STUDY PROTOCOL Open AccessEngaging Pediatric Intensive Care Unit (PICU)clinical staff to lead practice improvement: thePICU Participatory Action Research Project(PICU-PAR)Jean-Paul Collet1,2,3, Peter W Skippen1,3,4, Mir Kaber Mosavianpour1,3, Alexander Pitfield1,4, Bubli Chakraborty7,Garth Hunte6, Ronald Lindstrom8, Niranjan Kissoon1,2,3 and William H McKellin5*AbstractBackground: Despite considerable efforts, engaging staff to lead quality improvement activities in practice settings is apersistent challenge. At British Columbia Children’s Hospital (BCCH), the pediatric intensive care unit (PICU) undertook anew phase of quality improvement actions based on the Community of Practice (CoP) model with Participatory ActionResearch (PAR). This approach aims to mobilize the PICU ‘community’ as a whole with a focus on practice; namely, tocreate a ‘community of practice’ to support reflection, learning, and innovation in everyday work.Methodology: An iterative two-stage PAR process using mixed methods has been developed among the PICU CoPto describe the environment (stage 1) and implement specific interventions (stage 2). Stage 1 is ethnographic descriptionof the unit’s care practice. Surveys, interviews, focus groups, and direct observations describe the clinical staff’sexperiences and perspectives around bedside care and quality endeavors in the PICU. Contrasts and comparisons acrossparticipants, time and activities help understanding the PICU culture and experience. Stage 2 is a succession of PARspirals, using results from phase 1 to set up specific interventions aimed at building the staff’s capability to conduct QIprojects while acquiring appropriate technical skills and leadership capacity (primary outcome). Team communication,information, and interaction will be enhanced through a knowledge exchange (KE) and a wireless network of iPADs.Relevance: Lack of leadership at the staff level in order to improve daily practice is a recognized challenge that facesmany hospitals. We believe that the PAR approach within a highly motivated CoP is a sound method to create the socialdynamic and cultural context within which clinical teams can grow, reflect, innovate and feel proud to better servepatients.Keywords: PICU, ICU, Quality improvement, Community of practice, Participatory action research, Distributed leadership,Engagement, Learning community, Reflective practice, ChildrenBackgroundNational reports have highlighted undesirable practicevariations leading to sub-optimal healthcare, medicalerrors and adverse patient outcomes [1-6]. Numerousinterventions to standardize practice have been developedbased on well-established theories and techniques[7-11]. However, despite considerable efforts nationwide,the change in practice quality has been ‘frustratinglylow’ [6], and staff engagement in practice settings to leadquality improvement activities is a persistent challenge.Over the past decade, the Pediatric Intensive CareUnit (PICU) at British Columbia Children’s Hospital(BCCH), as member of the Canadian Pediatric CriticalCare Collaborative, experienced a track record of suc-cesses following the quality improvement approach withrepeated Plan-Do-Study-Act (PDSA) cycles. In 2009, theProvincial Health Services Authority, which governsBCCH, adopted the LEAN methodology [12-18] in an* Correspondence: mcke@mail.ubc.ca5Department of Anthropology, University of British Columbia, 6303 NWMarine Drive, Vancouver, BC, CanadaFull list of author information is available at the end of the articleImplementationScience© 2014 Collet et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited. The Creative Commons Public Domain Dedicationwaiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwisestated.Collet et al. Implementation Science 2014, 9:6http://www.implementationscience.com/content/9/1/6effort to improve quality and efficiency to the overallhealthcare system. LEAN Leaders and the PICU teamcompleted 23 Rapid Process Improvement Workshops(RPIWs) over four years. While the initial efforts wereimpressive, longer-term success was mixed becausegains from only 9 of 23 RPIW projects (about 40%)have been sustained at one year. The reason wasattributed in part to the way RPIW projects wereimplemented with unit and organizational leadershipselecting isolated component deficiencies withoutseeking input and engagement of the broad team ofcare providers and staff.Considering the complexity of the PICU environment,a third phase of QI was conceived based on theCommunity of Practice (CoP) model with ParticipatoryAction Research (PAR) [19-26] to support reflection,learning and innovation in everyday work [23-26].RationaleThe reason for this approach is based on the premisethat in complex adaptive systems [27-35] such as thePICU, one approach to promote changes consists of mo-bilizing the community as a whole with a focus on prac-tice; namely, to create a CoP [23,24]. Within our CoP,we aim to create the conditions of a collaborative, re-flective and innovative experiential system [36] that willenable collective discussions around daily practice issuesand finding solutions for improvement by integratingtacit-explicit knowledge [20]. Further, there is a growingbody of literature that supports the active engagement ofall staff when undertaking change management initia-tives in hospitals, particularly in relation to patient safetyand quality [35,37,38]. As Lindstrom suggests, ‘front-lineownership of the problems and, more importantly,collective solutions has highlighted the importance ofand effectiveness of distributed leadership’ [37]. Andwhen it concerns staff engagement, Van de Ven observesthat there are three key principles regarding the gapbetween what the theory says and what actually happensin practice [39]:1. Translating knowledge into practice requires a muchbetter understanding of how to engage stakeholdersand communicate across stakeholder knowledgeboundaries;2. Scientific knowledge and practical knowledge aredistinct, and that a more pluralist view is required,one that allows complementary perspectives tounderstand reality;3. Knowledge production is a problem because it hasreflected scientific inquiry that has often notengaged stakeholders other than researchers; thus,the research is not appropriately grounded inday- to-day realities and generates little impact.Consequently, there is a pressing need for staff engage-ment. Consistent with a PAR approach, this involvesengaging and working with, not on or for, clinical staff –those closest to and on the frontline of healthcare.PAR approachPAR is comprised of three basic elements [20,40]:1. Participation broadens who participates in theresearch process [22], in this case multiplestakeholders comprised of researchers, clinicaland managerial decision-makers, and also patients/family members;2. Action, which is emphasized over just generatingnew knowledge [22];3. Research focuses on perspectives locally definedby, e.g., decision-makers; shares power betweenresearchers and decision-makers; expands thepurview of knowledge generation from academiato the community; and realigns the researchers’role from directing to facilitating the process [22].This set of elements constitutes the collaborative, col-lective nature of PAR. As Greenwood and Levin assert,this is ‘co-generative inquiry because it is built on pro-fessional researcher-stakeholder collaboration and aimsto solve real-life problems in context. Co-generativeinquiry processes involve trained professional researchersand knowledgeable local stakeholders who work togetherto define the problems to be addressed, to gather andorganize relevant knowledge and data, to analyze theresulting information, and to design social change inter-ventions. Together these partners create a powerful re-search team’ ([41] p. 54). Unlike traditional research, PARdeliberately intervenes in the research setting [21]. This isan important distinction. However, scholars such asGreenwood and Levin caution against PAR as a short-term intervention; rather, it is a ‘continuous and participa-tive learning process’ ([20] p.18). This is an importantpoint, which is explicitly factored into our study and thecontinuous nature of QI and which underscores theimportant role of CoPs.Research questionsOur PAR approach addresses four research questions:1. How do frontline clinical staff and decision-makersidentify and conceptualize improving the quality ofcare a patient receives?2. Currently, in what ways do frontline clinical staffengage in QI activities?3. What resources, supports, enablers and capacity arerequired for frontline clinical staff to increase theirengagement in QI to lead and conduct projects?Collet et al. Implementation Science 2014, 9:6 Page 2 of 10http://www.implementationscience.com/content/9/1/64. What specific strategies are required to address theneeds identified by frontline clinical staff?Overall research goalTo sustain success in improving everyday practice in thePICU by building and supporting staff ’s individual andcollective capacity to conduct QI projects in an environ-ment of distributed leadership.Specific objectives1. To determine frontline clinical staff ’s engagement,collective capacity and ability to conduct projectsfor daily practice improvement.2. To develop and support staff ’s collective engagement,ability and leadership to improve practice.3. To ensure sustainability of the change within thecontext of a new PICU culture of distributedleadership.4. To generate a PAR practice QI framework overthe course of the study.DesignAn iterative two-stage PAR process employing mixedmethods will be used to explore the PICU environmentand to implement specific interventions within the con-text of an emerging CoP. The overall research approachis schematically illustrated in Figures 1 and 2 with theevaluation framework presented in Figure 3.Stage 1 Describe the collective action of the PICU usingartifacts, perceptions and patterns in order to offer anaccount of action in everyday practiceWe will be employing several strategies during the ob-servational phase of the study that will provide us withan ethnographic description of the practice of care. Wewill be enquiring into the experiences and perspectivesof clinical staff around bedside care and quality en-deavors in the PICU. Contrasts and comparisons acrossparticipants, time and activities, will help understandingPICU culture and experience. The mixed methods forPICU system description include both qualitative andquantitative methods:PARStudy & PlanTake ActionReflect- Frontline staff leadership & engagement?- Satisfaction?- Functional networks- Staff-led QI training projects –success? Challenges?- Use of communication infrastructure and dashboard?- Activity of the PQIS Team?EducationCommunicationReflection on practiceCoaching/MentoringConducting 15 staff-led specific QI training projects PAR SPIRAL INTERVENTIONCollect & AnalyzeEvidenceFigure 1 PICU-PAR project description – PAR spiral and specific interventions.Collet et al. Implementation Science 2014, 9:6 Page 3 of 10http://www.implementationscience.com/content/9/1/6Qualitative methodsObservationWe will observe communication and patterns of inter-action among the PICU staff, patients, and their parents[42-44] and evaluate effects of these patterns on thestaff ’s ability to engage and lead during routine activitiessuch as handover rounds and the orientation of newstaff. All QI activities conducted in the past three yearswill be reviewed and described with regard to the type ofproject, initiator, external support, staff involvement,output and sustainability. This information will be dis-cussed with staff as a way to engage discussion aboutimprovement and changes.InterviewsIndividuals and groups will be engaged in semi-structuredinterviews that will enable staff to contextualize the issuesof quality improvement with concrete examples from theirown patterns of practice. Some families will also be inter-viewed in regard to communication with staff and to ob-tain their perception of care quality. The interviews willalso address the issue of ‘readiness to change’ [45-47] toinform optimal choice of possible interventions [48].Finally, during the semi-structured interviews, participantswill be asked to diagram their own networks of individualswhom they recognize as leaders in different areas of qual-ity improvement. This will contribute to the quantitativeanalysis of social networks within the PICU [49-53] andhelp in developing future interventions strategies.Focus groupsThe dynamic aspect of group discussions will be used toidentify solutions. One set of focus groups will be multidis-ciplinary, composed of individuals representing differentdisciplines, while the second set will be discipline specific.At least one focus group will be conducted with families.Quantitative methodsStaff engagement surveyEngagement will be assessed using the ‘Employee EngagementQuestionnaire’ [54], a validated tool that has been widelyInvolving frontline staff in improving practice quality & safetySITUATION ANALYSEDINTERVENTION DEVELOPEDINTERVENTIONSTUDIEDINTERVENTIONADAPTEDINTERVENTION PLANNEDPARTEAM WORKKnowledge users: clinicians, policy makers, QI expertsResearchers: anthropologist, psychologist, QI, clinical researchEngagementLeadershipEvaluationEvaluationEvaluationBASELINEFINALImprovementImprovementImprovementFigure 2 PAR description with teamwork process and reflection for improved adapted intervention.Collet et al. Implementation Science 2014, 9:6 Page 4 of 10http://www.implementationscience.com/content/9/1/6used to assess staff engagement [54], including on ayearly basis in the PICU for the past three years.Staff leadership surveysFrontline staff leadership will be assessed with twoinstruments.1. The ‘Healthcare Team Vitality Instrument’ (HTVI):This is a short survey developed by the Institute ofHealthcare Improvement as part of the initiativeTransforming Care at Bedside. It measures teamvitality with an emphasis on front-line staff em-powerment and engagement, perception of a workenvironment, supportive, effective communication,and team collaboration. The HTVI has been vali-dated [55] and is widely used in North America, asit has been in our PICU during the last three years,therefore providing a good reference point againstwhich changes can be assessed.2. Leadership abilities: Leadership capacity will beassessed by employing a questionnaire inconjunction with the LEADS in a CaringEnvironment framework [56], which is being usedby BCCH to assess leadership abilities andimprovement during specific interventions.Network analysis surveysSocial network analysis will be used to identify existingnetworks in the unit. Social network information is im-portant to understand the unit’s functioning, patterns ofcommunication, and for planning effective interventionsto support organizational changes at a later stage [57,58].Network analysis can also be repeated over the course ofthe project to assess changes over time [49-53,57].Stage 1 SamplingAll PICU staff will be invited to participate. Surveys willbe offered to everyone, while personal interviews andfocus group discussions will identify specific sub-groupsbased on categories of personnel and seniority in theunit. Ethnographic observations and network analysiswill also require specific sampling strategies.Problem:Improving practice is not part of daily workGoal: To improve quality of patient care in PICU by developing a model of frontline staff engagement and leadership to improve care practice quality as part of their daily workLogic Model-Evaluation FrameworkPriorities Inputs Activities Outputs OutcomesEVALUATION 1-Continue current quality improvement activities1-Conducting PICU-led QI activities for Action-Learning1- More projects conducted = more staff trained1-Number of PICU- led and conducted QI projects w/o external support (autonomy)6-Ensuring sustainability of the change6-Develop a sustainability plan3-Increase staffs’ capability to identify situation to improve4--Increase the staffs’ capability to conduct QI interventions through Action Research Learning process.5-Increase staff engagement and leadership for QI 5-Participatory Action Research (PAR) to build frontline staff collective capacity and capability to engage in improve care practice and play leadership role3- Staff engaged in leading QI initiative good mentorship = learn how to improve practice4- Communication infrastructure for education and QI support 5- Educational material6-Sustainability plan (activity # 6)6-PICU QI structure sustainability1. Individually: Capacity lead QI interventions2. Collectively: Capacity to improve the system3. A new care practice model Long term outcome:2-Understanding PICU current state re. QI:- Persons’ experience and background- Workplace characteristics- Current and past QI activities- Attitudes towards QI- Current QI organizational structure- QI Knowledge and skills3-Existing documents regarding previous QI activities4-Documenting current QI capacity 2-Strengths and weaknesses regarding staff engagement & capability in improving practice as part of daily work2-Number of staff able to lead QI activities 3- Staff’s engagement knowledge, and leadership4- Staff’s perception & satisfaction5- Staff’s attitude re. quality and safety2-Ethnographic assessment: interviews; surveys; observations & focus groups7-Knowledge Transfer7- PICU KT plan7-Develop a KT plan 7-KT plan (activity # 7)2-Expertise: anthropology; QI, leadership, evaluation 3-Partnership : - Hospital: support from Teaching & Learning- PICU team- Partners in care (families)- Pediatrics Dept- InspireNet- Royal Roads Univ.- CFRI5-imPROVE structure (PHSA) support4-Funding6-BC Patient Safety and Quality Council 7-CanadianAssociation of Pediatric HeatlhCenters (CAPHC)1-Exisging trained staff and QI resourcesFigure 3 Logic Model for PICU-PAR project evaluation framework.Collet et al. Implementation Science 2014, 9:6 Page 5 of 10http://www.implementationscience.com/content/9/1/6Stage 1 data analysis1. Observation: Audio transcripts (and possibly videorecordings) will be analyzed using conversationalanalysis methods [50,59] to identify the distributionof knowledge and collaboration among membersof the staff [60], to augment topics for qualitativeinterviews, and as a vehicle for reflection andself-assessment [44,61-63].2. Interviews and focus groups: The semi-structuredinterviews will be transcribed, and entered intoATLAS-ti, a qualitative database for coding. We willemploy techniques from conversational analysis[59,64] and cognitive linguistic discourse analysis[65] to describe the themes and identify informationschemas in the interview data [66]. The analysisprocess will be continuous; new data will becompared to existing ones to identify emerginginformation (themes or ideas) with the ultimateobjective to create a body of knowledge regardingthe PICU attitudes and thoughts. The interviewiterative process will continue until saturation isreached, defined as no new information or themesabout a specific topic in three consecutive interviews[67-69], and clear assertions can be made regardingthe topic studied. For focus groups, the analysis willalso consider the dynamics of the discussions andthe role of the individuals.3. Survey questionnaires: Because one importantobjective is communicating results to studyparticipants and decision-makers, special attentionwill be given to descriptive statistics, using pointestimates and 95% confidence intervals, as well asfrequency tables and graphs.4. Social network analysis: We will use UCINET 6 [70]to illustrate the different relational and structuralmeasures. Sociograms, other forms of networkdiagrams, and matrices will depict the relationshipand structure of social and informational networkson the unit. Before-after intervention will becontrasted with regard to the type of measuresidentified.Stage 2 The PAR spiralsBuilding the communication structureOne specific consideration is the development of a posi-tive environment to facilitate and encourage reflectionsabout actual care practices and possible improvement.With appropriate coaching and support, a CoP can beexpected to develop progressively [24,71], enriched bythe participation of all PICU members who will bringtheir specific insight on practice and improvement. Twostaff members who have extensive expertise in this areawill lead a ‘PICU QI Support Team’, which will be astructuring element to guide and support the CoP’sfunctions. Intervention in the context of PAR will alsobenefit from a solid knowledge exchange (KE) that facili-tates communication among frontline staff, decision-makers, quality leaders and researchers in a continuous,multi-directional way. A web-based dashboard will visu-ally describe the PAR project’s status and the evolutionof each QI change project. Periodic reports, posters andnewsletters will circulate and remain posted on the studydashboard to facilitate discussions with the whole team.The acquisition of iPAD tablets will enable flexibleaccess to the study dashboard and create new communi-cation streams with nurses who primarily stay at thepatient’s bedside. The iPAD tablets will also be used fordata collection during quality change interventions –with automated generation of run charts and controlcharts when new data are entered.Identifying and implementing specific QI interventionsSpecific interventions will be incorporated accordingto the needs/gaps identified during Stage 1 and con-textual factors. The main educational/training featureswill be ‘action learning’ [72] through engaging PICUstaff leading their own QI change interventions. Withappropriate coaching, mentoring and communication,we expect care providers both to individually engagein everyday practice improvement, and collectivelylead changes toward a new emerging practice culture.Webinars and educational documents will be storedin a library with remote access through the studydashboard.Continuous evaluation and reflectionContinuous evaluation is an integral part of a PAR torefine, modify and adjust the specific interventions asper action research principles [25,26]. Staff capabilities,skills, leadership, and engagement will be assessed regu-larly, with a main review conducted every six months toguide the PAR cycles. Relevant performance metrics willbe used to visualize results and share the informationwith staff. New interventions will then be identified andrefined for a new PAR spiral.Project outcomesThe different outcomes are structured according to theproject’s evaluation framework presented in Figure 3.The following is a list of key areas that are expected toshow effects, with related indicators.1. PICU frontline staff ’s current state and actualfunctioning with regard to quality improvement:QI definition and model used; QI practice andexperience; Existing QI knowledge and support;Perceived barriers and facilitators for QI.Collet et al. Implementation Science 2014, 9:6 Page 6 of 10http://www.implementationscience.com/content/9/1/62. Capacities developed to support frontline staff toengage and lead actions to improve daily practice:Development of educational and training materialsas well as an interactive communication structurethrough the study dashboard; Development of newPICU and hospital policies that encourage and valuestaff ’s reflections about care practice andimprovement.3. New frontline staff capabilities to initiate andconduct practice improvement interventions: Staffengagement in QI initiatives will be assessedthrough the following: number of projects developedand completed by bedside clinicians, changes inengagement and leadership survey scores, andnumber of visits to the quality dashboard. Theultimate goal of these changes is to affect the qualityof patient care, but this outcome cannot bemeasured within a three-year span.4. Frontline staff capabilities and willingness to becomeleaders in reflective practice: PICU staff self-efficacyand self-confidence leading changes as well assatisfaction with the new culture of distributedleadership will be assessed through interviewsand focus groups.Analysis of the impactThe CoP-PAR impact will be assessed through qualita-tive and quantitative approaches assessing before-afterchanges in important outcomes (engagement, collectiveactivities, leadership, and barriers). Factors associatedwith behavioral changes will be identified through multi-variate modeling. The analysis will also consider the highinter-variables correlations and the numerous tests toprevent misinterpretation of the study results.Knowledge exchange (KE) and sustainabilityDuring the course of the study, the KE and dashboardmaterial will ensure rapid exchange of relevant informa-tion among PICU staff. On-going knowledge transferactivities will be planned for specific stakeholders,guided by Lavis’s questions [73], to inform them of thestudy results.The main strength of our sustainability plan lies in a‘Training the trainers’ process, with new QI experts sup-porting and training other staff. It also includes: keepingthe KE website and dashboard active for communicationsupport after the study; maintaining access to local supportresources that remain through the project’s partners; or-ganizing future coaching and leadership sessions to assiststaff ’s capacity to retain their skills and develop new ones.EthicsUniversity of British Columbia ethics review board hasreviewed and given approval for the project. The mainethical issue was related to the need to clearly separatemanagers from clinical staff in organizing sampling andinterviews. Although managers will help, they will not beinvolved in identifying and contacting frontline workers.3monthsBuilding PARTeamworkAnalyzing PICU SystemPAR InterventionPAR EndingCONTINUOUSEVALUATIONResearchers Knowledge UsersSustainability Knowledge TransferPAR Spiral5 months20 months3months15 monthsFigure 4 PICU-PAR Research Plan with stages and timeline.Collet et al. Implementation Science 2014, 9:6 Page 7 of 10http://www.implementationscience.com/content/9/1/6Timeline and study statusThis is a three-year project with six months for the initialPICU system description (phase 1) and the remaining fordeveloping the PAR approach employing three to fourspirals. During the last six months of the project, KE andsustainability will become a high priority. The projectstarted in August 2012 with meetings of care providers,PICU managers, hospital executive and managers andresearchers. Our methodology was developed over fourmonths, and ethical approval was granted two monthslater. The initial interviews and surveys commenced April2013. Figure 4 shows the study timeline.DiscussionLack of staff leadership to improve daily practice is arecognized challenge that faces many hospitals. We be-lieve that the PAR approach within active and motivatedCoPs is a sound methodology to create the social cultureand context within which clinical teams can grow,reflect, innovate and feel proud to better serve patients.CoP is a recognized model for communities to learnand improve [23,24,74,75]. In particular, the sharing ofexperiences and ideas among CoP members is a sourceof reflection that enriches learning and subsequentchanging practice [23,24,76].However, two systematic reviews on CoP did not findstrong evidence supporting its positive effect [77,78].One reason may be the way in which CoPs generallydevelop and function with voluntary participation andinformal communication around practical experience(social networks type); the primary objective being toimprove healthcare, not to publish. Consequently, we donot yet know the actual effect of CoPs toward improvingpractice.Our PAR approach enables the co-generation ofpractical data to facilitate future development. Thisis supported by different socio-cultural theories suchas Situated Learning [24,79], Activity Theory [80-85],and Distributed Cognition [62,81], all linked to thebroad concept of ‘Learning Communities’ and rooted toVygotsky’s landmark work on social learning [84,85]. Weexpect that, in the context of PAR, the CoP will identifygaps, defects and challenges and will also generateinnovative solutions to be implemented through a se-quence of improvement spirals [25,26,86-88].With this approach, we aim to better understand howto increase PICU frontline staff engagement and leader-ship to improve care, and to assess the overall impact ofintroducing a PAR process within a well-structured CoP.Throughout the study, we expect to obtain a deeperinsight into the CoP development process, including KEand sustainability. Our goal is to create a PAR approachand framework that is generic enough to be used inother units yet flexible enough to meet the units’characteristics and specific needs. If successful, weexpect this framework to be adopted and refined inother units at BCCH and other pediatric centres acrossCanada.Developing the PICU PAR approach was also a learn-ing experience for all participants. It is the product of amultidisciplinary team of clinicians, managers, policy-makers, and researchers who all desire to activelyparticipate. Stakeholder commitment constitutes thefoundation on which PAR can emerge; it requires partic-ipants to dedicate time, to be open-minded to under-stand/accept the different points of view, to accept adegree of uncertainty and to be interested in the process.Stakeholder commitment, therefore, can be taken as asign of interest and trust in the project and mutualrespect among team members. From our experience, wesee the PAR development process as a unique opportun-ity to strengthen the PICU team.AbbreviationsBCCH: British Columbia children’s hospital; CoP: Community of practice;HTVI: Healthcare team vitality instrument; KE: Knowledge exchange;PAR: Participatory action research; PICU: Pediatric intensive care unit;PDSA: Plan-do-study-act; PHSA: Provincial health services authority;RPIW: Rapid process improvement workshops.Competing interestsNo competing interests to declare.Authors’ contributionsJPC WM PS MKM AP initially developed the concepts with contributionsfrom BC GH RL NK. JPC wrote the initial draft with contributions from allauthors during the editing. All authors provided comments and approvedthe final version.AcknowledgementsWe wish to acknowledge the strong support from BC Children’s HospitalExecutive; PHSA imPROVE Program for Quality Improvement; PHSA Learningand Development; PHSA nursing facilitator; Inspire-Net; the BC Quality andSafety Council; Department of Pediatrics, University of British Columbia; Childand Family Research Institute (CFRI); The Canadian Association of PediatricHealth Centers (CAPHC); Mojgan Gitimoghaddam for logistical support anddashboard development. Dr Collet is partly supported by a CFRI clinician-investigator award; Dr Lindstrom is a Henri M Toupin Research Fellow inHealth Systems Leadership at Royal Roads University.Author details1Department of Pediatrics, University of British Columbia, 4480 Oak Street,Vancouver, British Columbia, Canada. 2British Columbia Children’s Hospital,4480 Oak Street, Vancouver, British Columbia, Canada. 3Child and FamilyResearch Institute, 950 West 28th Avenue, Vancouver, British Columbia,Canada. 4Pediatric Intensive Care Unit, BC Children’s Hospital, 4480 OakStreet, Vancouver, British Columbia, Canada. 5Department of Anthropology,University of British Columbia, 6303 NW Marine Drive, Vancouver, BC,Canada. 6Emergency Department, Saint Paul’s Hospital and Department ofEmergency Medicine, University of British Columbia, 1081 Burrard Street,Vancouver, British Columbia, Canada. 7Provincial Health Services Authority,1380 Burrard Street, Vancouver, British Columbia, Canada. 8School ofLeadership Studies, Royal Roads University, 2005 Sooke Road, Victoria, BritishColumbia, Canada.Received: 2 October 2013 Accepted: 27 December 2013Published: 8 January 2014Collet et al. Implementation Science 2014, 9:6 Page 8 of 10http://www.implementationscience.com/content/9/1/6References1. Baker GR, Norton PG, Flintoft V, et al: The Canadian adverse events study:the incidence of adverse events among hospital patients in Canada.CMAJ 2004, 170(11):1678–1686.2. Brennan TA, Leape LL, Laird NM, et al: Incidence of adverse events andnegligence in hospitalized. Qual Saf Health Care 2004, 324:145–152.3. Leape LL, Brennan TA, Laird NAN, et al: The nature of adverse events inhospitalized patients results of the Harvard medical practice study II.N Engl J Med 1991, 324(6):377–384.4. Vincent C, Neale Graham WM: Adverse events in British hospitals:preliminary retrospective. BMJ 2012, 322(7285):517–519.5. Wilson RM, Runciman WB, Gibberd RW, Harrison BT, Newby LHJ: Thequality in Australian health care study. Med J Aust 1995, 163(9):458–471.6. Institute of Medicine: Crossing the quality Chasm a new health system for the21st Century. Washington, DC: National Academy Press; 2001.7. Grol R, Berwick DM, Wensing M: On the trail of quality and safety inhealth care. BMJ 2008, 336(7635):74–76.8. Grimshaw J, Eccles M, Thomas R, et al: Toward evidence-based qualityimprovement. Evidence (and its limitations) of the effectiveness ofguideline dissemination and implementation strategies 1966–1998.J Gen Intern Med 2006, 21(2):S14–S20.9. Freeman C, Todd C, Camilleri-Ferrante C, et al: Quality improvement forpatients with hip fracture: experience from a multi-site audit. Qual SafHealth Care 2002, 11(3):239–245.10. Ring N, Malcolm C, Coull A, Murphy-Black T, Watterson A: Nursing bestpractice statements: an exploration of their implementation in clinicalpractice. J Clin Nurs 2005, 14(9):1048–1058.11. Barwick MA, Peters J, Boydell K: Getting to uptake: do communities ofpractice support the implementation of evidence-based practice?J Am Acad Child Adolesc Psychiatry 2009, 18(1):16–29.12. BC Ministry of Health: Lean in British Columbia’ s health sector: annual report2010-11]; 2011. [http://www.llbc.leg.bc.ca/public/pubdocs/bcdocs2012/517937/2010_2011.pdf]13. McCulloch P, Kreckler S, New S, Sheena Y, Handa A, Catchpole K: Effect of a“Lean” intervention to improve safety processes and outcomes on asurgical emergency unit. BMJ 2010, 341(1):c5469–c5469.14. Johnson JE, Smith AL, Mastro KA: From Toyota to the bedside: nurses canlead the lean way in health care reform. Nurs Adm Q 2012, 36(3):234–242.15. Poksinska B: The current state of lean implementation in health care:literature review. Qual Manag Health Care 2010, 19(4):319–329.16. Vats A, Goin KH, Villarreal MC, Yilmaz T, Fortenberry JD, Keskinocak P: Theimpact of a lean rounding process in a pediatric intensive care unit.Crit Care Med 2012, 40(2):608–617.17. Toussaint JS, Berry LL: The promise of lean in health care 2013.Mayo Clin Proc 2013, 88(1):74–82.18. Powell AE, Rushmer RK, Davies HTO: A systematic narrative review of qualityimprovement models in health care; 2008. [http://www.healthcareimprovementscotland.org/his/idoc.ashx?docid=ba9d888b-8837-4d57-8842-b0493b80a2a5&version=-1]19. Cornwall A, Jewkes R: What is participatory research? Soc Sci Med 1995,41(12):1667–1676.20. Greenwood DJ, Levin M: Introduction to action research: social research forsocial change. Thousand Oaks, CA: Sage; 1998.21. Herr K, Anderson GL: The Action Research Dissertation: a guide for studentsand faculty. Thousand Oaks, CA: Sage; 2005.22. Reason P, Bradbury H: Editorial. Action Res 2008, 6(1):5–6.23. Wenger E: Communities of practice and social learning systems.Organization 2000, 7(2):225–246.24. Wenger E, McDermott RA, Snyder WC: Cultivating communities of practice: a guideto managing knowledge. Cambridge, MA: Harvard Business School Press; 2002.25. Robin M: Principles for participatory action research. Adult Educ Q 1991,4((3):168–187.26. Riel M: Understanding action research. In Center for collaborative actionresearch. Pepperdine University; 2010. [http://www.actionlearning.com.au/Classes/ActionResearch/Articles/Pepperdine-UnderstandingActionResearch.doc]27. Ellis B, Herbert SI: Complex adaptive systems (CAS): an overview of keyelements, characteristics and application to management theory.Inform Prim Care 2011, 19(1):33–37.28. Ellis B: Complexity in practice: understanding primary care as a complexadaptive system. Inform Prim Care 2010, 18(2):135–140.29. Plsek PE, Greenhalgh T: Complexity science: the challenge of complexityin health care. BMJ 2001, 323(7313):625–628.30. Plsek PE, Wilson T: Complexity science Complexity, leadership, andmanagement in healthcare organisations four articles. BMJ 2001,323(7315):746–749.31. Glouberman S, Zimmerman B: Complicated and complex systems: Whatwould successful reform of medicare look like?. Toronto: Commission on theFuture of Health Care in Canada; 2002. [http://publications.gc.ca/collections/Collection/CP32-79-8-2002E.pdf]32. Tan J, Wen HJ, Awad N: Health care and services delivery systems ascomplex adaptive systems: commun ACM. 2005, 48(5):36–44.33. Zimmerman B, Lindberg C, Plsek P: Edgeware: insights from complexityscience for health care leader. Irving TX: VHA Publishing; 1998.34. Begun JW, Zimmerman B, Dooley K: Health care organizations as complexadaptive systems. In Advances in health care organization theory. Edited byMick SM, Wyttenbach M. San Francisco: Jossey-Bass; 2003:253–288.35. Zimmerman B, Reason P, Rykert L, Gitterman L, Christian J, Gardam M:Front-line ownership: generating a cure mindset for patient safety.Healthc Pap 2013, 13(1):6–22.36. Orlikowski WJ: Knowing In practice: enacting a collective capability indistributed organizing. Organ Sci 2002, 13(3):249–273.37. Lindstrom R: Leadership needs to shift in the health system: threeemerging perspectives to inform our way forward. Healthc Pap 2013,13(1):48–54.38. Baker GR, Denis J-L: Medical leadership in health care systems: fromprofessional authority to organizational leadership. Public Money Manag2011, 31(5):355–362.39. Van de Ven A: Engaged scholarship: a guide for organizational and socialresearch. New York: Oxford University Press; 2007.40. Wallerstein N, Duran B: The conceptual, historical, and practice roots ofcommunity based participatory research and related participatorytraditions. In Community-based participatory research for health. Edited byMinkler M, Wallerstein N. San Francisco, CA: Jossey-Bass; 2003:27–52.41. Greewood DJ, Levin M: Reform of the social sciences and of universitiesthrough action research. In The Sage handbook of qualitative research. 3rdedition. Edited by Denzin NK, Lincoln YS. Thousand Oaks, CA: Sage; 2005:43–64.42. Cicourel AV: Cognitive overload and communication in two healthcaresettings. Commun Med 2004, 1(1):35–43.43. Iedema R, Mesman JCK: Visualizing health care practice improvement:innovation from within. Oxford: Radcliffe Publishing Ltd; 2013.44. Carroll K, Iedema R, Kerridge R: Reshaping ICU ward round practices usingvideo-reflexive ethnography. Qual Health Res 2008, 18(3):380–390.45. Norcross JC, Krebs PM, Prochaska JO: Stages of change. J Clin Psychol 2011,67(2):143–154.46. Di Noia J, Prochaska JO: Dietary stages of change and decisional balance:a meta-analytic review. Am J Health Behav 2010, 34(5):618–632.47. Lam CS, Wiley AH, Siu A, Emmett J: Assessing readiness to work from astages of change perspective: implications for return to work. Work J PrevAssess Rehabil 2010, 37(3):321–329.48. Dickson G, Lindstrom R, Black C, Van der Gucht D: Evidence-informed changemanagement in Canadian healthcare organizations. Ottawa, ON: CanadianHealth Services Research Foundation; 2012.49. Carrington PJ, Scott J, Wasserman SM: Models and methods in social networkanalysis. Cambridge: Cambridge University Press; 2005.50. Hogan B, Carrasco JA, Wellman B: Visualizing personal networks: workingwith participant-aided sociograms. Field Methods 2007, 19(2):116–144.51. Scott J: The Sage handbook of social network analysis. Thousand Oaks:Calif. Sage Publications; 2011.52. Walker ME, Wasserman S, Wellman B: Statistical models for social supportnetworks. Sociol Methods Res 1993, 22(1):71–98.53. Barry W, Janet S, Dimitrina D, Laura G, Gulia Milena HC: Computer networksas social networks: collaborative work, telework, and virtual community.Annu Rev Sociol 1996, 22:213–238.54. Harter JK, Ph D, Schmidt FL, Killham EA, Asplund JW: Q 12® Meta-Analysis.Gallup Inc: Omaha NB; 2006.55. Upenieks VV, Lee EA, Flanagan ME, Doebbeling BN: Healthcare TeamVitality Instrument (HTVI): developing a tool assessing healthcare teamfunctioning. J Adv Nurs 2010, 66(1):168–176.56. CCHL: Canadian college of health leaders. 2013 [http://www.cchl-ccls.ca].57. Thomas WV: Social networks and health: models, methods and application.New York: Oxford University Press; 2010.58. Cross R, Laseter T, Parker A, Velasquez G: Assessing and improvingcommunities of practice with organizational network analysis. TheCollet et al. Implementation Science 2014, 9:6 Page 9 of 10http://www.implementationscience.com/content/9/1/6Network Roundtable at the University of Virginia; 2004. [http://www.robcross.org/pdf/roundtable/formalizing_communities_of_practice.pdf]59. McKellin WH, Shahin K, Hodgson M, Jamieson J, Pichora-Fuller K: Pragmaticsof conversation and communication in noisy settings. J Pragmat 2007,39(12):2159–2184.60. Hutchins E: The distributed cognition perspective on human interaction.In In roots of human sociality: culture, cognition and interaction. Edited byEnfield NJ, Levinson SC. Berg: Oxford; 2006:375–398.61. Henry SG, Fetters MD: Video elicitation interviews: a qualitative researchmethod for investigating physician-patient interactions. Ann Fam Med 2012,10(2):118–125.62. Iedema R, Long D, Forsyth R: Visibilising clinical work: video ethnographyin the contemporary hospital. Health Sociol Rev 2006, 15(2):156–168.63. Iedema R, Merrick T, Rajbhandari D, Gardo ASA: Viewing the taken-for-granted from under a different aspect: a video-based method in pursuitof patient safety. Int J Multiple Res Approaches 2009, 3(3):290–301.64. Heritage J: Conversation analysis and institutional talk. In Handbook oflanguage and social interaction. Edited by Fitch KL, Sanders RE. Mahwah, NJ:Lawrence Erlbaum Associates; 2005:103–47.65. Waugh LR, Fonseca-Greber B, Vickers C, Eröz B: Multiple empiricalapproaches to a complex analysis of discourse. In Methods in cognitivelinguistics. Edited by Gonzalez-Marquez M, Mittelberg M, Coulson S, SpiveyM. Amsterdam: John Benjamins; 2007:120–148.66. Agar M, Hobbs J: How to grow schemata out of interviews. In Directions incognitive anthropology. Edited by Dougherty J. Urbana: University of IllinoisPress; 1985:413–431.67. Hofstede SN, Marang-van de Mheen PJ, Assendelft WJJ, et al: Designing animplementation strategy to improve interprofessional shared decisionmaking in sciatica: study protocol of the DISC study. Implementation Sci2012, 7(1):55.68. Francis JJ, Johnston M, Robertson C, et al: What is an adequate samplesize? Operationalising data saturation for theory-based interview studies.Psychol Health 2010, 25(10):1229–1245.69. Guest G: How many interviews are enough?: an experiment with datasaturation and variability. Field Methods 2006, 18(1):59–82.70. Borgatti SP, Everett MG, Freeman LC: UCINET Software. In Ucinet forwindows: software for social network analysis. Cambridge, MA:Harvard-Analytic Technologies; 2002.71. Wenger E, White N, Smith J: A review of digital habitats: stewardingtechnologies for communities. Portland OR: CPsquare; 2009.72. McGill I, Beaty L: Action learning. A practitioner’s guide. 2nd edition. London:Routledge; 2013.73. Lavis JN, Lomas J, Hamid M, Sewankambo NK: Assessing country-level effortsto link research to action. Bull World Health Organ 2006, 84(8):620–628.74. Mann KV: Theoretical perspectives in medical education: past experienceand future possibilities. Med Educ 2011, 45(1):60–68.75. Bleakley A: Broadening conceptions of learning in medical education: themessage from teamworking. Med Educ 2006, 40(2):150–157.76. Soubhi H, Bayliss EA, Fortin M, et al: Learning and caring in communities ofpractice: Using relationships and collective learning to improve primarycare for patients with multimorbidity. Ann Fam Med 2010, 8(2):170–177.77. Li LC, Grimshaw JM, Nielsen C, Judd M, Coyte PC, Graham ID: Use ofcommunities of practice in business and health care sectors: asystematic review. Implementation Sci 2009, 4:27.78. Ranmuthugala G, Plumb JJ, Cunningham CF, Georgiou A, Westbrook J,Braithwaite B: Communities of practice in the health sector: a systematicreview of the peer-reviewed literature. Sydney: University of New South Wales,Australian Institute of Health Innovation; 2010.79. Lave J: Situating learning in communities of practice. In Perspectives onsocially shared cognition. Edited by Resnick LB, Levine JM, Teasley SD.Washington, DC: American Psychological Association; 1991:63–82.80. Engestrom Y: Developmental studies of work as a testbench of activitytheory: the case of primary care medical practice. In Understandingpractice: perspectives on activity in context. Edited by Lave J. Cambridge:Cambridge University Press; 1993:64–103.81. Nardi BA: Studying context: a comparison of activity theory, situatedaction models, and distributed cognition. In Context and consciousness:activity theory and human-computer interaction. Edited by Nardi B.Cambridge: Mass:MIT; 1996:69–102.82. Engeström Y, Sannino A: Whatever happened to process theories oflearning? Learning. Cult Social Interact 2012, 1(1):45–56.83. Mercer N, Howe C: Explaining the dialogic processes of teaching andlearning: the value and potential of sociocultural theory. Learn Cult SocialInterac 2012, 1(1):12–21.84. Engeström Y, Sannino A: Studies of expansive learning: foundations,findings and future challenges. Educ Res Rev 2010, 5(1):1–24.85. Nicolini D: Practice theory, and organization: an introduction. Oxford: OxfordUniversity Press; 2013.86. McTaggart R: Participatory action research: issues in theory and practice.Educ Action Res 1994, 2(3):313–37.87. Rains JW, Ray DW: Participatory action research for community healthpromotion. Public Health Nurs 1995, 12(4):256–261.88. Dickens L, Watkins K: Action research: rethinking Lewin. Manag Learn1999, 30(2):127–140.doi:10.1186/1748-5908-9-6Cite this article as: Collet et al.: Engaging Pediatric Intensive Care Unit(PICU) clinical staff to lead practice improvement: the PICU ParticipatoryAction Research Project (PICU-PAR). Implementation Science 2014 9:6.Submit your next manuscript to BioMed Centraland take full advantage of: • Convenient online submission• Thorough peer review• No space constraints or color figure charges• Immediate publication on acceptance• Inclusion in PubMed, CAS, Scopus and Google Scholar• Research which is freely available for redistributionSubmit your manuscript at www.biomedcentral.com/submitCollet et al. Implementation Science 2014, 9:6 Page 10 of 10http://www.implementationscience.com/content/9/1/6