Family Practice Vol. 21, No. 3, 290-298
Family Practice Vol. 21, No. 3 © Oxford University Press 2004, all rights reserved.
How research-conscious GPs make decisions about anticoagulation in patients with atrial fibrillation: a qualitative study
Westerhope Medical Group, 377, Stamfordham Road, Westerhope, Newcastle upon Tyne NE5 2LH and a School of Population and Health Sciences, Medical School, Newcastle upon Tyne, UK
E-mail: toby{at}tobylipm.demon.co.uk
Received 8 April 2003; Revised 12 November 2003; Accepted 7 January 2004.
Lipman T, Murtagh MJ and Thomson R. How research-conscious GPs make decisions about anticoagulation in patients with atrial fibrillation: a qualitative study. Family Practice 2004; 21: 290–298.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Background. Delays in the implementation of research findings have been of particular concern during the last decade. New findings, such as the use of warfarin to prevent stroke in patients with atrial fibrillation, have been found to be implemented patchily in clinical practice, in both primary and secondary care.
Objective. The pupose of the study was to explore how GPs with an active interest in research or evidence-based medicine (EBM) make decisions about anticoagulation in patients with atrial fibrillation.
Methods. Semi-structured interviews with GPs about their experiences in managing patients with atrial fibrillation were recorded on audio-tape, transcribed and analysed using the ‘Framework’ method. A constructivist approach was taken to analysis and interpretation.
Results. Eleven interviews were included in the analysis. Two key themes, ‘evidence’ and ‘professional role’, were identified. No two respondents had the same perception of the evidence, which was influenced by experience, attitudes and a variable knowledge of the literature. Recent publications about the effectiveness of aspirin compared with warfarin, and the publication Clinical Evidence were the most frequently mentioned sources of evidence. GPs with confidence in EBM skills described giving highly detailed explanations to patients and having a great commitment to shared decision making, even if this resulted in patients declining treatment. For this reason, they also expressed antagonism towards prescriptive clinical guidelines. Hospital doctors were seen as exerting a powerful influence on decisions, as being ‘disease-centred’, difficult to challenge and poor at communicating.
Conclusions. Decision making about anticoagulation is complex and is determined by a socially constructed view of the evidence strongly influenced by the GP's professional role.
Keywords. Anti-coagulation, atrial fibrillation, decision making, EBM, GPs.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Delays in the implementation of research have been of particular concern during the past decade.1 A good example is chronic atrial fibrillation, which carries an annual 3–6% risk of stroke, which is 5–7 times greater than that of controls with sinus rhythm.2 Since the late 1980s, evidence has accumulated that anticoagulation with warfarin produces substantial risk reduction both in patients with no previous history of stroke [odds ratio (OR) 0.47, 95% confidence interval (CI) 0.28–0.80],3 and in those with a previous history of stroke or transient ischaemic attack (TIA) (OR 0.36, 95% CI 0.22–0.58).4
Despite early advocacy by opinion leaders,5 there is evidence of reluctance to consider investigation and treatment for atrial fibrillation in general practice.6–8 Howitt and Armstrong9 found that patients' unwillingness to take warfarin was a major factor in limiting its use. Most studies of physicians' treatment decisions are based upon case vignettes, but a few have directly questioned physicians about their perceived barriers.10 A study of why GPs do not implement evidence11 used focus groups to identify barriers to its use. We decided to turn this question around and asked "how do GPs perceive evidence and integrate it into their practice?" rather than "what stops them using evidence?"
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
We took a constructivist approach, which sees knowledge as ‘created in interaction among investigator and respondents’ in order to achieve an understanding of the world of the person or group under study as they see it (the ‘emic’ view).12 In this case, the ‘world’ was that of GPs who were attempting to use research evidence in their practice and of the principle investigator (TL).
In order to maximize ‘information richness’,13 we purposively sampled14 GPs who were either active members of NoReN (the Northern Primary Care Research Network, i.e. they had conducted research and presented it at the NoReN Annual Research Presentation Day, November 2000) and/or had been participants or tutors at evidence-based practice workshops. GPs were identified from NoReN and Northern and Yorkshire Evidence-Based Practice Workshop databases, contacted by E-mail, given an outline of the proposed research and its background, and invited to contact TL if they were willing to be interviewed. It should be noted that TL would himself have been included in this sampling frame.
Respondents were invited to search their databases to identify patients15 and to use computerized or manual records as an aide-mémoire during the interviews. These were not to be seen by the interviewer, both to preserve confidentiality and to ensure that the respondents' views of the encounters were not influenced by possibly differing opinions of the interviewer. Other methods of triangulation were excluded for the same reason (in order to obtain a GP-centred view). TL carried out semi- structured interviews (Box 1), inviting respondents to describe encounters with patients with atrial fibrillation in which the decision of whether or not to prescribe warfarin was discussed, and to explore what influenced these decisions, including (but not exclusively) research evidence.
| BOX 1 Interview plan Could we start by you telling me about a consultation with a patient with atrial fibrillation that you remember particularly well (it doesn't matter if they were prescribed warfarin or not) Could you tell me about a(nother) very straightforward consultation, where you thought the patient should have warfarin and the patient readily agreed? Have you seen any patients who (situations not already covered)? How do you use evidence in practice? Do you talk about uncertainty with the patient? How much do the patient's own situation and needs affect your decisions? To what extent do you think you make your own assessment of these as against the patient's expressed wishes? How do you deal with a situation where you don't agree with the patient's decision? Could you give an example of a particular patient? Is there anything else you'd like to mention?
|
Interviews were recorded on audio-tape and fully transcribed. Transcripts were reviewed and corrected by TL usually within a week of the interviews, and preliminary notes were made as a running check on the emergence of new themes until ‘saturation’ was achieved, i.e. no new themes were emerging and old themes were being repeated. Analysis was based on the ‘Framework’ method (Box 2)16 and done by TL with regular reflexive discussion with MM throughout the interpretive process. We purposely delayed coding, indexing, charting and mapping until the interviews were complete in order to minimize the potential influence of one respondent on another through the interviewer/analyst. We use Potter and Wetherell's criteria for validity in qualitative research: (i) reflexive analysis, by explicitly encouraging challenges to the developing interpretation through the involvement of more than one researcher in the analysis; and (ii) plausibility, by presenting substantial extracts of the data the reader can assess the plausibility of the interpretation.17
| BOX 2 Analysis and interpretation using the Framework method Familiarization (listening to tapes and reading and re-reading transcripts) Identification and coding of themes, categories and subcategories, and construction of a thematic framework Indexing transcripts by electronically applying codes to the transcripts using Qualitative Data Analysis software (N.Vivo v1.1) Charting (lifting themes from their original context and rearranging them according to the appropriate thematic reference) Mapping and interpretation (re-examining the indexed and charted data, transcripts and memos made throughout the study to bring them together in a coherent interpretation of the data) From Ritchie and Spencer, 1994.16
|
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Respondents
Seventeen GPs were contacted, of whom 12 were interviewed. One interview was discarded because of poor recording quality. Two respondents were female and the ages ranged from 33 to 55 years, with a median age of 43. Five worked full-time (nine clinical sessions), three worked seven sessions and the rest six, five, three and two sessions, respectively. Two had been tutors at evidence-based medicine (EBM) workshops and six others had some formal EBM or critical appraisal training. Seven had formal academic appointments, and 11 were research active at the time of the interviews.
Content
Respondents gave detailed accounts of a large number of factors that they perceived to have influenced decisions about their own patients, in addition to how they saw the evidence. The thematic framework that emerged (Table 1) groups discussion of evidence under the broad theme ‘evidence’ and the other matters under the theme ‘professional role’. The categories are similar to those identified by Freeman and Sweeney.11
|
Evidence
Nine GPs referred to identifiable papers or articles, sometimes discussing them in depth. Four used the publication Clinical Evidence18 to set standards or as a basis for in-practice education. All were aware of the debate about the merits of warfarin compared with aspirin and perceived that the evidence was constantly changing. Five cited a systematic review that had been published within 3 months of the interviews.19 Of these, two stated that warfarin was definitely superior to aspirin, one that aspirin was ‘better than nothing’ and two that they now doubted the superiority of warfarin.
"I think that the recent leader in the BMJ that was sort of saying ‘hang on, hang on, you know we are pushing this too hard is aspirin just as good’ unnerves one after you've spent three years, I would think, moving towards ‘right we are going to warfarinise them’ erm ... that bothers me." (GP6)
While some expressed anxiety about the changing evidence as in the above extract, others expressed relief that giving aspirin rather than warfarin could now be justified:
"... it maybe makes me think ‘Well okay if somebody doesn't want warfarin we can give them aspirin’ so maybe it does colour me in thinking it's not the end of the world if you don't go on to warfarin." (GP10)
Applying evidence
Most respondents discussed the difficulty of applying evidence from clinical trials in primary care, partly because of their inclusion criteria.
"... if someone comes to you with atrial fibrillation you want to know, if he's the average man in the street, what am I best to treat him with and that's ... that's not answered by studies that have 80% exclusion rates." (GP4)
There was some ambiguity in the approach to guidelines. Three respondents were broadly in favour of guidelines (GP5, below), one applied them ‘flexibly’ and three were largely hostile, (GP2, below).
"... you can look at the British Heart Foundation guidelines and say ‘Well aged over 75, erm. hypertension, heart failure, these are the people that might benefit’ you know their risk of stroke is going to be quite high and should they not be anticoagulated?" (GP5)"... the problem that guidelines to me, they are very dichotomous if you like, they're saying ‘this blood pressure is 165, he should be on warfarin’, ‘It's 155, he shouldn't be on warfarin’ you know there's got to be a cut off." (GP2)
The Decision Analysis in Routine Treatment Study (DARTS)20 used a computer program to calculate individual risks of patients with atrial fibrillation and incorporate their values in a decision analysis during the consultation. Three respondents had been recruited for the study and found it helpful.
EBM skills
Confidence in using the skills of EBM varied from, at the most basic level, an awareness that ‘this is what the evidence shows’ to the ability to carry out Medline searches, critically appraise papers and calculate number needed to treat (NNT). Five respondents mentioned aspects of EBM skills, including that the concepts were sometimes ‘difficult’ (GP4, below), but none described the full EBM process of question forming, searching for information and applying it to individuals.
"... I still have difficulty with trying to picture what an NNT means to me in practice ... I know that people view sort of up to sort of NNTs of 20 and 30 as being quite important but I don't know why it is that much, I can't put that into context." (GP4)
Interacting with patients
Two GPs described minimal explanation of risk and benefit to patients (GP10, below), five gave some detail and three described detailed and explicit explanation (GP9, below).
"... I don't use risk tools, facts and figures it's a ... more of a generality of erm.. ‘It's just going to reduce your risk of stroke and it should prolong your life’ and erm ... .... you know maybe I shouldn't but I don't have ... I don't use NNTs [number needed to treat]" (GP10)"... so we have had a discussion about risk and I quoted an NNT of 7 in patients with ... in a high risk group over a year to prevent one cerebrovascular event." (GP9)
They also talked about explaining the hazards of the treatment.
The idea that the patient's idea, concerns and expectations should be a central concern of consultations in general practice21,22 was expressed in one form or another by all the respondents as patient-centred consulting, as opposed to concentrating on the GP's perspective.
"I think it's a personal decision rather than a right decision ... that they have to throw into the pot ... their life and how they feel about life, how they feel about death, how they feel about illness and all those sort of things to actually throw that into the equation, just the fact that you can prevent something doesn't mean that you should ... " (GP8)
Shared decision making23 was described by all except one GP.
"I like to advise ... identify what the patient thinks they need, what I think they should have, and then if it's acceptable we come to an agreement and we try to take it forward, it's negotiation, try more and more nowadays to do that in a consultation ... " (GP1)
Some GPs expressed equanimity with regard to patients who declined the offer of warfarin.
"I've had one patient who has refused to go on warfarin ... because he doesn't want to take it, he doesn't want to have that commitment to taking medication, to being monitored you know to possibly having the side effects that he might have, even though he is aware of the potential benefits, now if that's ... if that's his decision that's fine by me, it's not my life, I'm not taking warfarin for the rest of my life you know." (GP11)
In contrast, others described trying to persuade patients to take warfarin and a sense of failure if they thought the patient had made the wrong decision.
"Well it's my fault for putting it in a way that ... that perhaps didn't put the benefits and harms quite in context. I've made a mistake in how I've described the risk for that individual patient because they made a decision which I think is probably the wrong decision ..." (GP4)
All GPs recognized that patients might decline warfarin because of strongly held personal beliefs. Such beliefs included the identification of warfarin with ‘rat poison’ (GP1, below) mentioned by six respondents, dislike of medication (four respondents) and toleration of risk (three respondents).
" ‘Hang on doctor’ he said ‘That's the stuff they use to poison rats with isn't it’, I said ‘That's ... well it is and how do you ... ’ and he said ‘Oh I was a rat catcher for the council, I don't want that because I've seen too many things happen to rats with that’." (GP1)
The influence of hospital doctors
The power of hospital doctors was mentioned by all but three of the respondents. Four said that GPs were reluctant to challenge hospital doctors' judgement (GP5, below), four that patients believed hospital doctors more than GPs, and six that it was usually hospitals that initiated warfarin.
"... the one thing that surprised me looking through this case note review is just how big an influence secondary care is still having on the decision as to what people have, whether they get anticoagulated or not and how difficult it is in practice to change that, the patient perceptions and how we fear we may be held liable if we reverse the decision." (GP5)
Nine respondents discussed referral to hospital. Two referred routinely for anticoagulation, two referred to support their recommendation and four referred complex cases, or for cardioversion. Five respondents commented that they saw hospital doctors as treating the disease for its own sake rather than taking into account the patient's values and needs. They presented this as incompatible with the general practice philosophy of patient-centred consulting.
Four respondents said that hospitals' communication with patients was poor (GP1, below), four that it was poor with the practice (GP7) and two that it was generally good.
"... when he did come back to see us he did ask questions about would this be for life er ... would he notice any difference, what are the risks, which made me ... anxious that either the hospital had not explained the risk benefits or they had and he hadn't understood them ..." (GP1)"... the main problem we have is the communication of patients coming out of hospital and what the latest INR was and when it was checked and when it should be next checked ... and it's improving, but still you know we get faxed through flimsy discharges that you can't read ... ‘Patient Warfarinised, latest INR 2.3’ or whatever and that's it or ‘Patient Warfarinised please re-check’." (GP7)
Safety and harm
There were a number of accounts relating to the safety of warfarin, and the potential for harm if monitoring systems broke down, or the patient was not competent to manage their medication. Three respondents described critical incidents resulting from one or both of these problems, including a death. Six respondents who described practice- or community-based International Normalized Ratio (INR) monitoring (pharmacist led in four cases) expressed low concern about safety. Invariably, those expressing high concern about safety lacked clear protocols or systems for initiation of anticoagulation and INR monitoring, often compounded by confusion about who was ultimately responsible
"Decision making for who goes on warfarin is taken often by one person, monitoring of warfarin is taken by another person and in our practice people are monitored in 5 different systems, alright and er ... ongoing responsibility for patient education is non-existent ... the potential risks of warfarin to me are so large in terms of errors basically." (GP2)
Eight GPs discussed patients' competence to manage warfarin, and their own responsibility for the consequences, of whom six were moderately or highly concerned about it.
"... if I do give him a stroke or make him ... have to send him up to have haematuria investigated I will know that's the direct result of what I did and that weighs very heavily with me and I think it weighs very heavily with a lot of other ... other doctors." (GP12)
Partners
Nine respondents mentioned interactions with their partners. They saw themselves as innovators and teachers, and occasionally expressed anxiety about how their partners would view their activities (GP1, below).
"... what would my other partners think and say about me anticoagulating patients but in fact they were all very supportive and realised it was the way forwards ..." (GP1)
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This study demonstrates that the process of making decisions about anticoagulation, whilst including research evidence, is strongly influenced by many other factors which also influence GPs' perception and interpretation of the research itself. Decisions emerge from a complex social process only partly influenced by a rational or objective evaluation of the risks and benefits of treatment.
Respondents' perceptions of the interviewer influence interview interactions, and this needs to be considered carefully in interpreting interview data.23 In this study, the interviewer was a member of the same professional group as the respondents and, as an EBM tutor, would have been included in the same sampling frame. Shared assumptions may have led to some issues not being explored. For example, GP1 and GP9 (both EBM tutors) said very little about evidence-based practice. It is possible that they might have explained more to an interviewer whom they perceived did not share their experience of evidence-based practice. Conversely, others might not have mentioned odds, risk and NNT if they had been interviewed by a social scientist whom they might have expected to be neither familiar with nor interested in these concepts. As a final point, the analysis and interpretation of the data must have been influenced by the researcher's own interest in EBM.
A recent theoretical model of an evidence-based clinical decision is shown in Figure 1.24 Here ‘clinical expertise’ is seen as the skill of bringing together consideration of the patient's clinical circumstances, evidence of effectiveness regarding management options, and the patient's preferences and likely actions.
|
This study showed the decision-making process as being determined by the interaction of three actors–GPs, hospital doctors and patients–and influenced to a greater or lesser extent by the evidence, which is seen as being mediated through doctors. The weight exerted on the decision by each actor varied, and there is an almost infinite number of pathways through the process, which is represented in diagrammatic form in Figure 2.
|
A common perception was that GPs tried to explain and negotiate with patients, while hospital doctors tended to impose a decision ‘treating the disease rather than the patient’.
"... the idea that there is a treatment of a condition therefore you must treat patients with this condition with this treatment is what a lot of hospital doctors think ... whereas I think in general practice people often are prepared to take a step back and say ‘Well yes, there is a treatment there but is it in the best interests of this patient, is it going to fit in with the rest of you know ...’ it's the holistic view ..." (GP11)
However, the picture is more complicated than this. Figure 2 shows GPs influencing patients by a combination of persuasion, shared decision making, GP- or patient-centred consulting, and explanation, with the GPs' input being modified by their view of the evidence. The weight and quality of each component varied from one GP to another (and possibly from one consultation to another). Some GPs described quite strong levels of persuasion while others were content to explain their view of the risks and benefits and assist the patient to choose. Equally, while some GPs described negotiating the decision, others seemed to be describing exerting pressure on patients to accept responsibility for the decision or, conversely, to agree to the choice the GP thought most beneficial. This process could be rendered irrelevant if a patient refused the treatment because of strongly held personal beliefs (such as the former rat catcher) or accepted it because a hospital doctor had already made the decision.
GPs' views on the ‘right’ decision were influenced only in part by the evidence and the patient's situation. The practice and opinion of hospital doctors and their colleagues, the availability of reliable INR monitoring, their experience (or in most cases lack of it) of a critical event, and their view of the patient's competence and their own responsibility for possible adverse events all influenced the view they expressed. The formation of a view on the evidence itself is illustrated in Figure 3.
|
No two GPs used exactly the same combination of sources of evidence, but even when they referred to the same sources, their interpretations varied. For example. the Taylor et al. review19 was claimed to support the use of aspirin rather than warfarin by some, but others stated that it had not changed their view that warfarin was more effective. None of the respondents who cited it had picked up on the controversy it caused in academic circles, nor the reasons for it (disputed inclusion criteria25), and none had formally critically appraised it. The common factor among those who were persuaded of the effectiveness of aspirin was that they had experienced serious critical events (including the death of a patient) and lacked confidence in the INR monitoring systems available to them. It appeared that not only were clinical decisions influenced by the interactions illustrated in Figure 2 (we would have expected that), but interpretation of the evidence was also modified by them. Thus the rational process of EBM in Figure 1 was not found in real life general practice, even where GPs were motivated to use it.
This goes some way to refuting the darker prognostications of some commentators, who saw EBM as threatening the social and personal domains of general practice.26 On the contrary, in this study, those GPs with most experience of EBM were also most willing to engage with the agendas of patient-centred consulting and shared decision making, and were most relaxed if patients declined treatment, provided they were convinced that the patient understood the consequences of the decision. Although this might be seen as a good thing, it raised some potential conflicts. These GPs showed signs of challenging the expert and professional power of hospital doctors, and of rejecting the clinical effectiveness agenda of ‘treating the disease’. They were willing to question the applicability of guidelines to individual patients and believed that patients would sometimes accept higher levels of risk than suggested by guideline treatment thresholds, an insight that is confirmed by Devereux et al.27 These reactions are consistent with the idea that practising EBM raises the expert power of its practitioners28 but may conflict with performance management agendas, such as are suggested by the proposed new GP contract.29
This study examined the viewpoint of an atypical, one might say ‘vanguard’, group of GPs, some of whom are likely to represent as close an approximation to evidence- based practice as can be achieved in routine practice. How far the findings are transferable to typical GPs would need further research to explore, although it is likely that many of the influences, constraints and situations described are commonly experienced in general practice.
Conclusions
Implementing research evidence was more complex than is suggested by current models of EBM, not least because there was no consistent view of the evidence. Opinions about the evidence were influenced by GPs' experience and professional environment. The most important of these was the perceived safety and reliability of INR monitoring. Where a straightforward view that the evidence favoured warfarin was taken, respondents reported less sharing of decisions and more potential conflicts with patients. Where a more sophisticated, probabilistic view was taken (as with the EBM tutors and some others), GPs were more willing to share decisions and happier to accept refusal of treatment by patients, but this led them to challenge both guidelines and the views of hospital doctors.
|
Acknowledgments |
|---|
We thank the 12 GPs who gave generously of their time to prepare for and participate in the interviews, and Norma Cardill, who transcribed the tapes. TL's time was funded by an NHSE Northern & Yorkshire Research Training Fellowship. TL conceived the original idea for the research and developed the research question. RT advised on the background and evidence regarding atrial fibrillation and the prevention of stroke. MJM advised on qualitative methodology and theory. TL and MJM designed the qualitative research. TL carried out the interviews and analysis. MJM and RT were involved throughout the research in discussing and reviewing the data analysis. TL wrote the first draft of the paper and is the guarantor. MJM and RT critically reviewed the paper and suggested revisions. All authors approved the final version of the paper.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1 Haines A, Jones R. Implementing findings of research. Br Med J 1994; 308: 1488–1492.
2 Petersen P. Thromboembolic complications in atrial fibrillation. Stroke 1990; 21: 4–13.
3 Benavente O, Hart R, Koudstaal P, Laupacis A, McBride R. Oral anticoagulants for preventing stroke in patients with non-valvular atrial fibrillation and no previous history of stroke or transiant ischaemic attacks (Cochrane Review). The Cochrane Library 1999; issue 2. Oxford: Update Software Ltd.
4 Koudstaal PJ. Anticoagulants for preventing stroke in patients with nonrheumatic atrial fibrillation and a history of stroke or transient ischemic attacks (Cochrane Review). The Cochrane Library 1995; Oxford: Update Software Ltd.
5 Sweeney KG, Gray DP, Steele R, Evans P. Use of warfarin in non-rheumatic atrial fibrillation: a commentary from general practice. Br J Gen Pract 1995; 45: 153–158.[Web of Science][Medline]
6 Sudlow M, Rodgers H, Kenny RA, Thomson R. Population based study of use of anticoagulants among patients with atrial fibrillation in the community. Br Med J 1997; 314: 1529–1530.
7 Oswald N, Bateman H. Applying research evidence to individuals in primary care: a study using non-rheumatic atrial fibrillation. Fam Pract 1999; 16: 414–419.
8 Lip GY, Golding DJ, Nazir M, Beevers DG, Child DL, Fletcher RI. A survey of atrial fibrillation in general practice: the West Birmingham Atrial Fibrillation Project. Br J Gen Pract 1997; 47: 285–289.[Web of Science][Medline]
9 Howitt A, Armstrong D. Implementing evidence based medicine in general practice: audit and qualitative study of antithrombotic treatment for atrial fibrillation. Br Med J 1999; 318: 1324–1327.
10 Bungard TJ, Ghali WA, Koon KT, McAlister FA, Tsuyuki RT. Why do patients with atrial fibrillation not receive warfarin? Arch Intern Med 2000; 160: 41–46.
11 Freeman AC, Sweeney K. Why general practitioners do not implement evidence: qualitative study. Br Med J 2001; 323: 1100.
12 Guba EG, Lincoln YS. Competing paradigms in qualitative research. In: Denzin NK, Lincoln YS (eds) Handbook of Qualitative Research. Thousand Oaks (CA): Sage; 1994: 105–117.
13 Kuzel AJ. Sampling in qualitative enquiry. In: Crabtree BF, Miller WL (eds) Doing Qualitative Research. Newbury Park: Sage Publications; 1992: 31–44.
14 Marshall MN. Sampling for qualitative research. Fam Pract 1996; 13: 522–535.
15 Sudlow M, Rodgers H, Kenny RA, Thomson R. Identification of patients with atrial fibrillation in general practice: a study of screening methods. Br Med J 1998; 317: 327–328.
16 Ritchie J, Spencer L. Qualitative data analysis for applied policy research. In: Bryman A, Burgess R (eds) Analysing Qualitative Data. London: Routledge; 1994: 173–194.
17 Potter J, Wetherell M. Analysing discourse. In: Bryman A, Burgess B (eds) Analysing Qualitative Data. London: Routledge; 1994.
18 Barton S. Using clinical evidence. Br Med J 2001; 322: 503–504.
19 Taylor FC, Cohen H, Ebrahim S. Systematic review of long term anticoagulation or antiplatelet treatment in patients with non-rheumatic atrial fibrillation. Br Med J 2001; 322: 321–326.
20 Thomson R, Parkin D, Eccles M, Sudlow M, Robinson A. Decision analysis and guidelines for anticoagulant therapy to prevent stroke in patients with atrial fibrillation. Lancet 2000; 355: 956–962.[CrossRef][Web of Science][Medline]
21 Pendleton D, Schofield T, Tate P, Havelock P. The Consultation: An Approach to Learning and Teaching. Oxford: Oxford University Press; 1984.
22 Neighbour R. The Inner Consultation. Lancaster: Kluwer Academic Publishers; 1987.
23 Richards H, Emslie C. The ‘doctor’ or the ‘girl from the University’? Considering the influence of professional roles on qualitative interviewing. Fam Pract 2000; 17: 71–75.
24 Haynes RB, Devereaux PJ, Guyatt GH. Physicians' and patients' choices in evidence-based practice. Br Med J 2002; 324: 1350.
25 Mant J, Fitzmaurice D, Murray E, Hobbs FDR, Lip GYH. Inclusion criteria determine results of review. Br Med J 2001; 323: 233.[Medline]
26 Sweeney K. How can evidence-based medicine help patients in general practice? [editorial]. Fam Pract 1996; 13: 489–490.
27 Devereaux PJ, Anderson DR, Gardner MJ et al. Differences between perspectives of physicians and patients on anticoagulation in patients with atrial fibrillation: observational study. Commentary: varied preferences reflect the reality of clinical practice. Br Med J 2001; 323: 1218.
28 Lipman T. Power and influence in clinical effectiveness and evidence-based medicine. Fam Pract 2000; 17: 557–563.
29 Anon. New GMS Contract: Investing in General Practice. London: General Practitiioners Committee, BMA. The NHS Confederation; 2003.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S DeWilde, I M Carey, C Emmas, N Richards, and D G Cook Trends in the prevalence of diagnosed atrial fibrillation, its treatment with anticoagulation and predictors of such treatment in UK primary care Heart, August 1, 2006; 92(8): 1064 - 1070. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. A. Lorenz, G. W. Ryan, S. C. Morton, K. S. Chan, S. Wang, and P. G. Shekelle A qualitative examination of primary care providers' and physician managers' uses and views of research evidence Int. J. Qual. Health Care, October 1, 2005; 17(5): 409 - 414. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||