PUBLISHED VERSION Gill Harvey and Alison Kitson PARIHS revisited

14 Pages·2016·0.71 MB·English

PUBLISHED VERSION

Gill Harvey and Alison Kitson
PARIHS revisited: from heuristic to integrated framework for the successful
implementation of knowledge into practice
Implementation Science, 2015; 11(1):33-1-33-13

© 2016 Harvey and Kitson. Open Access This article is distributed under the terms of the
Creative Commons Attribution 4.0 International License
(http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and
the source, provide a link to the Creative Commons license, and indicate if changes were
made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this
article, unless otherwise stated.

Published version http://dx.doi.org/10.1186/s13012-016-0398-2

PERMISSIONS

http://creativecommons.org/licenses/by/4.0/

13 April, 2016

http://hdl.handle.net/2440/98344
HarveyandKitsonImplementationScience (2016) 11:33
DOI10.1186/s13012-016-0398-2
DEBATE Open Access
PARIHS revisited: from heuristic to
integrated framework for the successful
implementation of knowledge into practice
Gill Harvey1,2* and Alison Kitson1,3
Abstract
Background: The Promoting Action onResearch Implementation inHealthServices, or PARIHS framework, was first
published in1998. Since this time,work has been ongoingto further develop, refine and test it.Widely used as an
organising or conceptualframework to helpboth explain and predictwhy theimplementation ofevidence into
practice is or is not successful, PARIHS was one of the firstframeworks to make explicit the multi-dimensionaland
complex nature ofimplementation as well as highlighting thecentral importance ofcontext.Several critiques of
theframework have also pointed out its limitations and suggested areas for improvement.
Discussion: Building on thepublishedcritiquesand a number ofempirical studies, this paper introducesa revised
version of the framework, called theintegrated or i-PARIHS framework. The theoretical antecedentsof the
framework are described as well as outlining the revised and new elements, notably, therevision of how evidence
is described; how theindividualand teams are incorporated; and how context is further delineated. We describe
how the framework can be operationalised and draw on case study data to demonstrate thepreliminary testingof
theface and content validity of therevised framework.
Summary: This paper is presented for deliberation and discussion within theimplementation science community.
Responding to a series of critiques and helpful feedback ontheutility of the original PARIHS framework,weseek
feedback ontheproposed improvementsto the framework. We believe that the i-PARIHS framework creates a
more integrated approach to understand thetheoretical complexity from which implementationscience draws its
propositions and working hypotheses;thatthenew framework is more coherentand comprehensive and atthe
same time maintains it intuitive appeal; and thatthe models of facilitation described enableitsmore effective
operationalisation.
Keywords: PARIHS, i-PARIHS, Implementation framework, Facilitator role, Facilitation
Background introduced or facilitated (F) into practice. Each of these
In 2008, the PARIHS group published a paper in Imple- dimensions was further subdivided into a number of
mentation Science that summarised the work over the sub-elements that needed to be considered in order for
previous10yearsindevelopingandrefiningthePARIHS implementation tobesuccessful[2,3].
(Promoting Action on Research Implementation in The 2008 paper outlined three linked areas of work in
Health Services) framework [1]. From its inception, developing PARIHS, namely, conceptual development
PARIHS argued that successful implementation (SI) of [4–6], empirical testing and refinement [7] and the de-
evidence into practice was a function of the quality and velopment of reliable measures to diagnose and evaluate
type of evidence (E), the characteristics of the setting or an organisation’s readiness for change and the effective-
context (C) and the way in which the evidence was ness of that change [8–10]. It concluded by identifying a
number ofchallengesincludingtheneedformoretheor-
*Correspondence:[email protected] etical work on the conceptual framework, the need to
1SchoolofNursing,UniversityofAdelaide,Adelaide,SA5005,Australia
set up more rigorous ways to develop and test the
2AllianceManchesterBusinessSchool,UniversityofManchester,Manchester,UK
Fulllistofauthorinformationisavailableattheendofthearticle diagnostic and evaluative methodologies and associated
©2016HarveyandKitson.OpenAccessThisarticleisdistributedunderthetermsoftheCreativeCommonsAttribution4.0
InternationalLicense(http://creativecommons.org/licenses/by/4.0/),whichpermitsunrestricteduse,distribution,and
reproductioninanymedium,providedyougiveappropriatecredittotheoriginalauthor(s)andthesource,providealinkto
theCreativeCommonslicense,andindicateifchangesweremade.TheCreativeCommonsPublicDomainDedicationwaiver
(http://creativecommons.org/publicdomain/zero/1.0/)appliestothedatamadeavailableinthisarticle,unlessotherwisestated.
HarveyandKitsonImplementationScience (2016) 11:33 Page2of13
instruments based on elements of PARIHS and the need or i-PARIHS framework. This paper describes the re-
to agree upon the practical contents of a facilitator vised framework, outlines the new elements and ex-
training programme that would equip facilitators to plains why the changes have been made. Within this
know how to operationalise the framework. Subse- discussion, we draw on empirical data from three case
quently, both PARIHS team members [11–13] and other studies of implementation (summarised in Table 1).
research teams [14, 15] have been involved in studies to The paper then describes how the i-PARIHS frame-
evaluate and refine the framework further. These stud- work can be operationalised and summarises the
ies have reinforced some of the conclusions of the 2008 underpinning theoretical antecedents of the frame-
paper and identified some additional issues for consider- work. We conclude the paper by raising some ques-
ation. For example, Helfrich and colleagues [14] under- tions for further consideration and outlining plans for
took a critical synthesis of the literature on PARIHS and future research and development activity.
identified a number of perceived limitations to its effect-
ive utilisation. These included the lack of evidence from
prospective implementation studies on its effectiveness; Main Text
lack of clarity between elements and sub-elements of The main reasons for re-visiting the original PARIHS
the framework; a predominant focus on the facilitation frameworkincluded:
role rather than the facilitation process and the lack of
a clear definition of what successful implementation ac- (cid:1) Theoriginalframeworkfailed toaddresskey
tually was. Building on this review, a revised PARIHS dimensions,including theintended targetsfor
framework was put forward, including a detailed diag- implementation andthe wider external context
nostic tool based on the refined elements and sub- (social, political,policy and economic)inwhich
elements of the framework [15]. implementation occurs[14,18,19]
A repeat search in 2014 using the same databases and (cid:1) Growing evidenceonthe keyroleindividuals play in
search terms as the review by Helfrich and colleagues in theimplementation process[12]
2010 [14] identified over 40 more papers that reported (cid:1) Increasedinterest and awarenessofrelevanttheories
applying PARIHS [16]. This indicates continuing interest thatcanand shouldinformimplementation
inusingthe frameworkandreinforceswhat Helfrich and strategies [23–25]
colleagues observedinterms of theframework’s intuitive (cid:1) Recognitionofthe diverse waysinwhich people
appeal and relevance to the real world setting. However, wereapplying PARIHS,not simply toguidethe
prospective studies remain limited. One exception to implementation ofmore conventionalresearch
this is a prospective study on peri-operative fasting, evidenceintheformofclinicalguidelinesor
which used PARIHS to design a pragmatic trial to test evidencesummaries,buttoinform andevaluate
the effectiveness of the introduction of guidelines to im- developmentsinpracticemoregenerally[26]
prove practice [12, 17]. From their analysis, the authors
suggested that an additional weakness in the framework Based on our analysis of these issues, we are pro-
was the failure to acknowledge the central role of indi- posing the revision of the key constructs of evidence,
viduals in determining the process and outcomes of im- context and facilitation and suggesting the addition of
plementation, mediated through individual interactions a new construct termed ‘recipient’. The original
with and influence on the evidence and context dimen- PARIHS framework was expressed as a simple equa-
sions of PARIHS. Useful findings have also emerged tion (Table 2). Critics have rightly pointed out that
from reviews that have compared PARIHS to other im- we did not define what successful implementation
plementation frameworks and models. Tabak and col- meant [14, 15]. In our revised, i-PARIHS framework,
leagues reviewed over 60 models and frameworks and successful implementation is primarily specified in
suggestedthatPARIHS lacked afocus onthe system and terms of the achievement of implementation/project
policy level of implementation [18]. Flottorp and col- goals and results from the facilitation of an innova-
leagues also undertook a review of frameworks and their tion with the recipients in their (local, organisational
findings indicated that PARIHS failed to pay attention to and health system) context (Table 2). The core con-
the individual health professional and the wider social, structs of the i-PARIHS framework are facilitation,
politicalandlegalcontext ofimplementation[19]. innovation, recipients and context, with facilitation
Our own ongoing application of the framework in im- represented as the active element assessing, aligning
plementation studies (see, for example, [11, 13, 20–22]) and integrating the other three constructs. As illus-
together with critiques and evaluations of the frame- trated, a number of other characteristics of successful
work by other research teams has led us to create a implementation are proposed, reflecting the multi-
refined version of PARIHS. It is called the integrated dimensional nature of the constructs.
H
a
rv
Table1Threeimplementationcasestudies ey
a
Implementationstudy Innovation Recipients Context Facilitation Implementationoutcomes nd
K
1.Improvingthe Startingpoint:existingdata Generalpracticeteams Practiceswereworking Facilitationteamssetup, Beforeandafterstudy its
identificationand indicatingprevalencelevels recruitedtoparticipateinan toapay-for-performance comprisingamixofinternal design on
managementof ofCKDinthelocalpopulation improvementcollaborative; system;CKDwaspartof andexternalnovice,experienced/ Recordedprevalenceof Im
chronickidney werelowerthanwouldbe eachteamrequiredtohave thissystem;hence,there expertfacilitators,supportedby CKDincreasedby1.2%in p
le
disease(CKD)in expected multi-disciplinarymembership wasanincentivetoimprove clinicalleadersandprojectmanagers 30participatingpractices m
primarycare Nationalclinicalguideline Sponsorshipfromseniorleaders Widerchangesoccurringin Facilitationmethodsusedincluded (n=1863additional en
prerceosemnmtinegndeavtiidoennscfeo-rbased iSnomtheerpersimistaarnycheeeanltchoucanrteerseedttiantg raenladtimonantoagtehmeeonrgtaonfisation ccoolnlatebxotraatsisveesslemaernnitn,gPleavne-Dntos-,Sltoucdayl-Act pidaetnietniftiesdw)icthomCpKDaredto tation
identifyingandmanaging alocallevel,e.g.frompractice generalpractice (PDSA)cycles,auditandfeedback, anationalincrease S
c
CKD colleagueswhodidnotrecognise benchmarkingofdata of0.2% ie
n
Stakeholdergroupconvened CKDasapriority,were andregularpracticevisits Managementofblood ce
toconsidertheevidenceand uncomfortabledisclosingto pressureimprovedinline (2
thelocalpopulationdata; patientsordidnotfeel withnationalguidelines 01
6
identified2targetsfor sufficientlyinvolved from34to74%(cohort1) ) 1
improvement and58to83%(cohort2) 1
:3
[21] 3

2.Improving Startingpoint:4evidence-based Facilitatorswereencouragedto Contextualchallengesina Internalnovicefacilitatorstrained ClusterRCTshowedno
continencecare recommendationsforpractice establishimprovementteams numberofhomescaused andsupportedbyexternalexpert differencebetweencontrol
inanursing identifiedfromaninternational withinthenursinghome bychangeofmanagement facilitators andinterventionwardson
homesetting clinicalguidelinebytheproject Somedifficultiesinconvincing andreorganisation Internalfacilitatorsencouragedto primaryoutcomemeasure
stakeholdergroup colleaguesthatimprovements Cultureofmanaging partnerwithabuddy—somedid ofoverallcomplianceto
Recommendationswere incontinenceoflong-term incontinenceratherthan andothersdidnot continencerecommendations
discussedandreviewedby residentswaspossible promotingcontinence Majorityofexternalsupport [11,85]butsignificant
facilitatorsandasetof Inputfromcontinencenurse Positiveimpactofexternal providedvirtually improvementsonanumber
commonauditcriteria specialist inspection/accreditation Facilitationmethods:jointtraining, ofsecondaryoutcomes
agreed Useofpatientstoriestohighlight monthlyteleconferencemeetings, and1ofthe4specific
theneed/potentialforimprovement auditandfeedbackandPDSAcycles recommendations
Gate-keeperroleofnursinghome Internalevaluation
manager demonstratedvariable
achievementofkeyaudit
targetsbyparticipatingsites
[45]
3.Improving Startingpoint:evidencereviewto Organisationwideapproach Contextualissuesto Experiencedinternalfacilitators SteppedwedgeRCT[86]
nutritionalcare identifythreeinterventionstobe adopted,withseniorleadership benegotiatedatan supportedbyexternalexpert demonstratednodifference
ofolderadults implementedaspartoftheproject supportandcommunication organisationallevel facilitators inweightlossafter1week
inanacute Combinedthethreeinterventions strategyinplace relatedtothe Internalfacilitators betweeninterventionand
caresetting (nutritionalscreening,nutritional Dietitianspreviouslytriedto infrastructureand recruitedwardlevelclinical controlwards
supplementsandredtraysystem) introduceimprovementsbut resourcesrequiredto championstoworkwiththem Improvementnotedonkey
intoanimprovementbundle unabletosecurebuy-in enableimplementation, Facilitationmethods:staff auditmeasuresrelatingto
Formedpartofan e.g.providingfridgesat informationandeducation nutritionalscreening,
inter-disciplinaryteaminthis wardlevel,financingthe programmes, provisionofnutritional
projectwithinvolvementof purchaseofnutritional auditandfeedback supplementsanduseof
otherclinicalcolleaguesand supplements,issuesof redtraysforpatients
P
otherdepartmentssuchas supplyandstock requiringassistancewith a
g
cateringandsupplies management feeding[46] e
3
o
f
1
3
HarveyandKitsonImplementationScience (2016) 11:33 Page4of13
Table2FromPARIHStoi-PARIHS(adaptedfrom[16])
‘Successfulimplementation’intheoriginalPARIHSframework ‘Successfulimplementation’intherevisedi-PARIHSframework
SI=ƒ(E,C,F) SI=Facn(I+R+C)
SI=successfulimplementation SI=successfulimplementation
ƒ=function(of) Achievementofagreedimplementation/projectgoals
E=evidence Theuptakeandembeddingoftheinnovationinpractice
C=context Individuals,teamsandstakeholdersareengaged,motivatedand
F=facilitation ‘own’theinnovation
Variationrelatedtocontextisminimisedacrossimplementation
settings
Facn=facilitation
I=innovation
R=recipients(individualandcollective)
C=context(innerandouter)
Theinnovationconstruct register, to improve the management of patient blood
The original PARIHS construct of evidence adopted a pressuretoevidence-basedtargets.
broad view of evidence, comprising information from re- This process of aligning external explicit evidence with
search, alongside clinical, patient and local experience localprioritiesandpracticeisanimportantwayofenhan-
[6].Ini-PARIHS,wehavefurtherextendedthe construct cingthecompatibilityofaproposedchange,asrecognised
to embrace a more explicit view of how the characteris- intheinnovationliterature[33–35].Forthesereasons,we
tics of knowledge affect its migration and uptake in have re-labelled the construct ‘innovation’, incorporating
different settings. This includes the more emergent, in- Rogers’seminal work on the diffusion of innovations [33,
ductive ways in which evidence is generated from 34] and other key studies on the nature of innovation
practice as, for example, within practice development withinandoutsidehealthcare[35,36].Wearguethatevi-
initiatives in nursing and healthcare [27–29]. Our prop- dence is one type of knowledge and (new) knowledge is
osition is that people rarely take evidence in the original thesubstancethatneedstobeintroducedinordertogen-
form of a systematic review or clinical guideline and dir- erate change and improvement. The characteristic of the
ectly apply it within an implementation project rather knowledge creates a set of conditions that make it more
they incorporate evidence in a number of different ways, or less likely to be recognised and applied. This phe-
which typicallyinvolves adapting the original evidencein nomenoniswelldescribedinRoger’sDiffusionofInnova-
some way to suit their particular situation, a process de- tionsTheory[33,34],forexample,intermsofthelikelyfit
scribedbysomeas‘tinkering’[30]wherebyexplicitknow- of the new knowledge with existing practice, the relative
ledgeisblendedwithtacit,practice-basedknowledge. advantage it presents and potential trialability. We are
Thisisclearlyapparent inoneofthe caseswe drawon therefore proposing ‘innovation’ as a central construct
in this paper, namely a project to improve the identifica- within the i-PARIHS framework but with an explicit
tion and management of chronic kidney disease (CKD) focus on sourcing and applying available research evi-
in a healthcare region in England [21, 31]. Aware of the dence to inform the innovation. Table 3 summarises
potential to improve CKD, the team leading the project the main characteristics of the innovation to be con-
accessed a recently produced national clinical guideline sidered in implementation.
on the identification and management of CKD in pri-
mary care [32]. However, rather than setting out to ‘im- Therecipientconstruct
plement the guideline’, a number of prior processes were This is a new construct, added in response to consistent
put in place. Firstly, a local stakeholder group compris- feedback that insufficient attention had been paid in the
ingpatientrepresentatives,cliniciansfromacuteandpri- original framework to the actors involved in implemen-
mary care, researchers and managers was established to tation. Although reviews and empirical studies applying
consider the evidence and agree on the priorities at a PARIHS have emphasised the importance of the individ-
local level. This involved taking into consideration ual on implementation processes and outcomes [12], we
existing policies and practice at the local level, includ- areproposingrecipientsasaconstructthatencompasses
ing the CKD related measures in the national pay-for- the people who are affected by and influence implemen-
performance systeminprimarycareandthelocalratesof tation at both the individual and collective team level.
achievementontheseindicators.Fromthestakeholderde- This extension enables the i-PARIHS framework to con-
liberations, a decision was made to distil the evidence sider the impact individuals and teams have in support-
from the guideline into two overarching aims related to ing or resisting an innovation. We have elected to
improving the identification of CKD patients within a consider recipients at both an individual and collective
practice populationand,once identified and ona practice level as alongside research highlighting the importance
HarveyandKitsonImplementationScience (2016) 11:33 Page5of13
Table3Characteristicsoftheinnovation,recipientsandcontexttobeconsideredwithinthei-PARIHSframework
Innovation Recipients Context
Underlyingknowledgesources Motivation Locallevel:
Clarity Valuesandbeliefs Formalandinformalleadershipsupport
Degreeoffitwithexistingpracticeandvalues Goals Culture
(compatibilityorcontestability) Skillsandknowledge Pastexperienceofinnovationandchange
Usability Time,resources,support Mechanismsforembeddingchange
Relativeadvantage Localopinionleaders Evaluationandfeedbackprocesses
Trialability Collaborationandteamwork Learningenvironment
Observableresults Existingnetworks Organisationallevel:
Powerandauthority Organisationalpriorities
Presenceofboundaries Seniorleadershipandmanagementsupport
Culture
Structureandsystems
Historyofinnovationandchange
Absorptivecapacity
Learningnetworks
Externalhealthsystemlevel:
Policydriversandpriorities
Incentivesandmandates
Regulatoryframeworks
Environmental(in)stability
Inter-organisationalnetworksandrelationships
of individuals in supporting or resisting change [12, 37]; from acontinence nursespecialist. Inonenursinghome,
there is good evidence to suggest that groups or teams the staff responsible for facilitating implementation col-
of individuals have an important role in determining the lected stories from residents about their experience of
uptake of new knowledge in practice. This is particularly living with incontinence, which provided avery powerful
evident in studies that have been undertaken on com- motivational tool to convince their colleagues of the
munities of practice and the notion of collective ‘mind- need tochange.
lines’ influencing the uptake (or not) of evidence in As these examples illustrate, the people involved in
practice[38–40]. implementation, including their views, beliefs and estab-
The CKD case study illustrates one way in which ac- lished ways of practice, can significantly affect the ease
tors at the local level can influence the course of imple- of introducing an innovation or change. A wide range of
mentation, as one of the challenges encountered was stakeholderspotentiallyfit into the constructwehavela-
whether practice staff perceived value in ‘labelling’ pa- belled ‘recipients’ including patients and clients, clinical
tients with CKD. This was particularly the case for older staff and managers. It is also apparent that the relation-
patients as some General Practitioners (GPs) and prac- ship between the innovation and the recipients is in
tice nurses viewed declining renal function as a natural many ways an inter-dependent one. Given this set of cir-
part of ageing and believed that disclosing a diagnosis of cumstances, part of the facilitator’s role at the recipient
CKD could cause unnecessary anxiety in patients. By levelinvolvesassessingtheactualandpotentialboundar-
adopting this approach, opportunities to improve self- ies that exist and the ways in which these barriers might
management and overall management of cardiovascular exert an influence during implementation [43]. Table 3
disease and to address the issue of increased susceptibil- identifies the main characteristics of the recipients at the
itytoacutekidneyinjurywere potentiallymissed[41]. individualandcollectivelevel.
A second case study which elucidated the key role of
recipients focused on the implementation of evidence- Thecontextconstruct
based recommendations for the management of contin- Context remains a core construct within i-PARIHS but
ence in a nursing home setting [42]. Focusing on goals with a wider focus on the different layers of context,
to improve the assessment and attainment of continence from the micro through the meso and macro levels, that
amongst residents, a key area was addressing care staffs’ can act to enable or constrain implementation. In the
strongly held views as to whether such goals were PARIHS framework, we defined context in terms of re-
achievable, particularly where residents had been man- sources, culture, leadership and orientation to evaluation
aged as ‘incontinent’ over prolonged periods of time. and learning; however, we did not delineate between the
This required a significant amount of effort to change immediate local context and the wider organisational
the mindset amongst nursing home staff about achieving context. Furthermore, we did not explicitly consider the
continence. Various strategies were helpful in this re- impact that the wider health system—the external con-
gard, including input, support and practical guidance text—could have on implementation processes and
HarveyandKitsonImplementationScience (2016) 11:33 Page6of13
outcomes. Thesemesoandmacrolevelcontextualfactors and a set of strategies and actions (the facilitation pro-
havebeen recognised as important considerations, for ex- cess) to enable implementation. The i-PARIHS frame-
ample, in other implementation frameworks such as the work therefore locates the success or otherwise of
Consolidated Framework for Implementation Research implementation upon the ability of the facilitator and
[44]andinreviewsofmodelsandtheoriesofimplementa- the facilitation process to enable recipients within their
tion [19]. We have also observed their influence in our particular context to adopt and apply the innovation by
ownempiricalresearch.Forexample,inthecasestudyon tailoring theirinterventionappropriately.
promotingcontinence,insomeofthecountriesstudied,a We have adopted this position for a number of rea-
focus on continence formed part of the external accredit- sons, both experientially and empirically based. Tracing
ation system for nursing homes. This created a driver for the history of facilitation as a concept in healthcare [5],
introducing changes to improve the management of con- thereisatraditionofapplying facilitator rolestosupport
tinence at a local level [45]. Similarly in the case study of the implementation of changes in practice. From the
CKD, the presence of CKD indicators in the pay-for- introduction of facilitators to promote primary care
performance system in primary care created an incentive prevention programmes in the 1980s [47], the use of fa-
forimprovement[21]. cilitatorsinprimarycarehasbecome commonplace,par-
In a third case study that focused on the prevention ticularly supporting the implementation of change
and reduction of weight loss amongst older patients in through quality improvement methods [48–51]. A 2012
an acute hospital setting, a number of contextual factors systematic review of practice facilitation in primary care
were important, particularly at the organisational level concluded that practices supported by a facilitator were
[46]. The implementation project introduced three 2.76 times more likely to adopt evidence-based clinical
evidence-informed interventions, one of which was the guidelines [52]. Within the 23 studies reviewed, facilita-
provision of oral nutritional supplements for older pa- tors employed a number of different facilitation strat-
tients at risk of malnutrition. However, the reality of egies, in particular audit and feedback (used in 100 per
making these supplements available at the point of care cent of studies) and interactive consensus building and
delivery required the agreement of financial support to goal setting (91 per cent use), alongside reminders, tai-
make the supplements, and the fridges to store them in, loring to context and quality improvement tools such as
available in the ward setting. Furthermore, negotiations Plan-Do-Study-Act (PDSA) cycles. This concurs with
with the stocks department were needed to address the our own experiences of applying facilitation to support
issue of stock supply and management. These are typical the development of standards, audit and quality im-
of the sort of organisational context issues that have to provement in nursing and health care [53, 54]. More re-
be consideredwithin theprocessofimplementation. cently, the use of facilitation has been evaluated in a
Consequently, in the i-PARIHS framework, we have number of other settings. For example, the NeoKIP
made a distinction between the layers of inner and outer (Neonatal Knowledge into Practice) trial evaluated the
context, where inner context includes both the immedi- effectiveness of facilitation as a knowledge translation
ate local setting, whether a ward, unit, hospital depart- intervention for improved neonatal health and survival
ment or primary care team, and the organisation within [55]. Using lay members of the community who received
which this unit or team is embedded. Outer context re- training in facilitation techniques such as PDSA and
fers to the widerhealth system inwhich theorganisation group consensus building, the study demonstrated a re-
is based and reflects the policy, social, regulatory and duced neonatal mortality of 49 % in the third year of the
political infrastructures surrounding the local context. intervention [56]. In the United States Veterans Health
Table 3 illustrates the differentiation of inner and outer Administration, a number of studies have demonstrated
contextat themicro, mesoandmacrolevels. the benefits of using facilitation to support the imple-
mentation of evidence into clinical practice (for example
Thefacilitationconstruct [57, 58]), whilst in the UK, facilitators have been
As with the original PARIHS framework, facilitation employed to support the implementation of evidence-
remains a core construct. However, we emphasise facili- based vascular care [20, 21], as described in the CKD
tation as the active ingredient within i-PARIHS by posi- casestudy.
tioning it differently to the other main constructs of To fulfil the role effectively, facilitators have to be able
innovation, recipients and (inner and outer) context to function in a flexible and responsive way to tailor
(Table 2). We propose that facilitation is the construct their approach to the particular issue, setting and people
that activates implementation through assessing and involved; hence, our proposition that facilitation com-
responding to characteristics of the innovation and the prises the active element of implementation. However,
recipients (both as individuals and in teams) within their asthe casestudiesinTable1illustrate,evidencefromef-
contextual setting. This requires a role (the facilitator) fectiveness studies of facilitation is mixed. This likely
HarveyandKitsonImplementationScience (2016) 11:33 Page7of13
reflects the fact that facilitation itself is a complex inter- and provide mutual support, supplemented by support
vention, involving one or more individuals in the role of from external, expert facilitators in the co-located uni-
facilitator, applying a combination of improvement and versity. In all three cases, the methods employed by fa-
team-focused strategies to enable and support change. cilitators typically involved improvement approaches
In some cases, facilitators are internal to the implemen- such as Plan-Do-Study-Act cycles and audit and feed-
tation setting; inothers,theyareexternaland sometimes back, underpinned by project management. This helped
a combination of internal and external facilitators is to address key issues such as establishing clear goals,
used. Studies that reportprocessevaluation alongside ef- demonstrating the potential for improvement, providing
fectiveness data demonstrate the importance of having regular feedback and trialing changes on a small-
the right individuals in the role with the right level of scale—all important factors in terms of securing and
skills, knowledge, support and mentoring [59, 60]. This maintainingstaffmotivationandcommitment.
highlightstheneedtoconsiderissuesoffacilitatorrecruit- The facilitator needs to have a sound understanding of
ment, selection, preparation and development when de- the nature of the innovation being introduced (the focus
signing and conducting implementation studies that and content of implementation), the individuals and
employ facilitation as an intervention. These are issues teams that have to enact the change (the recipients) and
that we have taken into consideration in our proposed theenvironmentinwhich they work(the local, organisa-
operationalisationoffacilitationwithini-PARIHS. tional and health system context). This essentially in-
volves thinking about what is to be implemented, who
Howthei-PARIHSframeworkisactioned with and where. Facilitation provides the how compo-
AconsistentcriticismoftheoriginalPARIHS framework nent ofimplementation.
was that it was difficult to operationalise [15]. In devel- In order to help the facilitator understand the dy-
oping the i-PARIHS framework, we have used ongoing namic nature of implementation, we have chosen to
empirical research from our own and other teams’appli- represent the i-PARIHS framework as a continuous
cation, development and evaluation of PARIHS to spiral which starts with a focus on the innovation and
present a practical model of facilitation (see for example the recipients, moving out to the different layers of
[11, 21, 22, 45, 55, 61]). This has led to the development context (inner context at local and organisational
of a preliminary Facilitator’s Toolkit utilising quality im- level and outer context at wider system and policy
provement and audit and feedback methods and also a level). Figure 1 summarises what the facilitator looks
more structured approach to the identification, training at within each of these levels and also summarises
and development of facilitators within and across sys- the sort of activities they need to undertake; in other
tems [62, 63]. (For a more detailed description of the fa- words, what they have to be able to do. This effect-
cilitation model and toolkit, see [63]). Specifically, we ively involves progressing from a focus on the more
are proposing a facilitation pathway from beginner or specific, concrete aspects of implementation to ad-
novice facilitator to experienced and expert facilitator, dressing the contextual factors and barriers that are
assuming different roles in the process of implementing likely to influence the trajectory of the implementa-
and researching the implementation of new knowledge tion journey. Our hypothesis is that the further out
into practice [62]. into the spiral the facilitator moves, the greater the
Positioned as the active ingredient, facilitation is level of experience and skill they will need. This in
undertaken by one or more trained facilitators, who help turn suggests that whilst novice facilitators may be
to navigate individuals and teams through the complex able to support locally focused implementation pro-
change processes involved and the contextual challenges jects (in terms of working with a local team to plan
encountered. Facilitators can either be internal to the and undertake the project), they are likely to need the
system, external to it or a combination of both, as the support of a more experienced facilitator to assess
three case study examples illustrate, with a mix of and negotiate some of the more challenging barriers
internal-external and novice-experienced-expert combi- or contextual factors they may encounter.
nations. This reinforces that there is not a single right This leads us onto another important consideration
way to apply facilitator roles; however, there are clear about the need for facilitators to work within a sup-
benefits in mechanisms that provide support and men- portive network, ideally mentored and supported by
toring to new or less experienced facilitators. In case 1, peers and more experienced colleagues, as is evident in
this was achieved through having teams of novice and the case study examples. Depending on the scale of the
experienced facilitators working together and by bring- change being considered and how it is set up, there
ing in novice internal facilitators to build local capacity could be a team of facilitators, each supporting a num-
for facilitating implementation. In case 3, facilitator pairs ber of units or areas. In some cases, a facilitator role
were formed to role model inter-disciplinary working may be combined with another role, such as a clinical
HarveyandKitsonImplementationScience (2016) 11:33 Page8of13
Fig.1Thefacilitationroleandprocess
leader, quality improvement coordinator, knowledge bro- being overwhelmed [62]. Whilst a formal infrastructure
ker or project manager. The specific title of ‘facilitator’ might not exist, the individuals concerned should be
is not important. The crucial thing is that the individ- encouraged to seek opportunities for support and guid-
uals function as facilitators in that they actively use fa- ance, for example, by establishing a ‘buddy’ relationship
cilitation methods and processes to enable and optimise with others in a similar role or identifying a more expe-
implementation. In some cases—and particularly where rienced facilitator to mentor them. Organisations com-
facilitation is part of another role—the facilitator may mitted to knowledge translation and implementing
feel like a lone agent. However, given the scope and innovations in healthcare ought to reflect on the infra-
complexity of the role, this is not a desirable situation structure they have to enable facilitation capabilities and
and can result in individual feelings of isolation and skills to flourish. Otherwise, there is a danger of setting
Table4Novice,experiencedandexpertfacilitators(adaptedfrom[62])
Experience Focusoffacilitation
Novicefacilitator Workingunderthesupervisionofanexperiencedfacilitator
Focuson:
Whataninnovationis;whatevidenceinformstheinnovationandhowtoassessandapplyit
Readinesstochangeatalocallevel
Whatmotivatesindividualsandteamsandhowteamsworkeffectively
Whatcontextis;whatimpactcontexthasonimplementationatalocalandorganisationallevel
Identifyingandengagingkeystakeholders
Planning,implementing,measuringandembeddingchange
Experiencedfacilitator Workingunderthesupervisionofanexpertfacilitator
Focuson:
Indepthunderstandingandknowledgeoftheorganisationororganisationstheyareworkingwith
Awarenessofcompetingtensionsandhowtomanagetheseinrelationtoimplementinginnovationandchange
Indepthunderstandingofindividualandteammotivation,teamdynamicsandproductivity
Experiencedandknowledgeableinlocalcontextevaluation
Abletoassesssystem-wideactivitiesandinfluenceactions
Awareofwidercontextualissuesandconfidentintermsofnegotiatingboundariesandpoliticaltensions
Expertfacilitator Expertfacilitatoroperatingasaguideandmentortootherfacilitators
Focuson:
Coordinatingandsupportingnetworksofexperiencedandnovicefacilitators
Workingwithhealthsystemstoimproveimplementationsuccess
Workingacrossacademic,serviceandotherorganisationalboundariestointegratefacilitationandresearchactivity
Developingandtestingtheoriesofimplementation,innovationandfacilitation
Evaluatingimplementationandfacilitationinterventionstogeneratenewerknowledge
Refiningandimprovinglearningmaterialsandmentoringprocesses
Runningworkshopsandadvancedmasterclassesonfacilitationapproaches
HarveyandKitsonImplementationScience (2016) 11:33 Page9of13
people up to fail without the requisite level of acquired, interpreted and applied in a way that is con-
preparation, skills and support. Table 4 summarises the sistent with the i-PARIHS framework; in turn, it would
main descriptors of novice, experienced and expert also provide a theoretical perspective that could inform
facilitators. orexplain theinnovation anditsimpact.
A final point in relation to the facilitation construct For a facilitator thinking about how to work with indi-
within i-PARIHS is that in presenting our description viduals and groups, the i-PARIHS framework again
of facilitation and outlining the key ingredients within points them towards a number of different theories.
the facilitator’s skill repertoire, there may be a sugges- These include theories of innovation, reflecting the
tion that the process is logical and sequential. The real- inter-connection between an innovation and the people
ity, however, is very different; the interrelationship who have to use it. For example, Rogers highlighted the
between the innovation, the recipients and the multiple importance of understanding different groups within the
layers of contexts is often unpredictable, fluid and it- intendedaudienceforinnovationandhowtheyarelikely
erative. Experienced facilitators learn how to manage to react, as well as making use of peer-to-peer conversa-
this uncertainty and keep individuals and teams on tions and credible, trusted teachers and leaders to bring
track. people on board with the change [33, 34]. Issues relating
(Note: Additional file 1 provides a more detailed illus- to adopter characteristics are also reflected in theTheor-
tration of the facilitator’s focus and activities at the level etical Domains Framework [65, 70] where motivation is
of the innovation, the recipients and the multiple layers considered alongside factors such as role and identity,
of context; Additional file 2 outlines a set of reflective goals, behavioural regulation, beliefs and capabilities and
questions that facilitators can use to think about key is- consequences. Weiner’s theory of organisational readi-
sues within thedifferentdimensionsofimplementation.) ness to change [71] proposes that readiness depends on
collective behaviour change linked to two key factors,
Theunderpinningtheoreticalantecedentsofthei-PARIHS described as change commitment (wanting to change)
framework and change efficacy (able to change). Insights into these
Another criticism of PARIHS was the lack of detail types of theories help to inform the way that facilitators
around its theoretical foundations. Unlike other frame- structure their interventions to achieve the behavioural
works such as the Knowledge-to-Action (K2A) Frame- change that is required for successful implementation.
work [64] and theTheoretical Domains Framework [65], Equally,they provide researchand evaluation teamswith
which identify with a particular theoretical perspective a set of parameters to frame studies of implementing
to explain implementation (planned change and behav- evidence-basedinnovationinpractice.
ioural change, respectively), PARIHS claimed an eclectic Theories that inform our views about the context of
provenance of relevant theories and philosophical per- implementation are rich and varied, particularly focusing
spectives [1]. In our deliberations with i-PARIHS, we on issues of organisational complexity and how organi-
have continued with the theoretical eclecticism but have sations learn and use new knowledge. Again this is
tried to present it in a more coherent way [16]. The rea- consistent with the multi-dimensional perspective of im-
son for doing this is twofold: first, it helps the facilitator plementation that the i-PARIHS framework adopts and
to understand the theoretical antecedents of the issues embedded beliefs about reflective and responsive learn-
they are dealing with, and second, it helps research and ing. Included in this mix are theories related to com-
evaluation teams to create a theoretical framework plexity [72, 73], absorptive capacity [74] and learning
around one or more particular aspects of the implemen- organisations [75] as well as theories related to leader-
tation process they wish to explore in greater detail. Our ship and organisational culture [76]. Other theories re-
identification of relevant theories is necessarily selective; late to how innovation and change can be sustained in a
however, we have sought to identify those theories that system. Again, there are a number of theories that at-
reflect the core constructsofinnovation, recipients,con- tempt to explain this phenomenon. One that has been
text and facilitation and that are consistent with our applied in healthcare is normalisation process theory
overarching view of implementation as iterative, negoti- [77, 78], which acknowledges the interaction of actors
ated and relational. Thus, if a facilitator or a research (recipients) within their context and proposes four con-
team studying implementation was interested in under- structs titled coherence, cognitive participation, collect-
standing what aspects of the evidence influenced its up- ive action and reflexive monitoring as the generative
take and use, the i-PARIHS framework would point mechanisms required to routinely embed innovations. A
them in the direction of theories around experiential further set of theories relevant to the study of context
learning [66], situated learning [67], evidence-based are economic and political theories that govern the exter-
practice [68] and innovation [34, 36, 69]. This would nal environment, including theories of regulation, market
provide insights into the means by which knowledge is economy,financialincentivesandcontracting[79].