Family Practice Vol. 18, No. 4, 399-406
© Oxford University Press 2001
Referrals and relationships: in-practice referrals meetings in a general practice
Department of General Practice and Primary Care, St George's Hospital Medical School, 6th Floor, Hunter Wing, Cranmer Terrace, London SW17 0RE and
a Group Analytic Therapist, Twickenham, Middlesex, UK.
Rowlands G, Willis S and Singleton A. Referrals and relationships: in-practice referrals meetings in a general practice. Family Practice 2001; 18: 399–406.
Received 2 October 2000; Revised 2 February 2001; Accepted 12 March 2001.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Background. GP referrals to secondary care are an important factor in the cost of running the NHS. The known variation in referral rates between doctors has the potential to cause tension within primary care which will be exacerbated by the latest reorganization of primary care and the trend towards capitation-based budgets. The importance of postgraduate learning for GPs has been recognized; continuing professional development is moving towards self-directed practice-based learning programmes. Educational interventions have been shown to alter doctors' prescribing behaviour. This, together with the pressure on accounting for referral activity, makes the prospect of improving, and possibly reducing, referral activity through educational interventions very attractive.
Objectives. This study complemented a randomized controlled trial (RCT) which investigated whether an intervention of the type which had reduced prescribing costs would have a similar effect on referral activity.
Methods. The context of the study, description of the characteristics of the practice and the issues seen as important by the doctors and practice manager were identified through preliminary semi-structured interviews. The practice then held a series of educational in-practice meetings to discuss referrals and issues arising from referrals. The audio- and videotaped transcripts were interpreted using content and group dynamic analysis. Participants commented upon our preliminary findings. In addition, we used dimensional analysis to induce a preliminary theory describing the effect of the intervention on this general practice which enabled us to review the findings of the parallel RCT. The educational value of the meetings and the learning needs of the participants were also assessed.
Results. Our complementary study showed no alteration of practice referral rates following the educational intervention. The qualitative study, unencumbered by the assumptions inherent in the development of the hypothesis tested in the RCT, highlighted the complexity of decision making in general practice and the likely impact of historical background and a variety of internal and external pressures on this self-directive educational intervention. The practice members described the individual and group learning needs identified as a result of the meetings.
Conclusion. The findings of this study raise important questions for developing practice-based learning. The outcomes of self-directive interventions in practices will be influenced by internal and external events both past and present. Such outcomes may be qualitative and difficult to measure. They are likely to differ from outcomes seen when interventions are applied to groups of doctors who are not all members of the same practice.
Keywords. Changing behaviour, meetings, referrals.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The decision on whether and when to refer a patient to hospital is a complex one,1,2 that varies between doctors3,4 and, in view of its complexity, is likely to vary for individual doctors at different times. Such variations represent a large impact on the costs to the NHS.5 The latest restructuring of primary care into Primary Care Organiztions (PCOs) may generate stress and conflicts within and between practices.6 The additional trend towards capitation-based budgets7 has the potential to compound this as PCOs and the doctors within them struggle to define budgets and referrals policies.
As a contribution towards reducing such conflict, we sought to test an educational innovation designed to reduce in-practice referral rates. It was based on a study showing that self-directive small group educational meetings held between doctors from different practices reduce the frequency and cost of prescribing by GPs8 and a pilot study indicating that such meetings might exert the same effect on doctors' referral rates when held in one practice.9 In addition, Davis10 has shown that peer discussion, rehearsal of communication skills and patient-centred approaches are effective educational strategies that can lead to a change in clinical behaviour. The current restructuring of continuing professional development for health professionals recognizes this and will focus in future on self-directed practice-based learning programmes.11
The formulation of a hypothesis to test in a randomized controlled trial (RCT) required several assumptions based on the previous work described above. We assumed firstly that referral and prescribing decisions share common influences that would respond to a previously proven intervention in similar ways and, secondly, that applying the intervention in a practice setting rather than in groups of doctors brought together for the purpose of the study would not alter the effect of the intervention.
The qualitative arm of the study was unencumbered by those assumptions necessary in the formulation of a hypothesis that could be tested in the RCT. It allowed scope for an in-depth focus on one practice with a detailed description of the intervention and its effect on the group as it appeared to us. The self-directive nature of the intervention allowed for the description of a range of natural outcomes related to the referral process that occurred in this practice at this point in its development, illuminating findings that would otherwise have remained hidden. In particular, it highlighted the complexity of practice dynamics and a number of issues related to the secondary care referral process that the rigid structure of the RCT would have been unable to accommodate.
Our aims were: (i) to conduct a qualitative study running parallel with an RCT to investigate the effect of in-practice meetings on practice referral rates; (ii) to conduct an in-depth review of the issues relating to the referrals process within the study practice; (iii) to describe the intervention and the outcomes of the intervention in the study practice; (iv) to compare the change in referral rate in the study practice before and after the referrals meetings with the RCT findings; (v) to induce a theory to explain the outcomes of the intervention as seen in the study practice; (vi) to describe the learning needs identified by participants as a result of the meetings; and (vii) to discuss the relevance of the intervention in the light of the move towards practice-based educational programmes.11
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The study took place between July 1996 and April 1997.
Setting and subjects
Recruitment to this qualitative study took place in parallel with recruitment for the RCT and is described elsewhere.12 The practice chosen for the qualitative project was the one sited nearest to GR's place of work in order to facilitate her attendance at the practice meetings.
The practice was a four-partner practice that trained both GP registrars and medical students. The patient population (11 000) consisted of mainly social classes I, II and IIIa; the practice was not in receipt of deprivation payments. The practice was a second wave fundholder; at the time of the study, it had been a fundholding practice for 3 years.
Internal practice context
This practice recently had undergone a period of instability and uncertainty brought about by major personnel changes, with doctors variously leaving, joining and on long-term leave. The situation had stabilized in the months before the study; one practice member said she saw the referrals meetings as a good opportunity for the practice to ‘regroup’. This practice had a complex system of practice meetings with clear areas of responsibility. Decision making appeared to us to be linked to three main factors: the structure and types of the various meetings; the communication between the senior practice members responsible for decision making (doctors and practice manager); and the ways in which decisions were made and carried out. Each senior practice member had differing views on the ways in which each of these factors operated; the ‘reality’ of the situation differed between them (Fig. 1
).
|
External local health service context
At the time of the study, 85% of the GPs within the Health Authority in which the practice is sited were fundholding; this resulted in difficulties in ensuring co-ordinated long-term commissioning of hospital and community services in the local area. One large local district general hospital was in danger of closure. The next general election would be held within a year; the future of the fundholding system was uncertain and the nature of any new systems which might replace it were as yet unknown.
Design
The study design is detailed in Figures 2 and 3. Preliminary interviews ‘sensitized’ GR to the practice and allowed the doctors and manager to give their views on how the practice functioned. These views and GR's knowledge of the local and national NHS environment formed the basis of our interpretation of the social location of the practice in terms of its past history, current context and future plans, both as an organization in its own right and as part of the local and national NHS.
|
|
Referrals meetings
The intervention required the practice GPs and manager to hold a series of four self-directive Postgraduate Education Allowance (PGEA)-approved ‘referrals meetings’. The practice were asked to leave at least 1 week between each meeting. GR attended these meetings as a non-participant observer. The doctors were asked to bring copies of all the routine secondary care referrals they had made during the week before the meeting and discuss the referrals and any issues arising from them. The meetings differed from the usual meetings held in the practice which were well structured and specific; secondary care referrals had not been discussed prior to this project. Given that the GPs were used to working together as a group, we believe that the results emanating from the meetings reflect the subject matter rather than the effect of the group meeting together.
Preliminary analysis
GR and SW performed content and group dynamic analyses separately and compared findings after each meeting.
Content analysis.. The factors relating to referral decision making discussed by the group were noted. The framework used was based on one devised by King et al.3 during a qualitative project looking at individual doctor's decision-making processes. As the analysis progressed, we found additional factors emerging from the data. The framework therefore was expanded to allow these to be included. These additions were modified and refined inductively as the meetings progressed. The framework enabled us to code small (30 second) units of data to give a detailed description of the contents and interactive quality of the meetings. We then identified ‘kinds of interaction’ occurring in the meetings and categorized the units of data within five ‘kinds of interaction’ groups. After the analysis of the final meeting, all four meetings were re-analysed using the expanded framework.
Group dynamics.. GR and SW wrote initial reports. We looked at how the group and the members within it behaved and interacted in each meeting. As with the content analysis, we did this separately and then compared our findings, leading to a consensus in interpretation.
Development of the preliminary theory
Our data collection, preliminary analysis and feedback meetings produced a large volume of data which illustrated different aspects of the practice and the impact of the referrals meetings. Different analytic techniques which we used to view the data from different perspectives had produced apparently contradictory results (see below). We used dimensional analysis to utilize the data to formulate a preliminary theory to explain the outcomes we had observed. Dimensional analysis is an analytic technique developed from grounded theory13 by Schatzmann.14 Using the technique requires the researchers to identify themes arising from the data and use them as ‘dimensions’ through which to view the data. Dimensions are abstract concepts that provide parameters for the purpose of describing the data. Examples in this practice were ‘views on good practice’, ‘development of clinical care’ and ‘communication and relationships’. The researchers jointly negotiated which dimension produced the most coherent and cogent picture of the impact of the intervention on the practice (the ‘best fit’).14 This dimension then became the key perspective from which the rest of the data was viewed and organized.
Practice feedback on the meetings and the preliminary analysis
The views of the participants on the meetings and the developing analysis were collected at several points in the study. This was done by means of participant diaries completed individually after each meeting, PGEA evaluation forms after the final meeting and a focus group 2 months later. At the focus group, the preliminary analysis was fed back to the practice for comments and discussion, and the learning needs identified through the evaluation forms were discussed.
Main outcome measures
- (i) Description of the context of the study;
- (ii) description of the in-practice meetings and the outcomes arising from them in the study practice;
- (iii) description of the individual and group learning needs identified by the participants;
- (iv) comparison of the impact of the referrals meetings on the practice referral rates with those found in the parallel RCT (see below);
- (v) comparison of the assessment by the participants of the value of the meetings with those found in the parallel RCT (see below); and
- (iv) a preliminary theory to explain the outcomes of the referrals meetings in this practice.
- (ii) description of the in-practice meetings and the outcomes arising from them in the study practice;
Comparison of quantitative outcomes with the RCT Doctors' evaluation of the value of the meetings. . The doctors gave a mean evaluation score of 8.67 (maximum score 10: 0 = no value, 10 = great value). This compares with a mean evaluation score of 6.62 from the intervention group practices in the RCT.
Effect of the meetings on practice referral rates.. The practice referral rate in the 16 weeks after the referrals meetings was 0.17 referrals/100 consultations more than in the 16 weeks before the meetings. This compares with a referral rate change of –0.08 seen in the intervention group in the RCT.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Content analysis
The time spent by the group discussing factors from the original framework devised by King et al.3 and the five other ‘kinds of interaction’ groups we induced are shown in Figure 4
. Well over half the time in each meeting was spent exchanging information on clinical and administrative issues. As the proportion of time spent on areas other than ‘exchange of information’ was so small, we have shown this using a logarithmic scale. We saw remarkable consistency in the use of the meeting time to share information. There was relatively little time spent discussing factors from the other categories; we thus felt unable to interpret this further.
|
Group analytic analysis
In contrast to the consistency found by content analysis, the group analytic analysis showed tremendous differences between the meetings as the intervention progressed. We felt that the ways in which the group interacted changed and developed. The following abbreviations are used below: P, participant; M, meeting number; page followed by paragraph (of audiotape transcript), e.g. p4:6 = page 4 paragraph 6; VC, video counter: time (minutes) from start of meeting on videotape; and FG, focus group to discuss initial analysis.
Meeting 1.. In the first meeting, the group showed high anxiety about the idea of presenting cases, exposing their clinical behaviour to their peers and about challenging each others' clinical decisions. SW was particularly struck by the tension evident within the group "(P1's) explanation (also) had a ring of defensiveness about it" (SW analysis M1 p3:4) "It seemed to me that the group members were indirectly voicing their own anxieties" (SW analysis M1 p3:6) and wondered whether a series of ‘horror stories’ about awful things that had happened to patients reflected this. Tension was evident in the participants' body language (VC 20.00–21.00, 26.00–28.00 and 35.00–36.00) and in the participant diaries:
- Q: What was difficult about today's meeting?
- A: (P1 ): "Initial feelings that referrals might be thought unnecessary."
- (P2): "I . . . was aware of my high referral rate compared to others."
Meeting 2.. This was more relaxed (‘good co-operative atmosphere’; doctor participant diary). We noted that there was still anxiety in this meeting and related it to the first examples of constructive criticism of each other's referral decisions. In this meeting, there was a move towards group cohesion: when one doctor became anxious, she was supported by two other doctors as evidenced by body language and verbal support "it's always difficult with (ophthalmology) isn't it". We noted that this meeting was chaotic, with participants arriving late, leaving early and entering and leaving the meeting at intervals. We interpreted this as an unconcious fear of change (SW).
Meetings 3 and 4.. In meetings 3 and 4, the group moved on to more mature and creative exchanges such as exploring the effect of involving patients in the referral decision-making process. The group also showed evidence of ‘individuation’15 as participants felt able to express their individuality in terms of their referral behaviour whilst remaining a cohesive group. Doctors showed openness and vulnerability together with a willingness to involve each other in discussions about the referral decision-making process. One doctor was open about his feelings of uncertainty (about the appropriateness of a referral) but made it clear that he wanted advice from the group before taking the next step (SW analysis).
Practice feedback meetings.. The participants accepted the research team's initial analysis although one doctor felt that whether or not the meetings were relaxed and open related more to how doctors felt on a particular day (than on the progression of the meetings).
Participants' identification of individual and group learning needs.. These fell into two main categories; information sharing and personal and group development.
- (i) Information sharing. Participants' valued the opportunity to share information and learn from each other in areas of both clinical expertise "having core time for clinical discussions is refreshing" (FG p1:11) and administrative information "I'm still learning about where (other doctors) send (patients) . . . I think (that) information . . . is useful". They also recognized areas where they as a group lacked expertise and needed outside teaching, "at one of the meetings . . . everyone was uncertain about knees . . . we will get someone in to talk to us about (knee problems)" FG p6:5. They also wanted to learn more about the clinical outcomes of their referrals, "it's more difficult to pick up the outcomes of the referral . . . some aspects of the outcome would be helpful . . .".
- (ii) Personal and team development. Several of the doctors commented on the value of small group work, "better work is done (in meetings like these)" (FG p5:6) and the self-reflective learning engendered during the meetings "it starts the process of you being more critical about your own criteria for referrals . . . ". (FG p6:7). They also recognized the value of team support in difficult clinical decisions "it's . . . supportive . . . to talk about cases and not always be working in isolation . . .". (FG p2:4).
- (ii) Personal and team development. Several of the doctors commented on the value of small group work, "better work is done (in meetings like these)" (FG p5:6) and the self-reflective learning engendered during the meetings "it starts the process of you being more critical about your own criteria for referrals . . . ". (FG p6:7). They also recognized the value of team support in difficult clinical decisions "it's . . . supportive . . . to talk about cases and not always be working in isolation . . .". (FG p2:4).
Development of the preliminary theory
We identified the key dimension in this practice as ‘communication and relationships’. This arose as an issue during the preliminary interviews when it became apparent that for each of the participants the ‘reality’ of communication, relationships and decision making differed (Fig. 1
). As the meetings progressed, the doctors explored their relationships and communication with each other and with their local specialist services. They also explored the ways in which specialists communicate and relate to each other and to GPs. Our choice of this as our ‘key dimension’ is supported by the enhanced openness and individuation seen within the group as the intervention progressed and in the fact that the con-tent analysis (Fig. 4) showed that most of the time in each meeting was spent communicating and sharing information.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This study only took place in one practice. Consequently we do not suggest that our findings can be generalized without further work in more practices. Our purpose in describing this work is to illustrate several important issues that arose in conducting the study which, we believe, may be witnessed by others researching in the complex area of changing behaviour through self-directed learning.
The hypothesis tested in our complementary study was based on previous research. This had shown that small group work involving doctors from different practices using a self-directive educational intervention had a measurable effect on the quality and cost of GP prescribing.8 In this study, we tested the effect of a similar intervention on GP referral behaviour. Formulation of our hypothesis required two assumptions. The first of these was that prescribing and referral decisions share common influences and will share common responses to interventions. In this practice, however, the GPs showed very personal styles of referral behaviour that they carefully defended. The second assumption inherent in the hypothesis was that applying an educational intervention to practices rather than to groups of doctors from different practices brought together for the purpose of the meetings would not alter its impact. Given the findings of this study, we feel that this is very unlikely to be true. The participants from this practice were a longstanding and mature group; it would seem reasonable to assume that they would benefit from self-directed learning in the way that the previous research suggests, namely that such learning would generate some form of standardized behaviour or at least some demonstrable change in behaviour. What we found, however, was that the practice as a complex organization was influenced by shared internal and external factors both past and present. The practice was not just a collection of individuals; the group was more than the ‘sum of the parts’.16 It is reasonable to assume that each practice will be unique, with specific characteristics and issues. The self-directive nature of the intervention enabled the group to use the meetings to explore issues of concern to them. Jetten17 has shown that well-established groups who are confident in their group identity will tend to accept differences in group behaviour rather than negotiate a new group norm. The reasons behind the ‘negative’ findings of the complementary RCT can be summarized by the comments of one of the participants during a feedback session when he said ". . . we agreed to differ . . .".
Despite the recognition by the participants that their rates of referral to hospitals were unlikely to change, the value of such small group learning, well known from previous work, was highlighted by the group. They saw the value of pooling information and expertise within their group and highlighted areas where they needed outside help to improve their knowledge. The value of a supportive protected environment in discussing difficult decisions made in professional isolation was also recognized.
We believe that this finding is important in terms of both establishing what information is needed for change and how change can be enforced. The notion that change simply requires the acquisition of new knowledge through peer-generated education would appear too simplistic, particularly when applied to well-established groups such as general practices. Andragogic educational principles18 that form the foundation of GP education provide empowerment to the learner not imposition of external controls. Thus the use of such techniques to generate change will always be problematic if the learner (or in this case the group) is resistant to that change.
We would suggest that our conclusion has important implications for the government's new policy of practice-based professional education programmes,11 and the expectations associated with such a development. The change generated by educational initiatives based in practices may lie in areas or forms different from those expected. In this instance, education had an impact in that the participants changed their understanding about referral practice and also modified their relationships with one another, but did not change their referral rates as measured in the RCT. In complex areas such as GP style, measurement of change may need to reflect a complex interaction of factors that are difficult to evaluate and quantify through the classic methodology of RCTs. The outcomes of such interventions are likely to be very valuable but difficult to measure and evaluate with currently available tools.
|
Acknowledgments |
|---|
Funding for the present study was provided by the NHS Executive South Thames Research and Development Directorate. Dr Rowlands is funded by the South Thames Research Network (STaRNet).
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1 Dowie R. General Practitioners and Consultants: A Study of Outpatient Referrals. London: King Edward's Hospital Fund for London, 1983.
2
Wilkin D, Smith A. Explaining variation in general practitioner referrals to hospital. Fam Pract 1987; 4: 160–169.
3
King N, Bailey J, Newton P. Analysing general practitioners' referral decisions 1. Developing an analytical framework. Fam Pract 1994; 11: 3–8.
4 Roland MO, Bartholomew J, Morrell DC, McDermott A, Paul E. Understanding hospital referral rates: a user's guide. Br Med J 1990; 301: 98–102.
5 Crombie DL, Fleming DM. General practitioner referrals to hospital: the financial implications of variability. Health Trends 1988; 20: 2053–2056.
6 Department of Health. The New NHS. Modern, Dependable. London: HMSO, 1997.
7
Majeed FA, Malcolm L. Unified budgets for primary care groups. Br Med J 1999; 318: 772–776.
8 Harris CM, Jarman B, Woodman E et al. Prescribing—A Suitable Case for Treatment. London: The Royal College of General Practitioners, 1984.
9 Rowlands GP. ‘Within practice’ review of referrals significantly alters referral rates—a pilot project. Br J Gen Pract 1996; 47: 124–125.
10
Davis DA, Thomson MA, Oxman AD, Haynes B. Changing physician performance. A systematic review of the effect of continuing medical education strategies. J Am Med Assoc 1995; 274: 700–705.
11 Chief Medical Officer. A Review of Continuing Professional Development in General Practice. London: HMSO, 1998.
12 Rowlands GP, Sims J, Kerry S. Minding the gap—small group work and its effect on GP referrals. Submitted.
13 Glaser BG, Strauss AL. The Discovery of Grounded Theory: Strategies for Qualitative Research. New York: Aldine, 1967.
14 Schatzman L, Maines DR (eds). Dimensional analysis: notes on an alternative approach to the grounding of theory in qualitative research. In Social Organization and Social Processes. New York: Aldine de Gruyter, 1991: 303–314.
15 Mahler M. On Human Symbiosis and the Vicissitudes of Individuation. USA: International Universities Press, 1968.
16 Foulkes SH. Therapeutic Group Analysis. London: George Allen and Unwin Ltd, 1964.
17 Jetten J, Spears R, Manstead ASR. Intergroup norms and intergroup discrimination. J Personality Soc Psychol 1996; 71: 1222–1233.[Web of Science][Medline]
18 Knowles M. The Modern Practice of Adult Education: From Pedagogy to Androgogy. New York, Cambridge Books, 1980.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  Y. J. F. M. Jansen, M. M. E. Foets, and A. A. d. Bont The contribution of qualitative research to the development of tailor-made community-based interventions in primary care: a review Eur J Public Health, June 26, 2009; (2009) ckp085v1. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||