Using the behavior change wheel to identify barriers to and potential solutions for primary care clinical guideline use in four provinces in South Africa

RESEARCH ARTICLE Open Access Using he beha io change wheel o iden i y ba ie s o and po en ial solu ions o p ima y ca e clinical guideline use in ou p o inces in Sou h A ica Tama a K edo 1,2* , Sa a Coope 1,3 , Ambe Ab ams 1 , Jocelyn Mulle 1 , Jimmy Volmink 4 and Salla A kins 5,6 Abs ac Backg ound: Clinical p ac ice guidelines isk ha ing li le impac on heal hca e i no e ec i ely implemen ed. Theo y in o med, a ge ed implemen a ion may maximise hei impac . Ou s udy explo ed ba ie s o and acili a o s o guideline implemen a ion and use by Sou h A ican p ima y ca e nu ses and allied heal hca e wo ke s in ou p o inces in Sou h A ica. We also p oposed in e en ions o add ess he issues iden i ied. Me hods: We used quali a i e esea ch me hods, comp ising ocus g oup discussions using semi-s uc u ed opic guides. Se en ocus g oup discussions we e conduc ed (48 p o ide s) in ou Sou h A ican p o inces (Eas e n Cape, Wes e n Cape, Kwazulu-Na al, Limpopo). Pa icipan s included mos ly nu ses, die icians, den is s, and allied heal h p ac i ione s, om p ima y ca e acili ies in u al and pe i-u ban se ings. The analysis p oceeded in h ee phases. Fi s ly, wo analys s conduc ed induc i e hema ic con en analysis o de elop hemes o da a. This was ollowed by i ing eme gen hemes o he Theo e ical Domains F amewo k and inally o he associa ed Beha iou Change Wheel o iden i y ele an in e en ions. Resul s: Pa icipan s a e knowledgeable abou guidelines, gene ally us hei c edibili y and a e ecep i e and mo i a ed o use hem. Guidelines a e seen by nu ses o p o ide con idence and eassu ance, as well as p o essional au ho i y and independence whe e doc o s a e sca ce. Ba ie s o guideline use include: inadequa e sys ems o p in ed book dis ibu ion, insu icien and subs anda d pho ocopies, linguis ic inapp op ia eness (e.g. complica ed language, lack o summa ies, una ailable in local languages), unsuppo i e audi ing p ocedu es, limi ed in ol emen o end-use s in guideline de elopmen , and pa chy aining ha may no il e back o all p o ide s. Fu u e aspi a ions iden i ied include: imp o ing he design ea u es o guidelines, accessible places o ind guidelines, making digi ally- o ma ed e sions a ailable, mo e supplemen a y ma e ials (e.g. pos e s) o suppo pa ien engagemen , accessible clinical suppo ollowing aining, and in- acili y aining o all p o essional cad es o ensu e ai access, simila le els o capabili y and in e disciplina y consis ency. Conclusions: Sou h A ican p ima y ca e nu ses and allied heal h p ac i ione s ha e high le els o mo i a ion o use guidelines, bu ace many sys emic ba ie s. We used he Beha iou Change Wheel o sugges ele an , implemen able in e en ions add essing iden i ied ba ie s. This heo y-in o med app oach may imp o e clinical guideline implemen a ion and impac heal hca e o Sou h A ica. Keywo ds: Quali a i e esea ch, Clinical p ac ice guidelines, Implemen a ion, P ima y ca e, Focus g oups, Theo e ical domains amewo k, Beha iou change, Quali y imp o emen * Co espondence: [email p o ec ed] 1 Coch ane Sou h A ica, Sou h A ican Medical Resea ch Council, Cape Town, Sou h A ica 2 Di ision o Clinical Pha macology, Facul y o Medicine and Heal h Sciences, S ellenbosch Uni e si y, Cape Town, Sou h A ica Full lis o au ho in o ma ion is a ailable a he end o he a icle © The Au ho (s). 2018 Open Access This a icle is dis ibu ed unde he e ms o he C ea i e Commons A ibu ion 4.0 In e na ional License (h p://c ea i ecommons.o g/licenses/by/4.0/), which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided you gi e app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o he C ea i e Commons license, and indica e i changes we e made. The C ea i e Commons Public Domain Dedica ion wai e (h p://c ea i ecommons.o g/publicdomain/ze o/1.0/) applies o he da a made a ailable in his a icle, unless o he wise s a ed. K edo e al. BMC Heal h Se ices Resea ch (2018) 18:965 h ps://doi.o g/10.1186/s12913-018-3778-2 Backg ound In e na ionally, high-quali y, e idence-in o med clinical p ac ice guidelines (CPGs) a e ecognised as essen ial quali y imp o emen ools [1–3]. CPGs ha e a ange o pu poses, in ended o s anda dise ca e, imp o e i s qual- i y and sa e y, dec ease cos s, and imp o e pa ien ou - comes [3,4]. They o e a ‘one-s op shop’ o end-use s, by p o iding syn hesised in o ma ion om sys ema ic e iews ega ding bes p ac ices [5]. Howe e , despi e g owing a ailabili y o CPGs, i no used, hey canno impac on he quali y o he ca e ha is deli e ed. Sou h A ica has long been de eloping CPGs, mos p onounced du ing he pos -apa heid pe iod when CPGs we e conside ed impo an ools o ed ess in- equi y, s anda dise ca e and p omo e cos -e ec i e ca e o all. Many CPG de elopmen playe s ha e been iden- i ied: na ional go e nmen , p o essional socie ies, hospi- als and clinics all con ibu e acco ding o hei needs and agendas [6,7]. Howe e , despi e de elopmen and dis ibu ion o CPGs, heal h ou comes emain poo , and gene ally wo se han expec ed gi en he pe capi a heal h spend ela i e o o he simila middle-income coun ies [8,9]. As CPGs aim o op imise ca e, and ye ca e appea s no o be op imally deli e ed, i may be help ul o unde s and he ba ie s o CPG implemen a- ion and use [10,11]. We know he e a e no ‘magic bulle s’ o imp o ing CPG implemen a ion [12,13]. Sys ema ic e iews sug- ges many po en ial implemen a ion s a egies, such as audi and eedback, ou each educa ion and key opinion leade s [14]. A ailable e idence sugges s ha ailo ed, mul i- ace ed app oaches may do be e han gene ic and single- ocused in e en ions [13,14]. Se e al p agma ic ials o CPG implemen a ion o lung heal h, Human Immunode iciency Vi us (HIV) and b oade p ima y ca e ha e been conduc ed in Sou h A ica, inding some imp o emen s when educa- ional ou each is used [15–17]. I is he e o e possible ha , when used, CPGs may imp o e heal h ou comes. I we be e unde s ood when and how CPGs a e used by Sou h A ican p ima y ca e p o ide s, hen CPG de el- ope s may design e idence-in o med s a egies o enhance enable s and o e come ba ie s. The Theo e ical Domains F amewo k (TDF) is a use ul app oach o iden i ying acili a o s o and ba ie s o be- ha iou change, and o de eloping ailo ed in e en ions when implemen ing CPGs [18]. Unde s anding how bes o enhance heal hca e p o ide s’use o CPGs equi es conside a ion o he complex in e play o clinician and pa- ien beha iou s, en i onmen al con ex and social in lu- ences. The TDF in ends o in eg a e heo ies o beha iou change, and b idge heal h psychology, o ganisa ional he- o y and heal h se ices, p o iding a heo e ical basis o implemen a ion esea ch [19]. Se e al s udies ha e used he TDF o e alua e heal hca e implemen a ion challenges o o design heo y-in o med implemen a ion s a egies. Examples include hand hygiene, child en’s heal h checks, human papilloma i us accina ion, den al in ec ions, and lowe back pain [10,11,20–22]. Some o hese explo a- ions u he in o med he design o complex in e en- ions o esea ch o public heal h p og ammes [20]. U ilising he TDF, his s udy aimed o explo e p ima y ca e heal hca e p o ide s’pe spec i es ega ding he con ex , po en ial ba ie s o and enable s o CPG use in ou p o inces in Sou h A ica. Based on he indings, and d awing on concep s om he Beha iou Change Wheel (BCW), his s udy also sough o p o ide ecom- menda ions o po en ial in e en ions o imp o e CPG usage and implemen a ion. Me hods Theo e ical amewo k We used a quali a i e s udy design, including semi-s uc- u ed ocus g oup discussions (FGDs). The o e a ching con- cep ual amewo k used o his a icle was he Theo e ical DomainsF amewo k(TDF) ha also o medpa o ou analysis p ocess desc ibed below. The TDF p o ides a basis o unde s and beha iou s heo e ically and he e o e a ge p ocesses mos likely o implemen desi ed change [19,23]. The 14 domains o TDF ha e been u he mapped on o he Capabili y, Oppo uni y, Mo i a ion –Beha iou al model (COM-B model), a ‘beha iou sys em’model which seeks o encapsula e he condi ions in e nal o indi iduals and hose wi hin hei social and physical en i onmen necessa y o achie ing speci ied beha iou al a ge s [18]. Th ee essen ial condi ions: Capabili y, Oppo uni y, and Mo i a ion (COM-B) a e a he co e o his sys em, which posi s ha hese componen s in e ac o gene a e beha iou , which in u n in luences hem in a back-and- o h cycle. These com- ponen s o m he hub o wha is e med a ‘Beha iou Change Wheel’(BCW), a ound which a e a numbe o in e en ions which may be implemen ed a he indi idual (e.g. educa ion and aining), o policy le el (e.g. legisla ion o iscal mea- su es) o enable he COM-B elemen s [18]. The BCW is a p ac ical ool ha can be applied in implemen a ion esea ch o mo e om iden i ying ba ie s and enable s o aligning hese wi h ailo ed in e en ions [24]. De ini ions o he COM-B domains, how hey map o he TDF and o he BCW in e en ion unc ions a e shown (Table 1). S udy se ings Sou h A ica has a popula ion app oaching 57 million and a heal h sys em in es ed in p ima y heal hca e [25–29]. The coun y is cu en ly s i ing o uni e sal heal h co e age, publishing a Whi e pape (2015) desc ibing as- pec s o he Na ional Heal h Insu ance sys em [30]. Finan- cial ede alism is in place in which na ional go e nmen de elops s a egies, policies and clinical CPGs; and K edo e al. BMC Heal h Se ices Resea ch (2018) 18:965 Page 2 o 12 p o incial go e nmen s implemen CPGs, some imes a e adap a ion, o heal hca e acili ies ( om egional, o dis- ic , o communi y heal hca e acili ies) [8]. Sampling and ec ui men Sou h A ica is a la ge and di e se coun y. We he e o e selec ed ou o he nine p o inces o ep esen a spec um o p ima y heal hca e se ings: Wes e n Cape, Kwazulu-Na al, Eas e n Cape and Limpopo p o inces. Each p o ince is di e en in e ms o popula ion size and densi y, economic de elopmen , heal hca e spending and esou ces, and heal h ou comes (Table 2). While he Wes e n Cape, Eas e n Cape and Limpopo ha e simila popula ion sizes, he Wes e n Cape is be e unded, and Table 1 Links be ween COM-B, Theo e ical Domains F amewo k and Beha iou Change Wheel in e en ion unc ions COM-B model Theo e ical Domains F amewo k Beha iou Change Wheel In e en ion unc ions Mo i a ion De ini ion: all hose cogni i e p ocesses ha di ec beha iou , including habi ual p ocesses, emo ional esponding, as well as analy ical decision-making. Re lec i e mo i a ion P o essional/ social ole and iden i y Educa ion, pe suasion, modelling Belie s abou capabili ies Educa ion, pe suasion, enablemen Op imism Educa ion, pe suasion, modelling, enablemen Belie s abou consequences Educa ion, pe suasion, modelling In en ions Educa ion, pe suasion, incen i isa ion, coe cion, modelling, enablemen Au oma ic mo i a ion Rein o cemen T aining, Incen i isa ion, coe cion, En i onmen al es uc u ing Emo ion Pe suasion, incen i isa ion, coe cion, modelling, enablemen Capabili y De ini ion: he indi idual’s psychological and physical capaci y o engage in he ac i i y conce ned, and includes ha ing he necessa y knowledge and skills. Physical capabili y Physical skills T aining Psychological capabili y Knowledge Educa ion Cogni i e and in e pe sonal skills T aining Memo y, a en ion and decision p ocesses T aining, en i onmen al es uc u ing, Enablemen Beha io al egula ion Educa ion, aining, modelling, enablemen Oppo uni y De ini ion: all he ac o s ha lie ou side he indi idual ha make he beha iou possible o p omp i . Physical oppo uni y En i onmen al con ex and esou ces T aining, es ic ion, en i onmen al, es uc u ing, enablemen Social oppo uni y Social in luences Res ic ion, en i onmen al, es uc u ing, modelling, enablemen Table 2 Key heal h and demog aphic indica o s by Sou h A ican p o ince a Indica o Yea P o ince WC KZN EC LPP A ea as a % o o al a ea o Sou h A ica 2011 10.6 7.7 13.8 10.3 Popula ion 2016 6,279,730 11,065,240 6,996,976 5,799,090 Popula ion % by p o ince 2016 11.3 19.8 12.6 10.4 GDP pe capi a (USA) 2010 8.69 4.77 3.65 4259 Educa ion le el (% popula ion wi h no schooling) 2015 1.5 6.7 6.1 9.8 Po e y p e alence ( ood po e y line) 2011 23.2 37.4 40.5 41.5 Popula ion % dependen on public sec o 2016 75.96 88.22 90.13 91.58 Heal h as % o o al expendi u e 2000 30.0 26.7 20.9 17.8 Pe capi a public sec o heal h expendi u e 2015 4242.5 3623.1 3304.4 2957.7 Li e expec ancy a bi h 2010 68.0 52.9 53.8 63.6 Adul mo ali y a e (p obabili y o dying be ween 15 and 60 yea s) 2010 26.6 52.8 52.2 37.7 Unde 5 mo ali y a e 2015 23.1 57.8 59.6 36.6 a Adap ed om Sou h A ican Heal h Re iew 2017 [27] K edo e al. BMC Heal h Se ices Resea ch (2018) 18:965 Page 3 o 12 has highe educa ional le els, lowe le els o po e y and a highe li e expec ancy han he o he s. Kwazulu-Na al has he la ges popula ion size, a high po e y p e alence and poo li e expec ancy, despi e heal h expendi u e ap- p oaching ha o he Wes e n Cape. O he ac o s be- sides a ailable unds, a e likely o play a ole in his ega d, including high p e alence o in ec ious diseases, such as HIV [25]. Wi hin each p o ince, we a ge ed wo public sec o p ima y ca e clinics, one u al and one u ban o pe i-u ban. While we in ended o conduc eigh FGDs, we comple ed se en due o delayed access in he Wes e n Cape. To iden i y clinics, we con ac ed he p o incial esea ch di ec o a es and colleagues wo k- ing in he p o inces o guidance. All heal hca e p o- ide s wo king a clinics, ega dless o cad e, we e in i ed o pa icipa e (Table 3). Da a collec ion and managemen The FGDs enabled us o explo e collec i e expe iences o CPG use a he on line o heal hca e deli e y. This me hod is sui ed o explo ing complexi y su ounding CPG use wi hin he con ex o li ed expe iences, in ways ha encou age pa icipan s o engage ac i ely wi h he esea ch opic [31,32]. Se en FGDs we e held om No embe 2015 o Augus 2016. G oup sizes anged om h ee o ele en pa icipan s and las ed om 60 o 90 min. A o al o o y-eigh p o- ide s pa icipa ed. P ima y ca e p o ide s who ook pa included nu ses, occupa ional he apis s, physio he apis s, die icians, den is s, o al hygienis s and medical doc o s. The FGDs we e guided by a semi-s uc u ed opic guide which explo ed he ollowing opics: he con ex o CPG awa eness and use; speci ic CPGs used (and e- quency o usage); access o CPGs; gene al iews and ex- pe iences o using speci ic CPGs; pe cep ions o ba ie s o and enable s o CPG use; and ecommenda ions o s a egies ha migh add ess cu en ba ie s o use. The guide was lexible o ensu e ha pa icipan s could ex- p ess wha was impo an o hem, and so lea nings om p e ious FGDs could be cla i ied and p obed u he in subsequen FGDs. The FGD guide was no based on he TDF, bu a he sough o unde s and nu ses’pe cep ions abou and expe iences wi h using CPGs on hei own e ms and hei own meaning ames. The TDF was used du ing he analysis s age o help analyse and o ganise he da a as desc ibed below. FGD acili a o s ecei ed aining in acili a ion echniques. All FGDs we e conduc ed in pai s; mem- be s o he esea ch eam (all emales) ook u ns o acili a e. FGDs we e eco ded digi ally. Re lec ions and sum- ma ies we e w i en a e FGDs o cap u e insigh s. Ini- ial coding and hema ic analysis we e conduc ed a e each FGD o guide he sampling p ocess and o ensu e da a sa u a ion. FGDs we e ansc ibed e ba im, and ansc ip ions we e e iewed o accu acy by he esea ch eam (TK, TM). A ew pa icipan s including a lay counsello and en y le el nu se chose o sha e hei iews using hei mo he ongue which was no English. A esea ch eam membe assis ed o ansla e hese sho sec ions o us o include in he analysis. Da a we e s o ed elec onically on passwo d-p o ec ed compu e s; a mas e lis and consen o ms we e s o ed in a locked cabine o which only he p ojec lead had access. Analysis We used an i e a i e, hema ic con en analysis ap- p oach [31,33]. Speci ically, wo esea che s ead he ansc ip s (TK, SA) and ag eed on he gene al mean- ing and cen al issues p esen ed. One esea che (TK) hen e- ead ansc ip s, pe o ming open coding e- la ed o gene al ques ions posed, including con ex , use, ba ie s o and enable s o CPG use, ex ac ing he ela ed quo es [34]. Quo es we e hen u he ex- amined (TK, SA) o mani es and la en meanings [35]. A his poin , we sea ched o concep ual ame- wo ks ha migh help us be e unde s and and o ganize he da a. The TDF was conside ed o p o- ide a use ul model in his ega d, enabling us o en- capsula e he indi idual and con ex ac o s ha acili a e and /o hinde CPG use ha we saw eme - ging om he da a. The model was also deemed alu- able o acili a e he subsequen ansla ion o ou indings in o ac ionable ecommenda ions o in e - en ions which a ge speci ic ba ie s. This model has been used success ully by o he s o e alua e heal hca e implemen a ion challenges and o design heo y-in o med implemen a ion s a egies [20,21]. Ha ing examined indi idual quo es o mani es and la en meanings, wo esea che s (TK, SC) hen used he TDF o u he ca ego ise he da a. In pa icula , speci ic quo a ions and hei meanings we e ma ched o he 14 domains wi hin he TDF. The wo e- sea che s pe o med he ma ching independen ly, and Table 3 Schedule o Focus G oups Loca ion Discipline Numbe o ocus g oups (pa icipan s) Wes e n Cape Nu ses, den is s, heal h p omo ions o ice 1(n=6) Eas e n Cape Nu ses 2 (n= 12) Limpopo Den is s, o al hygienis , occupa ional he apy, physio he apy, die ician, counsello s, da abase adminis a o 2(n= 17) Kwa-Zulu Na al Doc o s, nu ses, quali y assu ance o ice , den is , physio he apis , counsello s 2 (n = 12) K edo e al. BMC Heal h Se ices Resea ch (2018) 18:965 Page 4 o 12 subsequen ly discussed hese wi h each o he and he hi d esea che (SA) o each ag eemen and esol e unce ain ies. Each quo a ion was coded o a leas one TDF domain, bu some we el could be coded in o wo o h ee domains. In he case o he la e , judgmen s we e made abou which speci ic domain he quo es should be ca ego ized, in a manne ha cap u ed he meaning o indi idual quo es and i ed wi h he b oade hemes ha we e eme ging. Once ou indings we e aligned wi h he TDF domains and associa ed COM-B sys em, hen p oceeded o map he indings on o he espec i e in e en ion unc ions o gene a e ecommenda ions based on he BCW [24]. The p ocess o de eloping ecommenda ions was in o med by he me hods used by Michie and col- leagues o link hei analysis o he a ge ed beha - iou s o app op ia e in e en ions o con olling obacco and educing obesi y [18]. Rigou C edibili y was ensu ed h ough de ailed cap u ing and desc ip ion o ou app oach o sampling, da a collec ion, da a managemen , analysis and in e p e a ion [35]. Con- side a ion o issues ega ding e lexi i y and ans e abil- i y we e conside ed h oughou he p ocess. Quo a ions we e chosen o p o ide eade s he oppo uni y o in e - p e da a, es ablish con i mabili y and o show he ich- ness o he da a. Complemen a y esea ch compe encies and expe iences among all esea che s in luenced da a in e p e a ion and s eng hened s udy igou . Resul s Mos pa icipan s we e nu ses; wo we e doc o s a one FGD in Kwa-Zulu Na al (Table 3). Al hough we col- lec ed limi ed demog aphic da a, we obse ed ha hose in u al acili ies had wo ked o a longe ime and li ed in he a ea, whe eas a he mo e u ban acili ies, pa ici- pan s we e gene ally younge , mo e ecen ly appoin ed and po en ially mo e mobile. In his sec ion we epo he po en ial enable s o and ba ie s o CPG use in e ms o he COM-B domains o ‘Mo i a ion’( e lec i e, au oma ic), ‘Capabili y’(psycho- logical, social) and ‘Oppo uni y’(social, physical) (Table 1)[23], and e lec on and unpack he TDF ca - ego ies wi hin hem. Mo i a ion –Re lec i e and au oma ic Mo i a ion includes beha io s co esponding o e lec i e mo i a ion and hose ha a e mo e au oma ic o habi ual. We epo on bo h e lec i e and au oma ic mo i a ion as hey include issues o emo ion, p o essional iden i y, be- lie s abou capabili y and consequences. S ikingly, ac oss all FGDs, he o e whelming majo i y o pa icipan s exp essed mo i a ion o use CPGs. CPG use appea ed o e oke a ange o posi i e emo ional esponses, pa icula ly amongs nu ses. Sen imen s included ‘ eassu ing’,inspi ing ‘con idence’and p o iding a sense o au onomy o ‘inde- pendence’. The la e was pa icula ly p onounced in mo e u al se ings, wi h ew doc o s: I makes [allows] us o be in line wi h he doc o s, i makes us doc o s ou sel [sic], so i means you will be in- dependen (Nu se_LPP_ u al). Addi ionally, CPGs we e pe cei ed as use ul ools o engage he communi y, sha e in o ma ion and p o ec heal hca e p o ide s’p o essional in eg i y, which u - he mo i a ed use: E en i he e is a complain among he communi y membe s ha we ha e mismanaged his clien , so we say, I ha e managed his clien … h ough he guidelines and we show him he guidelines (Nu se_EC_ u al). O e all, CPGs we e pe cei ed as c edible sou ces. Nu ses and allied heal hca e p o ide s in se e al clinics desc ibed ha ing i s -hand expe ience o CPGs imp o ing pa ien ca e. One pa icula ly signi ican example ci ed was ha o HIV, whe e CPGs had changed apidly as he ield o HIV ca e changed in Sou h A ica. P o ide s desc ibed ha ing seen pa- ien s ansi ion om dying p io o he a ailabili y o HIV CPGs, o pa ien s li ing wi h HIV a e CPGs we e imple- men ed. This unde sco ed o hem he pe cei ed alue ha using CPGs b ing: I ’s wo king, because when we wan o ind ou ou s a is ics, people hey a e now…[HIV] nega i e… hey ha e go ARV’s [an i e o i als] and hey a e ine… (Nu se_KZN_ u al). Compa ed o nu sing s a , he link be ween CPGs, p o essional iden i y and enablemen seemed lesse o doc o s, as one doc o sugges ed: I mus con ess, we doc o s a e no e y good a seeing his is wha he guidelines says. This is he way I do hings and hen you go on. I ’s no jus he e bu i you go o ano he place you’ll ind he same hing. (Doc o _KZN_pe i-u ban). Capabili y –Knowledge and skills Capabili y includes knowledge, unde s anding, decision-making and skills as undamen al d i e s o beha iou . A con- sis en na a i e amongs pa icipan s was ha know- ledge o CPGs was no a ba ie o usage. Pa icipan s con eyed conside able awa eness o CPGs, wi h many naming se e al ha we e in egula , pe haps e en daily use. In addi ion o knowledge, emembe ing and deciding o use CPGs was no pe cei ed as a ba ie . Some pa icipan s e en oiced cu iosi y abou why we K edo e al. BMC Heal h Se ices Resea ch (2018) 18:965 Page 5 o 12 would conduc esea ch on some hing ha was so ob- iously pa o ou ine clinical ca e. While some pa icipan s desc ibed using CPGs o ‘each and e e y pa ien ’, o he s sugges ed ha hey we e mos likely o use CPGs in pa icula ins ances. Tha is, hey ended o use CPGs when aced wi h an un amilia clinical case o a change in he ecommenda ions ha spa ked cu iosi y, and equi ed lea ning: …wha makes me wan o ead some o hem is because I came ac oss such a pa ien , and I didn’ know wha o do hen I go back o ead. Tha is wha makes me wanna ead, o he wise I don’ hink I’ll jus si down and ead he guideline (O alheal h_LPP_pe i-u ban). Despi e hei own knowledge, pa icipan s exp essed an impo an gap in CPG awa eness amongs pa ien s and he public. Many el ha inc easing public awa eness o CPGs was impo an o success ul CPG implemen a ion. Tha is, a mo e heal h-li e a e and empowe ed public was pe cei ed o encou age accoun abili y o heal hca e p o- ide s. Se e al app oaches o aising public awa eness we e p oposed, including engaging jou nalis s, use o adio, ele ision and social media: Maybe when you’ e lis ening o [ he] adio and eading news, hey should in oduce his change e e ywhe e, be- cause e en [ he] pa ien s should know (O alheal h_LPP_ pe i-u ban). Ano he signi ican gap iden i ied by pa icipan s was aining in CPG usage. T aining was pe cei ed as an essen ial ool o ‘keep ab eas ’o ‘ge up o speed’ wi h CPG con en . I was also conside ed impo an o enhancing clinical p ac ice and ensu ing ha all disciplines ‘will be on he same le el’and hus p e en - ing a ‘clash o in o ma ion’.While aining was unani- mously pe cei ed as necessa y o p o icien CPG usage, pa icipan s we e undecided abou he se ing in which aining should ake place. Speci ic eedback abou he p os and cons o on-si e aining and o -si e wo kshops we e p o ided, which a e de ailed in Addi ional ile 1. Though aining was conside ed key o CPG use, many pa icipan s el ha skills building h ough aining was inadequa e. T aining, ega dless o whe he p o ide s we e om u ban o u al se ings, was conside ed insu icien o pa chy, no co e ing all opics and no inclusi e o all clinical disciplines. This inadequacy was pe cei ed o esul in CPGs which a e ‘ha d o in e p e ’and hus s a ha - ing o ‘s uggle’on hei own o use CPGs p ope ly. The managemen p ocess o deciding who would a - end wo kshops was also desc ibed as non- anspa en and un ai , wi h ‘no consis ency’su ounding a end- ance.Thus, while pa icipan s we e ca ego ical abou heneed o mo e aining, heissueo howbes odo his emains complex. Oppo uni y –Social and physical Oppo uni y includes bo h physical oppo uni y and social oppo uni y. Social oppo uni y conside s he social in lu- ences ha may impac CPG use. While his domain did no gene a e subs an ial discussion amongs pa icipan s, wha eme ged consis en ly, pa icula ly in u al acili ies, was he alue o suppo i e social and p o essional sys ems as enab- ling quali y clinical ca e and CPG use. These sys ems, in- cluding in ol emen o non-go e nmen al o ganisa ions, and associa ed cohesi e eams and s ong leade ship, we e pe cei ed o enable he cul u e o CPG use. So i ’s eam wo k ha ma e s, i you a e wo king as a eam you do (Nu se_EC_pe i-u ban). Whe eas we ound gene ally suppo i e social and p o essional en i onmen s, he physical en i onmen eme ged as a conside able obs acle o CPG use. This do- main gene a ed ex ensi e discussion, wi h se e al sub- hemes eme ging, namely: he need o adap o local con ex ; heal h sys em challenges; access o CPGs; CPG design needs; and digi al CPGs. In addi ion o desc ibing hese ba ie s in g ea dep h, pa icipan s om all disci- plines also p o ided p ac ical ecommenda ions o how hese con ex ual ba ie s migh be add essed. CPGs being insu icien ly adap ed o local con ex s eme ged as a key issue. Gi en he di e si y in a la ge coun y like Sou h A ica, he con ex in which CPGs a e used may di e by p o ince. Some CPG ecommen- da ions we e expe ienced as ‘no p ac ical’and no ap- p op ia e o local heal hca e con ex s. Many ag eed ha o CPGs o become ‘some hing ha can eally apply o us’and ha ‘ac ually wo ks o sui he PHC [p ima y heal h ca e]’,heal hca e p o ide s should be pa o CPG de elopmen p ocesses. Heal h sys em challenges eme ged as ano he majo ba ie o CPG implemen a ion. The abili y o ope a ion- alise CPG ecommenda ions was desc ibed as signi i- can ly hinde ed by ‘no budge ’,‘slow p ocu emen ’, o he lack o equipmen whe e s a simply ‘don’ ha e he ma- chine’. S ock ou s o medicines was highligh ed as an issue: when he e is a ecommenda ion and he medica ion is no he e…we a e s uck (Nu se_LPP_ u al). Rela edly, p ima y ca e clinic p essu es we e pe cei ed o limi p o ide s’abili y o p ope ly ead CPGs. All cad es desc ibed ha he ‘long queues ou side’and he ime needed o ‘page and page’ h ough a CPG was no easible du ing a consul a ion. Pa icipan s also iden i ied ba ie s ela ed o he design, layou and language o CPGs, and made K edo e al. BMC Heal h Se ices Resea ch (2018) 18:965 Page 6 o 12 sugges ions o how hese migh be imp o ed o en- hanceCPGuse(Addi ional ile2). Many spoke abou he leng hy na u e o CPGs and he ‘big ja gon Eng- lish’,which limi ed unde s anding and use. They exp essed a wish o ‘much mo e use iendly’CPGs, including using ‘sho di ec i e’and mo e simple lan- guage, and inco po a ing ‘summa ised’ e sions, mo e de ini ions, local e nacula and supplemen a y ools (e.g. pos e s) o aid unde s anding and suppo pa- ien engagemen . A doc o sugges ed ha , as people maybe ‘ isual lea ne s’,use o mo e a ac i e and ap- pealing o ma s, such as g aphics, cha s, and colou , would enhance CPG use. Colou -coding in one o he p ima y ca e CPGs (PC101) was desc ibed as e ec - i e, as one nu se said, i ‘keeps you on he oes’ (EC_pe i-u ban). Poo access o good quali y and up- o-da e CPGs ma e ialised as an especially pe inen physical ba ie o CPG usage. Many pa icipan s, pa icula ly hose in u al se ings, p o ided de ailed na a i es abou how ‘ha d o each’CPGs we e. Many desc ibed how hey equen ly ‘ge hem la e’o ha e access o ‘only one copy’in hei clinics. O he s spoke abou he way in which CPGs a e o en s o ed inaccessibly ou side o consul ing ooms, while o he s highligh ed he poo sys ems ha exis o CPG e sion con ol, ul ima ely esul ing in ‘con usion’and ou da ed in o ma ion. Fu - he mo e, i eme ged ha e en when CPGs a e a ail- able, hey a e equen ly o sub-s anda d quali y: They make copies and pages a e missing, he a angemen o he pages, [i ] becomes bulkie and all hese hings. So ha ’s a p oblem, I mean people don’ eally ge he eal hing, a ep in o make a copy and make you own. (Doc o _KZN_pe i-u ban). Nume ous pa icipan s, bo h u al and u ban, highligh ed ha many o hese ba ie s a ound access would be add essed i CPGs we e a ailable digi ally. They explained ha access o digi al CPGs would en- able hem o ead hem in hei own ime, no only du ing consul a ions, which would in u n make keep- ing up- o-da e easie . They also sugges ed ha i would imp o e knowledge ans e a e wo kshops, educing issues ela ed o in o ma ion sha ing. Addi ionally, many belie ed ha digi al CPGs would esul in all heal hca e p o ide s ecei ing CPGs in a imely manne and u he suppo in- acili y capaci y building when new CPGs we e dissemina ed. Despi e gene al ag eemen ha digi al CPGs may a- cili a e usage, a numbe o complexi ies associa ed wi h his medium eme ged. Some pa icipan wonde ed whe he use o digi al CPGs in on o pa ien s would gene a e nega i e pa ien pe cep ions, who migh belie e ha heal hca e p o ide s a e ‘busy on Wha sapp’,acces- sing o he nonwo k- ela ed con en , o ha hey lack knowledge. A he same ime, while some pa icipan s had CPGs on hei phones, including he CPG app o elec onic books, his was a mino i y, and mos ly seen in pe i-u ban acili ies. Mos clinics did no ha e in e ne access ei he ia compu e s a ions o wi eless in e ne , and heal hca e p o ide s did no consis en ly ha e sma phones, da a and in e ne access h ough o he means. This was pa icula ly e iden in he mo e u al clinics whe e in a FGD o 11 s a , one nu se epo ed ha ing opened a pe sonal email accoun , and e en ha was a ecen de elopmen . Al hough pa icipan s in he Wes - e n Cape FGD desc ibed ha ing pe sonal in e ne access, hey sugges ed ha limi ed phone memo y, high da a cos s and he need o download CPGs a hei own ex- pense was a ba ie . Thus, use o digi al CPGs was de- sc ibed o come wi h i s own se o access issues, and while e iden ly desi able, emains aspi a ional om p o- ide s pe spec i es. Implica ions o policy and p ac ice: Theo y in o med in e en ions The ba ie s mos o en exp essed by pa icipan s we e ela ed o he en i onmen al con ex , esou ces and aining needs. We hus used he BCW app oach o map he mos ele an in e en ion unc ions o add ess hese speci ic ba ie s, as shown in Fig. 1 [18]. In his ma ix we p o ide speci ic sugges ions o possible in e en ions o inc ease use o Sou h A - ican p ima y ca e CPGs. The e o e, om ou indings, ‘physical and psycho- logical capaci y’, in pa icula poo ly suppo ed aining was a ba ie o CPG use; and mos s ikingly, he ‘phys- ical oppo uni y’, in ha he en i onmen al con ex and a ailable esou ces we e subs an ial challenges o CPG use. Based on ou esul s, he ollowing in e en ion unc ions a e sugges ed ha align he COM-B domain, beha iou al ba ie s and possible in e en ions: –T aining - impa ing skills ( o example wo kshops, on si e men o ing and supe ision, pos - aining suppo ) –Educa ion –inc easing knowledge o unde s anding abou speci ic CPG ecommenda ions (e.g. wo kshops, pos wo kshop suppo and clinical suppo ) –En i onmen al es uc u ing –changing he physical en i onmen (e.g. making he CPGs mo e accessible h ough di e en o ma s, g ea e design conside a ion, summa ized simple language, mo e appealing ools ha suppo implemen a ion ha help engage pa ien s such as pos e s and algo i hms; K edo e al. BMC Heal h Se ices Resea ch (2018) 18:965 Page 7 o 12 ensu ing supply chain unc ioning and access o medicines and equipmen ; building ICT in as uc u e and c ea ing digi al access) –Enablemen –inc easing means and educing ba ie s o inc ease capabili y o oppo uni y ( o example, his may include audi and eedback, clinical suppo and eam building). Discussion This s udy explo ed he pe spec i es o p ima y ca e heal hca e p o ide s, wo king in public sec o clinics, e- ga ding he con ex , po en ial ba ie s o and enable s o CPG use in ou di e se p o inces in Sou h A ica. We in es iga ed hese issues h ough he lens o he TDF, in o de o ca ego ise he ba ie s and enable s in e ms o COM-B: Capabili y, Oppo uni y, and Mo i a ion. Enable s –mo i a ion, knowledge and social oppo uni y. Ou indings e ealed ha p ima y heal hca e nu ses and allied heal h p ac i ione s gene ally us he c ed- ibili y o CPGs and a e highly mo i a ed o use hem. CPG usage was pe cei ed o be associa ed wi h a ange o posi i e emo ional and p o essional consequences, ex- pe iences ha ha e been desc ibed elsewhe e as po en- ial bene i s o CPGs [3]. In addi ion, knowledge o CPGs, along wi h egula use, we e epo ed by mos heal hca e p o ide s, a inding o impo ance, gi en ha bo h knowledge abou and posi i e a i ude owa ds a CPG a e ac o s ha ha e been iden i ied as po en ially enhancing CPG implemen a ion [36]. Pa icipan s in ou s udy, pa icula ly nu sing s a , emphasised he impo ance o cohesi e eams, clinical supe ision and s ong leade ship o enable CPG use. This co obo a es epo s om highe income se ings, which desc ibe he impo ance o socio-beha iou al ac- o s, such as pee suppo , o enhance CPG use [36–38]. S udies in Sou h A ica ha e e ealed ha suppo and supe ision o heal hca e p o ide s a e cu en ly inad- equa e. Fo example, a quali a i e s udy in which allied heal h p ac i ione s and heal h manage s we e in e - iewed, ound a lack o suppo o allied heal h p ac i- ione s in hei p ac ice [39,40]. Simila ly, a ecen su ey among p ima y heal hca e nu ses sugges s ha many el unsuppo ed by supe iso s o p o ide bes quali y clinical ca e [37,41]. Agains his backd op, and in ligh o he indings om ou s udy, enhancing CPG use in Sou h A ica necessi a es de eloping cohesi e p o essional eams and building clinical suppo o p ac i ione s. Ba ie s –Physical capabili y (skills and aining) Despi e appa en knowledge o CPGs and mo i a ion o use hem, pa chy and non-inclusi e aining in CPGs eme ged as an impo an ba ie o hei usage. Lack o equisi e skills and sel -e icacy a e epo ed ba ie s o CPG implemen a ion [36]. The pa icipan s in ou s udy conside ed skills building in CPGs essen ial o enhan- cing p o iciency o use CPGs p ope ly, ensu ing simila le els o capabili y and knowledge amongs heal hca e p o ide s, and o acili a ing s anda dised use ac oss disciplines. Howe e , exac ly how his aining should be deli e ed eme ged as a complex issue, wi h pa icipan s sugges ing ad an ages and disad an ages o bo h on- o o -si e ain- ing. Ou pa icipan s alked abou o -si e educa ional mee ings, on-si e educa ional ou each and suppo i e Fig. 1 Ma ix o COM-B Model ba ie s and sugges ed in e en ion unc ions. This igu e ep esen s a Ma ix o ba ie s ha we e iden i ied om pa icipan s and he po en ial in e en ions o o e come hem, as guided by he BCW. The ma ix is colou coded and all blue colou ed a eas ep esen whe e he COM-B domain aligns wi h he in e en ion unc ions. The da ke he shade o blue, he mo e pe inen he need o an in e en ion, in ligh o ou indings K edo e al. BMC Heal h Se ices Resea ch (2018) 18:965 Page 8 o 12 clinical audi s as desi able. In an o e iew o sys ema ic e iews, se e al skills building s a egies o implemen ing heal h sys ems in low- and middle-income se ings we e epo ed wi h a ying le els o e ec i eness, including p ac ice acili a ion, educa ional ou each, audi and eed- back, educa ional mee ings, and local opinion leade s [14]. In Sou h A ica, he e ha e been se e al ials o educa- ional ou each o nu se-d i en p ima y ca e e alua ing CPG implemen a ion [15,17]. As such, we ha e suppo - ing e idence ega ding easibili y o his app oach o managing co-mo bidi y, and in some s udies, e idence o e ec i eness o ube culosis and HIV CPG implemen a- ion [15,17,42,43]. Thus, while his s udy e ealed a clea s a ed need o inc eased skills building, he bes means o p o iding his in Sou h A ican p ima y ca e migh con- side using a combina ion o me hods o enable CPG up- ake and use. Ba ie s - physical oppo uni y (en i onmen al con ex and esou ces) While o he COM-B cons uc s eme ged as enable s o CPG usage, ‘physical oppo uni y’ma e ialised as he mos subs an ial ba ie , wi h pa icipan s highligh ing nume ous con ex ual issues ha hinde e ec i e CPG use. These may be u he unde s ood as pe aining o wo aspec s, he CPG i sel o he en i onmen al con ex in which CPGs a e implemen ed. Rega ding he CPG i sel , pa icipan s pe cei ed ha usage o CPGs is hinde ed signi ican ly when hei con en is imp ac ical o implemen and linguis ically inapp op i- a e; when CPG design ea u es a e no use - iendly; and i he e a e inadequa e CPG supplemen a y ools (e.g. pic- u es) o no summa ised e sions. This esona es wi h a e iew o di e en ea u es o ensu ing CPG ‘implemen - abili y’, oge he wi h a suppo ing checklis o CPG de- elope s o conside [44]. These esou ces sugges ha speci ic ea u es o CPGs a e likely o enhance hei usage, including s uc u ed ecommenda ions; p o iding con- ex ual in o ma ion ega ding clinical cases; explici e- sou ce implica ions; and suppo ing algo i hms and clinical ools [44,45]. In e ms o he physical en i onmen , se e al ac o s we e iden i ied as c i ical obs acles o CPG implemen a- ion. In pa icula , a lack o necessa y equipmen and e- po ed budge a y and supply cons ain s, including s ock ou s o medicines, we e a conce n and pe cei ed o be ela ed o poo dis ic o p o incial managemen sys- ems. These heal h sys em challenges a e well desc ibed in he coun y, including a ecen quali a i e s udy in which access o equipmen o medicines posed se ious challenges o deli e y o heal h se ices o bo h use s and p o ide s o ca e [6,9,26,46,47]. In ou s udy, in- adequa e sys ems o dis ibu ion o p in ed CPGs and CPG- ela ed ci cula s, as well as poo CPG e sion and quali y con ol, appea ed o impac upon CPG use. Taken oge he , his collec ion o en i onmen al issues was seen by pa icipan s o esul in CPGs ha a e e- quen ly una ailable, inaccessible, o a subop imal quali y and/o di icul o implemen . While hese ba ie s eme ged ac oss he di e en s udy se ings, hey ap- pea ed o be pa icula ly pe inen and heigh ened in u al a eas. Aspi a ional enable - digi al access o CPGs Pa icipan s consis en ly sugges ed ha making digi ally- o ma ed CPGs and associa ed echnologies (e.g. in e - ne , compu e s, lap ops) a ailable was a key s a egy o inc ease CPG access and use. Digi al CPGs we e sug- ges ed o ed ess many o he con ex ual challenges hey cu en ly ace, such as lack o su icien CPG ha d copies o poo e sion con ol. The e is g owing e idence ega ding he ole o handheld de ices o suppo CPG use. A sys ema ic e iew epo ed ha doc o s and nu ses using a CPG on a handheld de ice may inc ease access o in o ma ion, adhe ence o a CPG and suppo o diagnosing condi ions, in compa ison o pee s using pape -based esou ces [48]. Howe e , despi e his p omising e idence, esul s emana e p edominan ly om high-income se ings whe e access and a ailabili y o echnologies a e di e en o hose in low- and middle-income se ings. The e o e, despi e in e es in his a ea and as -g owing oppo uni ies in echnology, cu en da a cos s, lack o in as uc u e, in e ne o de ices, pa icula ly in u al se ings, p esen majo challenges o his becoming a eali y, as e ealed in ou s udy. Implica ions o policy and p ac ice: S a egic heo y in o med in e en ions o o e come ba ie s Gi en he limi ed esou ces o in es in CPG implemen- a ion in many se ings, ensu ing ha he in e en ions bes ma ch he issues and ba ie s ha eme ge is a a- ional app oach. We iden i ied ha in es men o implemen ing p ima y ca e CPGs should conside en i - onmen al es uc u ing, enablemen , and aining and educa ion (Fig. 1). T aining and educa ion is al eady a majo means o deli e ing in o ma ion o p ima y ca e ia egional ain- ing cen es and esponsible dis ic aining pe sonnel. Howe e , he esul s o his and o he s udies, sugges speci ic adap a ions and enhancemen s need o be con- side ed and implemen ed [39,40] such as enhanced in- acili y aining and pos - aining clinical suppo . An- o he in e en ion unc ion is enablemen . Gi en he mo i a ion o heal hca e p o ide s o use CPG, u he enablemen using e idence-based s a egies, such as cons uc i e clinical audi and eedback, clinical suppo K edo e al. BMC Heal h Se ices Resea ch (2018) 18:965 Page 9 o 12