Prof Ronald I Henbest
MD, CCFP, M CI Sc (Fam Med) The Dept of Familv Medicine: Medunsa P O B o x 2 2 2
Medunsa, 0204
Curriculum vitae
Ronald J Henbest u.as born in Edmonton, Albena (Canada) u'herc he qualified in 1974 u'ith a BSc in Maths and Psrchology and in 1978 lith an MD from the Unir.ersin'of A l b c n a . H e t h c n c o m p l c t e d n * o , c a . i postgraduate studv ( residcncr') in Famill Medicinc lith the Depanment of Familv Medicine at the Unirersiw of Wesrern dtrrtario (Canada) and obtained his CCFP frorn the College of Family Physicians of Canada. Ron joined the Department of Familv Medicine at Medunsa in 1980. He has a panicular interest in the doctor-patient interaction and its imponance for healing. He returned to the Universin'of Western Onrario in 1984 to take their Master of Clinical Sciencc Degree in Familv Medicine (MCISc), rvhich emphasizes patient care, teaching and learning, and research. His thesis on Patier-rt-Centred Care involved the develonment ofa method for measuring patient-cintredness and testing it against patient outcomes. In 1989, Ron returned to his home ciw, Ednonton, fbr a period of 2l months \\,'here he rvas engaged as an associate professor in the Depanmcnt of Familv Medicine ar the Universin'of Albena.
During this time, he also completed funher training in systemic family therapy. In October 1990, Ron returned, lith his life Iudy 2n4 four vear old son Benji, this tirne as associate professor and depuw head ofthe Department of Familv Medicine at Medunsa.
Measuredf A Review of the Measuremenr of the Interpersonal Aspects of Care - Prof Ronald J Henbest
Sururnatry
Tbe iwpottance 0f the d.octor-paient relationsbip bas been increasingh, rccognized, apecially d.uring the past three decadts. With this rengniti.on has folhwed the nced for adzquate
m.ectsLtrerntnt instwuwnts -tpith wbich to stady it.
This pnper ra/iews B nuthods that hat e been dtsigrued t0 ,no&sare the dnctor- patient intera.ction in fawily prattice. In
addition, to prorifu appropriate barhgruund and percpective, twr nther groups of insawments are drsa,ibed;
n&rilsly, those da,ehped 14,the client- cenned therapists and tbose czrrl?unruly referccd to as s,steruatic interaction - analysis m*hod.s. Both wts of
insfiaments are oldzr and hat'e been wsed ettms'fuely. Hovever, thq, lyt t *,
doigwd specif.calll, for the context of tbe d.octor-pathnt interafiion &nd a.re sbzwrl to be inadrquate for its ,uarr.surerutnt.
Tbus, there is a wed to dmehp nmt metbod.s to drscribe and analyze the doc tnr-patient in teraction in' fawity pradice. Some of the newet" mcthods bave demans*ated validity and. reliability and one bas eyen sbown an associ.atinn between the dnctor-patfun t intera.ctiln and pntient outcomtl Tbe worh connnues.
S Afr Faw Pract 1991; 12: 17482
KEYWORDS:
Physicians, Family; Physician-
Patient Relationship;
Evaluation Studies: Research
174 SA Familv Practice Mer. l99l
Introduction
The realizatiolt that the doctor-patient relationship is of tremendous
inrportance is not a ne\\r one. In 1927 Sir Francis Peabodr'\\?s so bold as to declare that: "Thc significancc of the intimatc personal relationsl-rip betu,'cen phvsiciar"r and patient cannot be too stronghr ernphasized, fbr in an extraordinarilv large nurnbcr of cases both diagnosis and treatment are directlv depcndent on it, and the failurc of the voung phvsician to establisl'r this relationship accounts lbr much ot his ineffectivencss in thc care ofpatients".t
Fortv r,cars latcr the fbllori'ing
statement appeared in an editorial ir-r the Lancet: "Care of the doctor- patient rclation has for too long been left to chance; because ofits
importancc to gencral pracrice ir rlust no\\r be examined. defined. ar-rd taught, fbr onlv then can it bc practiced eflectir,'clv."' Indeed, during the past three decades much
examination, dcfinition and teaching of the doctor-parienr relationship has occurred;3.a.5'u and the need to
demonstrate a relationshio betn'een process and outcomes hai been emphasized.t.t But to do that, rr'e need to be able to measure the doctor-patient i ntcraction.
The prin"ran'purpose of this paper is to review the methods that have been developed specifically to measure rhe doctor-patient interaction in familv practice arrd the reader ma1'u'ish to turn directly to that secrion. Ho\\.
ever, in order to provide appropriate background and perspective, I shall first brieflv describe nvo older and widelv used groups of instruments;
namel,v, those developed by the client- centred therapist and those produced bv the svsternatic interaction analvsts.
SA Huisar-tspr:rktlk Mei i99l
Genuiness. Unconditional Positive Regard and Empathy:
The Measurement Instruments of the Client-Centred
Therapists
These instruments consist of a number of scale s for ratine the charactcristics or conditio.-ns offcrcd b,v thc thcrapist to his client and are based on the work of Carl
R o g c r S . e t o ' r t ' r z ' r e
The characteristics first described and now referred to as tl"le core
dimensions, werc: cmpathy, unconditional positive regard, and genuiness. Measurement scales for these dimensions wcre first described by Truax in 196I.'a Each scale consists of a number of defined stages along a continuum. For example, the genuiness scale includes at its lowest lcvel, such descriptions as "... there is exolicit evidence ofa verv considcrable discrcpancy betrvecn his (the therapist's) expcriencing and his current verbalizations;" whcreas, at the
Don't settle for what is measurable; go on trying to measure what you really want to know
highest level, "therc is an openness to experience and fecling by the
therapist of all types - both pleasant and hurtful - without traces of dcfcnsiveness or rctrcat into
profbssionalism... "ts As rescarch progrcssed several ncw dimensions including respect, concreteness, self- disclosure, confrontation, and immediacv were added.t6't8
The use of thesc scales, particularly those of the core dimensions has been extensivc. Most typically, the scales have been applied to brief samples of three minutes duration cxcerpted from audio-tape recordings of psychotherapy. Reliabilities for the thrcc core dimensions (expressed as Pearson Correlations) have ranged from 0,42 to 0,79 for the empathy scale, from 0,23 to 0,84 for the
In an extraordinarily large number of cases the diagnoses as well as the treatment are directly dependent on the doctor-patient relationship
unconditional positive regard scale, and from 0,20 to 0,62 for the genuiness scale as reported by Truax and Mitchell in a review of more than thirry-five studies.ra
At present, the measurement scales are being used in a modified form in human development progams. re'20 These scales have not to my
knowledge been applied to any great extent to the doctor-patient
interaction in family practice, altl"rough Lehman used the empathy, warmth, and genuiness scales as outcome measures in a studv called.
"Intake History and Physicai Examination and Its Association with the Doctor-Patient Relationship".2l
Bales, Roter, Stiles and Katz:
Systematic Interaction Analysis
Systematic interaction analysis, as the term implies, refers to the use of a system for analyzing the interaction.
Inui et al" have characterized a
t 7 5 S A F a m i l v P r a c t i c e M a y l 9 9 l SA Huisanspraktyk Mei l99l
system of analysis as consisting of the following four elements:
J.) an observational strategy, 2) a specific process ofinterest, 3) an exhaustive taxonomy for
categorizing encounter events, ancl
4) an operational approach to measuring these events.
Four representative systems of interaction analysis will be briefly mentioned here.
Probably the best known and most widely applied system of interaction analysis to the doctor-patient interaction is that of Bales' Interaction Process Analysis.'3 Researchers who have applied Bales' method to clinical interactions include: Korsch et al,2a Francis et al,'s Davis.26 and Stewart.27.28 In this system, the observational strategy is left to the discretion of the
researcher, who may code from transcripts, audio-tapes, or first-hand observation. The unit of analysis is the smallest speech segment that can be assigned a classification.
Empathy
IJnconditional positive regard Genuiness
Communication units are classified according to a set of 12 mutually exclusive categories. Six of the categories reflect the affective nature ofthe encounter (show solidarity, tension release, agrees, disagrees, shows tension, antagonism); the other six categories document the information exchange process of the
interaction ( gives. suggestion, opinion, orientation, asks for.
suggestron, oprnron, onentauon).
Reliability of this method has been found to be high with 91,60lo inter- observer agreement for unitizing and 82,4o/o agreement for categorizing."
Bales' system deals best with the relationship aspect of the encounter in that it is sensitive to the feelings of the doctor and patient toward each other. However, its classification of information transfer is inadequate for the doctor-patient interaction and the system as a whole does not take context rnto account.
Roter's method2e for interactional analysis is a modification of Bales' System that was originally devised to study patient question-asking behaviour. Eight categories are used to document information transfer:
affect is measured by a global rating scale. As with Bales' system, the immediate context of the interaction is not measured and sequence is not taken into account.
Stiles' Verbal Response Mode System3o.3' was designed specifically for the analysis ofdyadic exchanges.
His taxonomy of verbal responses is meant to define a mutually exclusive and exhaustive set ofeight basic modes. Each mode is intended to describe the particular function that the verbal act was meant to achieve.
The unit of analysis is the ufterance, which Stiles defines as "the
grammatical equivalent of one psychologic-unit of experience." This system best describes the ways in which information is transferred in the encounter; affect is omitted as is any analysis ofcontent or context of the interaction.
Katz's system of Resource Exchange Analysis33 defines resources as
important goods or services (such as information, greeting, diagnosis), and categorizes the modes of exchange as being either initiating or responding. The unit of analysis is the interact, which consists of the initiation of a response by one communicator and the response to it by the other. This system ehphasizes the content and the social context of the doctor-patient interaction and has the additional advantage ofa conceptually appealing unit of analysis. The interact potentially allows for analysis of interaction sequence. Problems with the resource exchange system include difficulty in the designation ofinteracts and the fact that some interaction sequences
Assessing the consultation has to do with the interaction between doctor and patient - not only each one's behaviour in isolation
do not fit easily into the resource exchange model of initiating and response modes. Two recent reviews of systematic interaction analysis"'3n have emphasized the need for the development of new ways of analyzing consultations that are specific for the context ofthe interaction. None of the
representative systems of analysis reviewed (Bales', RoteCs, Stiles' and Katz's) are ideally suited to the doctor-patient encounter. None of them for example, adequately identifies a patient's effort to raise an issue which may be critical for determining the patient's real reason for attendance. This would be a
176 SA Family Practice May l99l SA Huisanspraktyk Mei l99I
particularly serious deficiency for an instrument being used to assess the interpersonal process in family pracuce.
The Doctor-Patient
Interaction: Its Measurement in Family Practice
I shall now examine 8 specific aftempts to describe and analyze the doctor-patient interaction in family pracuce.
(l) Balint et al,3s noting the very complex difficulties inherent in att€mpting to rate the doctor- patient relationship, limited themselves to, "a very crude assessment based on the general practitioner's own judgment." 36 For this purpose they first used a five-point scale: hot war, uneasy truce, peace, friendly
negotiations, and mutual trust.
However, even this proved to be too complicated, and they reduced their scales to three points: negative, peace, and posltlve.
(2) As part of a larger study of doctor-patient communication, Gozzi, Morris and Korsch3"
devised a method of analysis to learn how two people in a medical setting help and hinder one another in expressing themselves. Three sets of categories were developed. The facilitation categories designated statements which one person made that were in accord with the preceding comments of the other person. The blocking categories included statements which were not in accord with the preceding ones or statements that fit one of ten specifically
( 3 )
dcfi ned doctor-block cateeories.
E,xamolcs of doctor-block categories lvere the doctor intcrrupting the paticnt or ignoring the patient's comments.
The third category was fbr interactions u4rich could not bc coded as facilitations or blocks.
The rating for each doctor- paticnt interaction was calculated from the numbers of doctor blocks and facilitations to yield a pcrcentagc of blocks per visit.
This mcthod of analvsis has been applied to a studv of82 doctor- patient interactions.3t Critcrion validity fbr the method was supported in that those interactions which had scorcd high in doctor blocks, showed proportlonatclv more negatlve aflbctivc statements by the doctor as determined by Bales' slrstem of analysis. No indication of
reliability testing was rcported.
Bvrne and Longa devcloped a mcthod for a detailcd analvsis of rhc doctor's behaviour in thc consultation. Their method ir-rvolves the use of checklists, comprised of specific behaviours, fbr e ach of two major categories of doctor bchaviour: doctor- ccntred behaviour and patient- centred behaviour. The doctor- centred category includcs such behaviours as asking closed qucstiorls, directing, and giving infbrmation. Paticr-rt-centred behaviours includc the use of broad questions, reflecting, and accepting patient ideas or feelings. A third category, labelled ncgative bchaviour, is used to classifv behaviour that rcjccted or dcnied thc patient in some way.
In addition, Byrne and Long
dcvcloped a scoring system that identifies a doctor's basic consultation style.a The manner in which this method of analyzing consultations was developed (involving the detailed examination of the rccordings of
over 2 500 doctor-patienr interviews) grants it considcrable validity. As far as I am aware, therc are no rcports ofothers u s i n g th i s m c t h o d in t h c i r research and no reports of reliability testing. One drawback to the scoring method is that it requires a transcript ofthe consultation bccause ofthc dctail reouired bv each of the two disiinct chlcHists that have to be completed.
A consultation represents just one interaction of the many which make up the doctor- patient relationship
(4) Bain,".3e through thc detailed study of 480 audio-tape recordings of his own consultations, devcloped a method of analysis using tcn categories (five cach for
physician and patient) to classily the verbal content ofthe doctor- Datient interaction. He latcr increased the number of cate- gorics to six cach.ao'ot The catcgories for classif ing the doctor's vcrbal behaviour are:
social exchange, facilitation, asking questions, medical problem resolutions (answers regarding medical problems or treatment), response to social problems, and instruction. The
177 SA Famil,v Practicc Mav 1991 SA Huisartspraktyk Mei I99l
( s )
categorles lor patlcnts arc:
prcsentation of symptoms, answenng qucstlons, resPonse to instruction, problcm-related expressions, questions, and social exchangc. This method does not allow sequencc to be analyzed, nor do thc categories selected by Bain secm to allow for an assessmcnt of the extent to which a consultation is oatient-ccntred othcr than in thc verv gcncral scnse of who does the most talking.
An intenicrv rating scale fbr general practice has been designcd and tested by Verby ct a1.42'43 Thc scale consists of 17 items. Eleven of tl-rcse items could be classified as statements of intervicw techniqr.rc, three of the items describe nonverbal aspects of communication, and the remaining three relatc to verbal content. Each item is scored on a fbur point scale. The itcms for this rating scale were carefully chosen and the scale would appear to be a valid way of asscssing paticnt centrcd care.
Floweve r, this method of rating a consnltation does not allow fbr the examination of any particular interaction seouence betwecn doctor and paiient, and it
requires either direct observation or the usc ofvideo rccordings to assess thc nonverbal items.
Pendleton et aloa has dcvised a consultation rating scale based on 7 tasks for the consultation.
These tasks are:
(1) to define the reasons for the paticnts' attendance, (2) to consider othe r problems, (3) to choose (r,r'ith the patient) ( 6 )
an appropriate action for each problern,
(4) to achieve a shared understanding ofthc problems with the patient, (5) to involve the patient in the
management of the problems,
(6) to use time and resources appropriatell', and (7\ to establish or maintain a
relationship with the patient which helps to achier,rc the other tasks.
Each part ofeach task is rated seDaratelv such that there are 14 items on'thc scale. each
consisting of nvo opposing statements linkcd bv a line. The rater olaces a mark on the line to show-how much he agrees with the opposing statements. For example, the first item consists of the statements: nature and history of problcms inadequately dcfined,/nature and history of problems adequately defi ned.
Thc l4 items provide for a comprehensive assessment of the consultation taking into account both the interpersonal and the technical skills ofthe physician.
This scale is specificallv intended to provide feedback to trainccs.
Thi theoretical base for thc instrument is well articulated. No studies have been rcported widr it to date.
Reccntly, the patient-centred clinical model developed at the Universiw of Western Ontario,as has bccn defined in operational terms and a method devised specifically for scoring patient- centredness.* This method u'ould qualifl as a system of
interaction analysis as defined by Inui et al." The observational strategv employed may be that of direct observation, audio-tape, or audio-video recording. Its process ofparticular intcrest is that ofthe doctor's verbal
response to what the patient sa,vs.
Its taxonomy for categorizing encounter events involves firstlv- classif,,i ng the patient's
communications as expectations, feelings, fears and prompts, and secondlv classifying the doctor's response as to whether he acknowledges (or not) and cuts off(or not) the patient's communication. In addition, the facilitating behaviours of the ohvsician are listed. All the terms ,t.a fo. categorizing events are operationally defined. The
scoring system involves assigning
of sequence to some extent, in that the patient's communication is recorded verbatim and then the doctor's resDonse to that
communicaiion is classified.
However. this method is still in its developing stages and needs further work. Firstly, more categories for classifying patient communications would be helpfirl (for example, there is no provision for classif ing a patient's thoughts about his illness). A second related problem is that the expectation category is defined very broadly (it includes e!'er,whing that is not a feeling or a fear). Thirdly, a broader range ofcategories of doctor responses representing varying degrees of patient- centredness might allow the instrumcnt greater sensitivity in differentiating between
physicians with regard to this factor. A fourth problem is that of different raters recordine different patient statements so that it is diflerent aspects ofthe same intervier.r'that are being scored. A fifth area that needs attention is the mechanics of the scoring system. As it stands, each category ofpatient behaviour is given equal weight so that different numbers of patient offers in differcnt categories results in unequal weighting of those offers. Because it is verv diffrcult to weigh individual
- patient cues according to their potential significance, it would seem better, at least for the present, to assign equal weight to the doctors response to each of these cues.
(8) A modification of the method piloted by the University of
Non-verbal behaviour ls an important part of th€ doctor- patlent communlcatron
a score from I to 4 to each ofthe five catcgories of physician behaviour; namell', his responses to cach ofthe four categories of Datient communications
( expectations, feelings, fears and prompts) and his facilitating behaviours which are listed separately. This method would appear to be the most specific method thus far developed for measuring the concept of patient- centredness in that it focuses directly on the patient's agenda, as defined in terms of patient expectations, feelings and fears.
It also provides for the analysis ( 7 )
l,7B SA Familv Practice Mav l99I SA Huisarrsprakwk Mci 1991
Western Ontario, Henbest et alo'a" has attempted to solve some of these problems and in an initial trial, hai demonstrated not only validity and reliability but also an association with patient outcomes.
This method assessed the doctor- patient interaction specifically in terms of its Datient-centredness with oatient-centredness defined as care in which the doctor responds to the patient in such a way as to facilitate the patient's expression ofall his or her reasons for coming to the doctor, including symptoms, expectations, thoughts, and feelinss. A score sheet is used to ,..o.J the offers (any potentiallY significant response) made bY the patient and to score the doctor's iesponse(s) to each offer on a 4 point scale as follows: (0) if the docto. ienored the ofFer altogeth-er, ( 1) if closed
responses were used, (2) if open ended responses were given, and (3)if the doctor specifically facilitated the expression ofthe patient's expectations, thoughts, or feelinqs. In addition to dcmonstiating validity and reliabiliw. this method has also shown sensitivity in that the oatient-centred scores varied iignificantly among the practitioners studied and also varied significantly for the rcsponscs to different categories of patient offers; symptoms received the most patient-centred responses followed by thoughts and expcctations, with feelings receiving the least patient-centred resDonses. The method was also found to be practical in that it was cheao. could be used in a variew of situations. and showed
potential for being very time effective as the score for the first two minutes of the consultation was highly correlated with the patient-centred score for the entire consultation (rs : 0,806 p : 0,001).47 PerhaPs most imoortant of all. this method has been able to demonstrate that the doctor's responses to a patient do makc a diffcrence: parient- cenfredness was folnd ro be associated with the doctor having ascertained the paticnts reasons for coming and with resolution of the patient's conccrns.a8
Discussion
The measurement of the doctor- patient relationship will be discussed in terms of three major properties:
validity, reliability, and practicality.
Vnlid.iry
Four main approaches to the study of validity are commonly distinguished:
face validity, content validity, concurrent or criterion validity. and construct validity.so
It would seem that for a method for analyzing the consultation in family practice to have face validity, that ihree things would have to be taken into account. First. the categories used for classifying and scoring behaviours should be appropriare to the doctor-patient intcraction.
Second, the method should allow for interaction sequence to be identified and evaluated. In order to assess the appropriateness of what the
oractitioner savs or does it is i-portn.rt to know what has immediately gone on before. For example, a statement that sounds facilitating when taken in isoiation, may be entirely inappropriate when
considered in context. Third, the context ofthe interaction itself; that is, as part ofan ongoing doctor- patient relationship, needs to be taken into account.
The measurement instruments used by the client-centred thcrapists are inadequate for assessing the family practice consultation primarily because they do not cover the full range of family doctor behaviours or responsibilities. The instruments developed by the interaction analysts also lack face and content validity in that the categories used are not specific for the doctor-patient interaction and on the whole, they do not allow for the assessment of the doctor's specific response to a patient's specific offer or complaint.
The newer methods, developed specificaily to assess the doctor- patlent lnteractlon, appear to nave more approprlate categorles ror classifiiing behaviours as well as broader ranges of items assessed. The categories used by Byrne and Long,o Bain,38 Vcrby et al,at Pendleton et al,aa and the University of Western Ontario (UWO),46 all seem well suircd to the flamily practice consultation and cover most of its content. However, only the methods described by Gozzi et al", and Henbesr et al ot allow fbr intcraction sequence to be taken into account.
The llWOou method has
demonstrated criterion validity using appropriate categories of Bales' System for comparison.an
The method developed by Henbest et al, in initial trials, has also
demonstrated criterion validity, both with the UWO method and with the Empathy Scale used by the client- centred therapists.aT In addition, it has demonstrated construct validity
f 79 SA Family Practice May 1991 SA Huisartspraktyk Mei 1991
in that patients who had experienced interviews that were assessed as being more patient-centred were
significantly more likely to report that their reasons for comins had been ascertained and that thEy had felt really understood by the doctor, than patients who had experienced interwiews assessed as less oatient- centred.
An important, but difficult reality to take into account, is the fact that most often in family practice, a consultation represents just one interaction of many that make up the doctor-patient relationship. None of the methods described in this paper orovide a measure of the overall ielationship.
One funher validiry issues requires mention, the measurement of nonverbal as well as verbal communications. Nonverbal behaviour is recognized to be an important part of communication,st'5' but much more work is necessary in order to develoo a method to assess it in the context ofthe doctor-patient interaction. Only the method described byVerby et alo'takes nonverbal behaviours into account and only to a small extent.
Reliability
Reliability or reproducibility is a measure of agreement between repeated measurements on the same subjectoe. Two aspects would seem important here:
(f) the recording ofthe consultation, and
(2) the scoring of the consultation.
Methods that require recording of the consultation in some way include those described by Byrne and Long,a
Bain3s, U\41O,o6 and Henbest et alo'.
Bain provides some evidence for the reliabl[ty of his method by showing that two independent raters had put the same number of units of expression in each offive
caiegories,tt but it is not clear that rhe same units of exoression were placed in the same categories.
Henbest et alaT found that two raters independentlv recorded over 807o of the patient offers identified by the other rater.48 The main difference between the raters was the extent to which they grouped the patients offers , especially symptoms,
together. Reliability of the scoring of
Longa requires transcripts. The other methods oresented can be scored either from direct observation or from audio-tapes. The time required for scoring either matches the length of the consultation if direct
observation is used or may be longer if audiotapes are used (which allows for replay for careful scrutiny of the interaction). Of note. the method developed by Henbest et al,a' in an initial trial, showed a high positive correlation berween the oatient- centred score determined from the first two minutes of the audiotape of the consultation and the score determined from the entire tape of the consultation. The use of a two- minute score would prove very useful especially in large scale studies.
Conclusions
Encouraging work has been done towards the develooment of a measurement instrument suitable for assessing the doctor-patient interaction in familv Dractice with initial studies demonitrating validity, reliability, and practicality for some methods.
It would seem especially important that the assessment of the
consultation pay attention to the interaction between the doctor and patient, rather than just observing the doctor's or Datient's behaviours in isolation.
At present, no methods are available for measuring the total doctor-patient relationship, rather than single consultations.
In\964, Pellegrino, in an article titled, Pntient Cnre - Mystical Rewarch or Researchable Mystiqwe?, stated that;
"Investigators seem to have settled for what is measurable instead of
Encouraging progress has been made towards developing practical, valid and reliable measurement lnstruments
consultations has been reported for the method described by Verby et ala' (Pearson Correlation Coef{icients 0.87 and 0.80 for inter- and intra- rater reliabilities respectively) UWOa6 (Pearson r : 0,687 ; 0,835; and 0,803 for inter-rater reliabilities), and for Henbest et al,o' (Spearman Rank Correlation Coefficients rg :0,91;
p : 0 , 0 0 1 ; a n d r g : 0 , 8 8 ; p : 0 , 0 0 2 for inter- and intra rater reliabilities respectively).
Practicnkty
The main practical considerations are the resources and time reauired for measurement. The measurement of nonverbal behaviour, for example Verby et al,a'requires either direct observation or video equipment. The comoletion of detailed checklists such as thbse compiled by Byrne and
l B 0 S A F a m i l y P r a c t i c e M a y I 9 9 l SA Huisartspraktyk Mci t99t
mcasllring u'hat thev lr,ould rcalll' like to kno$'."s'
Tlrc studv of thc doctor-patient rclationship is an important task, but not an casv onc. Lct us l"lot settlc for sirnplv measuring thc measurablc.
Rathcr, lct us continuc to strivc torvards developing thc means for measuring what we would rcally like to knou'; namel)/, what makcs for an efTectivc doctor-paticnt intcraction, one that is satisfying for both doctor and oatient and that leads to
impiovcd paticnt outcomcsf
Acknowle dge ments
I n'ould like to thank Prof Mike Breunan, Prof Wil\1lc Wcston, l)r Moira Stovart, Prof Martirr llass, and Prof Sarl Fchrsen firr manl' helpfr.rl discr.rssior.ts conce rning thc cloctor- paticllt intcractiou.
References
l. Pe:rbody FW.'fhe Carc of thc Patient.
I A M A 1 9 2 7 : 8 8 : 8 7 7 .
2. Anonvmous. Patient l)cnranci ( Eclitorial ).
I - a n c e t 1 9 6 7 ; l : 4 8 4 .
3. llalint M. Thc l)octor, His Paticnt arrd tlre Illness. 2ncl cd. London: Pitman Medical, 1964.
4. Bvrne PS, l,ong BFII-. Doctors Talking to Plticr.rts: A Studv of thc Verbal Behaviour of General Prirctitioners Oonsulting in their Surgcrics. London: HMSO, 1976.
5. Br<>snc K, Freeling P. Tlrc l)octor-Paticnt Rclationship. Edinburgh: (lhurchill Livingstone, 1976.
6. Stoeckle IL) cd. Encounters Betu'ccn Patients and Doctors. Cambridge: MIT Prcss, 1987.
7. Itcndk:ton D. Dtttor-Paticrrt
Communicatior.r: A Rcvierv. In Pendleton l). Hasle r I ecls. l)octor-Patient Commltt.ticirtiorrs. l,<>ndolr: Academic P r c s s , l 9 l i 3 : 5 - 5 3 .
8. Bass Ml, Buck C, Turncr L. Primary Carc Phvsician Actior.ts Influcncing Paticnt Outcomcs. UWO Lontlon, Ontrrio l9li4.
(Final Rcport ol ()M 697 to the C)ntxrio Ministry of Hcalth).
9. Rogers CR. Thc Proccss of Therapl'. I (lcrnsult Psvchol 1940; 4: 16l-4.
10. Rogers CR. Cor.urselling and Psvclrothcrlpv: Ncrve r Concepts in Practicc. lloston: Houghton Miffin Companr', 1942.
I L Rogcrs CR. (llicnt-Centrccl Tlrerapr': Its Current Practicc, Inrplications and'Iheory.
Cambridgc, Mass: fuverside Prcss, 1951.
12. Rogcrs CR. The Ncccssarv and Sufficicnt conditions of -l'herapeutic
Personalit.v C l r a n g e . J C o n s u l t P s y c h o l 1 9 5 7 ; 2 1 : 9 5 . r 0 3 .
13. Rogers CR. A Theorv ofThcrapy, Pcrsonalitv and Intcrpcrsonal
Reletionships, as l)cveloped in thc Client- Centrecl Frames,ork. In: Koch S cd Psvcholog.v: A Studl'of a Science, Vol 3.
Formulatiolrs of the Person and the Social (lontext. Neu' York: McGas.Hill B<xrk Company, 1959: I84-256.
14. Tmax CB, Mitchcll Klv[. Research on Cenain Therapist Interpersonal Skills in Rclatiou to Proccss and Outcome. In:
I3ergin AE, Crargield SL cds. Handbook of Ps-vchotherapy and Behaviour Change:
An Empirical Analvsis. Nerv York: )ohn Wilcv & Sons, l97l: 299-344.
15. Truax CIl. Mitchell I(M. Rcscarch on Certain Therapisr Interpersonal Skills in Relation to Process and Outcome. In:
lle rgirr AE, Crargicld SL eds. Handbook of Psvchothcrap.v and Bchaviclur Changc:
An Erlpirical Anall'sis. New York: fohn W i l e v & S o n s . l 9 7 I : . 3 1 6 .
16. Tmax Cll, Carkhuff RR. Torvard Efltctivc Couuselling and Psl,chothcrapy': Iraining and Practice. Ohicago: Aldine Publishing C o , 1 9 6 7 .
17. CarkhufiRR, llcrcnscnt B(i, Bcyond Counsclling and Therapv. Ncw York:
Holt, funehart and Winston, 1967.
18. Carkhuff RR, Hclping and Human Rclations: A Primer for La.v and Prof'essional Hclocrs. Vol I. Sclcction and Training. Neu' Ybrk: Holt, Rinehart and Winston, 1969.
19. Gazda GM. Waltcrs R'P. Childers WC.
Human Rclations Developmcnt: A Manual fbr thc Health Scicnccs. doston: Allvn arrd ISacon Inc, 1975.
20. Gazda, Asbury FR, Balzer FJ, Childers WC, Walters RP. Humrn Relatkrns Develoomcnt: A Menual fbr Educators.
3rd cd Boston: All,vn and Bacon Inc, 1984.
2L Lchman RM. Intakc Histo4'and Pht'sical Examination and its Assocation rr'ith the Doctor'Paticnt Relationship (Thcsis).
London, Ontario: Thc Universit"v of Wcstcrn Ontario 1984. l53pp
22. hlui TS, Caner WB, I(ukull WA, Haigh
\rH. Outcomc- Bascd Doct<>r-Paticnt Intcraction Analvsis. I Comparision of 'l'cchniqucs.
II Identit\'ing F,ffectivc Provider and Paticnt Bchaviour. Med Care 1982; 20: 535'549, 550-66.
23. Bales RF. Interaction Process Anall'sis: A Mctl-rod fbr the Stud.v of Small Groups.
Cambridge, Mass: Addision'Wcslc"v, 1950.
24. I(orsch BM, Gozzi EK, Francis V. Gaps in Doctor-Patient Communication: Doctor' Patient Intcraction and Satisfhction.
Pcdiatrics 1968; 42: 855-7 l.
25. Francis V, I(orsch BM, Morris M). Gaps in l)octor-Paticnt Comnrunication.
Paticnts' Response to Mcdical Advice. No E n g l I Med 1969;280: 535-40.
26. Davis MS. Variation in Paticnts' Comoliance rvith L)octors' Ordcrs:
Medical Practicc and Doctor'Patient Interaction. Int ) Psychiatrv Med l97l;2:
3 l - 5 4 .
27. Stewan MA. Paticnt (lharactcristics u'hich arc Related to thc l)octor-Patient Intcraction. I Farn Pract 1984; 1:30-6.
28. Stcwart MA. What is a Successful l)octor- Patient Intcrvicwl A Stud,v of Intcractions and Outcomes. Soc Sci Mcd 1984; 19:
t 6 7 - 7 5 .
29. Roter D. Patient Participation in thc Paticnt'Provider Intcraction. ( Doctoral I)isscrtation). Baltin.rorc: f ohn Hopki rrs Univcrsity School of Hygiene and Public Hcakh 1977.
30. Stilcs WB. Verbal Responscs-Modes and Dimcnsions of Intcrpcrsonal Roles: A Method of l)iscussi<Jn Analvsis. I Pers Soc P s . v c h o l 1 9 7 8 ; 3 6 : 6 9 3 .
3I. Stilcs WI3, Putnam SM, James SA, Wolf MH. Dimension of Patient and Physician Rolcs in Medical Scrccning Intervicrvs.
Soc Sci Mcd 1979; l3:335-41.
32. Stilcs WB, Putnam SM, Wolf MH, famcs SA. Ve rbal Responsc Modc Profiles of
I8l SA Familv Practicc May l99I SA Huisartsorakn'k Mci l99l
l. Patients lnc{ Physicians in Medical Screer.ring Inten'iervs. I Mcd Educ 1979;
5 4 : 8 1 9 .
33. IQtz E, Gr,rrevitch M Tsiyond P, l)irnet B.
Doctor Pltient Exchanges: A Diagnostic Approach to Organization and Professions. Human Relations 1967 ; 22:
309 24.
34. Wirsserman RC, Inui TS. S1'sten-ratic Auralysis of Clinical Patient i-uturc R c s e a r c h . M e d C a r e 1 9 8 3 ; 2 1 : 2 7 9 ' 9 3 . 35. Balint M, Hunt L ]oyce D, Marinker M,
Woodcock J. Treatment or l)iagnosis: A Studv ofRepeat Prescriptions in General P r l . t i c e . T u r o n t o : l B L i p p i r r e o t t , l9 - 0 . 36. Balint M, Hunt l, loyce D, Marinker M,
Woodcock J. Treatment or Diagnosis: A Studl' ofRepeat Prcscriptions in General Practice. Toronto: lB Lippir"rcott, 1970:
39.
37. Gozzi EI! Morris Ml, I(orsch BM. Gaps i r r D o c r o r P a t i c r r t C o n r n t u r t i c a l i o r r : Implications fbr Nursing Practicc. Arn J Nurs 1969; 69:529-33.
38. Bain DlG. Doctor Patient Comrnunication in General Practicc Consultation. Med E d u c 1 9 7 6 ; 1 0 : 1 2 5 1 3 1
39. Ilain DJ(i. Patient Itrou,lcdge and tl-re Content of thc Consultation in General Practice. Med Educ 1977; 11:347 50.
40. Bain DlG. The Content of Physician,/
Patient Communicatior.r i n l-amil.v Practice.
J l ' a m P r a c t 1 9 7 9 8 : 7 4 5 - 5 3 . 4L Bain DlG. The Relationship lJcnr'een
Time ancl Clinical Management in Fan'ril1' P r a c t i c e . J F a n ' r P r a c t 1 9 7 9 ; B : 5 5 1 - 9 . 42. Verbv JE, Holden P, I)avis RH. Peer
Rcview of ConsLrltations in Priman' Care:
The Usc of Audiovisual Recordings. Br M e c l J 1979; 1: 1686 8.
43. Verby JE, Davis RH, Holden I'. A Studv of thc Inten'iewing Shills of Trainces Assistants in Gcneral Practice. Patient Counselling and Health Education 1980;
2ncl Quarter: 68-71.
44. Pcndletor-r D, Scl-rofield T, Tate P, Hal'elock P. The Consultation: An Approech to Learning and 'feaching.
Oxfbrd: Oxford Universit,-v Press, 1984.
45. Lcvcnstein JH, McCrackcn EC, McWhinney IR, Stewart MA, Bror.vn fB.
The Paticnt-Cemtred Clinical Method: A
Model lbr thc L)octor-Patient Intcraction in Fan.rilv Medicinc. l-am Pract 1986: 4;
2 4 3 0 .
46. Brou'n J, Stervan MA, McCrackcn E, McWhinnc.v IR, Levenstein IH. TIic I'aticnt-Cer.rtred Clinical Mctl-rod:
Definition and Application. Fam Pract 1 9 8 6 ; 3 : 7 5 ' 9 .
47. Henbest RJ, Stcwart MA. Patient Centrcclness ir.r the Consultation. 1.: A Mcthod fbr Measuremcnt. Firrnily Practice 1989, 6: 219'54.
48. Hcnbest RJ, Ster'r'art MA. Patient Ccntreciness in the consultatior.r. 2: Does it Really Make a l)iflcrencel Firrnily Practicc 1 9 9 0 1 7 : 2 8 3 3 . .
49. Stcr'vart MA. A Comparison of Trvo Methods of Analyzing Doctor Patient Communication i n Fan-ril.v Practicc.
NAICRG Presentation, Scattle, Washington, 1985.
50. Gelbach SH. Interpreting the Medical Literaturc: A Cliniciar-r's Guide. Toror.rto:
Collamore Press, 1982.
51. Larscn I(M, Smith (lI(. Asscssment of Nonverbal Communication in the Paticnt- I'hysician Inten'ieu'. I Fam Pract l98l; 12:
4 8 1 .
52. Byrrre I'S, Heath CC. Practitioncr's Use of Non Verbal Bchar,ior,rr ir.r Real
Consultatior-rs. I R Coll Gcn Pract 1980;
3 0 : 3 2 7 3 I .
53. Pellegrino ED. Paticnt Care Mystical Research or Rescarcl.rable M1'stiquci CIin Research 79 61', 12: 122.
1 8 2 S A F . r n r i l l P r . t e t i r e M a y l 9 9 l S A H r r i : . r r t : p r . r l t n l t M c i l o O l
