General Practice

Qualitative study of educational interaction between general practitioners and specialists

BMJ 1998; 316 doi: (Published 07 February 1998) Cite this as: BMJ 1998;316:442
  1. Martin N Marshall, lecturera
  1. a Institute of General Practice, Postgraduate Medical School, University of Exeter, Exeter EX2 5DW
  • Accepted 3 October 1997


Objectives: To identify the main barriers to effective educational interaction between general practitioners and specialists and to suggest ways of overcoming these barriers.

Design: Qualitative study using semistructured interviews and focus groups. A content analysis of the data was performed.

Setting: South and West Regional Health Authority.

Subjects: 12 general practitioner principals and 12 hospital consultants were sampled purposefully and underwent a semistructured interview; a further 16 general practitioners and 16 hospital consultants similarly sampled made up four focus groups representing the two branches of the medical profession.

Results: There was a mismatch between what the general practitioners wanted from specialists in educational terms, and what the specialists were providing. General practitioners wanted to learn information that was directly applicable to their clinical work and to use referrals as two way learning opportunities. They were not sufficiently explicit about their learning needs. Specialists preferred to concentrate on new developments in their subject and would benefit from learning different ways of teaching. The participants were willing to learn from each other. Three models of educational interaction were identified: traditional didactic lectures given by specialists to general practitioners, interactive clinically based teaching, and informal interaction based on referrals.

Conclusions: The two main branches of the medical profession have to address several significant problems before the full potential of teaching and learning together can be realised.

Key messages

  • There is a mismatch between what general practitioners want from specialists in educational terms and what specialists are providing

  • Specialists prefer traditional, formal teaching methods but general practitioners prefer informal problem oriented learning

  • General practitioners should define and express their learning needs to specialists

  • Education should be a two way process since this will help promote mutual understanding of different roles and functions within the medical profession

  • The two branches of the medical profession are willing to learn from each other


General practitioners and specialists have different but complementary knowledge and skills and potentially have much to learn from each other. The traditional hierarchical relationship which resulted in a one way transfer of knowledge from specialists to general practitioners1 is now being challenged as general practice continues to assert its unique identity.2 Nevertheless, many learning opportunities remain unrecognised or underused,3 and educational interaction between the two main branches of the medical profession has been inadequately researched.

Published work suggests that many general practitioners are dissatisfied with the current situation. They dislike didactic lectures,4 want more feedback from specialists about the quality and appropriateness of their referrals, and complain that the content of specialist teaching is often inappropriate.2 3 5 The educational benefit of specialist outreach clinics in general practitioners' surgeries is considered to be minimal,6 and only a few examples of successful joint initiatives to encourage two way education exist.7 8 To my knowledge, there is no research evidence to indicate what specialists want to teach or learn from general practitioners.

This study is part of a larger project investigating the changing relationship between general practitioners and specialists. Educational interaction was identified by the participants as important and problematic. The purpose of this study is to identify the main barriers to effective educational interaction and to suggest ways of overcoming them. Since it is an exploratory study of interpersonal and intraprofessional interaction, a qualitative approach was considered appropriate.

Subjects and methods

The study population included all general practitioners and specialists working in the South and West Regional Health Authority who had regular personal and professional interaction. Specialists with minimal contact with general practitioners, such as anaesthetists, were excluded. Data were collected using two types of qualitative interview, and all interviews were audiotaped and supplemented with field notes.

Semistructured interviews were conducted with a purposeful sample of clinicians from across the region. Subjects were selected using a qualitative sampling frame9 to ensure a broad spectrum of demographic and professional characteristics and were identified by snowball sampling techniques.10 Sampling ceased when categorical and theoretical saturation was achieved.10 In total, 12 general practitioners and 12 specialists were interviewed. The interview schedule was designed during an earlier part of the study11 and was applied flexibly in the way that most suited the interviewee.12

Focus groups–Four focus groups, each with four general practitioners and four specialists from a single locality, were conducted at the local postgraduate medical centre. Participants were selected using similar criteria to those for the semistructured interviews. The membership of the groups was carefully balanced according to the advice from an expert in group dynamics to ensure maximum productivity. I was facilitator, but the groups were largely self managed.13

Transcript analysis and interpretation–The tapes were fully transcribed, and these transcripts, together with the field notes and reflective comments, formed the raw data for further analysis. The processes of sampling, data collection, and data analysis were continuous and iterative. A content analysis10 14 was conducted using the computer program Atlas ti (T Muhr, Berlin, Germany, 1994). Once the data had been fully coded, annotated, and categorised I used network and matrix visual displays to aid interpretation.10 Such displays are easily created and modified with Atlas ti.

Quality assurance of data analysis and interpretation–The consistency (reliability) and confirmability (validity) of data analysis and interpretation was assessed using three techniques. Firstly, external validation of all stages of coding and interpretation of five randomly selected semistructured interview transcripts and two focus group transcripts was performed independently by three experienced qualitative researchers. The results were compared and there were no significant inconsistencies. Secondly, I discussed the interpretation of the data with four randomly selected doctors who had been interviewed with the semistructured questionnaire and with one member of each of the focus groups to compare my perspective with that of the subjects. This process of respondent validation again found no significant criticisms of my interpretation. Thirdly, the results were triangulated with different data sources within the study, with other data collecting methods used in the larger project (including key informant interviews11 and a Likert survey), with the available literature, and with data collected incidentally outside the formal interviews. I gave the most importance to results that were consistent between data sources and with different data collection methods.


Three themes were identified from the data. The first concerned ways of meeting the learning and teaching needs of the participants, the second considered barriers to achieving effective educational interaction, and the third identified ways of improving educational interaction.

Meeting educational needs

What do general practitioners want to learn from specialists?

The general practitioners had clear views on the content and style of teaching they wished to receive from their specialist colleagues. They wanted teaching to be directly related to their clinical work, and several suggested that specialists had little understanding of their learning needs or of the pressure that they were under to keep up to date:

“[General practitioners] have got to decide what their learning needs are and head for those, they haven't got time to do everything” (focus group 3).

Most general practitioners thought that “informal” learning centred on clinical cases was likely to be of greater use to them than formal lectures, and they liked to have easy access to telephone advice from specialists.

Specialists were criticised for not using referrals as “learning opportunities” (focus group 2) to their full potential and not being sufficiently explicit about the quality of referrals. One general practitioner referred to his own “specialist seeking behaviour” and claimed: “Unless you actually tell us that our behaviour might not be quite appropriate, we will carry on doing the same thing” (focus group 2).

For straightforward clinical issues, such as the management of hypertension, general practitioners were happy to be informed of current “best practice” and thought that to a certain extent it was the specialists' responsibility to keep them up to date. Discussion in focus groups showed that specialists' reluctance to assume this role was in part caused by their failure to reach consensus among themselves about what they wanted from general practitioners.

In general, however, general practitioners wanted to control their own educational agenda and to inform specialists of their learning needs:

“General practice is such a separate specialty really from the way that consultants work that there isn't always a lot of useful information that they can give really … I recognise that they have got distinct expertise that I haven't got, but it's not expertise that I need” (interview 16).

What do specialists want to teach general practitioners?

Most specialists responded willingly to requests to teach and seemed to be motivated by a deep, apparently unquestioned, sense of professional duty. Some wanted to improve the quality of the referrals that they received, but none thought that their input would reduce the number of referrals—indeed, they often said the opposite.

Specialists regarded lectures as the principal way of transferring information: few had given any serious consideration to alternative teaching methods. Only a few used their response to referrals as an opportunity to educate the referring general practitioner because they thought that this might be perceived as condescending or critical and because of the time required to formulate such replies. Most specialists disliked being interrupted by telephone calls from general practitioners, thought that outreach clinics had no educational benefits, and considered guidelines to be an ineffective way of transferring information to general practitioners.

General practitioners did not doubt specialists' desire to teach but questioned their teaching skills, and, when challenged in the group environment, the specialists accepted that their teaching responsibility was not underpinned by educational training:

“Just because you are a consultant you are automatically able to teach—that is an assumption that you get” (focus group 2).

Can general practitioners teach specialists?

In the semistructured interviews the two branches of the profession seemed to be locked in a traditional teacher-pupil hierarchy, with few general practitioners and even fewer specialists questioning whether education should be a two way process.

In the interactive environment of the focus groups, however, the general practitioners thought otherwise. They considered that they could teach their specialist colleagues about the benefits of a holistic approach, about teaching methods, and about communication skills. The specialists were sceptical that communication could be “taught,” but they recognised general practitioners' skills in the other areas and wanted to learn from them.

Barriers to effective educational interaction

The disparity between what general practitioners want to learn from specialists and what they think they are receiving from them is an important barrier to effective educational interaction, but other barriers were also identified. The participants thought that effective educational interaction could be built only on a foundation of mutual trust and respect since this made it easier and safer to make constructive criticisms. They thought that a desire to protect each other from litigation also inhibited critical feedback.

Increasing workload was considered to be an important barrier since this prevented clinicians meeting and led to them giving a low priority to their educational needs. Not only specialists were criticised for their poor availability: general practitioners were perceived to be increasingly absent from their surgery on half days, study leave, and holiday, and this meant that specialists were often unable to make contact with them about clinical problems.

Several of the specialists considered that the NHS internal market had influenced educational interaction. Indeed, one specialist questioned whether he should attempt to educate local general practitioners when the end result might be reduced referrals for the local trust hospital:

“I mean you could argue now to keep [general practitioners] as ignorant as possible, so they couldn't do a bloody thing, had to refer everything, we could have four or five consultants … I mean, one could be cynical now” (focus group 3).

The way forward

Two approaches to addressing these barriers were identified. The first considered active promotion of two way educational interaction and the second considered other ways of communicating and transferring information.

Promoting two way education

Participants thought that two way interaction was important for effective mutual education and that this could be achieved by general practitioners expressing clearly what they wanted from their specialist colleagues and by specialists developing greater educational expertise. The general practitioners considered that a true understanding would be achieved only by specialists spending time in general practice at some stage of their training, since they thought that specialists had little concept of the work of general practitioners. Most of the specialists accepted that this would be useful, but there was little consensus about the duration, timing, or content of the attachments.

Developing new ways of teaching and learning

The participants thought that limited time would always be a barrier to effective educational interaction and that other ways of transferring information had to be considered. The use of television links, electronic communication, and the internet was considered in detail in one of the focus groups. The merits were considered to be the potential speed of interaction, the time that email communication gives people to think in comparison with the telephone, and the potential to overcome the problem of availability. Only one specialist was unhappy with the suggestions because he believed that they would “dehumanise” interaction with his colleagues.


This study has identified the main barriers to effective educational interaction between general practitioners and specialists and suggests ways of overcoming them. Its qualitative approach enabled these issues to be explored in a way that would not have been possible with quantitative methods. The rigorous application of qualitative sampling techniques ensured a broad range of opinions, but the utility of the study depends not on the generalisability of the results but on their ability to provoke readers to compare their own experience with that of the participants.

This study confirms earlier findings that general practitioners dislike didactic lectures,4 want educational content to be more oriented towards their daily work,2 5 and consider that opportunities for mutual education are not being used to their full potential.3 For the first time, this study identifies the reasons for these problems and highlights a strong desire to base education on mutual feedback about the clinical practice of both branches of the profession.

The participants described three models of educational interaction. The first is based on traditional didactic lectures given by specialists to general practitioners. This was the least popular model among general practitioners but the preferred way for specialists. The second model entailed interactive sessions centred on clinical cases. This was a popular model for general practitioners and for the specialists who had used it, though they were few. The final model entailed informal and unplanned learning based on referrals. This was the preferred method for most general practitioners, who valued it as an opportunity for two way learning, but the least popular for specialists chiefly because it was so time consuming.

General practitioners and specialists in the South and West region are enthusiastic about working and learning together, but the two branches of the medical profession still have some barriers to overcome.


I thank the specialists and general practitioners who took part in this study, and Dr Keith Bolden for his advice on the composition of the focus groups. I thank my project supervisors, Professor Denis Pereira Gray, Professor David Phillips, and Professor Roger Jones, for their guidance.

Funding: The study was funded by a research training fellowship from the Research and Development Directorate of the South and West Regional Health Authority.

Conflict of interest: None.


Contributors: MNM designed, executed, and analysed this study. Professor David Phillips and Drs Keiran Sweeney and Adrian Freeman independently analysed the transcripts.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
View Abstract