lable at ScienceDirect

European Journal of Oncology Nursing 18 (2014) 619e625

Contents lists avai

European Journal of Oncology Nursing

journal homepage: www.elsevier .com/locate/ejon

Perceived quality of interprofessional interactions between physiciansand nurses in oncology outpatient clinics

Charlotte T. Lee a, *, 1, Diane M. Doran b, Ann E. Tourangeau b, 2, Neil E. Fleshner c

a Daphne Cockwell School of Nursing, Ryerson University, 350 Victoria Street, Toronto, Ontario, Canadab Lawrence S. Bloomberg Faculty of Nursing, University of Toronto, Toronto, Ontario, Canadac Division of Urology, Department of Surgery, University of Toronto, Toronto, Ontario, Canada

Keywords:Integrated care deliveryInterprofessional collaborative practiceNurse-physician relationshipsRelational coordination

* Corresponding author. Tel.: þ1 416 979 5000x799E-mail addresses: lee.charlotte@ryerson.ca, charlot

1 This study is a part of Charlotte T. Lee's doctoraCapital and Relational Coordination in Outpatient Clinwas supported by Co-Investigator Small Grant Award aResearch Unit, University of Toronto. Charlotte T. Lsupported by the following funding sources: GovernmToronto Foundation Graduate Scholarship in Sciencelowships from the Ontario Training Program in Healthas well, the Wilson Centre for Research in EducatiNetwork.

2 Ann Tourangeau is funded in part through a NursAward from the Ontario Ministry of Health and Long

http://dx.doi.org/10.1016/j.ejon.2014.06.0041462-3889/© 2014 Elsevier Ltd. All rights reserved.

a b s t r a c t

Purpose: To evaluate the perceived quality of interactions between nurses and physicians in oncologyoutpatient clinics.Methods: A cross-sectional, observational survey involving 250 physicians and nurses was conducted atoncology outpatient clinics at two regional cancer centres in the province of Ontario, Canada. Eligibleparticipants were identified by administrators and invited to complete a one-time survey questionnaire.Quality of interactions was assessed using a seven-item survey of relational coordination, which mea-sures two factors of interaction: supportive relationships and quality communication. Descriptive ana-lyses and multivariate analyses of variance (MANOVA) were conducted to assess potential differencesbetween the two study sites and the two professional groups.Results: Overall, nurses and physicians at both study sites rated their interactions highly (mean ¼ 4.32and 4.51 out of 5 for supportive relationships and quality communication, respectively). No difference ineither factor was reported between physicians and nurses at either study site, but the two study sitesdiffered significantly in both factors [F(2, 245) ¼ 7.54, p < 0.001].Conclusions: Overall, oncology nurses and oncologists at outpatient clinics rated their levels of inter-professional interaction highly. Contextual factors may have contributed to the high interaction scoresand different ratings between the two cancer centres. The finding that nurses and physicians reportedsimilar levels of perceived interactions suggests that relationships in these outpatient cancer clinics arehighly collaborative and collegial.

© 2014 Elsevier Ltd. All rights reserved.

Introduction

Teamwork is essential in healthcare, with multiple providersinvolved in every phase of a patient's illness trajectory. The per-formance of healthcare teams is known to strongly affect the

2; fax: þ1 416 979 5332.te.t.lee@gmail.com (C.T. Lee).l dissertation entitled ‘Socialics’. The dissertation projectt the Nursing Health Servicesee's doctoral education wasent of Ontario/University of& Technology; Training fel-Services and Policy Research,on at the University Health

ing Senior Career Researcher-Term Care.

quality of patient care in all medical subspecialties, includingoncology (Baggs et al., 1999; Manojlovich, 2010; Mukamel et al.,2006; Ponte et al., 2010; Friese and Manojlovich, 2012). As aresult, more researchers have been focussing on various aspects ofinterprofessional teamwork and collaboration.

This study was motivated by the awareness that high-qualityinteractions and relationships among care providers are crucialto improving interprofessional practice. Merriam-Webster (2012)defines interactions as “mutual actions or influences.” Theseinclude all types of contact among individuals (e.g., verbalcommunication, non-verbal communication, behavioural ex-changes) and are manifestations of interpersonal relationships.Research about interpersonal relationships builds on findings fromthe fields of psychology and sociology. Scholars have used varioustheories to explain different aspects of interpersonal relationships,and these have been applied to research about actions and in-fluences within healthcare teams. For example, social exchangetheory helps clarify leaderemember dynamics of a team (Brunetto

C.T. Lee et al. / European Journal of Oncology Nursing 18 (2014) 619e625620

et al., 2013), relational dialectics helps clarify role conflicts withinteams (Apker et al., 2007), and sensitivity to equity appears to bepositively correlated with an individual's agreeableness at work(Bing and Burroughs, 2001). With regard to teamwork, researchinvolving these psychological or sociological theories suggests thatan individual's mindset affects his/her actions, so positive inter-personal interactions (attitudes and behaviour) are vital to pro-mote collaboration. However, few studies have focused on theattributes of interpersonal interactions among healthcareprofessionals.

We conducted a literature review to assess current knowl-edge about interpersonal relationships, teamwork theories, andtheir applications in healthcare work environments. Within thefew relevant publications, two general frameworks were usedto integrate the concepts of interpersonal relationships andteamwork theories in healthcare: the nursing practice envi-ronment framework (Fox, 2000; Lake, 2002); and the inter-professional practice framework (D'Amour and Oandasan,2005; Interprofessional Education Collaborative Expert Panel,2011). Although the nursing work environment frameworkdoes not focus mainly on provider relationships, it includesnurseephysician relationships as one contextual factor. Pre-liminary evidence suggests that positive work environmentalfactors, as measured by the Practice Environment Scale (PES,which includes a measure of provider interactions), areassociated with a superior quality of care (Schmalenberg andKramer, 2009). The interprofessional practice framework(D'Amour and Oandasan, 2005; Interprofessional EducationCollaborative Expert Panel, 2011) helps clarify the factorsassociated with healthcare professionals' collaborative capacityat both pre- and post-licensure levels. This framework is basedon interactions among professionals and patients, and theseare expected to vary depending on the complexity ofcare. However, only a few studies have focused on providerrelationships or interactions beyond communication andof these, only one was conducted in an oncology setting usinga qualitative approach which limits generalizability (Hunt,1998).

Our review of the available literature revealed several addi-tional gaps and limitations. Most published studies that examineinterprofessional interaction have been conducted in inpatientsettings (Manojlovich, 2005; Nadolski et al., 2006; Reeves et al.,2009). Because collaborative attitudes and behaviours are influ-enced by professional culture and context (Hall, 2005), in-teractions among healthcare professionals likely vary betweeninpatient and outpatient settings because the structures andprocesses associated with these settings are different (e.g., staffingstructure, patient schedule). It is important to study interprofes-sional interaction in oncology outpatient clinics because of thecomplexity of care and toxicity of treatment. Additionally, currentresource constraints have meant that acutely ill cancer patients,who used to be admitted for disease management, are increas-ingly likely to be treated as outpatients. Increasing complexity ofcare, patient acuity, and patient volume make the functioning ofan outpatient oncology team crucial in providing sensitive careand preventing errors and redundancy, which are key indicatorsof quality healthcare.

Beyond the lack of studies conducted in outpatient settings,studies about interprofessional interactions are limited by meth-odological issues. For example, very few quantitative studies haveexamined interprofessional interactions beyond verbal commu-nication (Nadolski et al., 2006; Manojlovich, 2005). Quantitativetechniques can help advance knowledge in this area by usingreplicable measurement tools and generalizable findings. Addi-tionally, verbal communication is only one aspect of interpersonal

interactions. Among the studies that have examined actions andinfluences beyond verbal communication, most have used a globalmeasure of perceived collaboration (Lake, 2007; Schmalenbergand Kramer, 2009; Friese and Manojlovich, 2012). Althoughperceived collaboration is an integral element of quality in-teractions, global ratings cannot provide details about actions andinfluences among team members. Gittell et al.’s (2000) RelationalCoordination Survey (used in the current study) is the only vali-dated instrument that assesses measurements for interpersonalinteractions. The authors proposed and validated the idea thatquality communication and supportive relationships are key teamattributes that lead to superior performance. Quality communi-cation is characterised by the presence of timely, frequent, andproblem-solving communication, while a supportive relationshipis characterised by the presence of shared knowledge, sharedgoals, and mutual respect.

Finally, staff physicians are rarely included in surveys ofinterprofessional attitudes. This can hamper the validity ofresearch, because physicians are a key healthcare provider groupwho may have different perceptions about interactions withother professionals due to different roles and expectations. Validresearch about interprofessional perceptions should includemore than one professional group. In summary, existing knowl-edge of provider interactions at outpatient oncology clinics islimited by: a) a lack of empirical data in the outpatient envi-ronment; b) a lack of research utilising quantitative methodologywhich limits comparison; and c) the inclusion of only one pro-fessional group.

The present study addressed some of these gaps by quan-titatively evaluating the quality of interactions between nursesand physicians in two oncology outpatient clinics. Nurses andphysicians were selected for inclusion because they are thehealthcare professionals most often involved in providing careat outpatient clinics: other allied health professionals werenot included because they are not involved in every patientvisit.

The specific research questions addressed were: a) What arethe levels of perceived interaction between nurses and physi-cians, as reflected by perceived levels of supportive relationshipsand quality communication, in oncology outpatient clinics? b)Do nurses and physicians in oncology outpatient clinics differ intheir levels of perceived interaction, as reflected by theperceived levels of supportive relationships and qualitycommunication?

Methods

Study design, setting, and sample

This study was nested within a larger dissertation study thatwas conducted to validate a theoretical framework related tonurseephysician interactions in outpatient clinics (Lee, 2012).To address the current research questions, a cross-sectional,observational survey was conducted at the outpatient de-partments within two comprehensive, university-affiliatedcancer centres in the province of Ontario, Canada. Both cen-tres are located within the same metropolitan boundarieswithin the governance of the same regional cancer program.Centre A is an older institution that opened in the 1950s andcentre B opened approximately three decades later. At the timeof the study, both offered all modalities of cancer treatment forall kinds of cancer. Centre A was larger than centre B, but theyboth housed 10e20 radiation treatment machines. Centre Aadministered more than 25,000 chemotherapy treatmentsannually, and centre B administered more than 17,000

3 University of Toronto REB approval number: 24758 (centre A) and 24550(centre B). Institution protocol and approval number: 09-0702CE (centre A). Insti-tution protocol and approval number: 265e2009 (centre B).

4 This unique code allowed us to track individual participation or non-participation (for sending reminders). No participants in this study completedmore than one questionnaire.

C.T. Lee et al. / European Journal of Oncology Nursing 18 (2014) 619e625 621

treatments annually. Centre A treated more than 15,000 newpatients annually and centre B treated more than 10,000 newpatients annually. Centre A had a total of approximately400,000 patient visits per year, and centre B had approximately220,000. Both centres were amongst the 10 largest compre-hensive cancer centres in North America. Compared with centreB, centre A had more attending physicians (approximately 90versus 70) and more staff nurses (approximately 90 versus 45)employed in ambulatory services. Both centres employed asimilar model of care at the time of study.

Data were collected from a study sample that included staffphysicians (n ¼ 150) and staff nurses (n ¼ 100). More physicianparticipants than nurse participants were included because thisproportion is consistent with the physician to nurse ratio at thetwo study sites. The sample size was determined based onexisting standards regarding the power requirements for con-ducting a multivariate analysis of variance (MANOVA), the mainanalysis method in this study. A minimum of 44 subjects perprofession per site was needed to achieve 80% power to detect asignificant difference in the level of interaction, assuming amoderate effect size with an alpha of 0.05 (Guilford and Fruchter,1978).

Instrumentation

Interactions between nurses and physicians were measuredusing a previously validated, seven-item survey of relationalcoordination (Gittell et al., 2000). The goal of the emergingtheory of relational coordination is to extend the conventionaltheory of coordination (management of task dependencies), butthe survey of relational coordination has been used to assessrelational aspects of teamwork, such as interprofessional in-teractions (Meyer, 2010; Warshawsky et al., 2012). As noted,this instrument measures two factors of interaction: supportiverelationships and quality communication. As notedearlier, shared knowledge, shared goals, and mutual respectdistinguish supportive relationships, while timely, frequent,and problem-solving communication distinguish qualitycommunication.

Participants were asked to answer questions regarding thebehaviour and perceived beliefs of the other professional group ona five-point Likert scale, with higher scores indicating more pos-itive interactions. For example, nurse participants were asked,“How much does the physician in this clinic respect the role youplay in caring for patients?” The internal reliability for the rela-tional coordination scale (Cronbach's a ¼ 0.89) in the presentstudy compare favourably with previous reports (a ¼ 0.86 and0.87) (Hagigi, 2008; Gittell et al., 2008). Construct validity of thestudy instrument was assessed using confirmatory factor analysis(CFA). Results of CFA from study data support a two-factorstructure of interaction: communication factor (four items) andrelationship factor (3 items) [RMSEA ¼ 0.054 (90% CI: 0.03e0.08);CFI ¼ 0.98; SRMR ¼ 0.050]. Correlation and covariance betweenthese two factors were 0.71 and 0.14 respectively (p < 0.001),ruling out multicollinearity. Factor loadings ranged from 0.69 to0.89, and all were statistically significant. No modifications weremade to this two-factor model.

As is common in health services research, participants' de-mographics (age, gender, length of time at the institution, educa-tion attainment) were also collected.

Data collection

Participants were recruited from ambulatory departments intwo university-affiliated cancer centres between December 2009

and July 2010. Once approval was obtained from the relevantresearch ethics boards,3 questionnaires were sent to 122 nursesand 224 physicians, comprising 98% of the nurse and physicianstaff. The 2% (n ¼ 7) of nurses (n ¼ 4) and physicians (n ¼ 3)who were not approached were those who declined to providetheir names and/or email address for participating in anyresearch.

Study participant recruitment took place at the end of aregular staff meeting. Following a 5-min presentation about thestudy, each potential participant was provided with a studypackage that included a study description and informationsheet, a questionnaire, a return envelope with postage affixed,an unique code,4 and a URL to complete the survey electroni-cally as an alternative to filling out the paper questionnaire.Participation involved completing a one-time survey question-naire, which took approximately 10 min to fill out. The availableoptions for completing this survey ensured confidentiality andprivacy for both participants and non-participants. Follow-upreminders were sent to potential participants' work email andmailing addresses at approximately three, five, and sevenweeks after the initial distribution of surveys. Overall, 72.2%(n ¼ 250) of the potential participants completed thequestionnaire.

Data analysis

Following data verification and screening (e.g., checking formissing data, outliers, normal distributions), reliability of thestudy instruments was assessed using Cronbach's alpha.Descriptive statistics and baseline analyses consisted of adescription of the sample, central tendency, and an assessment ofbaseline differences. Analysis of variance (ANOVA) and c2 testswere used to assess baseline differences between sites and be-tween professions. Construct validity of the study instrumentwas assessed using confirmatory factor analysis (CFA), as re-ported in the Instrumentation section. MANOVA was also used toassess differences in perceived interactions between nurses andphysicians, and between centre A and centre B (independentvariables). The dependent variables analysed in MANOVA weresupportive relationships and quality communication, reflectingprovider interactions. Univariate findings were reviewed tofollow up the significant multivariate F test from MANOVA. Sta-tistical significance was set at 5%. Statistical analyses werecomputed using IBM SPSS® Statistics 20 for Windows (IBM,2011). CFA was performed using IBM AMOS 20.0.0 for Windows(Arbuckle, 2011).

Results

Sample characteristics

The dataset used for descriptive analysis included 100 nursesand 150 physicians (N ¼ 250), all experienced health professionalsworking in oncology outpatient care settings. Their mean age was48.2 years and the mean time they had been employed at theircurrent institution was 14.5 years (see Table 1).

Table

1Pa

rticipan

tdem

ographics(N

¼25

0).

Allcases

mea

ns(SD)

Allnurses

mea

ns

(SD)(n

¼10

0)

Allphysicians

mea

ns

(SD)(n

¼15

0)

Cen

treA(n

¼15

2)Cen

treB(n

¼98

)

Nurses

and

physicians

combined

(SD)

Nurse

participan

ts(SD)(n

¼63

)

Physician

participan

ts(SD)(n

¼89

)

Nurses

and

physicians

combined

(SD)

Nurse

participan

ts(SD)(n

¼37

)

Physician

participan

ts(SD)(n

¼61

)

Mea

nag

ein

years(SD)

48.18(9.89)

a

Ran

ge26

e70

49.90(10.60

)Ran

ge26

e67

47.08(9.73)

Ran

ge32

e73

48.34(10.32

)49

.64(11.25

)47

.37(9.54)

47.93(9.22)

50.35(9.52)

46.38(8.77)

%male

42.20%

3.00

%65

.97%

43.7%

3.20

%72

.7%

39.8%

2.7%

62.3%

Timeat

institution

(yea

rs)

14.52(9.97)

(ran

ge:0.10

e40

.6)

17.02(10.98

)(ran

ge:0.08

e40

.60)

16.06(10.01

)(ran

ge:0.75

e40

.58)

15.79(10.98

)18

.92(12.07

)13

.52(9.57)

12.46(7.66)

13.50(7.58)

11.84(7.70)

%em

ploye

dfull-time

82.00%

69.00%

90.70%

82.8%

68.3%

93.2%

84.7%

78.4%

88.5%

HighestEd

ucation

Receive

dRN

(n¼

95)

MD

(n¼

149)

eRN

(n¼

61)

MD

(n¼

88)

eRN

(n¼

34)

MD

(n¼

61)

College

diploma

e40

(42.10

%)

0e

30(49.2%

)0

e10

(29.4%

)0

Bachelors(includingMD)

e34

(35.80

%)

82(55.00

%)

e19

(31.1%

)50

(56.8%

)e

15(44.1%

)32

(52.5%

)College

diplomaan

dsp

ecialtynursingcertificate

e9(9.50%

)n/a

e4(6.6%)

n/a

e5(14.7%

)n/a

Bachelorsan

dsp

ecialty

nursingcertificate

e1(1.10%

)n/a

e0

n/a

e1(2.9%)

n/a

Masters/N

urse

Practition

er/equ

ivalen

te

11(11.60

%)

41(27.50

%)

e8(13.1%

)21

(23.9%

)e

3(8.8%)

20(20.4%

)

Doc

torate

e0

26(17.40

%)

e0

17(19.3)

e0

9(14.8%

)

a93

nursean

d13

3physicianresp

onden

tsresp

onded

tothis

item

(n¼

226).

C.T. Lee et al. / European Journal of Oncology Nursing 18 (2014) 619e625622

Based on information obtained from the College of Nurses ofOntario (2009) and Canadian Medical Association (2010), thestudy sample sub-groups resemble their respective general pro-fessional memberships, with some exceptions: the mean age ofboth nurses and physicians was slightly older, and physiciansgenerally had more advanced degrees than physicians in the gen-eral workforce. These exceptions are unsurprising, given thatnurses working in outpatient clinics often have more seniority andtherefore are generally older than nurses in inpatient settings, andstaff physicians at teaching institutions tend to be older due to theadditional clinical experience required to assume teaching roles inuniversity-affiliated institutions.

In terms of sample characteristics between the two sites(Table 1), ANOVA and c2 results suggest that the differences be-tween centre A and centre B were not significant for nurse orphysician participants. This includes the proportion of nurses whowere employed full time (68.3% versus 84.7%; c2 ¼ 1.19, df ¼ 1,p ¼ 0.28, 2-tailed) and the proportion of nurses who had a collegediploma (49.2% versus 29.4%) and a Master's degree (13.1% versus8.8%) as their highest level of professional education (c2 ¼ 1.95,df ¼ 2, p ¼ 0.38, 2-tailed).

Research question one: levels of perceived interactions betweennurses and physicians

Descriptive statistics revealed that all study participants re-ported high levels (mean ¼ 4.43, on a five-point scale) of perceivedrelational coordination (Table 2). Higher scores reflect a higherfrequency of positive interactions, so this finding implies that par-ticipants perceived their interactions with clinic colleagues aspositive.

Research question two: different perceptions between nurse andphysician participants

Results of a two-factor (profession, site) MANOVA (Table 3)revealed a significant main effect of site [F(2, 245)¼ 7.54, p < 0.001]on physicianenurse interactions. As shown in Table 2, participantsworking in centre B had significantly higher scores than centre A inboth components (factors) of interaction [communication: F(1,246) ¼ 10.88, p ¼ 0.01; relationship: F(1, 246) ¼ 13.06, p < 0.001].No differences were found in either factor between physicians andnurses.

Discussion

This study is one of the first to explore interprofessional in-teractions in an outpatient oncology setting using a validated two-factor instrument (supportive relationships and quality communi-cation) to measure the responses from both physicians and nurses,thereby addressing the research gaps caused by including partici-pants from only one profession, or by a more narrow focus oncommunication. The study yielded three key findings. First, theoverall perceived quality of interaction (as indicated by the level ofrating) was high among nurses and physicians in both oncologyoutpatient clinics. Second, nurses and physicians at both study sitesreported similar levels of perceived interaction as indicated by thenon-significant effect of profession in MANOVA. Third, a significantdifference appeared in perceived nurseephysician interactionsbetween the two study sites as indicated by the significant effect ofsite in MANOVA.

The ratings of study participants reflect a high quality of inter-action between nurses and physicians in both oncology outpatientclinics (overall mean ¼ 4.43 out of 5, between “often” and“constantly”, standard deviation ¼ 0.54). This finding is consistent

Table 2Means of relational coordination scores in two cancer Centers.a

All participants(N ¼ 250)

All nurses(SD) (n ¼ 100)

All physicians(SD) (n ¼ 150)

Center A (n ¼ 152) Center B (n ¼ 98)

Nurses andphysicianscombined (SD)

Nurseparticipants(SD) (n ¼ 63)

Physicianparticipants(SD) (n ¼ 89)

Nurses andphysicianscombined (SD)

Nurseparticipants(SD) (n ¼ 37)

Physicianparticipants(SD) (n ¼ 61)

Relationalcoordination(Composite score)

4.43 (0.54) 4.42 (0.49) 4.44 (0.57) 4.32 (0.59) 4.31 (0.51) 4.33 (0.64) 4.59 (0.41) 4.60 (0.39) 4.59 (0.43)

Relationalcoordination(Communicationcomponent)

4.51 (0.60) 4.52 (0.49) 4.51 (0.66) 4.41 (0.67)** 4.42 (0.50) 4.41 (0.77) 4.67 (0.42)** 4.68 (0.42) 4.66 (0.43)

Relationalcoordination(Relationshipcomponent)

4.32 (0.61) 4.29 (0.63) 4.34 (0.59) 4.21 (0.64)*** 4.17 (0.69) 4.23 (0.61) 4.49 (0.50)*** 4.49 (0.45) 4.49 (0.64)

SD ¼ Standard deviation.**p ¼ 0.01, 2-tailed ***p < 0.001, 2-tailed.

a Relational coordination and its two components are ranked on a 1 to 5 Likert-scale, with 5 representing most positive interaction.

C.T. Lee et al. / European Journal of Oncology Nursing 18 (2014) 619e625 623

with those of another study conducted in an outpatient setting(mean¼ 2.60 out of 3, between “some of the time” and “most of thetime”, standard deviation¼ 0.09) (Hagigi, 2008). To date, no studiesappear to have compared interprofessional interactions betweeninpatient and outpatient settings, but preliminary data fromoutpatient settings suggest that the perceived quality of interactionamong these healthcare providers is better than among healthcareproviders in inpatient settings, with means ranging from 3.74 to4.02 out of 5 (between “occasionally” and “often”) (Gittell et al.,2000; Gittell et al., 2008; Havens et al., 2010). Outpatient clinicshave unique attributes, which may explain the differences betweeninpatient and outpatient settings; these will be discussed below.The present study did not examine nurseephysician interactions ininpatient settings, so more research is needed to compare differ-ences between these two care settings.

Comparedwith the preliminary data from inpatient settings, therelatively high ratings of nurseephysician interactions in the pre-sent study suggest that nurses and physicians in both outpatientclinics were more engaged in teamwork and had highly collabo-rative relationships. Schmalenberg and Kramer (2009) developed atypology for studying nurseephysician relationships and foundthat collaborative and collegial nurseephysician relationships aremore prevalent in institutions that are rated as having nursingexcellence. If so, this maymean that the oncology outpatient clinicsincluded in the present study provided very good patient care.

The positive but non-significant finding between ratings ofnurses and physicians is supported by another study conducted inan outpatient setting, which reported similar ratings of nurse-ephysician interactions (Hagigi, 2008). However, given the lack ofresearch about outpatient care teams, it may be premature to ruleout potential discrepancies between the two professions. Researchabout nurseephysician interactions in inpatient settings has yiel-ded conflicting findings. A few studies have reported differences inperceived communication between the two professions (Larson,1999; Fox, 2000; Edmondson, 2003; Weinberg et al., 2009) inperi-operative and inpatient care settings. Manojlovich (2010)suggested this kind of discrepancy between physicians andnurses might be related to different assumptions about commu-nication, based on professional roles and responsibilities.Conversely, Schmalenberg and Kramer (2009) reported similarratings in perceived interdisciplinary interactions between physi-cians and representatives from other departments. They arguedthat recent clinical integration activities (e.g., regularly scheduledinterdisciplinary rounds) may have encouraged improved

interdisciplinary collaboration and relationships, resulting in moreconsistent ratings among different professionals. Further researchwill be needed to clarify these findings.

Finally, the present study revealed a between-site difference inperceived interactions, which supported the findings of a previousstudy (Gittell et al., 2000). This finding could be related tocontextual factors specific to each site. Contextual variations areknown to affect teamwork: Alt-White et al. (1983) reported that ininpatient settings, the communication process, the approach tocoordination, and a positive work environment are positivelycorrelated with nurseephysician collaboration. Additionally,Manojlovich (2005) suggested that a positive work environmentimproved nurses' perceptions about their communication withphysicians. The present study did not assess these variables, whichmay have differed between the two clinics. Further research mayhelp clarify the effects of various work environment characteristics(e.g., perceived autonomy, support and adequacy in staffing), but itis also important to maintain good response rates and preventsurvey fatigue: the Practice Environment Scale that assesses workenvironments consists of 49 items (Lake, 2002).

Similarities and differences in the organisational characteristicsof the two clinics were also compared with previous findings. Interms of similarities, both clinics offered the same modalities ofcancer treatment, housed a similar number of radiation treatmentmachines, and employed a similar model of care. In terms of dif-ferences, one site was older than the other by about three decades,had a higher patient volume, and employed more healthcare pro-viders. Few previous studies have explored how organisationalfactors influence interprofessional collaboration and interactions.SanMartín-Rodríguez et al. (2005) reported that power differences,adherence to collaboration logistics, and adherence to professionallogistics were some determinants of interprofessional collabora-tion. The present study did not assess these variables. Xyrichis andLowton (2007) reported that organisational support is one factorthat facilitates teamwork in primary and community care settings;again, the present study did not assess this variable. In sum, thebetween-site difference observed in the present study is not a novelfinding, but more research is needed to confirmwhich variables arerelated to the difference.

The highly positive interaction between nurses and physiciansin both clinics is an interesting finding. This high degree of inter-professional collaboration may reflect the increased demand forcollaborative competence in cancer care. Collaborative competence(e.g., the ability to communicate effectively; being knowledgeable

Table 3Multivariate and univariate effects of MANOVA.

Multivariate effects

Variable Pillai's trace F df Error df Significance

Profession 0.002 0.25 2 245 0.783Site 0.058 7.54 2 245 0.001a

Profession by Site <0.00 0.06 2 245 0.942

Univariate effects

Independent variable Dependent variable F df Error df Significance

Profession Communication component 0.053 1 246 0.817Relationship component 0.158 1 246 0.691

Site Communication component 10.878 1 246 0.001a

Relationship Component 13.061 1 246 <0.000a

Profession by site Communication component 0.012 1 246 0.912Relationship component 0.109 1 246 0.741

a Statistical significance a ¼ 0.05, 2-tailed.

C.T. Lee et al. / European Journal of Oncology Nursing 18 (2014) 619e625624

in one's role within a patient care team) is known to be important,especially as more professionals are becoming involved in oncology(Interprofessional Education Collaborative Expert Panel, 2011). Forexample, mental health professionals and social workers now helpmanage psychosocial and survivorship issues (Andersen et al.,2007; Rohan and Bausch, 2009; Giese-Davis et al., 2012). Therecent development of clinical pathways to help patients experi-encing distress, and the recent introduction of patient navigationmodels, have also expanded the network (and increased thecomplexity) of patient care teams. Within this context, healthcareproviders need to connect with each other: high-quality workplacerelationships, as signified by positive interpersonal interactions,enhance collaborative competence and facilitate collaboration andcoordination of care (Ponte et al., 2010).

Oncology nurses working in outpatient clinics face uniquechallenges with regard to collaboration, because the structuresand processes used in ambulatory care differ from those used ininpatient care. Ireland et al. (2004) surveyed more than 300ambulatory oncology nurses and found they used diverse kindsof care delivery models (functional, medical, and primary care).This kind of variation in patient care models does not occur ininpatient settings (Seago, 2001). The heterogeneous nature ofoutpatient oncology care requires unique management strategiesand/or practices. For example, delegation of outpatient care tasksis a key concern for ambulatory nurses (Schim et al., 2001;Ireland et al., 2004). The heterogeneity in outpatient care maybe related to a lack of a consistent framework for nursing prac-tice, and/or role confusion in this setting. No direct evidence isavailable to determine how ambulatory oncology nursing differsfrom other ambulatory settings or medical subspecialties, butcollaborative competence in outpatient oncology clinics appearsto differ from that in inpatient settings. More research is requiredto assess these differences and their causes. Additionally, trainingto improve interprofessional practices and collaboration shouldbe developed and implemented at both pre- and post-licensurelevels.

This study had several limitations. It used convenience sam-pling, which can involve sampling bias. Because the study samplewas not randomly selected, it might not represent the generalpopulation of all outpatient physicians and nurses, affecting thegeneralizability of the study findings. Generalizability of thefindings is also limited to professionals working in outpatientsettings, specifically in North American academic comprehensivecancer centres, which have similar work cultures and organisa-tional and professional hierarchies. The different sample sizesbetween the two clinics may have affected normality and homo-geneity in MANOVA, despite data screening to minimise these

risks. The self-reported survey design could also have increasedthe potential for common method variance or measurement bias.Analyses did follow the recommendations for minimising com-mon method variance (Spector, 2006), and every effort was madeto reduce measurement errors, such as using a previously vali-dated instrument with repeated testing to confirm validity andinternal consistency. Nevertheless, measurement bias cannot beruled out.

Conclusions

This is one of the first studies to explore interprofessionalrelationships in ambulatory oncology clinics by analysing howboth nurses and physicians rate their perceptions of interactionswith members of the other profession. The overall ratings fromall participants were high, indicating that both nurses and phy-sicians consider their interactions with members of the otherprofession to be of high quality. Nurses and physicians were quitesimilar in their perceptions of the nature of their interactions.However, perceptions differed significantly between the twostudy sites, raising questions about the effects of the workenvironment and organisational factors that may contribute tosuch differences. Future studies should assess how contextualfactors may influence collaborative competence, provider in-teractions, and ultimately the performance of patient care teams.The results of the present study help clarify nurseephysicianrelationships in outpatient oncology clinics, specifically the leveland patterns of perceived interactions between nurses andphysicians. This preliminary evidence will inform future researchon this topic.

Conflict of interest statement

Authors of this manuscript declare no conflict of interests.

Acknowledgement

The authors thank Dr. Glenn Regehr, Dr. Margaret Fitch and Dr.Richard Redman for providing feedback and guidelines on thisstudy. We also thank Dr. Jared Lessard andMs. Joyce Chung for theirtechnical assistance in preparing this manuscript.

References

Alt-White, A.C., Charns, M., Strayer, R., 1983. Personal, organizational and mana-gerial factors related to nurse-physician collaboration. Nursing AdministrationQuarterly 8, 8e18.

C.T. Lee et al. / European Journal of Oncology Nursing 18 (2014) 619e625 625

Andersen, B.L., Farrar, W.B., Golden-Kreutz, D., Emery, C.F., Glaser, R., Crespin, T.,et al., 2007. Distress reduction from a psychological intervention contributes toimproved health for cancer patients. Brain, Behavior and Immunity 21,953e961.

Apker, J., Propp, K.M., Zabava Ford, W.S., 2007. Negotiating status and identitytensions in healthcare team interactions: an exploration of nurse role dialectics.Journal of Applied Communication Research 33, 93e115. http://dx.doi.org/10.1080/00909880500044620.

Arbuckle, J.L., 2011. IBM SPSS Amos 20.0.0 User's Guide. SPSS, Chicago.Baggs, J.G., Schmitt, M.H., Mushlin, A.I., Mitchell, P.H., Eldredge, D.H., Oakes, D., et al.,

1999. Association between nurse-physician collaboration and patient outcomesin three intensive care units. Critical Care Medicine 27, 1991e1998.

Bing, M.N., Burroughs, S.M., 2001. The predictive and interactive effects of equitysensitivity in teamwork-oriented organizations. Journal of OrganizationalBehavior 22, 271e290 doi:10/1002/job.68.

Brunetto, Y., Shriberg, A., Farr-Wharton, R., Shacklock, K., Newman, S., Dienger, J.,2013. The importance of supervisor-nurse relationships, teamwork, wellbeing,affective commitment and retention of North American nurses. Journal ofNursing Management 21, 827e837. http://dx.doi.org/10.1111/jonm.12111/full.

Canadian Medical Association, 2010. CMA Masterfile. Available at: http://www.cma.ca/multimedia/CMA/Content_Images/Inside_cma/Statistics/historical/02SpecAge2010.pdf.

College of Nurses of Ontario, 2009. Membership Statistics. College of Nurses ofOntario, Toronto, ON.

IBM Corp Released, 2011. IBM SPSS Statistics for Windows. Version 20.0. IBM Corp,Armonk, NY.

D'Amour, D., Oandasan, I., 2005. Interprofessionality as the field of interprofessionalpractice and interprofessional education: an emerging concept. Journal ofInterprofessional Care 19, 8e20.

Edmondson, A.C., 2003. Speaking up in the operating room: how team leaderspromote learning in interdisciplinary action teams. Journal of ManagementStudies 40, 1419e1452.

Fox, E., 2000. An audit of inter-professional communication within a trauma andorthopaedic directorate. Journal of Orthopaedic Nursing 4, 160e169.

Friese, C.R., Manojlovich, M., 2012. Nurse-physician relationships in ambulatoryoncology settings. Journal of Nursing Scholarship 44, 258e265.

Giese-Davis, J., Waller, A., Carlson, L.E., Groff, S.L., Zhong, L., Neri, E., et al., 2012.Screening for distress, the 6th vital sign: common problems in cancer out-patients over one year in usual care: associations with marital status, sex, andage. BMC Cancer 12, 441.

Gittell, J.H., Fairfield, K.M., Bierbaum, B., Head, W., Jackson, R., Kelly, M., et al., 2000.Impact of relational coordination on quality of care, post-operative pain andfunctioning, and length of stay: a nine hospital study of surgical patients.Medical Care 38, 807e819.

Gittell, J.H., Weinberg, D.B., Bennett, A., Miller, J.A., 2008. Is the doctor in? A rela-tional approach to job design and the coordination of work. Human ResourceManagement 47, 729e755.

Guilford, J.P., Fruchter, B., 1978. Fundamental Statistics in Psychology and Education.McGraw-Hill, New York.

Hagigi, F., 2008. Evaluating Coordination as a Key Driver of Performance inAmbulatory Care Clinics. The Heller School for Social Policy and ManagementBrandeis University, Waltham, Massachusetts. OCLC. WorldCat No. 25971736.

Hall, P., 2005. Interprofessional teamwork: professional cultures as barriers. Journalof Interprofessional Care 1 (19 Suppl), 188e196.

Havens, D.S., Vasey, J., Gittell, J.H., Lin, W.T., 2010. Relational coordination amongnurses and other providers: impact on the quality of patient care. Journal ofNursing Management 18, 926e937.

Hunt, J., 1998. Empowering health care professionals: a relationship between pri-mary health care teams and paediatric oncology outreach nurse specialists.European Journal of Oncology Nursing 2, 27e33.

Interprofessional Education Collaborative Expert Panel 2011. reportCore Compe-tencies for Interprofessional Collaborative Practice: Report of an Expert Panel.(Washington, D.C).

Ireland, A.M., DePalma, J.A., Arneson, L., Stark, L., Williamson, J., 2004. The oncologynursing society ambulatory office nurse survey. Oncology Nursing Forum 31,E147eE156.

Lake, E.T., 2002. Development of the practice environment scale of the nursingwork index. Research in Nursing and Health 25, 176e188.

Lake, E.T., 2007. The nursing practice environment: measurement and evidence.Medical Care Research and Review 64, 104Se122S.

Larson, E., 1999. The impact of physician-nurse interaction on patient care. HolisticNursing Practice 13, 38e46.

Lee, C.T., 2012. Social capital and relational coordination in outpatient clinics: aninterprofessional analysis. Journal of Interprofessional Care 27, 81e87.

Manojlovich, M., 2005. Linking the practice environment to nurses’ job satisfactionthrough nurse/physician communication. The Journal of Nursing Scholarship37, 367e373.

Manojlovich, M., 2010. Nurse/physician communication through a sensemakinglens: shifting the paradigm to improve patient safety. Medical Care 48,941e946.

Meyer, R., 2010. Relationships Among Span, Time Allocation, and Leadership ofFirst-line Managers and Nurse and Team Outcomes (PhD dissertation). Grad-uate Department of Nursing, University of Toronto. Available at: http://hdl.handle.net/1807/24829.

Mukamel, D.B., Temkin-Greener, H., Delavan, R., Peterson, D.R., Gross, D., Kunitz, S.,et al., 2006. Team performance and risk-adjusted health outcomes in the pro-gram of all-inclusive care for the elderly (PACE). Gerontologist 46, 227e237.

Nadolski, G.J., Bell, M.A., Brewer, B.B., Frankel, R.M., Cushing, H.E., Brokaw, J.J., 2006.Evaluating the quality of interaction between medical students and nurses in alarge teaching hospital. BMC Medical Education 6, 23.

Ponte, P.R., Gross, A.H., Milliman-Richard, Y.J., Lacey, K., 2010. Interdisciplinaryteamwork and collaboration: an essential element of a positive practice envi-ronment. Annual Review of Nursing Research 28, 159e189.

Reeves, S., Rice, K., Conn, L.G., Miller, K.L., Kenaszchuk, C., Zwarenstein, M., 2009.Interprofessional interaction, negotiation and non-negotiation on general in-ternal medicine wards. Journal of Interprofessional Care 23, 633e645.

Rohan, E., Bausch, J., 2009. Climbing Everest: oncology work as an expedition incaring. Journal of Psychosocial Oncology 27, 84e118.

San Martín-Rodríguez, L., Beaulieu, M.D., D'Amour, D., Ferrada-Videla, M., 2005. Thedeterminants of successful collaboration: a review of theoretical and empiricalstudies. Journal of Interprofessional Care 19 (s1), 132e147. http://dx.doi.org/10.1177/1757177410389627.

Schim, S.M., Thornburg, P., Kravutske, M.E., 2001. Time, task, and talents in ambu-latory care nursing. Journal of Nursing Administration 31, 311e315.

Schmalenberg, C., Kramer, M., 2009. Nurse-physician relationships in hospitals:20,000 nurses tell their story. Critical Care Nurse 29, 74e83.

Seago, J.A., 2001. Nurse staffing, models of care delivery, and interventions. In:Making Health Care Safer: a Critical Analysis of Patient Safety Practices. Agencyfor Healthcare Research and Quality. University of California at San Francisco(UCSF)-Stanford University Evidence-based Practice Center, pp. 423e446.

Spector, P.E., 2006. Method variance in organizational research: truth or urbanlegend? Organizational Research Methods 9, 221e232.

Warshawsky, N., Havens, D., Knafl, G., 2012. The influence of interpersonal re-lationships on nurse managers' work engagement and proactive work behavior.Journal of Nursing Administration 42, 418e425.

Weinberg, D.B., Miner, D.C., Rivlin, L., 2009. 'It depends': medical residents' per-spectives on working with nurses. American Journal of Nursing 109, 34e43.

Xyrichis, A., Lowton, K., 2007. What fosters or prevents interprofessional team-working in primary and community care? A literature review. InternationalJournal of Nursing Studies 45 (1), 140e153.