European Journal of General Practice, 2014; Early Online: 1–7
Original Article
Strategies for diagnosing leg oedema in primary care: A qualitative study of GPs’ approaches
Eur J Gen Pract Downloaded from informahealthcare.com by University of Waterloo on 10/15/14 For personal use only.
Judith Diederich, Simone Hartel, Erika Baum & Stefan Bösner Department of Family Medicine, University of Marburg, Germany
KEY MESSAGES: · GPs use a broad variety of diagnostic approaches for patients with leg oedema. · Approaches can be grouped in ‘active’ and ‘passive’ behaviour. · Only a few of the GPs used clinical decision rules, and in this group some had strong reservations about the feasibility of these instruments.
ABSTRACT Background: The symptom leg oedema represents a broad range of possible underlying aetiologies. The background of leg oedema is multifactorial and usually the GP is the first contact point for patients presenting with this symptom. GPs rely on patient history and physical examination as their main diagnostic tools. Objective: To identify GPs’ diagnostic approaches and heuristics in patients presenting with leg oedema. Methods: Interviews with 15 GPs (20–30 min) using a semi-structured interview-guideline were conducted. GPs described their individual diagnostic strategies concerning all patients presenting with leg oedema they had prospectively identified during the previous four weeks. Interviews were taped and transcribed verbatim. Qualitative analysis was conducted by two independent raters. Results: GPs applied a variety of diagnostic approaches, which can be grouped in active and passive strategies. Active strategies comprised the use of decision rules and guidelines, Bayesian arguing, problem dichotomisation and discrepancy heuristics. Passive approaches included test of time, therapy as diagnosis, and taking patient assumptions into account. Conclusion: When dealing with leg oedema, GPs use prior information of individual patients in a specific way. There is a broad variety of diagnostic approaches that can be grouped in ‘active’ and ‘passive’ behaviour. Approaches mostly match with established diagnostic strategies in primary care. Keywords: General practice/family medicine, diagnosis/diagnostic research, qualitative designs and methods
INTRODUCTION Leg oedema is a common symptom among the general population. The Bonn Vein Study of 2003 reports a history of leg swelling for every sixth man (16.2%) and almost every second woman (42.1%) among German adults (1). More recent data show that 1.5% of all women above age 65 suffer from leg disorders (2). Additionally, varicosis ranks twelfth among documented chronic diagnoses of 4309 reasons for encounter in 34 (German) practices and represents 2.8% of all documented permanent diagnoses (3). Consequently,
leg oedema is a frequent cause of consultation in primary care (4). Leg oedema often has a multifactorial origin and general practitioners (GPs) face a broad range of underlying conditions (5). From the patient ’s perspective, a timely and precise diagnosis and treatment of the underlying aetiology is desirable as it may help to prevent an adverse course of chronic diseases such as heart failure, pulmonary hypertension, post thrombotic syndrome, kidney insufficiency, or lymphoedema. Additionally, GPs have to rule out potentially harmful or
Correspondence: Stefan Bösner, Department of Family Medicine, Karl-von-Frisch-Strasse 4, 35043 Marburg, Germany. Tel: ⫹ 49 (0)6421 28 65122. Fax: ⫹ 49 (0)6421 28 65121. E-mail: boesner@staff.uni-marburg.de (Received 18 July 2013; accepted 16 February 2014) ISSN 1381-4788 print/ISSN 1751-1402 online © 2014 Informa Healthcare DOI: 10.3109/13814788.2014.900535
Eur J Gen Pract Downloaded from informahealthcare.com by University of Waterloo on 10/15/14 For personal use only.
2
J. Diederich et al.
even life-threatening courses of disease such as acute heart failure, deep vein thrombosis or pulmonary embolism (6). For diagnosis in a primary care setting, patient history and physical examination remain the main diagnostic tools (4–6) and are embedded in several different diagnostic strategies (7). Besides classical approaches like spot diagnosis (7), pattern recognition (8) or Bayesian reasoning (9–11), GPs use also intuition (12) and gut feeling (13,14) in their clinical decision making. Simple heuristics (15), like the heuristic of discrepancy (16) or taking patients’ assumptions into account (17) are used as well. Aim is to investigate whether these or any other strategies are used in the further diagnostic work-up of patients presenting with unilateral or bilateral leg oedema in primary care.
METHODS Study design A qualitative research approach addressing our study question was used, to explore strategies and heuristics applied by GPs in a broad and open way. Single interviews were selected as the most appropriate technique for data collection as it created the best environment to talk with each GP about individual diagnostic strategies. The Ethics Committee of the Faculty of Medicine, University of Marburg (AZ 193/09) approved the study. Setting and data collection Fifteen GPs in Hessen and Saxony (Germany) were approached. The sample was chosen to ensure inclusion of male and female GPs practicing in urban and rural areas. The same researcher interviewed the GPs between April and September 2010. It was explained to the GPs that we were interested in understanding how they diagnosed leg oedema, and what role intuition and clinical experience played in their daily routine with these patients. To elucidate their diagnostic strategies, we asked the participating GPs to collect information of every patient presenting with leg oedema in their practice during the following four weeks (a minimum of three patients). After these four weeks, a second appointment for the interview was planned. The collected patient data were exclusively used by the GPs to aid recall during the interview (‘stimulated recall,’ a method that is frequently applied for examining the clinical reasoning of physicians), but were not presented to the interviewer. All final diagnoses presented by the GPs during the interview were based on the GPs’ own diagnostic work up and were not externally verified (18). We started the semi-structured interview with a
question about the number and the diagnoses of the patients collected by the respective GP. Afterwards, GPs were invited to recall the consultation of these patients. A semi-structured interview guideline was used to cover all relevant topics such as history taking, examination and further diagnostic investigations, identification of red flags, GPs’ diagnostic strategies and individual heuristics, dealing with uncertainty. We asked about central aspects of clinical history and physical examination, the use of scores or any other (also individual) algorithms, which helped the GPs in the differential diagnosis of specific cases. Additionally, GPs thoughts about the role of intuition and gut feeling were investigated. GPs’ reflections on their diagnostic reasoning or individual diagnostic strategies concluded the interview. Data analysis The interviews were taped and transcribed verbatim. To assist data handling, the material was transcribed and analysed using MAXQDA-10 (19). Our analysis can be categorized as a thematic survey, undertaken stepwise with each step informing the next (20,21). The first step was to develop a coding system (coding-tree) (22). The coding tree and coding was tested and refined within our qualitative working group. All interviews were coded by two independent raters (JD and SH); differences in coding were resolved by discussion. Data analysis was guided by already known theories and conceptions of diagnostic decision making in primary care. We observed a certain information saturation effect around the 10th to 11th interview, when repeatedly similar answers were given as before. Since we had already contacted the respective GPs, we nevertheless completed all 15 interviews.
RESULTS Table 1 shows the characteristics of the 15 interviewed GPs. GPs encountered 187 patients presenting with leg oedema during the four weeks prior to the interview. Table 1. Characteristics of German GPs (n ⫽ 15), being interviewed on their diagnostic work-up of patients presenting with leg oedema. Characteristic Gender (GPs)
Number
Male Female Years of clinical experience ⬍ 10 10–20 ⬎ 20 Practice location Urban Rural
11 4 2 6 7 5 10
Diagnosing leg oedema in primary care They all gave very individual and specific accounts of their diagnostic reasoning. General aspects and the role of intuition
Eur J Gen Pract Downloaded from informahealthcare.com by University of Waterloo on 10/15/14 For personal use only.
All GPs differentiated in their diagnostic reasoning between unilateral and bilateral oedema. Patient history, known comorbidities and results of the physical examination were the key elements of the diagnostic work-up; further technical examinations played a secondary role. Most of the interviewed GPs confirmed that the combination of intuition and clinical experience was very important in their clinical decision making. There are immediately intuitive thoughts when seeing [a patient with] leg oedema […], there might be something wrong. (A13: 70–71) GPs considered the own ‘feeling […] , concerning the patient ’s problem […] and the [number] of collected diagnostic hints…’ as essential. This means that I sense the patient, I observe how he shakes my hand, how he enters the room, or how he breathes. (A12, 93–96) GPs’ diagnostic strategies could be grouped into two categories, ‘active strategies’ and ‘passive strategies’ (Table 2). Hereby our definition of ‘active’ and ‘passive’ strategies refers to the GPs’ general focus regarding the diagnostic process when using these strategies: whether they proceeded actively and used all possible clinical investigations to find out causes of oedema (‘active’), or whether they stayed more observant in relation to patients’ complaints (‘passive’). Active diagnostic strategies Taking prevalence of disease into account. Normally being the first contact point of a patient, GPs took the basic epidemiology of leg oedema into account in their diagnostic reasoning. Prevalence of possible underlying disease was used strategically by five of the fifteen GPs. Table 2. Strategies used by GPs in the diagnostic work-up of patients presenting with leg oedema. Active strategies Taking prevalence of disease into account Clinical decision rules (scores/algorithms) Bayes-Theorem (accumulating probabilities) Hypothetico-deductive method Dichotomizing problems Familiarity versus discrepancy heuristics
Passive strategies Watchful waiting/test of time Therapy as diagnostic tool Taking patient assumptions into account
3
[First] I look if I can find anything palpable […] then a ranking follows according to “what is common is common” and “what is rare is rare.” […] And then I try out a few things. (GP13, 63–64) Participating GPs mentioned that they considered diseases of high prevalence (for example, cardiac oedema, oedema due to venous insufficiency) more often as a first hypothesis than diseases of low prevalence (like lipoedema, lymphoedema or others). Of course it matters whether [the oedema] is unilateral or bilateral. Bilateral thrombosis for example is much rarer than unilateral thrombosis. (GP13, 23–23) Clinical decision rules (scores, algorithms) and individual diagnostic schemes. Seven out of fifteen interviewed GPs used clinical decision rules as a diagnostic tool, but partly had strong reservations about the feasibility of these instruments. Scores are always subjective! They should be objective, but scores are basically developed for the wrong clientele, and do not apply for half of my patients, so they [the scores] lose evidence. (GP2, 138–138) GPs instead used individual concepts of diagnostic reasoning, subject to a large heterogeneity. Most GPs’ methods could be characterized as recurrent procedures, which were based on history and clinical examination. […] in this case I look, examine, touch, feel! Afterwards I decide which additional investigations I need. (GP1, 49–49) Well, I actually always follow a fixed order. First of all, physical examination, [and] if I have any suspicion about the heart I always do an ECG. […] If I do not see anything […] I also order blood tests. […] I always proceed in the same order. […] Heart, liver, thyroid, varicose veins—those are my four main categories […]. (GP3, 59–63) Diagnostic procedures of some GPs did not follow any pre-structured algorithm or scheme. How I proceed differs [from case-to-case]. (GP15, 23–23) These GPs adapt their approaches individually to each patient, taking the overall picture of patient ’s complaints, history, symptoms and the first impression the patient makes into account. Then I ask myself: are there chronic diseases? Does this come from the heart? Has he [the patient] always had one swollen leg? If I know this is unilateral, this is acute, this has not been noticed before, then I investigate further. […] If nothing is known, then I take a closer look at the
4
J. Diederich et al. leg, to see whether it might be lymphoedema or an oedema because of venous insufficiency. (GP11, 10–12)
Eur J Gen Pract Downloaded from informahealthcare.com by University of Waterloo on 10/15/14 For personal use only.
That is, I perceive the patient: I see how he shakes hands or how he walks into the room, or how he wheezes. [All this while] knowing the results of previous examinations. (GP12, 95–96) Bayes-Theorem (accumulating probabilities). When trying to find out the cause of the symptom ‘leg oedema,’ GPs used various active strategies to rule out (or in) possible diagnoses. Eight of fifteen interviewed GPs accumulated probabilities, a strategy known as the ‘Bayes Theorem.’ In this strategy, GPs form an early hypothesis about the cause of the symptom. Each further component of the patient ’s history and the clinical examination results in a given (post-test) probability of the suspected diagnosis. These probabilities are then summed-up to lead to acceptance or rejection of the early hypothesis. The likelihood of thrombosis increases with the risk factors […]. Risk factors are confinement to bed, surgery, immobility, previous thrombosis, smoking, a positive family history […]. […] all of these factors have to be checked. And, […] whether it [e.g. the swelling] is unilateral or bilateral […]. (GP13, 23–23) Usually it is the sudden onset [of the symptoms] […], unilateral, and more pronounced […]—which leads to a strong suspicion of thrombosis. Also if there is knocking pain in the heels or pushing pain in the calves […], there is often a matching history, for example, “Yesterday we came back from Austria by bus!” (GP8, 22–24) Hypothetico-deductive method. In other instances, GPs considered more than one potential diagnosis as a first hypothesis at the beginning of the consultation. Further items of the patient ’s history and findings from the physical examination together with further tests made one final aetiology much more likely than other differential diagnoses. Ten of the fifteen interviewed GPs used this diagnostic approach. [One] woman [with leg oedema] said: “I have pain in my right leg.” […] Well, the first impression was: obesity, pressure on the spinal cord, symptoms like sciatica. […] And then I did a Doppler ultrasound scan on such an obese patient. She did not have typical pressure points for acute thrombosis. […] The Doppler ultrasound scan was not positive for thrombosis. After all, it was still sciatica. (GP6, 25–25). Dichotomizing problems. Four out of fifteen interviewed GPs assigned history and clinical examination
findings to one of two opposite groups, following a strategy of dichotomizing. This method was used by GPs applying a chain of questions that led them to a final diagnosis (‘decision tree’). GPs differentiated between unilateral and bilateral oedema; chronic and acute condition, and whether there is accompanying pain or not. Additionally […] [it is relevant] whether the oedema increase at certain times of day or not […],whether there the swelling on both sides is similar or whether there are differences between the left and the right side. (GP7, 20–20) Basically I first distinguish whether oedema is unilateral or bilateral. […]. For every leg oedema, I also have a look at the veins […] and palpate them. [I ask myself] whether I can trigger tenderness on palpation of typical locations […]? [...] I take a look at the consistency of the oedema. Is it more of a tense oedema […], what is the dimension of the oedema, is it limited to the lower leg or does it concern the toes too? This way you can distinguish between whether it is lymphoedema or a venous problem. (GP10, 25–26) Heuristic of discrepancy. GPs knew most of their patients for years and used this information for diagnostic purposes. Three of the fifteen GPs, mentioned that discrepancy in behaviour or differences in the appearance of patients in comparison to previous consultations influenced them in their diagnostic decision making. Since I know the patients, I know how they behave in a given situation. Now, if I have a patient whom I know as smart and happy […], [and] if this patient comes in here and has really severe pain, then I only need to briefly glance at the leg—which is indeed swollen—then I refer him right away. But if this patient is known for constant complaining, and comes because of every little thing, then you have got to look at this a little closer. Of course, this patient can be seriously sick at times! (GP8, 49–50) Passive diagnostic strategies Watchful waiting/test of time. Eight out of fifteen interviewed GPs used the strategy of watchful waiting as a defensive strategy in their diagnostic process. One of these GPs said: That always depends on how [the symptoms] develop, […] how much time I have. In general practice […] you have to do “watchful waiting.” […] I rule out serious causes first […]. (GP14, 64–64)
Diagnosing leg oedema in primary care So if a serious disease is seen as unlikely, GPs abstain from further diagnostic procedures and reschedule the patient at regular intervals. This [...] is our advantage as general practitioners: we know the patients well, and we have time. […] We say “Come again next week.” At the hospital, the physicians don’t have enough time. (GP15, 70–70)
Eur J Gen Pract Downloaded from informahealthcare.com by University of Waterloo on 10/15/14 For personal use only.
Therapy as diagnostic tool. Ten out of fifteen GPs used this strategy. If the chosen therapy did not have an effect, GPs proceeded to more specific diagnostic interventions. And because she [the patient] had a relatively high pulse rate, I gave her a low dose of a beta blocker to see [how it worked]. And then […] after two hours […] the rate was down […]. (GP15, 46–47) Taking patient assumptions into account. Two out of fifteen interviewed GPs, purposefully asked patients about their personal perceptions in regard to their symptoms. Patients were invited to share their own suspicions, thoughts and explanations. This information influenced GPs with respect to their hypothesis and the following diagnostic steps. She [the patient] had said before that she had the same (disease) as her mother, which had been diagnosed before […] and now she had gotten all these dents. […] Principally this was the whole history. She said: “ This is a lipoedema, my mother already had the same!” She [the patient] came with the diagnosis. (GP12, 32–40)
DISCUSSION Main findings We identified GPs’ diagnostic approaches and heuristics in patients presenting with leg oedema. GPs used a broad spectrum of approaches, which could be categorized into two major groups, namely active and passive approaches (Table 2). Hereby GPs used prior information of individual patients in a specific way. Strengths and limitations The data presented have limitations. A different way to capture doctors’ diagnostic reasoning, e.g. videotaped interviews or direct observation, may have more internal validity. Yet, these methods are difficult to implement in a primary care setting. Using standardized patients would have been a possible alternative, however, practically difficult to implement for the symptom of leg oedema. In addition, this approach would not have captured the
5
diagnostic significance of long-term relationships between GPs and their patients. Contrarily, the prospective identification of patients we used in the interview allowed us to examine the diagnostic process in a defined clinical situation. We cannot rule out that GPs reinterpreted their diagnostic arguments afterwards, and may also have tried to change first intuitional diagnostic hypotheses into more rational thoughts during the interview. We tried to minimize this kind of bias by encouraging GPs to give their very individual accounts of diagnostic methods and strategies, heuristics or personal gut feelings. We stressed the lack of primary care research in this area, which helped to create an atmosphere of open and self-critical reflection. In the interview, we emphasized the widely accepted use of simple heuristics and ‘rules of thumb’ adapted to the working environment in general practice (23). As a result, GPs expressed their own thoughts and also frankly mentioned difficulties and failure so that we regard the influence of social desirability bias as very small. The generalizability of the results might have been affected by the fact that this study was conducted in only one country with its specific health system. However, we do not expect that diagnostic behaviour of GPs differs much in other European countries. Comparison with existing literature This study shows that clinical decision making is not merely a process of rational decisions (12). The GPs mentioned the importance of intuition in their diagnostic reasoning in different ways. Several studies have examined the efficiency of intuitional decision making in general practice and other fields of medicine (7,13,24,25). However, for the intuitional side of the ‘cognitive continuum’ between intuition and rationality, it is still hard to provide evidence, and intuitional decision-making processes are still underestimated in practice (12). Different diagnostic strategies and heuristics, the latter being the skill of coming to a good solution within a short period and with limited knowledge (15), were very helpful in the diagnostic reasoning of GPs concerning patients presenting with leg oedema. We have divided the GPs’ behaviour into two categories: ‘active strategies’ and ‘passive strategies.’ Wübker et al., studied GPs’ strategies in handling patients with tinnitus and divided the results into similar categories (26). Other authors preferred to focus on a chronological sequence of different diagnostic steps (7,27). Heneghan et al. grouped strategies into a three-stage model: initiation of diagnostic hypotheses, refinement of the diagnostic hypotheses and defining the final diagnosis (7). Some of our interviewed GPs used hypotheticodeductive strategies in their diagnostic reasoning. This is in line with findings of other researchers, which confirm
Eur J Gen Pract Downloaded from informahealthcare.com by University of Waterloo on 10/15/14 For personal use only.
6
J. Diederich et al.
the use of this strategy for GPs’ diagnostic reasoning in regard to a broad range of symptoms (7,28). Other authors pointed out the direct connection between the GPs’ initial hypothesis and the assessment of setting specific pre-test probability (29). Schneider et al. showed how non-awareness of prevalence in general practice leads to wrong diagnostic decisions (9). Watchful waiting was an often mentioned diagnostic strategy in our sample. GPs used this strategy mainly in the context of chronic illnesses, having ruled out serious disease beforehand. Some other studies indicate watchful waiting as a common strategy applied in general practice (7,27,30). Prior knowledge about the patient was important for the use of person specific discrepancy. Other authors described this heuristic as an important diagnostic criterion for chronic heart disease (16) or meningococcal infection in children (24). This may indicate that this strategy is a diagnostic tool for potentially serious diseases, which does not require specialized knowledge of the particular disease of concern, but of in-depth knowledge of the affected patients (16). Implications for research and practice Not all of the diagnostic strategies and heuristics described may be diagnostically helpful for the further work-up of patients presenting with leg oedema. To investigate this question further, we intend to evaluate the diagnostic effectiveness of different strategies in a cross-sectional study with a delayed-type reference standard focussing on the sub group of patients with bilateral leg oedema. This study will shed further light on medical decision making in the complex setting of primary care. Conclusion When dealing with leg oedema, GPs decided in their diagnostic reasoning between unilateral and bilateral oedema with patient history, known comorbidities and results of the physical examination being the key elements of the diagnostic work up. GPs used a broad variety of diagnostic approaches that can be grouped in ‘active’ and ‘passive’ behaviour and that mostly match with established diagnostic strategies in primary care.
ACKNOWLEDGEMENTS The authors thank all GPs who were willing to be interviewed for this study. Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
REFERENCES 1. Rabe E, Pannier-Fischer F, Bromen K, Schuldt K, Stang A, Poncar C, et al. Bonner Venenstudie der Deutschen Gesellschaft für Phlebologie. Phlebologie 2003;32:1–14. 2. Kühlein T, Laux G, Gutscher A, Szecsenyi J. Kontinuierliche Morbiditätsregistrierung in der Hausarztpraxis. Vom Beratungsanlass zum Beratungsergebnis. München: Urban & Vogel, S; 2008. p. 48. 3. Abholz H, Donner-Banzhoff N. Epidemiologische und biostatistische Aspekte der Allgemeinmedizin. In: Michael M. Kochen, editor. Allgemeinmedizin und Familienmedizin, 4th ed. Stuttgart: Georg Thieme Verlag KG; 2012. pp. 507–21. 4. Rabady S. Beratungsanlass Beinschwelllung: Differentialdiagnose in der Allgemeinmedizin. Z Allg Med. 2012;10:414–9. 5. Blankfield RP, Finkelhor RS, Alexander JJ, Flocke SA, Maiocco J, Goodwin M, et al. Etiology and diagnosis of bilateral leg edema in primary care. Am J Med. 1998;105:192–7. 6. Ely JW, Osheroff JA, Chambliss ML, Ebell MH. Approach to leg edema of unclear etiology. J Am Board Fam Med. 2006;19:148–60. 7. Heneghan C, Glasziou P, Thompson M, Rose P, Balla J, Lasserson D, et al. Diagnostic strategies used in primary care. Br Med J. 2009;338:b946. 8. Glasziou P, Rose P, Heneghan C, Balla J. Diagnosis using ‘test of treatment ’. Br Med J. 2009;338:b1312. 9. Schneider A, Dinant GJ, Szecsenyi J. Stepwise diagnostic workup in general practice as a consequence of the Bayesian reasoning. Z Arztl Fortbild Qualitatssich. 2006;100:121–7. 10. Pewsner D, Bleuer JP, Bucher HC, Battaglia M, Jüni P, Egger M. Der Intuition auf der Spur? Das Bayes’sche Theorem und die Diagnostik in der Grundversorgung. Schweiz Med Forum. 2001; 3:41–52. 11. Doust J. Diagnosis in general practice. Using probabilistic reasoning. Br Med J. 2009;339:b3823. 12. Hamm RM. Clinical intuition and clinical analysis: Expertise and the cognitive continuum. In: Dowie J, Elstein A, editors. Professional judgement—a reader in clinical decision making. Cambridge: Cambridge University Press; 1988. pp. 81–5. 13. Van den Bruel A, Thompson M, Buntinx F, Mant D. Clinicians’ gut feeling about serious infections in children: Observational study. Br Med J. 2012;345:e6144. 14. Stolper E, Van de Wiel M, Van Royen P, Van Bokhoven M, Van der Weijden T, Dinant GJ. Gut feelings as a third track in general practitioners’ diagnostic reasoning. J Gen Intern Med. 2011;26:197–203. 15. Gigerenzer G, Todd PM. Simple heuristics that make us smart. New York: Oxford University Press; 1999. 16. Hani MA, Keller H, Vandenesch J, Sönnichsen AC, Griffiths F, Donner-Banzhoff N. Different from what the textbooks say: How GPs diagnose coronary heart disease. Fam Pract. 2007; 24:622–7. 17. Goyder C, McPherson A, Glasziou P. Diagnosis in general practice. Self diagnosis. Br Med J. 2009;339:b4418. 18. Elstein AS, Shulman LS, Sprafka SA. Medical problem solving: An analysis of clinical reasoning. Cambridge, Mass.: Harvard University Press; 1978. 19. MAXQDA 10. Berlin: VERBI Software. Consult. Sozialforschung GmbH; 2010. 20. Sandelowski M, Barroso J. Classifying the findings in qualitative studies. Qual Health Res. 2003;13:905–23. 21. Kuckartz U. Einführung in die computergestützte Analyse qualitativer Daten. 3rd ed. Wiesbaden: VS Verlag für Sozialwissenschaften; 2010. 22. Kuckartz U. Qualitative Evalation—Der Einstieg in die Praxis. Wiesbaden: VS Verlag; 2007. 23. André M, Borgquist L, Foldevi M, Mölstad S. Asking for ‘rules of thumb’: A way to discover tacit knowledge in general practice. Fam Pract. 2002;19:617–22.
Diagnosing leg oedema in primary care
Eur J Gen Pract Downloaded from informahealthcare.com by University of Waterloo on 10/15/14 For personal use only.
24. Granier S, Owen P, Pill R, Jacobson L. Recognising meningococcal disease in primary care: Qualitative study of how general practitioners process clinical and contextual information. Br Med J. 1998;316:276–9. 25. Zuhlke H. II. Intuition in surgery-Is intuition in surgery learnable? ZentralblChir. 1999;124:55–63. 26. Wübker S. Hausärztliche Behandlung des Tinnitus aurium aus der Perspektive von Patienten und Hausärzten. Doctoral thesis. University of Göttingen 2007. Available at http://www.ediss. uni-goettingen.de/bitstream/handle/11858/00-1735-0000-
27. 28. 29. 30.
7
0006-AF36-6/wuebker.pdf?sequence ⫽ 1 (accessed 11 November 2013). Almond SC, Summerton N. Diagnosis in general practice. Test of time. Br Med J. 2009;338:b1878. Norman G, Barraclough K, Dolovich L, Price D. Iterative diagnosis. Br Med J. 2009;339:b3490. Summerton N. Diagnostic testing: The importance of context. Br J Gen Pract. 2007;57:678–9. Nazareth I, King M. Decision making by general practitioners in diagnosis and management of lower urinary tract symptoms in women. Br Med J. 1993;306:1103–6.
Original Article
Strategies for diagnosing leg oedema in primary care: A qualitative study of GPs’ approaches
Eur J Gen Pract Downloaded from informahealthcare.com by University of Waterloo on 10/15/14 For personal use only.
Judith Diederich, Simone Hartel, Erika Baum & Stefan Bösner Department of Family Medicine, University of Marburg, Germany
KEY MESSAGES: · GPs use a broad variety of diagnostic approaches for patients with leg oedema. · Approaches can be grouped in ‘active’ and ‘passive’ behaviour. · Only a few of the GPs used clinical decision rules, and in this group some had strong reservations about the feasibility of these instruments.
ABSTRACT Background: The symptom leg oedema represents a broad range of possible underlying aetiologies. The background of leg oedema is multifactorial and usually the GP is the first contact point for patients presenting with this symptom. GPs rely on patient history and physical examination as their main diagnostic tools. Objective: To identify GPs’ diagnostic approaches and heuristics in patients presenting with leg oedema. Methods: Interviews with 15 GPs (20–30 min) using a semi-structured interview-guideline were conducted. GPs described their individual diagnostic strategies concerning all patients presenting with leg oedema they had prospectively identified during the previous four weeks. Interviews were taped and transcribed verbatim. Qualitative analysis was conducted by two independent raters. Results: GPs applied a variety of diagnostic approaches, which can be grouped in active and passive strategies. Active strategies comprised the use of decision rules and guidelines, Bayesian arguing, problem dichotomisation and discrepancy heuristics. Passive approaches included test of time, therapy as diagnosis, and taking patient assumptions into account. Conclusion: When dealing with leg oedema, GPs use prior information of individual patients in a specific way. There is a broad variety of diagnostic approaches that can be grouped in ‘active’ and ‘passive’ behaviour. Approaches mostly match with established diagnostic strategies in primary care. Keywords: General practice/family medicine, diagnosis/diagnostic research, qualitative designs and methods
INTRODUCTION Leg oedema is a common symptom among the general population. The Bonn Vein Study of 2003 reports a history of leg swelling for every sixth man (16.2%) and almost every second woman (42.1%) among German adults (1). More recent data show that 1.5% of all women above age 65 suffer from leg disorders (2). Additionally, varicosis ranks twelfth among documented chronic diagnoses of 4309 reasons for encounter in 34 (German) practices and represents 2.8% of all documented permanent diagnoses (3). Consequently,
leg oedema is a frequent cause of consultation in primary care (4). Leg oedema often has a multifactorial origin and general practitioners (GPs) face a broad range of underlying conditions (5). From the patient ’s perspective, a timely and precise diagnosis and treatment of the underlying aetiology is desirable as it may help to prevent an adverse course of chronic diseases such as heart failure, pulmonary hypertension, post thrombotic syndrome, kidney insufficiency, or lymphoedema. Additionally, GPs have to rule out potentially harmful or
Correspondence: Stefan Bösner, Department of Family Medicine, Karl-von-Frisch-Strasse 4, 35043 Marburg, Germany. Tel: ⫹ 49 (0)6421 28 65122. Fax: ⫹ 49 (0)6421 28 65121. E-mail: boesner@staff.uni-marburg.de (Received 18 July 2013; accepted 16 February 2014) ISSN 1381-4788 print/ISSN 1751-1402 online © 2014 Informa Healthcare DOI: 10.3109/13814788.2014.900535
Eur J Gen Pract Downloaded from informahealthcare.com by University of Waterloo on 10/15/14 For personal use only.
2
J. Diederich et al.
even life-threatening courses of disease such as acute heart failure, deep vein thrombosis or pulmonary embolism (6). For diagnosis in a primary care setting, patient history and physical examination remain the main diagnostic tools (4–6) and are embedded in several different diagnostic strategies (7). Besides classical approaches like spot diagnosis (7), pattern recognition (8) or Bayesian reasoning (9–11), GPs use also intuition (12) and gut feeling (13,14) in their clinical decision making. Simple heuristics (15), like the heuristic of discrepancy (16) or taking patients’ assumptions into account (17) are used as well. Aim is to investigate whether these or any other strategies are used in the further diagnostic work-up of patients presenting with unilateral or bilateral leg oedema in primary care.
METHODS Study design A qualitative research approach addressing our study question was used, to explore strategies and heuristics applied by GPs in a broad and open way. Single interviews were selected as the most appropriate technique for data collection as it created the best environment to talk with each GP about individual diagnostic strategies. The Ethics Committee of the Faculty of Medicine, University of Marburg (AZ 193/09) approved the study. Setting and data collection Fifteen GPs in Hessen and Saxony (Germany) were approached. The sample was chosen to ensure inclusion of male and female GPs practicing in urban and rural areas. The same researcher interviewed the GPs between April and September 2010. It was explained to the GPs that we were interested in understanding how they diagnosed leg oedema, and what role intuition and clinical experience played in their daily routine with these patients. To elucidate their diagnostic strategies, we asked the participating GPs to collect information of every patient presenting with leg oedema in their practice during the following four weeks (a minimum of three patients). After these four weeks, a second appointment for the interview was planned. The collected patient data were exclusively used by the GPs to aid recall during the interview (‘stimulated recall,’ a method that is frequently applied for examining the clinical reasoning of physicians), but were not presented to the interviewer. All final diagnoses presented by the GPs during the interview were based on the GPs’ own diagnostic work up and were not externally verified (18). We started the semi-structured interview with a
question about the number and the diagnoses of the patients collected by the respective GP. Afterwards, GPs were invited to recall the consultation of these patients. A semi-structured interview guideline was used to cover all relevant topics such as history taking, examination and further diagnostic investigations, identification of red flags, GPs’ diagnostic strategies and individual heuristics, dealing with uncertainty. We asked about central aspects of clinical history and physical examination, the use of scores or any other (also individual) algorithms, which helped the GPs in the differential diagnosis of specific cases. Additionally, GPs thoughts about the role of intuition and gut feeling were investigated. GPs’ reflections on their diagnostic reasoning or individual diagnostic strategies concluded the interview. Data analysis The interviews were taped and transcribed verbatim. To assist data handling, the material was transcribed and analysed using MAXQDA-10 (19). Our analysis can be categorized as a thematic survey, undertaken stepwise with each step informing the next (20,21). The first step was to develop a coding system (coding-tree) (22). The coding tree and coding was tested and refined within our qualitative working group. All interviews were coded by two independent raters (JD and SH); differences in coding were resolved by discussion. Data analysis was guided by already known theories and conceptions of diagnostic decision making in primary care. We observed a certain information saturation effect around the 10th to 11th interview, when repeatedly similar answers were given as before. Since we had already contacted the respective GPs, we nevertheless completed all 15 interviews.
RESULTS Table 1 shows the characteristics of the 15 interviewed GPs. GPs encountered 187 patients presenting with leg oedema during the four weeks prior to the interview. Table 1. Characteristics of German GPs (n ⫽ 15), being interviewed on their diagnostic work-up of patients presenting with leg oedema. Characteristic Gender (GPs)
Number
Male Female Years of clinical experience ⬍ 10 10–20 ⬎ 20 Practice location Urban Rural
11 4 2 6 7 5 10
Diagnosing leg oedema in primary care They all gave very individual and specific accounts of their diagnostic reasoning. General aspects and the role of intuition
Eur J Gen Pract Downloaded from informahealthcare.com by University of Waterloo on 10/15/14 For personal use only.
All GPs differentiated in their diagnostic reasoning between unilateral and bilateral oedema. Patient history, known comorbidities and results of the physical examination were the key elements of the diagnostic work-up; further technical examinations played a secondary role. Most of the interviewed GPs confirmed that the combination of intuition and clinical experience was very important in their clinical decision making. There are immediately intuitive thoughts when seeing [a patient with] leg oedema […], there might be something wrong. (A13: 70–71) GPs considered the own ‘feeling […] , concerning the patient ’s problem […] and the [number] of collected diagnostic hints…’ as essential. This means that I sense the patient, I observe how he shakes my hand, how he enters the room, or how he breathes. (A12, 93–96) GPs’ diagnostic strategies could be grouped into two categories, ‘active strategies’ and ‘passive strategies’ (Table 2). Hereby our definition of ‘active’ and ‘passive’ strategies refers to the GPs’ general focus regarding the diagnostic process when using these strategies: whether they proceeded actively and used all possible clinical investigations to find out causes of oedema (‘active’), or whether they stayed more observant in relation to patients’ complaints (‘passive’). Active diagnostic strategies Taking prevalence of disease into account. Normally being the first contact point of a patient, GPs took the basic epidemiology of leg oedema into account in their diagnostic reasoning. Prevalence of possible underlying disease was used strategically by five of the fifteen GPs. Table 2. Strategies used by GPs in the diagnostic work-up of patients presenting with leg oedema. Active strategies Taking prevalence of disease into account Clinical decision rules (scores/algorithms) Bayes-Theorem (accumulating probabilities) Hypothetico-deductive method Dichotomizing problems Familiarity versus discrepancy heuristics
Passive strategies Watchful waiting/test of time Therapy as diagnostic tool Taking patient assumptions into account
3
[First] I look if I can find anything palpable […] then a ranking follows according to “what is common is common” and “what is rare is rare.” […] And then I try out a few things. (GP13, 63–64) Participating GPs mentioned that they considered diseases of high prevalence (for example, cardiac oedema, oedema due to venous insufficiency) more often as a first hypothesis than diseases of low prevalence (like lipoedema, lymphoedema or others). Of course it matters whether [the oedema] is unilateral or bilateral. Bilateral thrombosis for example is much rarer than unilateral thrombosis. (GP13, 23–23) Clinical decision rules (scores, algorithms) and individual diagnostic schemes. Seven out of fifteen interviewed GPs used clinical decision rules as a diagnostic tool, but partly had strong reservations about the feasibility of these instruments. Scores are always subjective! They should be objective, but scores are basically developed for the wrong clientele, and do not apply for half of my patients, so they [the scores] lose evidence. (GP2, 138–138) GPs instead used individual concepts of diagnostic reasoning, subject to a large heterogeneity. Most GPs’ methods could be characterized as recurrent procedures, which were based on history and clinical examination. […] in this case I look, examine, touch, feel! Afterwards I decide which additional investigations I need. (GP1, 49–49) Well, I actually always follow a fixed order. First of all, physical examination, [and] if I have any suspicion about the heart I always do an ECG. […] If I do not see anything […] I also order blood tests. […] I always proceed in the same order. […] Heart, liver, thyroid, varicose veins—those are my four main categories […]. (GP3, 59–63) Diagnostic procedures of some GPs did not follow any pre-structured algorithm or scheme. How I proceed differs [from case-to-case]. (GP15, 23–23) These GPs adapt their approaches individually to each patient, taking the overall picture of patient ’s complaints, history, symptoms and the first impression the patient makes into account. Then I ask myself: are there chronic diseases? Does this come from the heart? Has he [the patient] always had one swollen leg? If I know this is unilateral, this is acute, this has not been noticed before, then I investigate further. […] If nothing is known, then I take a closer look at the
4
J. Diederich et al. leg, to see whether it might be lymphoedema or an oedema because of venous insufficiency. (GP11, 10–12)
Eur J Gen Pract Downloaded from informahealthcare.com by University of Waterloo on 10/15/14 For personal use only.
That is, I perceive the patient: I see how he shakes hands or how he walks into the room, or how he wheezes. [All this while] knowing the results of previous examinations. (GP12, 95–96) Bayes-Theorem (accumulating probabilities). When trying to find out the cause of the symptom ‘leg oedema,’ GPs used various active strategies to rule out (or in) possible diagnoses. Eight of fifteen interviewed GPs accumulated probabilities, a strategy known as the ‘Bayes Theorem.’ In this strategy, GPs form an early hypothesis about the cause of the symptom. Each further component of the patient ’s history and the clinical examination results in a given (post-test) probability of the suspected diagnosis. These probabilities are then summed-up to lead to acceptance or rejection of the early hypothesis. The likelihood of thrombosis increases with the risk factors […]. Risk factors are confinement to bed, surgery, immobility, previous thrombosis, smoking, a positive family history […]. […] all of these factors have to be checked. And, […] whether it [e.g. the swelling] is unilateral or bilateral […]. (GP13, 23–23) Usually it is the sudden onset [of the symptoms] […], unilateral, and more pronounced […]—which leads to a strong suspicion of thrombosis. Also if there is knocking pain in the heels or pushing pain in the calves […], there is often a matching history, for example, “Yesterday we came back from Austria by bus!” (GP8, 22–24) Hypothetico-deductive method. In other instances, GPs considered more than one potential diagnosis as a first hypothesis at the beginning of the consultation. Further items of the patient ’s history and findings from the physical examination together with further tests made one final aetiology much more likely than other differential diagnoses. Ten of the fifteen interviewed GPs used this diagnostic approach. [One] woman [with leg oedema] said: “I have pain in my right leg.” […] Well, the first impression was: obesity, pressure on the spinal cord, symptoms like sciatica. […] And then I did a Doppler ultrasound scan on such an obese patient. She did not have typical pressure points for acute thrombosis. […] The Doppler ultrasound scan was not positive for thrombosis. After all, it was still sciatica. (GP6, 25–25). Dichotomizing problems. Four out of fifteen interviewed GPs assigned history and clinical examination
findings to one of two opposite groups, following a strategy of dichotomizing. This method was used by GPs applying a chain of questions that led them to a final diagnosis (‘decision tree’). GPs differentiated between unilateral and bilateral oedema; chronic and acute condition, and whether there is accompanying pain or not. Additionally […] [it is relevant] whether the oedema increase at certain times of day or not […],whether there the swelling on both sides is similar or whether there are differences between the left and the right side. (GP7, 20–20) Basically I first distinguish whether oedema is unilateral or bilateral. […]. For every leg oedema, I also have a look at the veins […] and palpate them. [I ask myself] whether I can trigger tenderness on palpation of typical locations […]? [...] I take a look at the consistency of the oedema. Is it more of a tense oedema […], what is the dimension of the oedema, is it limited to the lower leg or does it concern the toes too? This way you can distinguish between whether it is lymphoedema or a venous problem. (GP10, 25–26) Heuristic of discrepancy. GPs knew most of their patients for years and used this information for diagnostic purposes. Three of the fifteen GPs, mentioned that discrepancy in behaviour or differences in the appearance of patients in comparison to previous consultations influenced them in their diagnostic decision making. Since I know the patients, I know how they behave in a given situation. Now, if I have a patient whom I know as smart and happy […], [and] if this patient comes in here and has really severe pain, then I only need to briefly glance at the leg—which is indeed swollen—then I refer him right away. But if this patient is known for constant complaining, and comes because of every little thing, then you have got to look at this a little closer. Of course, this patient can be seriously sick at times! (GP8, 49–50) Passive diagnostic strategies Watchful waiting/test of time. Eight out of fifteen interviewed GPs used the strategy of watchful waiting as a defensive strategy in their diagnostic process. One of these GPs said: That always depends on how [the symptoms] develop, […] how much time I have. In general practice […] you have to do “watchful waiting.” […] I rule out serious causes first […]. (GP14, 64–64)
Diagnosing leg oedema in primary care So if a serious disease is seen as unlikely, GPs abstain from further diagnostic procedures and reschedule the patient at regular intervals. This [...] is our advantage as general practitioners: we know the patients well, and we have time. […] We say “Come again next week.” At the hospital, the physicians don’t have enough time. (GP15, 70–70)
Eur J Gen Pract Downloaded from informahealthcare.com by University of Waterloo on 10/15/14 For personal use only.
Therapy as diagnostic tool. Ten out of fifteen GPs used this strategy. If the chosen therapy did not have an effect, GPs proceeded to more specific diagnostic interventions. And because she [the patient] had a relatively high pulse rate, I gave her a low dose of a beta blocker to see [how it worked]. And then […] after two hours […] the rate was down […]. (GP15, 46–47) Taking patient assumptions into account. Two out of fifteen interviewed GPs, purposefully asked patients about their personal perceptions in regard to their symptoms. Patients were invited to share their own suspicions, thoughts and explanations. This information influenced GPs with respect to their hypothesis and the following diagnostic steps. She [the patient] had said before that she had the same (disease) as her mother, which had been diagnosed before […] and now she had gotten all these dents. […] Principally this was the whole history. She said: “ This is a lipoedema, my mother already had the same!” She [the patient] came with the diagnosis. (GP12, 32–40)
DISCUSSION Main findings We identified GPs’ diagnostic approaches and heuristics in patients presenting with leg oedema. GPs used a broad spectrum of approaches, which could be categorized into two major groups, namely active and passive approaches (Table 2). Hereby GPs used prior information of individual patients in a specific way. Strengths and limitations The data presented have limitations. A different way to capture doctors’ diagnostic reasoning, e.g. videotaped interviews or direct observation, may have more internal validity. Yet, these methods are difficult to implement in a primary care setting. Using standardized patients would have been a possible alternative, however, practically difficult to implement for the symptom of leg oedema. In addition, this approach would not have captured the
5
diagnostic significance of long-term relationships between GPs and their patients. Contrarily, the prospective identification of patients we used in the interview allowed us to examine the diagnostic process in a defined clinical situation. We cannot rule out that GPs reinterpreted their diagnostic arguments afterwards, and may also have tried to change first intuitional diagnostic hypotheses into more rational thoughts during the interview. We tried to minimize this kind of bias by encouraging GPs to give their very individual accounts of diagnostic methods and strategies, heuristics or personal gut feelings. We stressed the lack of primary care research in this area, which helped to create an atmosphere of open and self-critical reflection. In the interview, we emphasized the widely accepted use of simple heuristics and ‘rules of thumb’ adapted to the working environment in general practice (23). As a result, GPs expressed their own thoughts and also frankly mentioned difficulties and failure so that we regard the influence of social desirability bias as very small. The generalizability of the results might have been affected by the fact that this study was conducted in only one country with its specific health system. However, we do not expect that diagnostic behaviour of GPs differs much in other European countries. Comparison with existing literature This study shows that clinical decision making is not merely a process of rational decisions (12). The GPs mentioned the importance of intuition in their diagnostic reasoning in different ways. Several studies have examined the efficiency of intuitional decision making in general practice and other fields of medicine (7,13,24,25). However, for the intuitional side of the ‘cognitive continuum’ between intuition and rationality, it is still hard to provide evidence, and intuitional decision-making processes are still underestimated in practice (12). Different diagnostic strategies and heuristics, the latter being the skill of coming to a good solution within a short period and with limited knowledge (15), were very helpful in the diagnostic reasoning of GPs concerning patients presenting with leg oedema. We have divided the GPs’ behaviour into two categories: ‘active strategies’ and ‘passive strategies.’ Wübker et al., studied GPs’ strategies in handling patients with tinnitus and divided the results into similar categories (26). Other authors preferred to focus on a chronological sequence of different diagnostic steps (7,27). Heneghan et al. grouped strategies into a three-stage model: initiation of diagnostic hypotheses, refinement of the diagnostic hypotheses and defining the final diagnosis (7). Some of our interviewed GPs used hypotheticodeductive strategies in their diagnostic reasoning. This is in line with findings of other researchers, which confirm
Eur J Gen Pract Downloaded from informahealthcare.com by University of Waterloo on 10/15/14 For personal use only.
6
J. Diederich et al.
the use of this strategy for GPs’ diagnostic reasoning in regard to a broad range of symptoms (7,28). Other authors pointed out the direct connection between the GPs’ initial hypothesis and the assessment of setting specific pre-test probability (29). Schneider et al. showed how non-awareness of prevalence in general practice leads to wrong diagnostic decisions (9). Watchful waiting was an often mentioned diagnostic strategy in our sample. GPs used this strategy mainly in the context of chronic illnesses, having ruled out serious disease beforehand. Some other studies indicate watchful waiting as a common strategy applied in general practice (7,27,30). Prior knowledge about the patient was important for the use of person specific discrepancy. Other authors described this heuristic as an important diagnostic criterion for chronic heart disease (16) or meningococcal infection in children (24). This may indicate that this strategy is a diagnostic tool for potentially serious diseases, which does not require specialized knowledge of the particular disease of concern, but of in-depth knowledge of the affected patients (16). Implications for research and practice Not all of the diagnostic strategies and heuristics described may be diagnostically helpful for the further work-up of patients presenting with leg oedema. To investigate this question further, we intend to evaluate the diagnostic effectiveness of different strategies in a cross-sectional study with a delayed-type reference standard focussing on the sub group of patients with bilateral leg oedema. This study will shed further light on medical decision making in the complex setting of primary care. Conclusion When dealing with leg oedema, GPs decided in their diagnostic reasoning between unilateral and bilateral oedema with patient history, known comorbidities and results of the physical examination being the key elements of the diagnostic work up. GPs used a broad variety of diagnostic approaches that can be grouped in ‘active’ and ‘passive’ behaviour and that mostly match with established diagnostic strategies in primary care.
ACKNOWLEDGEMENTS The authors thank all GPs who were willing to be interviewed for this study. Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
REFERENCES 1. Rabe E, Pannier-Fischer F, Bromen K, Schuldt K, Stang A, Poncar C, et al. Bonner Venenstudie der Deutschen Gesellschaft für Phlebologie. Phlebologie 2003;32:1–14. 2. Kühlein T, Laux G, Gutscher A, Szecsenyi J. Kontinuierliche Morbiditätsregistrierung in der Hausarztpraxis. Vom Beratungsanlass zum Beratungsergebnis. München: Urban & Vogel, S; 2008. p. 48. 3. Abholz H, Donner-Banzhoff N. Epidemiologische und biostatistische Aspekte der Allgemeinmedizin. In: Michael M. Kochen, editor. Allgemeinmedizin und Familienmedizin, 4th ed. Stuttgart: Georg Thieme Verlag KG; 2012. pp. 507–21. 4. Rabady S. Beratungsanlass Beinschwelllung: Differentialdiagnose in der Allgemeinmedizin. Z Allg Med. 2012;10:414–9. 5. Blankfield RP, Finkelhor RS, Alexander JJ, Flocke SA, Maiocco J, Goodwin M, et al. Etiology and diagnosis of bilateral leg edema in primary care. Am J Med. 1998;105:192–7. 6. Ely JW, Osheroff JA, Chambliss ML, Ebell MH. Approach to leg edema of unclear etiology. J Am Board Fam Med. 2006;19:148–60. 7. Heneghan C, Glasziou P, Thompson M, Rose P, Balla J, Lasserson D, et al. Diagnostic strategies used in primary care. Br Med J. 2009;338:b946. 8. Glasziou P, Rose P, Heneghan C, Balla J. Diagnosis using ‘test of treatment ’. Br Med J. 2009;338:b1312. 9. Schneider A, Dinant GJ, Szecsenyi J. Stepwise diagnostic workup in general practice as a consequence of the Bayesian reasoning. Z Arztl Fortbild Qualitatssich. 2006;100:121–7. 10. Pewsner D, Bleuer JP, Bucher HC, Battaglia M, Jüni P, Egger M. Der Intuition auf der Spur? Das Bayes’sche Theorem und die Diagnostik in der Grundversorgung. Schweiz Med Forum. 2001; 3:41–52. 11. Doust J. Diagnosis in general practice. Using probabilistic reasoning. Br Med J. 2009;339:b3823. 12. Hamm RM. Clinical intuition and clinical analysis: Expertise and the cognitive continuum. In: Dowie J, Elstein A, editors. Professional judgement—a reader in clinical decision making. Cambridge: Cambridge University Press; 1988. pp. 81–5. 13. Van den Bruel A, Thompson M, Buntinx F, Mant D. Clinicians’ gut feeling about serious infections in children: Observational study. Br Med J. 2012;345:e6144. 14. Stolper E, Van de Wiel M, Van Royen P, Van Bokhoven M, Van der Weijden T, Dinant GJ. Gut feelings as a third track in general practitioners’ diagnostic reasoning. J Gen Intern Med. 2011;26:197–203. 15. Gigerenzer G, Todd PM. Simple heuristics that make us smart. New York: Oxford University Press; 1999. 16. Hani MA, Keller H, Vandenesch J, Sönnichsen AC, Griffiths F, Donner-Banzhoff N. Different from what the textbooks say: How GPs diagnose coronary heart disease. Fam Pract. 2007; 24:622–7. 17. Goyder C, McPherson A, Glasziou P. Diagnosis in general practice. Self diagnosis. Br Med J. 2009;339:b4418. 18. Elstein AS, Shulman LS, Sprafka SA. Medical problem solving: An analysis of clinical reasoning. Cambridge, Mass.: Harvard University Press; 1978. 19. MAXQDA 10. Berlin: VERBI Software. Consult. Sozialforschung GmbH; 2010. 20. Sandelowski M, Barroso J. Classifying the findings in qualitative studies. Qual Health Res. 2003;13:905–23. 21. Kuckartz U. Einführung in die computergestützte Analyse qualitativer Daten. 3rd ed. Wiesbaden: VS Verlag für Sozialwissenschaften; 2010. 22. Kuckartz U. Qualitative Evalation—Der Einstieg in die Praxis. Wiesbaden: VS Verlag; 2007. 23. André M, Borgquist L, Foldevi M, Mölstad S. Asking for ‘rules of thumb’: A way to discover tacit knowledge in general practice. Fam Pract. 2002;19:617–22.
Diagnosing leg oedema in primary care
Eur J Gen Pract Downloaded from informahealthcare.com by University of Waterloo on 10/15/14 For personal use only.
24. Granier S, Owen P, Pill R, Jacobson L. Recognising meningococcal disease in primary care: Qualitative study of how general practitioners process clinical and contextual information. Br Med J. 1998;316:276–9. 25. Zuhlke H. II. Intuition in surgery-Is intuition in surgery learnable? ZentralblChir. 1999;124:55–63. 26. Wübker S. Hausärztliche Behandlung des Tinnitus aurium aus der Perspektive von Patienten und Hausärzten. Doctoral thesis. University of Göttingen 2007. Available at http://www.ediss. uni-goettingen.de/bitstream/handle/11858/00-1735-0000-
27. 28. 29. 30.
7
0006-AF36-6/wuebker.pdf?sequence ⫽ 1 (accessed 11 November 2013). Almond SC, Summerton N. Diagnosis in general practice. Test of time. Br Med J. 2009;338:b1878. Norman G, Barraclough K, Dolovich L, Price D. Iterative diagnosis. Br Med J. 2009;339:b3490. Summerton N. Diagnostic testing: The importance of context. Br J Gen Pract. 2007;57:678–9. Nazareth I, King M. Decision making by general practitioners in diagnosis and management of lower urinary tract symptoms in women. Br Med J. 1993;306:1103–6.
