Family Practice Vol. 20, No. 3, 283-288
© Oxford University Press 2003
Communication |
GPs’ satisfaction with the doctor–patient encounter: findings from a community-based survey
a Centre for the Evaluation of the Effectiveness of Health Care (CeVEAS)
b GPs on behalf of the participants
c Modena Territorial Health Authority
d University of Modena and Reggio-Emilia, Italy.
Correspondence to Maria Monica Daghio, CeVEAS—Centro per la Valutazione dell’Efficacia dell’ Assistenza Sanitaria, Viale Muratori 201, 41100 Modena, Italy; E-mail: m.daghio{at}ausl.mo.it
Daghio MM, Ciardullo AV, Cadioli T, Delvecchio C, Menna A, Voci C, Guidetti P, Magrini N and Liberati A. GPs’ satisfaction with the doctor–patient encounter: findings from a community-based survey. Family Practice 2003; 20: 283–288.
Received 25 June 2002; Revised 2 October 2002; Accepted 13 January 2003.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Objective. The aim of this study was to explore the characteristics of the doctor–patient relationship from the GP’s point of view.
Methods. We performed a cross-sectional 1-day study in family practice. Thirty-three GPs volunteered to fill in a questionnaire at the end of each of 20 consecutive consultations on an index day. Six hundred and sixty-one patients (out of 665) participated in the study. Descriptive frequencies of GPs’ judgements about personal experiences during the consultations, and predictors of GP’s global satisfaction score on patient encounters were analysed.
Results. The mean age of the 33 GPs was 44.7 ± 3.6 years. Professional skills (62% of the GPs had no doubts on diagnosis, therapy or prognosis) and the quality of the human/interpersonal interaction were major determinants of GPs’ satisfaction in the patient–doctor relationship. Doctors felt professionally esteemed by 90% of their patients, and the median value of their global satisfaction score (matching the expectations from an ‘ideal patient’ to that experienced when meeting the real one) was very high (median 8, range 1–10). Nevertheless, GPs did not know if they were satisfied with the actual encounter with the patient in about one-third of the consultations.
Conclusions. Professional skills and quality of the human/interpersonal interactions are major determinants of GPs’ satisfaction in their professional activities.
Keywords. Cross-sectional study, family practice, personal satisfaction, physician-patient relationship.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Over the last few decades, a conflict between health and autonomy, between patients’ and doctors’ values and expectancies has emerged.1
Several models have been proposed for the patient– doctor relationship. The traditional model is the so-called disease-centred or doctor-centred model in which any medical action focuses on the disease rather than on the sick individual.2,3 Within this framework, the disease becomes the only issue and the patient’s experience is ignored or is seen as a hindrance to be removed.4,5 On the opposite side lies the patient-centred model,6,7 which assumes that doctors have the technical knowledge useful to solve patients’ problems as long as they attempt to obtain a full understanding of patients’ questions.8 In patient-centred medicine, the doctor’s agenda needs to be combined with that of the patient.9
The need for a patient–doctor relationship based on partnership recently emerged.10,11 According to this partnership-centred or shared model, the doctors and the patients share all the steps of the decision-making process, both expressing their views, preferences and values: "shared decision making, in which patients and health professionals join in both the process of decision making and ownership of the decision made, is attracting considerable interest as means by which patients’ preferences can be incorporated into clinical decision."12
In everyday practice, when several treatments with different effects on the patient’s quality of life are available, patients are to be offered options and active involvement in decision making in order to increase the effectiveness of the treatments.13,14
These arising issues involve each patient becoming more and more a co-leading actor in the doctor–patient encounter: the doctor has to become aware of the nature of the relationship occurring in the consultation.
A major complication for this process derives from the fact that each doctor keeps in his/her mind an ‘ideal patient’ image. The more the patient meets the requirements of his/her ‘ideal patient’ image, the more he/she is satisfied about the relationship experienced during the consultation; this occurs even if he/she has no awareness of the prevailing relationship model he/she embraces.
Doctors perceive themselves as at the front line. The dichotomy between what patients expect and the awareness of the limitations of medical interventions is often the source of pessimism, frustration and professional crisis following doctor–patient encounters.
We have examined these issues closely in the practice of GPs operating in our health district area. Therefore, we designed the present cross-sectional 1-day survey in order to explore the characteristics of the doctor–patient relationship from the GP’s point of view.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
We wished to conduct the survey among all the district’s GPs, but only 33 GPs out of 76 in the health district agreed to participate in the investigation. We set out the study specifically to explore GPs’ points of views, while patients’ experience was beyond our scope. Each of the 33 GPs agreed to fill in an anonymous questionnaire immediately after the consultation for each of 20 consecutive patients seen in an index day. The index day was the same for all the GPs in order to avoid information bias due to unknown confounders.
In order to calculate the statistical power needed to test our hypothesis, without losing anonymity, a mean number of 1000 patients in the charge of each GP was considered, thus corresponding on average to 30 consultations per day. Then, we calculated that 20 questionnaires were needed to obtain 67% of them with a successful satisfaction score (see below), at a 95% confidence level, and assuming a normal distribution.
All the questionnaires were put into a closed box and handed back to the study co-ordinating centre on the following day. The questionnaire was developed specifically for this study. It was formed by several sets of questions exploring cognitive (three items, section A), affective (five items, section B) and behavioural (10 items, section 0-C-D-E) issues linked to the doctor–patient relationship. The GP filled in the questionnaire after each consultation. The three main objectives of the questionnaire were: (i) analysis of the diagnostic–therapeutic process, i.e. satisfaction in terms of professional adequacy (‘cognitive’ objective); (ii) analysis of personal experiences during the consultation, i.e. satisfaction with communication skills or empathy (‘affective’ objective); and (iii) qualitative and quantitative analysis of the GPs’ satisfaction about the relationship behaviour and its association with the theoretical model (relationship behaviour/model; ‘behavioural’ objective). The quantitative analysis was performed using a score ranging from 1 (lowest) to 10 (highest possible satisfaction); the same scale used in Italian schools to evaluate performances.15
At the end of the study, 661 (out of 665) valid questionnaires were available, corresponding to a 99% participation rate. Doctors’ characteristics were collected anonymously, whereas no individual information on the patients seen by the doctors was recorded.
Data were analysed by means of the SPSS® statistical software package. We calculated the frequencies of all descriptive variables and contingencies tables; the sum of the percentages is >100 because multiple answers to each section/question were allowed. In addition, we used a one-way ANOVA with Scheffé multiple comparison test to compare the average final satisfaction score (dependent variable) among GPs (factor).
A cluster analysis was used to discriminate the GPs’ relationship behaviour/model in assigning the satisfaction score after each consultation for two main reasons: (i) our study design (a cross-sectional survey) did not allow the typical advantages of a randomized design to be obtained, in particular it did not avoid selection bias; (ii) due to the nature of our data, a cluster of 20 patients was seen by the same doctor and thus some internal correlation could not be anticipated and/or not accounted for. Non-hierarchical cluster analysis was used due to the large data set (n = 665) structured in both dichotomic and categorical variables. AutoClass® software detected two homogeneous clusters, 80% for the first class and 20% for second one. The clusters discriminate two different kinds of GP relationship behaviour/model associated with the score assigned to the observed patient. A simple linear regression was not sufficient to determine significant predictors due to low R2. Splitting data into two homogenous clusters allowed a better model to be built, with greater ability to explain strong predictors of GPs’ relationship behaviour/model. A second one-way ANOVA was performed to check variability. All the assump-tions of the analysis were checked. A Q–Q plot of score variable showed that values <3 (min = 1; max = 10) do not follow normal distribution criteria, therefore they were excluded from the analysis (<3%). In order to reduce the effects of the interactions among variables, they were included in the model where appropriate.16
Results are expressed as raw numbers and/or percentages for frequency data, and as mean ± SD for continuous data.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The GPs’ mean age was 44.7 ± 3.6 years (range 37–53 years); 61% were men and 91% worked in an urban setting. The frequencies of answers in the complete questionnaire’s are shown in Tables 1 and 2
.
|
|
The reasons for consultations were, in decreasing order, prescription of a drug or referral for examinations (37.8%), a new emerging health problem (36.2%), routine check-up consultations (34.6%) and counselling (8.9%) (Table 1
).
A perceived cognitive inadequacy was reported in fewer than a quarter of consultations: in 127 (19%) cases, doctors expressed doubts on diagnosis, and in 103 (16%) on treatments. Section A in Table 1 shows the breakdown of these answers. Another problem experienced in about one in five consultations was interference by a third party, which in most cases was a specialist (62%), and to a lesser extent (23%) a relative/friend (see section B, Table 1).
When the overall perception was explored, doctors reported that in most cases (90%) they felt that patients held them in high professional esteem (section C, Table 1).
Satisfaction appeared to be higher for human (60%, item D1, Table 1) than for professional (50%, item D2, Table 1) interactions with their patients. However, in over one-third of encounters, GPs were unable to state their satisfaction (answer ‘don’t know’, items D1 and D2, Table 1).
The GPs’ satisfaction score was high; in fact, the median across the whole sample of encounters was 8 (mean ± SD 7.31 ± 1.61; range = 1–10) and it was 
5 in only 10% of encounters (Table 2); note that according to Italian school grades, a score 
6 is needed to achieve a pass.
The results of cluster analysis suggest that two different kinds of GP’s relationship behaviour/model can be recognized according to the predictors of the global satisfaction score. Cluster 1 accounted for 75.0% (496 out of 661) of GPs’ relationship behaviour/model consultations, and cluster 2 for 18.9% (125 out of 661); 6.1% (40 out of 661) of data were not distributed according to these two clusters.
A GPs’ satisfaction score of <6 accounted for 8.7% according to cluster 1, and for 16.8% according to cluster 2 (Table 2). Overall, GP–patient relationships experienced according to cluster 1 were more satisfactory than those related to cluster 2.
Seven of thirty-three GPs were present only in cluster 1. Cluster 2 fits 30% better than cluster 1 at the goodness-of-fit analysis. In both clusters, GP score was predicted by items related to patient’s compliance with GP’s prescriptions (item B4, Tables 3 and 4). In cluster 1, the score was predicted by the interaction of human and professional satisfaction (item D1 interaction D2, Table 3) and in cluster 2, by both human and professional satisfaction (items D1 and D2, Table 4).
|
|
There were major differences in predictors between the two groups. In cluster 1, the GP’s score was also predicted by the patient’s respect for the doctor’s professional skills (item C, Table 3
), by less doubts on diagnosis, due both to patient’s self-reporting (item A1a, Table 3) and to the GP’s own doubts about the disease (item A1c, Table 3), and by not being bothered because patients demanded a solution to his/her problem at all costs (item B3, Table 3).
In cluster 2, GPs have less doubt on drug safety but more doubts on drug efficacy (interaction of items A2 and A2a, and item A2b, Table 4). Other factors predicting the GP’s satisfaction score were related to the reasons for the consultation (item 02 and 05, Table 4) and requests suggested by specialists (interaction of items B2 and B2b, Table 4).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
All three doctor–patient relationship models mentioned above, i.e. doctor-, patient- and partnership-centred (or shared model),1,7–11,14 are not mutually exclusive. The characteristics of doctor–patient interaction could depend on patients’ personal, social and cultural characteristics as well as on the characteristics and stages of the diseases, and the totality of beliefs or attribution processes of the subject about what causes the disease or how it could be avoided.3,5 These convictions may be very different from those of the doctor. The doctor–patient encounter could be affected by the doctor’s attitude to time and energy spent listening to the patient’s story and instances and to explaining exhaustively what should be done, including all benefits and harms related to therapeutic choices.5,8–9
We are aware that our findings might be affected by a potential selection bias since we did not use a random sample of the district’s GPs. Nevertheless, our study could highlight some interesting features of the patient–doctor relationship in general practice, and useful information on the leading model commonly used could be inferred.
Overall, our findings showed that both professional skills (
62% of the GPs had no doubts on diagnosis, therapy or prognosis, section A, Table 1) and the quality of the human/interpersonal interaction (section D, Table 1) are major determinants of GPs’ satisfaction in the patient–doctor relationship, so that doctors felt professionally esteemed by 90% (item C, Table 1) of their patients, and the median value of their global satisfaction score—comparing the expectations from an ‘ideal patient’ with that experienced when meeting the real patient—was very high (median = 8, range 1–10, Table 2). These findings suggest that both the disease- and the patient-centred models prevailed in ruling the doctor–patient relationships.
Our findings have also shown that a certain number of patient–doctor encounters deal with undefined or minor health problems, and not disease, such as consultation for just a prescription or referral for examinations (section 0, Table 1), or inappropriate needs (20%) suggested by a third party, i.e. colleagues and the mass media (section B, Table 1). In this case, the GP’s skills in diagnosis and therapy are not sufficient to handle the encounter, but something else is necessary such as listening, counselling and reassurance. This kind of encounter might be referred to as a partnership-centred or shared model.7
On the other hand, we also found that GPs frequently were unable to label the actual encounter with their patient, i.e. about one-third of them did not know if they were satisfied or not (item D1 and D2, answers ‘don’t know’, Table 1). At least two possible explanations can be proposed: (i) the GPs lack a predefined model of what an ideal ‘patient–doctor relationship’ should be; or (ii) all the three models briefly described above take place each time in general practice, probably as a result of a doctor’s need to play different roles in settings of a changing relationship.
The cluster analysis helped us to clarify the last statement. Overall, GP–patient relationships experienced according to cluster 1 were more satisfactory than those related to cluster 2. In fact, a ‘successful’ (>6) GPs’ satisfaction score was obtained by 91.3% in cluster 1, and 83.2% in cluster 2 (Table 2). In both behaviour clusters, the GPs’ satisfaction was predicted both by human and professional satisfaction and by patient’s compliance with the GP’s prescriptions (item B4, D1 and D4, Tables 3 and 4), thus confirming the above-reported general finding that both the professional skills and the quality of doctor–patient relationship were the major determinants of GPs’ satisfaction. The differential predictors between the two clusters showed that in cluster 1 the GP’s relationship behaviour/model was greatly influenced by his/her professional competence and self-esteem; in fact, it was predicted by both the patient’s respect for the doctor’s professional skill and their not demanding solutions at all costs because of a high level of confidence in the doctor, and by the doctor’s fewer diagnostic doubts relying on his/her own expertise (items C, B3 and A1, Table 3). In cluster 2, the GP’s relationship behaviour was influenced by his/her professional uncertainty and the patient’s low regard for his/her professional skills; in fact, cluster 2 was predicted differentially by fewer doubts on drug safety but more doubts on drug efficacy, by the routine practices of the consultation, and by the patient’s demands due to suggestions by a specialist (item A2s, 0s, B2s, Table 4). It could be useful to stress that these two clusters did not identify two groups of doctors but just two different GPs relationship behaviours/models implemented in varying types of consultations.
Several approaches had been formulated recently aimed at improving the patient–doctor relationship such as bridging, trading, logrolling and damage limitation. All these principles imply that the doctor has to put the patient first,17,18 i.e. the patient’s empowerment and his/ her right to know and see what is happening and what could or could not be done and what he/she could choose. This approach should be justified not only by humanitarian or ethical or legal reasons, but even because it has been shown that such a behaviour brings about therapeutic effects.19
"Luke Filde’s 19th century painting of a contemplative doctor alone with a sick child might now be replaced by a harassed doctor trying to park his car to get to a meeting on time."20 This statement probably best summarizes modern doctors’ struggles to interpret his/her changing role vis a vis their patients. It may be interesting in future studies to explore GPs’ perception about their professional roles compared with those of other health professionals such as specialists and academic clinicians.
|
Acknowledgments |
|---|
We are very grateful to the GPs participating in the study: L Azzolini, M Bacchelli, M Baroni, E Bedini, C Bellodi, AC Beltrami, B Bertacchini, M Bevini, G Carretti, A Di Fiore, C Fasdelli, G Feltri, F Iannaccone, G Gaglianò, G Gazzani, G Gualdi, G Loscalzo, A Losi, MG Lugli, R Lugli, L Mantovani, P Malavasi, A Morellini, V Pavarotti, A Ragazzoni, N Righi, C Rosselli, G Tassoni, G Tirelli, M Veratti and A Vincenzi. Special thanks to Dr Salvatore Panico (Naples University) for his helpful final suggestions, and to Drs Susanna Casari (past district head) and Manuela Lorenzetti (responsible for the GP district) for excellent field co-ordination. This work has been supported by public health authorities.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1 Emanuel EJ, Emanuel LL. Four models of the physician–patient relationship. J Am Med Assoc 1992; 257: 2221–2226.
2 Byrne PS, Long B. Doctors Talking to Patients. London: HMSO, 1976.
3 Engel GL. The need of a new medical model: a challenge for biomedicine. Science 1977; 196: 129–136.
4 Levenstein JH. Family medicine and the new science. S Afr Fam Pract 1988; 9: 11–17.
5 Butler NM, Campion PD, Cox AD. Exploration of doctor and patient agendas in general practice consultations. Soc Sci Med 1992; 35: 1145–1155.[CrossRef][Web of Science][Medline]
6 Balint M, Hunt J, Joice D, Marinker M, Woodcock J. Treatment or Diagnosis. London: Tavistock Publications, 1970.
7 Levenstein JH, McCracken EC, McWhinney IR, Stewart MA, Brown JB. The patient centred clinical method. A model for the doctor–patient interaction in family medicine. Fam Pract 1986, 3: 24–30.
8 Bellet PS, Maloney MJ. The importance of empathy as an interviewing skill in medicine. J Am Med Assoc 1991; 266: 1831–1832.
9 Marvel MK, Epstein RM, Flowers K, Beckman HB. Soliciting the patient’s agenda: have we improved? J Am Med Assoc 1999; 281: 283–287.
10 Charles C, Whelan T, Gafni A. What do we mean by partnership in making decisions about treatment? Br Med J 1999; 319: 780–782.
11 Coulter A. Paternalism or partnership? Br Med J 1999; 319: 719–720.
12 Coulter A. Partnerships with patients: the pros and cons of shared clinical decision-making. J Health Serv Res Policy 1997; 2: 112–121.[Medline]
13 Coulter A, Entwistle V, Gilbert D. Sharing decisions with patients: is the information good enough? Br Med J 1999; 318: 318–322.
14 Charles C, Gafni A, Whelan T. Decision making in the physician– patient encounter: reconsultationing the shared treatment decision making model. Soc Sci Med 1999; 49: 651–661.[CrossRef][Web of Science][Medline]
15 OCSE-CERI. Gli Indicatori Internazionali Dell’istruzione. Rome: Armando, 1994.
16 Rothman KJ, Greenland S. Modern Epidemiology, 2nd edn. Philadelphia: Lippincott-Raven, 1998: 737.
17 National Health Council. Putting Patients First. Washington (DC): NHC, 1995.
18 Michela JL, Wood JV. Causal attributions in health and illness. Adv Cogn–Behav Res Ther 1996; 5: 179–235.
19 Buetow S. Four strategies for negotiated care. J R Soc Med 1998; 91: 199–201.[Web of Science][Medline]
20 Smith R. Why are doctors so unhappy? Br Med J 2001; 322: 1073–1074.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M.-A. Landry, S. Lafrenaye, M.-C. Roy, and C. Cyr A Randomized, Controlled Trial of Bedside Versus Conference-Room Case Presentation in a Pediatric Intensive Care Unit Pediatrics, August 1, 2007; 120(2): 275 - 280. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||