Family Practice Vol. 16, No. 1, 33-38
© Oxford University Press 1999
How well do GPs and hospital consultants work together? A survey of the professional relationship
Institute of General Practice, Postgraduate Medical School, University of Exeter, Barrack Road, Exeter EX2 5DW, UK.
Marshall MN. How well do GPs and hospital consultants work together? A survey of the professional relationship. Family Practice 1999; 16: 33–38.
Received 10 June 1998; Accepted 7 October 1998.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Background. The professional relationship between GPs and hospital consultants (sometimes referred to as ‘specialists’) is important in a healthcare system based upon the generalist as the first point of contact for patients and the gate-keeper for hospital services. This relationship has been the subject of considerable interest over the years, but little empirical research.
Objectives. We aimed to compare the attitudes of GPs and specialists to key issues, and to produce a validated objective measure of their ability to work together.
Methods. We conducted a Likert-style survey based upon statements made in qualitative interviews with GPs and specialists working in the South and West of England. The questionnaire was modified and validated during a multi-stage pilot and was distributed to a stratified random sample of 800 clinicians.
Results. GPs and specialists demonstrate a good level of agreement, mutual understanding and respect, though there are significant differences between the two branches of the profession in terms of attitude towards financial parity and direct access to special investigations.
Conclusions. A measure of the ability of GPs and specialists to work together has been developed, with acceptable internal consistency and validity. It may be used in other geographical areas to assess a relationship which is central to the efficient and effective operation of the National Health Service.
Keywords. GP, professional relationship, specialist..
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The way that GPs and hospital consultants interact has important implications for any health care system in which the generalist is the first contact for patients and the point of access to relatively scarce and expensive specialist services.1 For this reason, the relationship between the two main branches of the medical profession has received a considerable amount of interest,2–6 though it has only rarely been systematically researched.
The traditional perceptions of the two branches of the profession were summarized by Horder4 20 years ago. He described GPs as being jealous of the status, facilities and income of their specialist colleagues and resentful that at the time there was no special training for their role. He describes specialists' perception that GPs dealt with minor problems, mostly of a psychological or social nature and rarely used or needed their medical knowledge. He claimed that many specialists said that a GP's main task was to sort out what was minor from what was major and to refer the latter to the hospital.
Whilst this description might have been exaggerated, detailed studies of the profession in the 1960s and 1980s did highlight major problems in the relationship.3,6 More recent research suggests that there might still be problems with mutual understanding and communication. Specialists complain about inadequate information7 and unnecessary referrals,1,8–10 whilst GPs have expressed dissatisfaction with lack of information,11–13 failure to take account of important psycho-social information9 and delays in communication.14 Others have suggested that the two branches of the profession have such different core values that lack of understanding is inevitable.13,15,16
In the last two decades, several factors have changed the relationship between the two main branches of the profession. Specialists are undoubtedly less autonomous and less powerful within the hospital environment than they have been in the past, as a result of increasing management control of their activities.17 In general practice, the quality of training has improved, as have the practice premises, available services and skills within the primary health care team. The GP's key role as patient's advocate has been enshrined in legislation,18 and the influence that they have as purchasers of hospital services has been said to result in a power shift within the profession.19
This paper is one part of a mainly qualitative study of the professional relationship in the South West of England. The aim of this stage of the study was to test out issues identified in the qualitative interviews on a random sample of clinicians in order to determine their generalizability and to produce a validated tool to measure the ability of GPs and hospital consultants to work together.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The study was conducted in the South and West Region of the United Kingdom between May 1996 and February 1997. In total, there were 3900 GP principals and 2470 hospital consultants working in the Region, though specialists with minimal daily contact with GPs, such as anaesthetists, were excluded from the study population.
Questionnaire development
Analysis of the data from the qualitative stages of the study suggested that some clinicians worked well together, whilst other relationships were markedly dysfunctional. A series of attitudinal statements were identified to reflect this range of attitudes and were drawn almost verbatim from the interview transcripts, thereby increasing their face and content validity. For the first stage of the pilot, 40 statements, some positively and some negatively worded, were used and piloted on a convenience sample of six GPs and six specialists, using a five-point Likert-style attitudinal questionnaire. The range of responses (‘strongly agree’, ‘agree’, ‘uncertain’, ‘disagree’ and ‘strongly disagree’) were scored from one (least positive attitude) to five (most positive attitude). Statements that were considered to be ambiguous or were poor differentiators between positive and negative attitudes were discarded, resulting in a bank of 20 statements that were piloted formally in a postal survey of a random sample of 30 GPs and 30 specialists identified from a published database of clinicians (Medical Telefax Guide, 1995).
The pilot data were analysed using SPSS for Windows (SDPSS Inc., 1993, Chicago, Illinois). Response bar charts were used to identify the items which were good differentiators between positive and negative attitudes, and an item analysis was performed20 to assess the correlation between the score for each item and the total score (minus the score for that item) for each subject. A correlation coefficient of greater than 0.5, with a P value of < 0.05, was considered acceptable. This resulted in a validated five-point Likert questionnaire with 12 items (Table 2
). Background data about the subjects were also requested.
|
The sample size needed to identify differences between groups was calculated from the pilot study (using P
0.05, power of 80%). Minor modifications to the layout and content of the questionnaire were made as a result of the pilot and the balance of positively and negatively worded items was assessed independently by three experienced researchers.
Distribution and validation of the final questionnaire
The questionnaire was distributed by post to a stratified random sample of 800 (approximately 1 in 8) GPs and specialists working in the South and West Region. Stratification was performed according to whether the subject was a GP or specialist and the type of speciality. A stamped addressed envelope was sent with a covering letter which assured confidentiality and a single reminder was sent to non-responders.
For ease of analysis, the different types of hospital speciality were reduced into three broad groups, physicians (which included paediatricians and psychiatrists), surgeons (which included obstetricians and gynaecologists) and pathologists/radiologists. The length of time qualified was classified into two groups, less than 20 years, or 20 or more years, which the pilot study suggested would divide the sample into two equally sized groups.
The characteristics of the questionnaire responders and non-responders were compared using the chi-square test. The internal consistency and the construct validity of the items and whole questionnaire were re-checked by correlating item scores with total scores, by constructing a Spearman's correlation matrix and by calculating the Cronbach's alpha coefficient.
Analysis of the final questionnaire
A comparison of the responses of the GP and specialist samples to each individual item was assessed graphically using bar charts and statistically using medians, interquartile ranges and the Mann–Whitney U test to compare the medians. Mean total scores were compared using the t-test for two independent samples and the one-way ANOVA for three independent samples. A factorial ANOVA analysis was performed to determine the proportion of variance in the total score which could be accounted for by each of the independent variables. Finally a principal component analysis with varimax rotation21 was used to indicate which statements examined similar aspects of the attitude towards working together. The eigenvalue was set as one and the component loading between item and factor was set at greater than 0.6.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Characteristics of responders
A useable response rate of 75.8% (606/800) was achieved after one reminder. Specialists were significantly less likely to respond to the questionnaire than GPs (chi-square 3.89, P = 0.05) and non-responders were significantly more likely to have been qualified for more than 20 years (chi-square 4.87, P = 0.03), less likely to be in teaching hospital consultant posts (chi-square 7.98, P = 0.005) and less likely to be members or fellows of the Royal College of General Practitioners (chi-square 8.39, P = 0.004).
Validation of the questionnaire
The scores for each item were statistically significantly correlated with the total score for each participant (correlation coefficient range 0.42–0.69, P < 0.001 in all cases) and each item was significantly correlated with other items in the questionnaire (Table 1). The Cronbach's alpha value for the questionnaire was 0.77. These results represent an acceptable level of internal consistency and construct validity for the questionnaire, as a tool for measuring attitude towards working together.22
|
Item responses
There was a statistically significant difference in the response of GPs and specialist to all items (Table 2
). However, the direction of response (generally agree or disagree) was similar for most of the questions. There was clear disagreement for the items relating to parity of income, where specialists thought they deserved to earn a higher NHS income than GPs and access to special investigations, where GPs wanted greater open-access than the specialists wanted them to have. Specialist opinion was more divided than GP opinion about the management of common chronic diseases, the use of purchasing by GPs to ‘get their own back’ on specialists and GPs' responses to specialist advice.
Total scores
There was also a significant difference between the total scores of GPs and specialists. Figure 1 and Table 3 demonstrate that the scores were generally high, but that GPs have a significantly more positive attitude to the professional relationship than specialists. Probability plots for specialist and GP scores separately and together demonstrated a high degree of normality for all distributions.
|
|
Higher positive scores were more likely to be obtained from those who had qualified less than 20 years ago, females, fundholding GPs, members or fellows of the Royal College of General Practitioners, and GP trainers. There was no significant relationship between the total score and type of speciality, teaching hospital status of the specialist or whether the GP referred to a teaching hospital (Table 3
). A factorial ANOVA analysis indicated that the professional status of the responders was the single most important determinant of total score (F = 163.83, P < 0.001). The only other significant factors were fundholding status (F = 8.539, P = 0.004) and membership of the RCGP (F = 9.013, P = 0.003), though these both had a relatively small influence on total score.
Factor analysis suggested two possible factors within the broad concept of attitude towards working together. The first consisted of four items (nos 1, 3, 8 and 12) which appear to relate to the perceived clinical competence of GPs. The second consisted of two items (nos 4 and 6) which appear to relate to non-clinical issues that were considered to motivate GPs' behaviour. However, the Cronbach's alpha values for the two factors were unacceptably low, at 0.62 and 0.37, respectively.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
A review of the literature appears to confirm the anecdotal impression that the relationship between the two main branches of the medical profession has been characterized by fundamental misunderstanding, lack of respect and professional jealousy. This has been the case since the earliest days of apothecaries and gentlemen physicians, before the formation of the medical profession in the middle of the nineteenth century. This study highlights some continuing differences in the expressed attitudes of GPs and specialists, but it also demonstrates a considerable positive attitude towards working together, reflected by the high mean score, and does not suggest a wide gulf between the two branches of the medical profession in most of the areas considered.
The apparent differences between the negative portrayal in the literature and the relatively positive picture in this study may be explained in a number of ways. It is possible that attitudes have changed and that previous work is out-dated. It could also be that previous studies were inaccurate, that researchers over-emphasized the problems and missed the positive features. Many of the previous studies were conducted by sociologists,3,6 who perhaps entered the professional field expecting to find a homogenous body, were surprised to find friction and therefore highlighted it. Clinicians working in the South-West of England might also be atypical in their desire to work together.
It is also possible that the chosen methodology for this study was inappropriate. A study of complex and sometimes sensitive inter-personal relationships is perhaps best conducted by utilizing only the qualitative research methods which were used to create the attitude statement. The combination of quantitative and qualitative methods have, however, been advocated by several authorities, in part as a way of triangulating the results to test their trustworthiness.23–25 The use in this study of a Likert-style survey based upon qualitative interviews supports this recommendation, though any variation in the findings of the different paradigms might have reflected the different contexts of the data collection methods.26 The experience of combining methods in this study suggests that quantitative surveys are more likely to highlight differences between groups, whilst naturalistic methods are more likely to explore the commonalties. This has important implications for researchers considering the most relevant methodological approach and for readers who might question the apparently contradictory conclusions drawn from different studies. The process of creating an internally consistent tool resulted in a questionnaire with a relatively small number of items. This not only potentially reduced the face validity of the questionnaire but also decreased the chance of identifying factors within the overall concept of ability to work together.
The high level of respect and co-operation between the two branches of the profession is reflected by agreement concerning the equal calibre necessary for specialist and generalist practice, and the acceptance that GPs should be able to influence the service provided by specialists. It is also demonstrated by the significant proportion of specialists who consider that GPs do not always have to follow their advice, an understanding that GPs are more than just filters for hospital services and a desire on the part of specialists to learn from and within general practice. Specialists and GPs disagree about financial parity and direct access to special investigations, but, perhaps surprisingly, there seems to be little desire on the part of GPs to allocate greater priority to the development of primary care services in preference to secondary care services.
This study had revealed a good level of mutual respect and understanding between GPs and hospital consultants in the South and West of England, and has produced a validated tool which could be used by others to measure the professional relationship. An effective working relationship influences professional morale and development, effective and appropriate use of resources and possibly also the quality of patient care. It should therefore be of interest to all those working in and with the medical profession.
|
Acknowledgments |
|---|
I would like to thank the specialists and GPs who took part in this study. The paper forms part of an MD thesis awarded by the University of London, in October 1997, supervised by Professor Denis Pereira Gray, Professor David Phillips and Professor Roger Jones. The study was funded by the Research Training Fellowship from the Research and Development Directorate of the South and West Regional Health Authority.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1 Westerman RF, Hull FM, Bezemer PD, Gort G. A study of communication between general practitioners and specialists. Br J Gen Pract 1990; 40: 445–449.[Web of Science][Medline]
2 Susser MW. Further thoughts on the future of medical practice outside the hospital. Lancet 1963; 1: 315–319.[Web of Science][Medline]
3 Stevens R. Medical Practice in Modern England. New Haven and London: Yale University Press, 1966.
4 Horder JP. Physicians and family doctors: A new relationship. J R Coll Gen Pract 1977; 27: 391–397.[Medline]
5 Higgins PM. The GP/Hospital interface. J R Coll Physicians Lond 1979; 13: 132–138.[Web of Science][Medline]
6 Honigsbaum F. The Division in British Medicine: A History of the Separation of General Practice from Hospital Care 1911–1968. London: Kogan Page, 1979.
7 Pullen IM, Yellowlees AJ. Is communication improving between general practitioners and psychiatrists? Br Med J 1985; 290: 31–33.
8 Rowland MO, Porter R, Matthews JG, Redden JF, Simonds GW, Bewley B. Improving care; a study of orthopaedic out-patient referrals. Br Med J 1991; 302: 1124–1128.
9
Grace JF, Armstrong D. Referral to hospital: Perceptions of patients, general practitioners and consultants about necessity and suitability of referral. Fam Pract 1987; 4: 170–175.
10
Newton J, Hutchinson A, Hayes V, McColl E, Mackee I, Holland C. Do clinicians tell each other enough? An analysis of referral communication in two specialities. Fam Pract 1994; 11: 15–20.
11 Harding J. Study of discharge communications from hospital doctors to an inner London general practice. J R Coll Gen Pract 1987; 37: 494–495.[Web of Science][Medline]
12 Muzzin LJ. Understanding the process of medical referral. Canad Fam Physician 1992; 38: 301–307.
13 Wood ML. Communication between cancer specialists and family doctors. Canad Fam Physic 1993; 39: 49–57.
14 Penny TM. Delayed communication between hospitals and general practitioners. Br Med J 1988; 297: 28–29.
15 Whitfield MJ. What do consultants think of general practice? J R Coll Gen Pract 1980; 30: 228–229.[Medline]
16 Whitfield MJ, Bradley MC. GPs and Consultants: Is there agreement on patient management? Bristol Medico-Chirurgical J 1989; 104: 75–78.
17
Browse N. Clinicians must lead. Br Med J 1996; 313: 1268.
18 Secretary of State for Health. Working for patients. Command 555. London: HMSO, 1989.
19 Klein R. The New Politics of the NHS. London and New York: Longman, 1995.
20 Oppenheim AN. Questionnaire Design, Interviewing and Attitude Measurement. Aldershot: Gower, 1992.
21 Norusis MJ. SPSS for Windows Professional Statistics, Release 6.0. Chicago: SPSS Inc, 1993.
22
Bland JM, Altman DG. Statistics notes: Cronbach Alpha. Br Med J 1997; 314: 572.
23 Miles MB, Huberman AM. Qualitative Data Analysis. 2nd edn. Thousand Oaks, London, New Delhi: Sage Publications, 1994.
24
Britten N, Jones R, Murphy E, Stacey R. Qualitative research methods in general practice and primary care. Fam Pract 1995; 12: 104–114.
25
Mays N, Pope C. Rigour and qualitative research. Br Med J 1995; 311: 109–112.
26 Denzin NK, Lincoln YS. Handbook of Qualitative Research. Thousand Oaks, London, New Delhi: Sage Publications, 1994.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||