content of an exercise training program to prevent lower-limb injuries. Clin 639bs_bs_query

J Sport Med 2015;25:221-9. 640bs_bs_query 572bs_bs_query doi:10.1016/j.jshs.2016.01.029.

22. Donaldson A, Finch CF. Applying implementation science to sports injury 641bs_bs_query 573bs_bs_query

prevention. Br J Sports Med 2013;47:473-5. 642bs_bs_query 574bs_bs_query References

bs_bs_query 575bs_bs_query 23. Padua DA, Frank B, Donaldson A, de la Motte S, Cameron KL, Beutler AI, 643

576bs_bs_query 1. Bahr R, Thorborg K, Ekstrand J. Evidence-based hamstring injury et al. Seven steps for developing and implementing a preventive training 644bs_bs_query

577bs_bs_query prevention is not adopted by the majority of Champions League or program: lessons learned from JUMP-ACL and beyond. Clin Sports Med 645bs_bs_query

578bs_bs_query Norwegian Premier League football teams: the Nordic Hamstring survey. 2014;33:615-32. 646bs_bs_query

579bs_bs_query Br J Sports Med 2015;49:1466-71. 24. Donaldson A, Lloyd DG, Gabbe BJ, Cook J, Finch FC. We have the 647bs_bs_query

580bs_bs_query 2. Hanson D, Allegrante JP, Sleet DA, Finch CF. Research alone is not programme, what next? Planning the implementation of an injury preven- 648bs_bs_query

581bs_bs_query sufﬁcient to prevent sports injury. Br J Sports Med 2014;48:682-4. tion programme. Inj Prev 2016;doi:10.1136/injuryprev-2015-041737; Pub- 649bs_bs_query

582bs_bs_query 3. Twomey D, Finch C, Roediger E, Lloyd DG. Preventing lower limb lished Online First 19 January 2016. 650bs_bs_query

583bs_bs_query injuries: is the latest evidence being translated into the football ﬁeld? J Sci 25. Rogers E. Diffusion of innovations. 5th ed. New York, NY: Free Press; 651bs_bs_query

584bs_bs_query Med Sport 2009;12:452-6. 2003. 652bs_bs_query

585bs_bs_query 4. Finch C. A new framework for research leading to sports injury prevention. 26. Nutbeam D, Harris E. Theory in a nutshell: a practical guide to health 653bs_bs_query

586bs_bs_query J Sci Med Sport 2006;9:3-9. promotion theories. 2nd ed. Sydney: McGraw Hill; 2004.p.23-7. 654bs_bs_query

587bs_bs_query 5. Donaldson A, Finch CF. Planning for implementation and translation: seek 27. McGlashan AJ, Finch CF. The extent to which behavioural and social 655bs_bs_query

588bs_bs_query ﬁrst to understand the end-users’ perspectives. Br J Sports Med 2012;46: sciences theories and models are used in sport injury prevention research. 656bs_bs_query

589bs_bs_query 306-7. Sports Med 2010;40:841-58. 657bs_bs_query

590bs_bs_query 6. Bartholomew LK, Parcel GS, Kok G, Gottilieb NH, Fernandez ME. 28. Donaldson A, Poulos RG. Planning the diffusion of a neck-injury 658bs_bs_query

591bs_bs_query Planning health promotion programs. An intervention mapping approach. prevention programme among community rugby union coaches. Br J 659bs_bs_query

592bs_bs_query 3rd ed. San Francisco: Jossey-Bass; 2011. Sports Med 2014;48:151-9. 660bs_bs_query

593bs_bs_query 7. van Mechelen W, Hlobil H, Kemper HC. Incidence, severity, aetiology and 29. Donaldson A. FootyFirst Program. [NOGAPS: National Guidance for Aus- 661bs_bs_query

594bs_bs_query prevention of sports injuries: a review of concepts. Sports Med 1992;14: tralian Football Partnerships and Safety]. June 17, 2015. Available at: 662bs_bs_query

595bs_bs_query 82-99. https://footyﬁrstaustralia.wordpress.com/footyﬁrst-program/ [accessed 663bs_bs_query

596bs_bs_query 8. Chalmers DJ. Injury prevention in sport: not yet part of the game? Inj Prev 01.2016]. 664bs_bs_query

597bs_bs_query 2002;8:22-5. 30. Malterud K. The art and science of clinical knowledge: evidence beyond 665bs_bs_query

598bs_bs_query 9. Klügl M, Shrier I, McBain K, Shultz R, Meeuwisse WH, Garza D, et al. measures and numbers. Lancet 2001;358:397-400. 666bs_bs_query

599bs_bs_query The prevention of sport injury: an analysis of 12,000 published manu- 31. Finch CF, Gabbe BJ, Lloyd DG, Cook J, Young W, Nicholson M, et al. 667bs_bs_query

600bs_bs_query scripts. Clin J Sport Med 2010;20:407-12. Towards a national sports safety strategy: addressing facilitators and 668bs_bs_query

601bs_bs_query 10. Schaalma H, Kok G. Decoding health education interventions: the times barriers towards safety guideline uptake. Inj Prev 2011;17:1-10. 669bs_bs_query

602bs_bs_query are a-changin’. Psychol Health 2009;24:5-9. 32. Donnelly CJ, Elliott BC, Ackland TR, Doyle TL, Beiser TF, Finch CF, 670bs_bs_query

603bs_bs_query 11. Wilson-Simmons R, O’Donnell L. Encouraging adoption of science-based et al. An anterior cruciate ligament injury prevention framework: 671bs_bs_query

604bs_bs_query interventions: organizational and community issues. In: Doll L, Bonzo S, incorporating the recent evidence. Res Sports Med 2012;20:239-62. 672bs_bs_query

605bs_bs_query Sleet D, editors. Handbook of injury and violence prevention. New York: 33. Sanson-Fisher RW, Campbell EM, Htun AT, Bailey LJ, Millar CJ. We are 673bs_bs_query + 606bs_bs_query Springer Science Business Media; 2007.p.511-25. what we do: research outputs of public health. AmJPrevMed 674bs_bs_query

607bs_bs_query 12. Finch CF, Gabbe B, White P, Lloyd D, Twomey D, Donaldson A, et al. 2008;35:380-5. 675bs_bs_query

608bs_bs_query Priorities for investment in injury prevention in community Australian 34. Kitzinger J. Qualitative research. Introducing focus groups. BMJ 1995; 676bs_bs_query

609bs_bs_query Football. Clin J Sport Med 2013;23:430-8. 311:299-302. 677bs_bs_query