Snyder ME, Zillich AJ, Primack BA, Rice K, McGivney MA, Pringle JL, Smith RB. Exploring successful

17 Pages·2010·0.26 MB·English

ARTICLE IN PRESS
Available online at www.sciencedirect.com
ResearchinSocialand
AdministrativePharmacyj(2010)j–j
Original Research
Exploring successful community pharmacist-physician
collaborative working relationships using mixed
methods
Margie E. Snyder, Pharm.D., M.P.H.a,1,*, Alan J. Zillich, Pharm.D.b,c,
Brian A. Primack, M.D., Ed.M., M.S.d, Kristen R. Rice, M.P.H.d,1,
Melissa A. Somma McGivney, Pharm.D.e, Janice L. Pringle, Ph.D.f,
Randall B. Smith, Ph.D.g
aDepartmentofPharmacyPractice,SchoolofPharmacyandPharmaceuticalSciences,PurdueUniversity,W7555Myers
Building,1001W.10thStreet,Indianapolis,IN46202,USA
bCenterofExcellenceforImplementingEvidence-BasedPractices,RoudebushVAMedicalCenter,Indianapolis,IN,USA
cDepartmentofPharmacyPractice,SchoolofPharmacyandPharmaceuticalSciences,
PurdueUniversity,Indianapolis,IN,USA
dDepartmentofMedicine,UniversityofPittsburghSchoolofMedicine,Pittsburgh,PA,USA
eDepartmentofPharmacyandTherapeutics,UniversityofPittsburghSchoolofPharmacy,Pittsburgh,PA,USA
fDepartmentofPharmacyandTherapeutics,ProgramEvaluationResearchUnit(PERU),
UniversityofPittsburghSchoolofPharmacy,Pittsburgh,PA,USA
gDepartmentofPharmaceuticalSciences,UniversityofPittsburghSchoolofPharmacy,Pittsburgh,PA,USA
Abstract
Background:Collaborativeworkingrelationships(CWRs)betweencommunitypharmacistsandphysicians
mayfostertheprovisionofmedicationtherapymanagementservices,diseasestatemanagement,andother
patient care activities; however, pharmacists have expressed difficulty in developing such relationships.
Additional work is needed to understand the specific pharmacist-physician exchanges that effectively
contribute to the development of CWR. Data from successful pairs of community pharmacists and
physicians may provide further insights into these exchange variables and expand research on models of
professionalcollaboration.
Financial isclosure: The authors declare no conflicts of interest or financial interests in any product or service
mentionedinthisarticle,includinggrants,employments,gifts,stock,holdings,orhonoraria.
Funding: This study was supported by the American Pharmacists Association Foundation and the Community
PharmacyFoundation.Also,Dr.ZillichwassupportedbyaResearchCareerDevelopmentgrantfromtheVeterans’
AffairsHealthServicesResearchandDevelopment(#RCD06-304-1).
Previous Presentations: This study was presented previously at the American Pharmacists Association Annual
Meeting, March 18, 2007, Atlanta, Georgia, and at the 2007 Eastern States Pharmacy Residency Conference, May
11,2007,Baltimore,Maryland.
* Correspondingauthor.Tel.:þ13176132315ext.338;fax:þ13176132316.
1 Atthetimeofthisresearch,DrSnyderwasCommunityPracticeResidentattheUniversityofPittsburghSchoolof
Pharmacy,andMs.RicewasM.P.H.CandidateattheUniversityofPittsburghGraduateSchoolofPublicHealthand
ResearchAssociate,UniversityofPittsburghSchoolofPharmacy.
E-mailaddress:[email protected](M.E.Snyder).
1551-7411/10/$-seefrontmatter(cid:2)2010ElsevierInc.Allrightsreserved.
doi:10.1016/j.sapharm.2009.11.008
ARTICLE IN PRESS
2 Snyderetal./ResearchinSocialandAdministrativePharmacyj(2010)1–17
Objective: To describe the professional exchanges that occurred between community pharmacists and
physiciansengagedinsuccessfulCWRs,usingapublishedconceptualmodelandtoolforquantifyingthe
extent ofcollaboration.
Methods: A nationalpool of experts in community pharmacy practice identified community pharmacists
engaged in CWRs with physicians. Five pairs of community pharmacists and physician colleagues
participated in individual semistructured interviews, and 4 of these pairs completed the Pharmacist-
Physician Collaborative Index (PPCI). Main outcome measures include quantitative (ie, scores on the
PPCI) and qualitative information about professional exchanges within 3 domains found previously to
influencerelationship development: relationship initiation, trustworthiness, androle specification.
Results:OnthePPCI,participantsscoredsimilarlyontrustworthiness;however,physiciansscoredhigher
on relationship initiation and role specification. The qualitative interviews revealed that when initiating
relationships, it was important for many pharmacists to establish open communication through face-to-
face visits with physicians. Furthermore, physicians were able to recognize in these pharmacists
a commitment for improved patient care. Trustworthiness was established by pharmacists making
consistent contributions to care that improved patient outcomes over time. Open discussions regarding
professionalrolesandanacknowledgmentofprofessionalnorms(ie,physiciansasdecisionmakers)were
essential.
Conclusions:Thefindingssupportandextendtheliteratureonpharmacist-physicianCWRsbyexamining
the exchange domains of relationship initiation, trustworthiness, and role specification qualitatively and
quantitativelyamongpairsofpractitioners.Relationshipsappearedtodevelopinamannerconsistentwith
a published model for CWRs, including the pharmacist as relationship initiator, the importance of
communication during early stages of the relationship, and an emphasis on high-quality pharmacist
contributions.
(cid:2)2010Elsevier Inc. Allrightsreserved.
Keywords: Pharmacists; Physicians; Collaborative working relationships; Pharmacist-physician collaborative index;
Community
Introduction and Doucette have proposed a conceptual model
for the development of pharmacist-physician
Therecentproliferationofmedicationtherapy CWRs(Fig.1).9TheCWRmodelwassynthesized
management (MTM) services offered through
from models of interpersonal relationships, busi-
MedicarePartD1,2hasputaspotlightonpatient
ness relationships, and collaborative care from
care opportunities for pharmacists, particularly nursing/physician relationships.10-15 This frame-
thosewhopracticeinthecommunitysetting.Ac-
work illustrates how individual, context, and ex-
tivities, such as community pharmacist-provided
change characteristics influence movement along
MTM and disease state management, are en-
a collaboration continuum, from stage 0 (profes-
hanced when an effective collaborative working
sional awareness) to stage 4 (commitment to the
relationship(CWR)existsbetweenthepharmacist CWR).9 Individual characteristics are those spe-
and the patient’s physicians. The potential bene-
cific to each collaborating professional, such as
fits of physicians and pharmacists working to-
ageandeducationalbackground.Contextcharac-
gether have been documented.3-7 Nevertheless,
teristics,suchastheproximityoftheprofessionals
communitypharmacistsstruggletoestablishrela-
and shared organizational structures, are associ-
tionships with physicians. Lounsbery et al sur-
ated with the practice site of the collaborators.
veyed 970 pharmacists from various outpatient
Exchangesarethepersonalinteractionsthatoccur
practice settings regarding their agreement with
between physiciansand pharmacists.
potential barriers in providing MTM services
UsingtheCWRmodelasaguide,Zillichetal
and found that community pharmacists were
demonstrated that, although select participant
morelikelythanpharmacistsinotherambulatory
andcontextualcharacteristicsinfluencedrelation-
settings to agree that establishing CWRs with
ship development, exchange characteristics are
physicianswas abarrier to service provision.8
the principal drivers in the development of
To assist practitioners and researchers inter- pharmacist-physician collaborations.16 In 2005,
ested in pharmacist collaborations, McDonough
ARTICLE IN PRESS
Snyderetal./ResearchinSocialandAdministrativePharmacyj(2010)1–17 3
Methods
Studydesign andparticipantrecruitment
The first step in studying the professional
exchanges occurring among highly collaborative
pharmacist-physicianpairsistoidentifyexamples
ofthesepairstoserveasresearchparticipants.In
qualitative research, participants are selected for
their familiarity with the concept in question19d
in this example, the professional exchanges that
have led to successful collaborations. Therefore,
a nonrandom, purposeful sampling technique
wasusedforparticipantidentificationandrecruit-
ment.19Inthisstudy,becausetheobjectivewasto
learnspecificallyabouttheexchangesoccurringin
uniquely collaborative examples (rather than
among ‘‘typical’’ cases of pharmacists and physi-
cians), purposeful sampling was used to identify
only highly collaborative pharmacist-physician
pairs in community settings.
Fig. 1. Model for physician-pharmacist collaborative To identify community pharmacist-physician
working relationships. Reprinted with permission from pairs who have established highly successful
American Pharmacists Association (APhA). Copyright collaborations, ‘‘community pharmacy experts’’
APhA. (n¼178)fromthroughouttheUnitedStateswere
contacted, and each was asked to identify 1-2
Zillich et al found that these exchanges can be community-based pharmacists whom they per-
grouped into 3 domains: relationship initiation, ceived to be engaged in an effective professional
trustworthiness, and role specification.17 The collaborationwithaphysiciancolleague.Commu-
extentofprofessionalcollaborationcanbequan- nitypharmacyexpertsweredefinedasindividuals
tified through the administration of the Pharma- who are well positioned to be knowledgeable
cist-Physician Collaborative Index (PPCI), a 14- about a variety of community pharmacist practi-
item Likert scale that measures collaboration tioners, particularly those in their respective geo-
within the 3 exchange domains.16-18 This quanti- graphic area. Experts were not provided with
tative measure, however, does not reveal the spe- specific guidelines or definitions for ‘‘successful,’’
cificexchangesthathaveoccurredtoreachahigh ‘‘effective,’’ or ‘‘highly collaborative’’ relation-
level of collaboration. ships;rather,itwasanticipatedthatexpertscould
The purpose of the present study was to identify uniquely collaborative community phar-
describetheprofessionalexchangesthatoccurred macistsbasedontheirfamiliaritywithcommunity
between community pharmacists and physicians pharmacypracticeandthe‘‘typical’’relationships
engagedinsuccessfulCWRs,usingtheaforemen- that exist between community pharmacists and
tionedconceptualmodelandtoolforquantifying physicians. Similar recruitment approaches using
the extent of collaboration among the profes- pharmacy leaders to identify innovative commu-
sionalsasguides.Insightsfromthisstudymayas- nitypharmacistshavebeendescribedbyotherau-
sist researchers interested in understanding thors whohavestudied theserelationships.20,21
collaborative care models and pharmacists inter- Contacted ‘‘experts’’ included select faculty
ested in developing collaborations in their prac- from colleges/schools of pharmacy (n¼102),
tice, while further validating the CWR model pharmacy clinical services managers (n¼15),
proposed by McDonough and Doucette.9 To the and leaders of major pharmacy organizations
authors’ knowledge, this is the first study to ex- (n¼61).Thepharmacyfacultyincludedexperien-
plore,quantitativelyandqualitatively,theprofes- tiallearningprogramdirectorsfromeachcollege/
sional exchanges occurring among pairs of school and others engaged in community-based
community pharmacists and physicians engaged initiatives. These faculties were selected because
inhighlyCWRs. of their familiarity with community pharmacist
ARTICLE IN PRESS
4 Snyderetal./ResearchinSocialandAdministrativePharmacyj(2010)1–17
preceptors. Clinical service managers and phar- pharmacist informants included age and years in
macy association leaders were included because practice, education and training, size of commu-
oftheirfamiliaritywithpharmacistsprovidingdi- nity served, weekly hours of pharmacy operation
rect patient care services as employees of their and time spent each week on dispensing versus
company (for the former) or as active members patient care, position title, number of prescrip-
in the organization (for the latter.) A total of 47 tions filled per hour, number of pharmacists/
national ‘‘experts’’ responded to the queries and technicians on duty at one time, academic affili-
identified87communitypharmacistsforpotential ation, type and location of pharmacy, and cur-
inclusion in the study. Of note, some of these rently established patientcareservices.
pharmacistswereidentifiedbymorethan1source As described earlier, the PPCI is a validated
(eg, a faculty member and an association leader 14-itemLikertscaletoolthatquantifiestheextent
mayhaveidentified the sameindividual). of practitioner collaboration within the exchange
Tobeincludedinthisstudy,these87pharma- domains of relationship initiation, trustworthi-
cists had to be currently engaged in an active ness, and role specification. The PPCI provides
community-based patient care practice, with at a summary score from 14 to 98, with a higher
leastsometimedevotedtonondispensing,patient score indicating a greater extent of collabora-
careactivities(althoughtheexactamountoftime tion.16-18AlthoughtheintentionwastousePPCI
was not defined). For the purpose of this study, scoresasaguideforchoosingwhichpractitioners
community-based practices (ie, traditional com- tointerview(ie,invitinginterviewsfirstfromprac-
munitypharmacies)wereconsidereddistinctfrom titioners with the highest scores), because of the
traditional ambulatory care, primary care, or small sample size, all responding practitioners
familymedicinepharmacistpracticesettings.This meeting study eligibility criteria were invited to
distinction was made to control for contextual participate in the interviews, and the PPCI was
factors (such as shared physical space) that may used to provide quantitative information about
influencethedevelopmentofpharmacist-physician the professional exchangesoccurring betweenthe
relationships. Second, the pharmacists needed to pharmacists andtheir physiciancolleagues.
have a collegial relationship with a physician, as When completing the PPCI, all pharmacists
initiallyperceivedbyexpertreferralandconfirmed were asked to use 1 physician as a frame of
by the pharmacist’s willingness to identify that reference for an effective professional collabora-
colleagueforpotentialparticipationinthestudy.It tion and provide that individual’s name and
wasnotrequiredthatthepractitionersbeengaged contact information electronically. These physi-
in any form of legal collaborative practice agree- ciancolleagues werethencontacted andaskedto
mentorotherpracticeprotocol.Finally,thephysi- complete similar survey tools: a background/de-
ciancolleaguesidentifiedbythepharmacistshadto mographic survey and the PPCI from the physi-
agreetoparticipateinanindividualsemistructured cianperspective(Appendix2).Demographicdata
interview.Thestudyprotocolwasapprovedbythe collected from physicians included age and years
University of Pittsburgh Institutional Review inpractice,educationandtraining,numberofto-
Board. tal hours worked per week and number of those
hoursspentonpatientcareactivities,sizeofcom-
munity served, physician work position, number
Datacollection
of patients seen per week, academic affiliation,
First, each of the identified pharmacists was andtypeandlocationofpractice.Eachphysician
contacted and asked to complete both an online was asked to complete the PPCI using his/her
survey of basic background and demographic pharmacist colleague as the frame of reference.
information along with the PPCI16-18 (Appendix Both practitioner participants were also asked to
1). A link to both surveys was included in the briefly describe, in their own words, their profes-
body of the e-mail. Survey responses were col- sional collaboration with the other professional
lectedinSurveyMonkey(cid:3),anInternet-basedsur- (either pharmacist orphysician).
vey tool. All informants gave electronic consent To gather qualitative information about the
beforecompleting thesurvey. specific interactions that occurred, these practi-
Demographic information was collected to tioner pairs were then asked to participate in
describe the participant and context characteris- individual interviews with study investigators.
tics that may influence the relationships being Both practitioners (pharmacists and physicians)
studied.Demographicinformationcollectedfrom in each pair were interviewed independently.
ARTICLE IN PRESS
Snyderetal./ResearchinSocialandAdministrativePharmacyj(2010)1–17 5
During the interviews, participants were asked these identified pharmacists, 24 provided consent
a standardized set of open-ended questions to and completed the online survey tools. Ten of
extract information about actions taken in the thesepharmacistswereexcluded,becausetheydid
exchange domains of relationship initiation, the not practice in a traditional community setting.
development of trustworthiness, and professional Two pharmacists were excluded for incorrectly
rolespecification(Appendices3and4).Interview completingthesurveytools(eg,notprovidingthe
questionsweretestedforfacevaliditythroughre- nameof a physiciancolleague). Twopharmacists
view by clinically trained pharmacist and physi- didnotrespondtotherequestforparticipationin
cian investigators. Pilot testing was performed the qualitative interview. Five pharmacists were
through individual interviews with a small group interviewedbutthenexcludedfromdataanalysis,
ofcommunity pharmacist practitioners. because their physician partner declined to par-
Interviews were conducted face-to-face at the ticipate. A total of 5 pharmacist-physician pairs
practitioner’s practice site, a common meeting completed individual semistructured qualitative
place,orbymeansoftelephone.Writteninformed interviews. One physician did not complete the
consentwasobtained fromall participants.Most background survey and PPCI, but their qualita-
oftheinterviewswereconductedbytheprincipal tive resultsare included in the followingsection.
investigator (M.E.S.), with the remaining inter-
Individual andcontext characteristics
views completed by a research assistant, trained
by the principal investigator. Interviews were Tables1and2summarizeindividualandcon-
audio (in the case of telephone) or audio- and textual characteristics of pharmacists and physi-
video-recorded when conducted face-to-face. In- cians obtained from the demographic surveys.
vestigatorstookfieldnotesduringeachinterview. All of the physicians providing information were
All interviews were transcribed verbatim and males,hadreceivedaDoctorofMedicine(incon-
reviewed byinvestigatorsfor accuracy. trast to a Doctor of Osteopathic Medicine and
most were in private practice. Most of the phar-
Dataanalysis
macists were malesand hadreceived a Doctorof
SPSS(version16.01;Chicago,IL)wasusedto Pharmacy (PharmD). Most pairs practiced in
calculatedescriptivestatisticsofdemographicvar- a relatively small community (ie, population of
iables and PPCI scores. A systematic approach less than 50,000). On average, the physicians
was used to evaluate the responses obtained were older and had been in practice longer than
from the semistructured interviews. Transcripts the pharmacists.
were entered into ATLAS.ti software (version
Exchangedomains
4.1;Berlin,Germany)forcontentanalysis.Acod-
ingschemewasdevelopedbasedonthetopicsad- AsummaryofPPCItotalanddomainscoresis
dressed in the semistructured interviews, and this provided in Table 3. On the trustworthiness do-
schemewas used to identify common themes dis- mainofthePPCI,pharmacistandphysicianpar-
cussed by participants. Transcripts were first ticipants appeared to score similarly; however,
read and coded independently by 2 investigators physicians appeared to score higher, on average,
(M.E.S. and K.R.), and a codebook was devel- than their pharmacist colleagues in the domains
oped to track and define variables throughout of relationship initiation and role specification.
the coding process. After independently coding Because of the small sample size, no statistical
each transcript, the investigators met to discuss comparisons were made. Table 4 triangulates the
coding decisions and finalize code assignments. qualitative (ie, representative quotations) and
Anydiscrepancieswereresolvedbygroupconsen- quantitative (PPCI scores) findings specific to
sus. Repeating themes under each question do- each exchange domain and pharmacist-physician
main are described later, with representative pair.
quotationsprovided.
Relationship initiation
Inthedomainofrelationshipinitiation,itwas
Results foundthatpharmacistsweretheprimaryinitiator
of these CWRs, and this role was acknowledged
Sample
by both types of practitioners during the inter-
Therewere87identifiedpharmacistsrepresent- views.Generally,pharmacistsapproachedeachof
ing a minimum of 29 states and Puerto Rico. Of the physicians in their geographic area (rather
ARTICLE IN PRESS
6 Snyderetal./ResearchinSocialandAdministrativePharmacyj(2010)1–17
Table1 Table2
Pharmacistcharacteristics(n¼5) Physiciancharacteristics(n¼4a)
Pharmacistcharacteristics Results Physiciancharacteristics Results
Ageinyears,median(range) 34(28-56) Ageinyears,median(range) 51(36-55)
Sex,no. Sex,no.
Men 4 Men 4
Women 1 Women 0
Yearsinpractice,median(range) 5(3.5-33) Yearsinpractice,median(range) 18(5-27)
Highesteducation,no. Education,no.
BS,pharmacy 1 DoctorofMedicine 4
PharmD 4 DoctorofOsteopathicMedicine 0
Residencytraining,no. 2 Communityserved,no.
!10,000 0
Communityserved,no.
10,001-49,999 3
!10,000 0
50,000-499,999 1
10,001-49,999 3
R500,000 0
50,000-499,999 1
R500,000 1 Typeofpractice,no.
Private 3
Typeofpharmacy,no.
Academic 1
Independent 3
Chain 2 Specialty,no.
Familypractice 2
Totalhoursworkedperweek,median 50(45-55)
Internalmedicine 1
(range)
Other 1
Hoursspentdispensing(weekly), 8(0-20)
median(range) Numberofpatientsseenperweek, 47(20-130)
Hoursspentonpatientcare(weekly), 25(16-30) median(range)
median(range) Hoursspentonpatientcareperweek, 34(20-65)
Pharmacistsonduty,median(range) 2(2-5) median(range)
Techniciansonduty,median(range) 4(2-5) Totalhoursworkedperweek,median 50(28-65)
Studenttrainingsite,no. 5 (range)
Residenttrainingsite,no. 3 Studenttrainingsite,no. 2
Residenttrainingsite,no. 1
Establishedclinicalservicesa,no.
Anticoagulation 0 a One physician did not complete background
Diabetes 4 survey.
Hypertension 2
Asthma 0
get the face-to-face. They are not going to refer
a Notmutuallyexclusive.
[patients] to someone they don’t know or they’ve
nevermetbefore.Evenifyouhaveallthecreden-
tialingintheworldtheyarenotgoingtodothat.
than targeting specific physicians). Initial conver- Soyouhavetogetinfrontofthem.Youhaveto
sations were usually (but not always) conducted tell them what your goals are .’’ Professionals
face-to-faceandoftenscheduledinadvancebythe whoultimatelybecamecollaboratorssharedasim-
pharmacist. Many of the participants described ilar value set in the sense that improved patient
the use of face-to-face visits as a mechanism for care was their primary motivator. Along with an
developing a ‘‘personal’’ relationship with their enhanced professional role for pharmacists, this
collaborator. One physician emphasized the im- wasoften theonlyexpectationthateitherprofes-
portanceoftheseencounters,‘‘.Rightnowinat sional had forcollaborating.
least one of the healthcare settings in which I Somepharmacistsscheduledthesepreliminary
work,Icouldnottellyouthenameoftheclinical meetings over lunch or dinner. Although the
pharmacist that gives me advice and I would not discussions clearly involved the physician, other
be able to recognize them if I was four feet from members of the office staff were sometimes pres-
them. So that to me is a bit of a problem.’’ A ent. Other mechanisms for establishing initial
pharmacist echoed this sentiment, ‘‘You have to contact with physicians included involvement in
ARTICLE IN PRESS
Snyderetal./ResearchinSocialandAdministrativePharmacyj(2010)1–17 7
Table3 useful in ensuring that recommendations are
ParticipantPPCIscoresa more likely to be accepted. As 1 physician de-
PPCIscore scribed, ‘‘I think that sometimes the way they
(possiblerange) Mean(cid:2)SD Range [pharmacists] frame their advice can be framed
in different ways. So frequently, especially if rec-
Pharmacists,n¼5
ommendations come in a written form, we’ll sit
Totalscore(14-98) 79.4(cid:2)10.2 62-88
Domain down and say, ‘Look if you could reframe this
Trustworthiness(6-42) 39.2(cid:2)3.1 35-42 in a different way in the future I think it would
Relationshipinitiation 16.2(cid:2)2.9 12-19 be easier for we physicians to respond to it con-
(3-21) structively rather than in a knee jerk fashion.’
Rolespecification 24.6(cid:2)6.9 15-32 So I think strategizing about ways to frame the
(5-35) recommendations canalso beveryhelpful.’’
Physicians,n¼4b
Totalscore(14-98) 89.8(cid:2)4.6 85-96 Trustworthiness
Domain In discussing the trust between these pharma-
Trustworthiness(6-42) 39.8(cid:2)1.7 38-42 cists and physicians, the conversations centered
Relationshipinitiation 20.3(cid:2)1.0 19-21 around actions taken by the pharmacist to gain
(3-21)
the physician’s trust (rather than the physician
Rolespecification 29.8(cid:2)2.9 26-33
needing to gain the pharmacist’s). Of primary
(5-35)
importance for establishing trust was the pro-
SD,standarddeviation. vision of high-quality clinical recommendations
a PPCI scores range from 14 to 98, with higher
that improved patient outcomes. Both profes-
scoresrepresentingamoreadvancedrelationship.
sionals commented on how seeing these positive
b Onephysiciandidnotcomplete.
outcomeswaskeytothesuccessoftheirrelation-
ship. These recommendations needed to be pro-
community organizations and sending written vided consistently over time to develop an
notes. Recognizing that relationships are built expectation for the level of care provided by the
overtimeandthatitwasimportanttotakethings pharmacists.
‘‘slow,’’ either with regard to asking for appoint- In addition to demonstrating competence, on-
mentswithphysiciansorinmakingdrugtherapy going communication with the physician (ie,
recommendations, were noted. One physician de- ‘‘keeping them in the loop’’) was very important.
scribedtheprocessheobservedinhispharmacist Onepharmacistdescribedthis,‘‘.he[thephysi-
collaborator,‘‘Well,hestartedslow.Hewas,this cian]needstoknowwhatwearedoingsoifheis
fellow,wasvery,verygoodatwinningconfidence queried by fellow physicians, the medical society,
and he took a little time to win mine. He just or ‘why are you supporting what they [pharma-
initially introduced himself. That was the first cists] are doing?’ He needs to know exactly what
thing. He knew I was busy and he knew he was we are doing . and vice versa, we need to
busy and then he just came back once or twice know that he is going to be able to supply us in-
a week for the first several months just to make formationorsupportatthelevelweneed.’’How-
sure things were going well. He didn’t try to lec- ever, unlike when the professionals were asked
ture me oreducate me or anything otherthan he about relationship initiation, this communication
knew the community and that was really did not usually happen face-to-face. More often,
obvious.’’ thiscommunicationoccurredbymeansofwritten
During these early conversations with physi- recommendations that were sent by fax for the
cians, the pharmacists explained their preferred physician to review at their convenience. As 1
role and the clinical services they offered. Some pharmacist described, ‘‘. making sure you keep
pharmacists also brought educational materials them [physicians] in communication. You know
thatwouldbeusedduringtheirvisitstopatients. like sending progress notes . just keeping them
Reflectingbackonthisfirstencounter,1pharma- in the loop because when you don’t keep them
cist also made the suggestion that it would have inthelooptheyreallywonderwhatisactuallygo-
been helpful to bring an example of the specifi- ing on.’’ The ability to vary communication
cally written documentation the physician can methods appropriately was also noted as impor-
expect to see after each patient visit. From the tantintherelationshipprocess,thatis,respecting
physician’s perspective, this may be especially the urgency of a situation and using clinical
ARTICLE IN PRESS
8 Snyderetal./ResearchinSocialandAdministrativePharmacyj(2010)1–17
judgment to respond with a phone call if neces- doing this and I haven’t had anybody do that)
sary,orifless urgent, the faxednote. sometimes things get changedwithout megetting
Preexisting relationships between the pharma- anactualrecommendationback.SoIrealizethat
cist and other community members or the phar- someone isreading italongtheway.I dothinkI
macist and the patient were also discussed. Some havethatresponsibilitytotakecareofthepatient
of the pharmacists were introduced to their soit doesn’tchange.’’
collaborator through another pharmacist and
commented that trust came more easily in these
situations. Physicians also commented that trust
Discussion
wasenabled,because the pharmacist hadperson-
allygottento knowthe patient. The process for identifying community phar-
macist-physicianpairsengagedineffectiveCWRs
Rolespecification was fruitful. Despite not providing experts with
Theinterviewsrevealedthatbothprofessionals a clear case definition of ‘‘effective’’ or ‘‘success-
weregenerallyinagreementregardingthepatient ful,’’thepharmacistPPCIscoreswerecomparable
carerole ofthe pharmacist as comparedwith the withthehighestscoresreportedinearlierstudies,
physician.Specifically,itwas notedthatpharma- indicating high levels of collaboration among the
cists ‘‘focus on the drug therapy’’ and provide identified sample.18 In addition, the physicians’
patient education whilesupporting the physician. PPCIscoreswerehigheracrosseachdomaincom-
Although physicians valued these professional pared with previously reported scores among
contributions, they emphasized that the role of a large, cross-sectional sample of primary care
thepharmacistwasquitedifferentfromtheirown physicians.16,17
roles.Physiciansreportedthatevenwhileengaged The qualitative exploration of the CWR ex-
in a highly CWR, they were still the team change domains revealed several exchanges that
members ultimately responsible for the patients’ occurred among the professionals when relation-
outcomes and, in that role, functioned as the ships were initiated, trust was established, and
decisionmakers. professional roles were clarified. Many of these
Pharmacists also described situations where findings (eg, pharmacist as initiator, the impor-
they have encountered physicians (not their phy- tance of communication at early stages of the
sician partner who was interviewed) who were relationship, and the emphasis on high-quality
resistant to collaboration. Resistance manifested pharmacist contributions) support the CWR
passively,aslackofphysicianresponsetorecom- model proposed by McDonough and Doucette9
mendations, and actively, as refusal to provide and provideopportunity for future study.
patientlaboratorydatainspiteofsignedmedical Specifically, the role of the pharmacist as
releases and hesitations to provide referrals for relationshipinitiatorhasbeendescribedinearlier
clinical services beyond patient education. When work examining these collaborations.20 The
asked how their patient care role was affected by study’s qualitative findings of the pharmacist as
theseencounters,thepharmacistsemphasizedthat theprimaryrelationshipinitiatorlikelyinfluenced
the professional service they provided did not thequantitativeresultsforthisexchangedomain.
vary. They stressed that, because the patient OntherelationshipinitiationdomainofthePPCI,
(rather than the physician) was their priority, physicians scored higher than pharmacists. Rec-
theyprovidedthesamelevelofcaretothepatient ognizing that relationship initiation depends
andcommunicationtothephysicianthattheydid largely on their actions, pharmacists may have
withphysicianswithwhomtheyhadahighlycol- been critical of their actions, resulting in lower
laborative relationship. The physicians then have scoresontheseitems.Ontheotherhand,thephy-
the choice whether to carry out or not to carry sicians who recognized that they would not have
out the recommendation. One pharmacist de- taken the initiative for relationship development
scribesthissituationinhispractice, ‘‘Myrespon- were pleased with the approaches taken by phar-
sibilityistotakecareofthepatient.Iwouldnever macists. Notably,themeanphysicianPPCIscore
letthatstopmefromdoingthat.Sotheymayget forthisdomainwas20.3outofa maximumpos-
upset that I keep sending them notes and recom- sible score of 21, suggesting high physician satis-
mendations, but I have found out even in those faction with the specific initiating behaviors
physicians who may not be responding (as long described by the pharmacists in this study. How-
as they’re not calling me and telling me to stop ever, other authors who studied CWRs from the
ARTICLE IN PRESS
Snyderetal./ResearchinSocialandAdministrativePharmacyj(2010)1–17 9
pharmacistperspectivehavenotfoundarelation- pairs)duringthequalitativeinterviews.Thiswar-
ship among exchanges within this domain and rantsfurtherstudy,becauseBrockandDoucette20
successful collaborations.21 Therefore, more collected data only from the pharmacist perspec-
workisneededtodeterminewhetherpharmacists tive, whereas the current study elicited both per-
initiatingrelationshipsinthemannerdescribedby spectives. Nevertheless, differences in
participants have greater success in developing perspectives of pairs engaged in high-level versus
collaborationsthanthoseusingother methods. lower-levelcollaborationswerenotassessed.Con-
Both the qualitative and quantitative findings sequently,moreworkisneededtounderstandthe
suggest that community pharmacists and physi- importanceoftheseexchangesondevelopingrela-
ciansengagedinhighlycollaborativerelationships tionships from the physician perspective.
view trustworthiness in a similar fashion. Both
professionals scored similarly on the PPCI and
Limitations
emphasizedsimilarcharacteristicsoftherelation-
ship that resulted in a high level of trust. In This study had a relatively low response rate;
particular,theimportanceofestablishinga‘‘track 26%ofthe‘‘experts’’respondedtotherequestfor
record’’ through consistent, high-quality contri- pharmacist identification, and fewer physicians
butions (by the pharmacist) to patient care was than pharmacists agreed to participate in the
emphasized bybothtypes of professionals. online surveys and interviews. This may be
For role specification, the qualitative findings because of ‘‘expert’’ misinterpretation of study-
suggested congruence between how each profes- inclusion criteria. For example, the first author
sionalviewedhisorherrespectiveroles.However, received several e-mails from ‘‘experts’’ stating
the discrepancy in PPCI scores between the thattheycouldnotidentifyapharmacist,because
pharmacists and physicians imply that pharma- the state they reside in does not allow legal
cists feel less strongly than physicians that both collaborative practice agreements. Furthermore,
professionals are mutually dependent on each alowerphysicianresponsemaybebecauseofthe
otherandthatpharmacistsareabletosuccessfully methodology used for recruitment, and/or the
negotiatetheirroleinpatientcare.Thisapparent lack of compensation for time associated with
discrepancybetweenthequalitativeandquantita- participation,becausetheinterviewsaveraged30-
tive results of this exchange domain warrant 60 minutes. Although an attempt was made to
further study, because other authors have found increaseparticipantenrollmentthroughfollow-up
that role specification is an important aspect of contacts with identified pharmacists and physi-
successful CWRs.16,21 cians,therewasnoattempttoidentifyandrecruit
To the authors’ knowledge, this is the first new pharmacist-physician pairs. This approach
study to triangulate PPCI scores with qualitative could haveresulted in a greater samplesize.
perspectives from both professionals. Several of Although analysis of the qualitative data re-
thefindingssupporttheworkofBrockandDouc- vealedrepeatingthemes,itisunknownifadditional
ette,20whoconductedanin-depthcasestudywith themeswouldhaveemergedwithagreaternumber
10 pharmacists engaged in varying levels of col- ofinterviews.Moreinterviewsalsomayhavepro-
laborationwithphysicians.Thepurposeofthein- vided greater insights into the qualitative and
quiry was to identify variables that distinguish quantitative discrepancies we found for the role
between highly collaborative and less collabora- specificationdomain.Furthermore,thedesignand
tive relationships. Similar to the current results, sampling strategy for this study only included
these authors found that the pharmacist was the participants with high levels of collaboration.
relationship initiator, that face-to-face communi- This study did not explore outliers who do not
cation was important, and that the relationships collaborate or struggle to collaborate. Future
developed over time.20 However, they did not studies among a group of low collaborators may
find that initiating behaviors, trust, conflict reso- provide additional information important in de-
lution, an assessment (by the physician) of the velopingrelationships.Inaddition,giventhesmall
pharmacists’ competence, or a history of a prior samplesize,inferencesfromthequantitativedata
relationship among professionals differentiated arespeculative.Futurestudieswithlargersample
between the pharmacists in high-level collabora- sizesshouldexplorevariationsinthemagnitudeof
tions versus those at a lower stage of the CWR. the difference between pharmacist and physician
In the present study, each of these exchanges PPCIscores.Finally,althoughsomeofthepartic-
was discussed by both professionals (across the ipantshavebeeninformallyexposedtothecurrent
ARTICLE IN PRESS
10 Snyderetal./ResearchinSocialandAdministrativePharmacyj(2010)1–17
Table4
PPCIscoresandparticipantquotations
RPhPPCI MDPPCI
Pair domainscore RPhquotations domainscore MDquotations
Relationship
initiationa
1 15 ‘‘Ireallythink.relationships 21 ‘‘Andwhathedidthatwasex-
havedevelopedbecauseof traordinarywashewentout
workingforacommongoal ofhiswaytogettoknowthe
andsowhenIsaythat,I patients.’’
meanpatientcaregoals..’’ ‘‘Andapersonwithgoodcom-
‘‘.thequalityoftheinterven- municationskills.andan
tionIamprovidingandI honestintenttodowhatis
provideitinaveryconcise rightforthepatient.’’
wayandthesamewayasifI ‘‘.Ithinkbeingabletoartic-
wasface-to-face.itispa- ulateyourrecommendation
tientfocused.’’ andinasuccinctwaysupport
yourrecommendationcomes
inveryhelpful.’’
2 12 ‘‘.Thereisasmallgroupof 21 ‘‘.webothhavedifferentjobs
physicians.sowepretty butwebothhaveanendgoal
muchhitthemall.’’[referring andthatistotakecareofthe
towhenvisitingphysiciansin patient.’’
thecommunity] ‘‘Justbeingreceptivetotheir
‘‘.wedidgotoeveryphysician commentsandtheirknowl-
intheareaanddidalittle edge.’’
breakfastorlittlelunchex-
plainingtothemthethings
thatwewouldbetryingto
do.’’
‘‘.we’reinapharmacythat
hasbeeninthecommunity
foralongtime.’’
‘‘.justmakingsuremyinten-
tionswerepositive.’’
3 16 ‘‘Sohisidealsandourideals 19 ‘‘Hecametomyofficeandsaid,
matchedverywell.’’ .thisiswhatIdoandthisis
‘‘.ourpatientfocusedactivi- whoIamandthisiswhere
tieswhenwedecidedwewere mypharmacyis.’’
goingtobeinvolvedinthat, ‘‘Thatwastheicebreakerandit
approachingphysicianswas madeitsoeasyafterthat.But
wherewestarted.’’ clearlyitwashisinitiative
‘‘.wehadalongstandingtra- thatmadeithappen.’’
ditioninthecommunity.’’ ‘‘Heissincere.Hedoesagood
‘‘Hisbottomlineishewantsto jobforthepatientsanditwas
improvepatientcare.’’ veryobvious.’’
4 19 ‘‘ThefirstthingthatIdidwasto 20 ‘‘Heestablishedcontactwithme
visitlike20or25physicians overthephoneandheex-
aroundthetown.Iwentto plainedtomethathewas
explaintothemtheprogram, planningtodevelopaplan
theservicethatwewere withdiabeticpatients.ThenI
planningtogive.’’ wenttothepharmacy,andhe
‘‘Iwoulddomorepersonal explainedtomehowwecan
contactwiththephysicians, workcollaboratively
notbyphone,notevenby together.’’
letter.’’