Argumentation and evidence

ABSTRACT. This essay explores the role of informal logic and its application in thecontext of current debates regarding evidence-based medicine. This aim is achievedthrough a discussion of the goals and objectives of evidence-based medicine and a reviewof the criticisms raised against evidence-based medicine. The contributions to informallogic by Stephen Toulmin and Douglas Walton are explicated and their relevance forevidence-based medicine is discussed in relation to a common clinical scenario: hyper-tension management. This essay concludes with a discussion on the relationship betweenclinical reasoning, rationality, and evidence. It is argued that informal logic has the virtueof bringing explicitness to the role of evidence in clinical reasoning, and brings sensitivityto understanding the role of dialogical context in the need for evidence in clinical decisionmaking.
KEY WORDS: argumentation, evidence-based medicine, informal logic, medical epistem-ology It cannot indeed be required that everything should beproved, since that is impossible; but one can see to it thatall propositions which are used without being proved, areexpressly stated as such, so that it is clearly known onwhat the whole structure rests.1 Medicine and health care are not traditionally associated with concernsfor argumentation and its evaluation. The strongest contribution of philos-ophy to medicine has been in the domain of bioethics. The advent ofevidence-based medicine (EBM) has raised issues relating to epistemo-logy, philosophy of science, and informal logic.
It has been argued that advances in informal logic have much to offer in clarifying the role of evidence in health care. Informal logic consists ofthe study of sound and unsound reasoning in natural language. Originallyconcerned with the detection of fallacies, modern informal logic consistsof a diverse and complex set of techniques used to analyze argumentationas it arises in practical life. Informal logic is concerned with the adequacyand sufficiency of reasons put forth to justify beliefs and actions.2 Theoretical Medicine 24: 283–299, 2003.
2003 Kluwer Academic Publishers. Printed in the Netherlands.
The potential contribution of informal logic to EBM has attracted scant scholarly attention. EBM provides a rationale and justification for the useof clinical research evidence in medical practice. A critical examinationof this process clearly entails the domain of informal logic. In this paperwe will outline two ways in which argumentation theory can contributeto the understanding of the use of evidence in health care contexts. Wewill draw from both traditional approaches in informal logic, in particular,Stephen Toulmin’s diagrams, and more recent developments as representedby Douglas Walton’s description of dialogical contexts. The analysis willillustrate the potential contributions of informal logic to understandingevidence-based medicine through the analysis of a common clinical casescenario, the management of hypertension. Finally we will conclude witha discussion of the relationship between informal logic, rationality, andevidence-based medicine.
BACKGROUND: WHAT IS EVIDENCE-BASED MEDICINE? Unveiled in 1992, EBM was bold in its proclamations: A new paradigm for medical practice is emerging. Evidence-based medicine de-emphasizes intuition, unsystematic clinical experience and pathophysiologic rationale assufficient grounds for clinical decision making and stresses the examination of evidencefrom clinical research. Evidence-based medicine requires new skills of the physician,including efficient literature searching and the application of formal rules of evidenceevaluating the clinical literature.3 The use of evidence in medicine intends to introduce a more rationaland foundational approach to the practice of medicine. Using the bestavailable research evidence would place clinical decision-making on amore objective basis. Furthermore, the use of research evidence in practicewould reduce unnecessary variations in practice, lessen arbitrariness in theuse of prescription medication and diagnostic testing, and eliminate theinfluence of values in decision making. As a consequence of this, betterpatient outcomes would result. As Haynes recently wrote: A fundamental assumption of EBM is that practitioners whose practice is based on anunderstanding of evidence from applied health care research will provide superior patientcare compared with practitioners who rely on understanding of basic mechanisms and theirown clinical experience.4 EBM is committed to explicit rules and definitions. There is a clearly articulated hierarchy of scientific evidence based upon study design. Pref-erence and greater credibility is given to studies with less apparent biassuch as systematic reviews, randomized control trials, and meta-analysis.
Evidence-based approaches have clearly grown in popularity and arenow mainstays in the curricula of most undergraduate and post-graduatemedical curricula.
However, despite the success and spread of the concept, EBM has notmet with universal approval. Critics of EBM have questioned its clin-ical applicability arguing that the approach ignores clinical judgement andexperience. Some argue that evidence-based approaches foster an inappro-priate reliance on epidemiology and statistical methodology, particularly adogmatic adherence to the randomized control trial (RCT). Others arguethat evidence-based approaches neglect the true underlying issue thatrelates to what and how physicians and health care workers know. Thatempirical studies have not shown conclusively the superiority of evidence-based approaches is regarded as an important and telling fault for a theorybased on the primacy of research evidence. Straus and McAllister havesummarized the misperceptions and limitations of EBM, arguing thatmany of the limitations of EBM are limitations inherent to the practiceof medicine.
Arguments have been directed at the very concept of evidence as articu-lated by EBM, suggesting that it is restrictive, stipulative, and overly relianton quantitative methods. A more inclusive model of evidence has beenproposed that stresses the importance of context in the creation of evidenceand places equal weight on qualitative and quantitative sources of evidenceas the priority given to research-based evidence over clinical knowledge inpractice, and the limitation of what constitutes evidence, in the currentevidence-based paradigm, to that which can be expressed as a probabilitystatement are problematic. Who defines evidence and the interrelationshipsbetween the power of those defining evidence and the range of interests andpractices excluded by this process raises critical issues related to the soci-opolitical context of evidence-based practice. Finally, how values, eitherexplicitly or tacitly stated, are to be integrated and evaluated in the processof the practice of EBM is unclear.
Values are crucial components of effective and appropriate health care and are elements of virtually every clinical decision. The existence of largegrey zones in clinical practice underscores the importance of eliciting and respecting patient values and openly acknowledging uncertainty. Recentwritings by leading advocates of EBM have emphasized the importanceof integrating patient values and clinical expertise in evidence-baseddecision-making. Despite acknowledging their importance, proponents ofEBM have provided few methods to determine how patient values and theexperience of clinicians are to be integrated with research evidence. Incontrast, a large volume of literature has been created to assess the validityof a wide variety of study designs and creating decision algorithms forrational diagnosis.
The epistemological and ethical tensions raised by EBM, then, are notinconsiderable. Advancing EBM requires the elaboration of a theory ofevidence. Edmund Pellegrino has noted that evidence enters any discoursein health care as a means of testing assertions and providing support forarguments. Since there is a dimension of persuasion inherent in the useof evidence, it has an inescapable moral dimension. Therefore, Pellegrinorecognizes the need to develop a theory of evidence that inquires intothe existence, nature, and kinds of evidence that exist. This theory ofevidence, though, also must take into account concepts of rationality andargumentation, which are assumed and not analyzed by current models ofEBM.
Evidence has a justificatory role. It is offered as a means of supporting conclusions or recommendations to act. The traditional concern for thecritical analysis of the adequacy of the relationship between justificationsand the conclusions they support arises from logic. Empirical studiessuggest that physicians require improvement in their reasoning skills.
Auclair et al. studied the ability of medical residents to detect fallaciespresent in discursive arguments. They found that 36–42% of fallacies wentundetected. This illustrative example may represent a biased picture andpaint a far bleaker portrait of medical reasoning skills than necessary.
However, it does provide grounds for reflecting on how argumentationtheory can contribute to medicine.
Recent commentators have pointed out the relevance of argument analysis, particularly the method of Toulmin diagramming, to the applic-ation of evidence-based practice. Horton argues that “the skill that phys-icians lack above all is the ability to reason successfully. By to reasonI mean interrogating a clinical argument to discover its weakness or thebasis of its validity.”5 He concludes that: “The argument is the fundamentalunit of medical thought.”6 Horton and Dickinson point out the important Figure 1. Adapted from Dickinson, H.D., 1998.11 role of a structured approach to the analysis of evidence and the way it isemployed in argumentation. The methodology set out by Stephen Toulminis used by both to exemplify the logical structure of how evidence entersinto medical reasoning.
In his book, The Uses of Argument, Stephen Toulmin outlines a prac-tical approach to the analysis of everyday arguments. His structural modelinvolves identifying and separating the various components of an argu-ment into a specific order and structure (as shown in Figure 1) so that anargument may be perspicuously appraised. The elements of this model,how they relate to each other and how this model can clarify the use ofresearch evidence in medical reasoning will be illustrated through the useof a clinical scenario.
Mrs. Smith visits her physician seeking advice about the best way to manage her hypertension. After performing a detailed history, physicalexamination, and lab work her physician decides that Mrs. Smith’s treat-ment will consist of hydrochlorothiazide, an anti-hypertensive medication.
In terms of the Toulmin model, the undecided issue is how to best treat her illness while the claim is that it should be treated with hydro-chlorothiazide. In its most simple form, an argument consists of a claim orconclusion to be established by the argument. Claims, on their own, lackany reasoned support to determine whether they are true or false despitethe fact that they may be true or false. When attempting to demonstratethe truth or probability of an argument’s claim, one may be asked: “Whatinformation have you got to go on?” Claims are usually supported byappeal to some facts or other considerations. Broadly considered these can be called data. Data can be a direct observation (e.g., the measurement ofelevated blood pressure), appeal to published literature, or any other formof positive or negative information.
Is there a way to justify the leap from the provided data to the claim? Supporting the step between data and the claim is the warrant. A warrantis distinct from the data and the claim and acts as a bridge between them.
Warrants are often implicit in arguments. With regards to Mrs. Smith, thewarrant supporting the decision to prescribe hydrochlorothiazide is thatit is effective in controlling blood pressure and preventing its long-termcomplications.
You’ve told Mrs. Smith that hydrochlorothiazide will lower her blood pressure. But she may respond “How do you know?” The range of possibleanswers available to you as her physician include the following: 1. Because I’m a doctor (intuition/authority)2. Because in my experience it works (experience)3. Because it interferes with a specific biochemical process that will 4. Because many well designed studies have shown that the drug is effective in lowering blood pressure (clinical science) Clearly, there is a wide variety of available warrants for use in an argument with some being stronger than others. EBM seeks to rein-force the need for type 4 warrants. When using evidence obtained froma study as backing to a warrant, we are improving our confidence inusing a particular warrant and thus using a stronger qualifier. Sincetaking hydrochlorothiazide does not necessarily guarantee adequate bloodpressure control, there is a need to qualify the warrant by stating thathydrochlorothiazide is usually effective.
Arguments are liable to rebuttal, which may be stated to indicate condi- tions in which the warrant is not applicable and consequently the claimcan be overturned. Mrs. Smith may have allergies or contraindications tohydrochlorothiazide; a more efficacious medication or treatment may exist,or her particular set of values, whatever they may be, prevent her fromusing the medication.
Dickinson argues that, when introduced in argument, information can either be “warrant-using” or “warrant-establishing.”7 Warrant-using infor-mation acts as the basis for a claim and attempts to answer “Whatinformation do you have to go on?” In the clinical context, warrant-usinginformation relates to the individual patient and is obtained through thepatient interview, physical examination and investigative tests. In the caseof Mrs. Smith, warrant-using information would include the measurementof high blood pressure and findings from the physical examination.
Warrant-establishing information serves as the backing or justifica- tion of the warrant used to make the leap from the data to the claim.
Essentially, this form of information is used to answer “How did you getthere?” In relation to evidence-based medicine, warrant-establishing infor-mation is typically derived from systematic research such as randomizedcontrolled trials and meta-analysis. In the case of Mrs. Smith, warrant-establishing information would include randomized controlled trials thatdemonstrate the effectiveness of hydrochlorothiazide in controlling bloodpressure levels and preventing the long-term complications of diabetes.
According to the Toulmin model of argumentation, clinical decisions require warrants even in the absence of research data that may certifythe warrant. Clinicians often find themselves in situations where there islittle, if any, research data available, or substantive disagreement about theinterpretation of the research data. This makes explicit the fact that in suchsituations a physician’s experience and intuition may potentially serve asthe backing of a warrant. For example, a clinician may argue that, eventhough hydrochlorothiazide is an effective medication, it is not an appro-priate choice in their experience as it poses concerns for potassium balanceor bothersome urinary frequency (noted adverse effects of the medication).
The Toulmin method can neither adjudicate conflicting interpretations of evidence, nor determine when claims derived from experience trumpthose of clinical research. The virtue of the Toulmin model is that it makesexplicit the relationship between evidence and inferences based on theevidence and therefore permits the existence of conflicts to be made clearto parties in dialogue.
The Toulmin model is a structural model, concerned with illuminatingthe architecture of arguments. However, medical decisions are acted outin context, often with practical decisions as the goal of the interaction.
The Toulmin model aims to be context independent. Recent scholarship inargumentation, however, modifies this aim.
In his book The New Dialectic: Conversational Contexts of Argument, Walton outlines a typology of argument contexts. In Walton’s view, argu-mentation is dialectical, that is, a fundamentally social process conductedin diverse contexts between individuals with potentially differing interests.
The New Dialectic, as described by Walton, pays close attention to thepragmatic dimensions of reasoning, and is particularly concerned withdescribing and explicating presumptive and defeasible reasoning.
The need for the new approach stems from the recognition that the context of argumentation strongly influences the adjudication of argumentadequacy and soundness. Previously, in the assessment of arguments, thepresence of a fallacious inference was regarded as invalid universally. This,though, neglects the fact that argumentation, in real life, occurs to servea variety of goals. What is warranted in one context can be regarded asunwarranted in another. Furthermore, dialogues are linked to communic-ative purposes that are largely concrete and practical in nature. Hencethere may not be a universal standard of adjudication of the sufficiencyof evidence.
Walton describes six essential contexts for dialogue and argumentation.
Table I summarizes the type of dialogue and the situation and goal of thedialogue. In what follows the application of these types of dialogue toclinical situations, the requirement of research evidence in that context, andthe role of values will be explored utilizing simple illustrative examples.
For the purposes of this analysis, eristic argument will not be discussed indetail.
Persuasion dialogue is likely ubiquitous in clinical practice. The processis dyadic and can involve a patient attempting to persuade the health careprovider or vice versa. However, there will also be a set of value commit-ments that are not reducible to empirical data and may not be expressed.
To return to our blood pressure example, suppose Mrs. Smith is reluctantto take medication while her physician believes it is in her best interest.
They each proffer their reasons for their perspective. Research evidencewill likely be offered in support of the disparate views, and likely willserve as the basis of the legitimacy of the physician’s standpoint.
It is easy to recognize that value issues are intimately entangled with empirical issues in this context. A patient can trump even a sophisticatedand compelling meta-analysis by declining to initiate therapy. The standardof adjudication is reasonable persuasion. The debate has a maieutic func-tion, that is, the opposition of perspectives is meant to elicit greaterexplicitness concerning the shortcomings of the rival perspective. Researchevidence is important for persuasive dialogue, but is neither necessary norsufficient.
The goal of an inquiry is to produce solid inferences, with clear conceptsof burden of proof articulated a priori. This context of argumentation maps well to that of systematic reviews, (and of the Cochrane collaboration inparticular) and the original conception of evidence-based practice. Theeffort is collective, exhaustive with clearly specified questions and criteriastipulated in advance to determine the acceptability of evidence. Consid-erable effort is made to ensure that all acceptable evidence is includedand evaluated according to the pre-established standards. It is the dialogiccontext where research evidence is necessary. In our example of hyper-tension it may start with a physician asking a clinical question aboutthe appropriateness of prescribing a particular therapy to Mrs. Smith;conducting a literature review and determining whether the results of thisinquiry apply to her situation. Another illustration would be a collaboratinggroup systematically reviewing the literature and pooling eligible random-ized trials to determine the effectiveness of an anti-hypertensive agent,either in comparison to a placebo or another anti-hypertensive medication.
The value commitments in this dialogue relate to the value commitmentsof good empirical inquiry.
In negotiation, commitment to the truth or falsity of premises is subor-dinate to the exchange, purchase, or movement of items of value. Theconcept of value has several meanings, but the most commonly understoodconnotation involves exchange value with some form of commensurabletoken such as money. Rather than the marshalling and adjudication ofthe burden of evidence, negotiation involves trade-offs and bargaining.
Negotiation is rooted in interests and not in the pursuit of truth per se.
It is important to recognize how evidence functions in a different role inthis context and does not assume a hierarchy or trump. The question inassessing evidence in this dialogue context is the reasonableness, fairnessand justice of the trade-off’s.
This type of trade off is illustrated by recent developments in concord- ance research. In this vision of the physician-patient relationship, adialogue is established and a process of partnership develops throughdiscussion. Negotiation is an important element of this as there may betradeoffs between the goals of the provider and the patient. Evidence ofthe effectiveness of a drug may be subordinated to the self-perception orlifestyle needs of the patient. For example, Mrs Smith and her physicianmay agree to a trial of weight loss, increase in aerobic exercise, or saltrestriction in exchange for forestalling the initiation of medical therapy.
Mrs Smith is willing to bear an increase of potential risks in order not toinitiate medication. Her physician can agree to this trade off if the patient iswilling to contemplate the use of medication if her blood pressure exceeds a certain value or if the self care measures fail to achieve therapeutic goals.
In this context, the decisions reached are arguably reasonable.
In information seeking dialogue one of the parties has a repository of skillsor information that the other party or parties does not possess. By definitionthis context is asymmetrical in that the roles are unequal. This context isfamiliar in health care and examples abound: physician-patient relation-ship, physician-physician relationships particularly between specialist andgeneralist physicians, student-teacher, administrator-administrated, expertconsultations, etc. Walton stresses that these relationships are usuallycollaborative and non-adversarial which may not always be the case inhealth care.
The main characteristic of this dialogue is that it is not necessarily truth seeking. The goal may not be the truth per se, but a reasonableenough exchange of information to support a decision. The evidenceinvolved in the exchange is required to solve a problem or carry out a task.
Consequently, the evidential standard is highly contextual and variable asopposed to the context of inquiry which is exhaustive and authoritative.
The information is determined to be satisfactory by mutual agreementrather than established criteria. Narrative evidence may be of more signifi-cance in this context than quantitative. In the sense that individuals seekhealth care information for their own purposes the context is also interestbased.
Using our hypertension example, a patient may simply come and con- sult a physician wondering whether they have high blood pressure. Asimple measurement, if normal, may reassure the patient and physicianthat all is well and no further information is required.
Deliberation relates to the considerations and use of evidence in the pursuitof a solution to a practical problem. It is aimed at coming to agreement ona course of action for joint implementation or decision making. Hence agreat deal of reasoning in health care is deliberative, from deciding on adiagnostic strategy, through treatment decisions, purchasing, and policy.
Deliberative reasoning occurs, then, at every level of health care. Evidenceis used as a means of assisting in deliberations about what steps to take.
The main characteristics of deliberative dialogue are the need to take action to solve some particular problem. In our example so far, instead asin the last scenario, the patient is revealed to have a high blood pressurereading. The patient and physician will then deliberate on what is the best course of action. The physician may recommend further sequential bloodpressure readings and inform the patient on the meaning and significanceof high blood pressure. The patient will deliberate and ask questions aboutthis course of action and together they come to an agreement about theplan for diagnosis and management.
Toulmin’s diagrams are effective in illustrating the warrant establishingnature of research evidence in argumentation and in making explicit therelationship between claims, their evidential support and highlights thesources of conflicting evidence claims. The typology of argument contextsestablished by Walton is useful in that it directs attention to the type ofdialogue in question and establishes that the need for evidence is relativeto that context of application. As a consequence, there is no invariant hier-archy of evidence that can be applied in each context, and the need topursue research evidence will therefore vary accordingly. Taken together,the role of research evidence in the logic of clinical decision makingbecomes more explicit. It also follows from this account that researchevidence is neither necessary nor sufficient for decision making as manyhealth care decisions, when regarded in the form of an argument, requireweighing normative as well as factual claims. An argumentation model canbe used for evaluating the strength of such arguments.
These models, particularly Walton’s approach, are well adapted to con- sidering how research evidence, values, and professional experience haveprobative weight in decision-making. Most discussions of medical dia-gnosis and reasoning have examined its relationship to models of scientificinference. Hence medical reasoning is regarded as analogous to induc-tion, the hypothetico-deductive method, or Popperian falsificationism.
However, it is clear that medical evidence has a restricted life span. It is thetransience of medical facts that makes the need for elements of evidence-based approaches, such as consulting the research literature, so pressing.
The inherent fallibility of medical knowledge indicates that the type ofreasoning employed by physicians is more provisional in nature, pragmaticin orientation and probabilistic in its expression.
Judgements and decisions in clinical medicine rest more on plausibility than certainty, that is, what seems to be true or appropriate in a given setof circumstances. Plausible inferences can carry probative weight that maybe quantitative (like probability statements and subject to the probabilitycalculus), or be expressed in qualitative or narrative terms (and hence notsubject to the probability calculus). Plausible inferences intend to provide a reasonable guide for sustaining a belief or justifying an action, but mayin fact turn out to be erroneous, and in need of revision (think of estrogenand cardioprotection in this regard). This is consistent with the nature ofmedical evidence. As demonstrated in the discussion of the Walton modelabove, what is plausible and reasonable to do in a clinical encounter isdetermined by the context of that clinical encounter, not by the existenceof research evidence.
The Walton model also emphasizes reasoning as a form of dialogue and hence a social undertaking. Walton, following Anscombe, identifiesfour contrasting types of reasoning: • Monolectical (critical appraisal searching)/dialectical (clinical situ- ation with patients and families, other views and priorities) • alethic (truth values)/epistemic (related to a knowledge base)• static/dynamic• practical based on goals and situation, context bound based on uncertainty or incomplete knowledge of one’s changing situation/theoretical: cognitive orientation/finding reasons for accepting truthor falsity of claims The process of evidence-based medicine so far, is monolectical, alethic, theoretical, and cognitive. It is to be carried out by the health care providerseeking to apply evidence in the clinical context, and is most likely tobe conducted outside the clinical encounter. The overarching concern isfor validity and the veracity of claims derived from the empirical clinicalliterature. However, this stringent concern for validity, as follows from theabove discussion, is most appropriate in one dialogic context. Much ofthe process of the practice of medicine is dialectical, dynamic, pragmaticand context bound. These reasoning approaches are not necessarily anti-thetical, but the role and status of research evidence is different in eachcontext. What is reasonable and justifiable in clinical practice may notalways turn out to be “evidence based.” The differences between the typesof reasoning and the standards appropriate to the adjudication of evidencewithin them has not been fully explored or acknowledged by proponentsof EBM. Conflating the two, though, is likely responsible for differencesin perspectives about the applicability of evidence-based approaches inpractice and a source of some of the criticisms leveled at EBM.
Explicitness is an essential element of the definition of EBM. It is important to be as explicit about the context and purpose of dialogue asit is to understand the type of evidence produced by research studies. Anargumentation framework can make explicit the stated assertions that linkpremises to conclusions as well as provide the grounds for the analysis of unexpressed assumptions. Critical appraisal provides a method to assessthe strength of claims arising from the empirical research literature.
The explication of context moves away from the original conception of evidence-based medicine. This model is more in keeping with the visionof “research enhanced health care.”8 It is also consistent with the movetowards recommending digested evidence sources and the existence of“evidence users,”9 as acting on the basis of a pre-assessed account of astudy is precisely to act on plausible grounds. Where the strong rhetoricof evidence-based medicine incited debate, polemic and rancor, the morepragmatic vision of the clinical encounter expressed by informal logic mayresonate with clinicians’ experience as it places patient values, clinicalexperience, and clinical research on equal grounds.
The account provided here will help to dispel notions that randomized trials or systematic reviews are required for each clinical decision. Clin-ical research should be paid due respect, and in other contexts we haveargued that such consideration is an integral component of virtuous prac-tice. However, there are contexts and decisions in which research evidenceplays very little or no role. In such circumstances physicians and patientsshould not fear that they are falling below standard.
Dr. Upshur is supported by a New Investigator Award from the Cana-dian Institutes of Health Research and a Research Scholar Award fromthe Department of Family and Community Medicine at the Universityof Toronto. Errol Colak was funded by a student assistantship fromHEALNet. This project was supported by a grant from HEALNet (HealthEvidence Application and Linkage Network), a member of the Networksof Centres of Excellence Program (1995–2002) which is a unique part-nership among Canadian universities, Industry Canada, and the federalresearch granting councils.
The authors would like to thank Shari Gruman for her expert attention to the manuscript and Jason Nie for his help providing and summarizingreferences.
1 Innocentius M. Bochenski, A History of Formal Logic (Notre Dame: University of NotreDame Press, 1970), pp. 282, 283.
2 For an overview of modern informal logic, see Leo Groarke, Stanford Encyclopedia ofPhilosophy. http://plato.stanford.edu/entries/logic-informal [February 4, 2003].
3 Evidence-based Medicine Working Group, “Evidence-based medicine: A NewApproach to Teaching the Practice of Medicine,” Journal of the American Medical Asso-ciation 268 (1992): 2420.
4 R.B. Haynes, “What Kind of Evidence is it that Evidence-based Medicine AdvocatesWant Health Care Providers and Consumers to Pay Attention To?” BioMed Central HealthServices Research 2 (2002): 3. http://www.biomedcentral.com/1472-6963/2/3.
5 R. Horton, “The Grammar of Interpretive Medicine,” Canadian Medical AssociationJournal 159 (1998): 245.
6 Ibid., 249.
7 H.D. Dickinson, “Evidence-based Decision-making: An Argumentative Approach,”International Journal of Medical Informatics 51 (1998): 75.
8 R.B. Haynes, P.J. Devereaux, and G.H. Guyatt, “Physicians’ and Patients’ Choices inEvidence-based Practice,” British Medical Journal 324, no. 7350 (2002): 1350.
9 G.H. Guyatt, M.O. Meade, R.Z. Jaeschke, D.J. Cook and R.B. Haynes, “Practitioners ofEvidence Based Care: Not All Clinicians Need to Appraise Evidence from Scratch but AllNeed Some Skills,” British Medical Journal 320, no. 7240 (2000): 954.
10 Adapted from D. Walton, “Argumentation and Theory of Evidence,” in New Trends inCriminal Investigation and Evidence, vol. 2 (Antwerp: Intersentia, 2000), p. 712.
11 Dickinson, cited in n. 7, above.
Auclair, F., J. Leroux, A. Weinberg, and J. Turnbull. “Logic in Medicine: A Need to Teach Avoidance of Fallacies.” Annals of the Royal College of Physicians and Surgeons ofCanada 30 (1997): 101–102.
Barry, C.A., C.P. Bradley, N. Britten, F.A. Stevenson, and N. Barber. “Patients’ Unvoiced Agendas in General Practice Consultations: Qualitative Study.” British Medical Journal320(7244) (2000): 1246–1250.
Britten, N., F.A. Stevenson, C.A. Barry, N. Barber, and C.P. Bradley. “Misunderstandings in Prescribing Decisions in General Practice: Qualitative Study.” British Medical Journal320(7233) (2000): 484–488.
Centre for Evidence-Based Medicine. “Levels of Evidence and Grades of Recommenda- tion.” http://cebm.jr2.ox.ac.uk/docs/levels.html [April 23, 2003].
Dickinson, H.D. “Evidence-Based Decision-Making: An Argumentative Approach.” Inter- national Journal of Medical Informatics 51(2–3) (1998): 71–81.
Dickinson, H.D. “Evidence-Based Medicine: A New Approach to Teaching the Practice of Medicine. Evidence-Based Medicine Working Group.” Journal of the American MedicalAssociation 268(17) (1992): 2420–2425.
Feinstein, A.R. and R.I. Horwitz. “Problems in the ‘Evidence’ of ‘Evidence-Based Medicine’.” American Journal of Medicine 103(6) (1997): 529–535.
Gorovitz, S. and A. MacIntyre. “Toward a Theory of Medical Fallibility.” Hastings Center Gray, J.A. “Evidence-Based Public Health: What Level of Competence Is Required?” Journal of Public Health Medicine 19(1) (1997): 65–68.
Guyatt, G.H., M.O. Meade, R.Z. Jaeschke, D.J. Cook, and R.B. Haynes. “Practitioners of Evidence Based Care: Not All Clinicians Need to Appraise Evidence from Scratch butAll Need Some Skills.” British Medical Journal 320(7240) (2000): 954.
Haynes, R.B. “What Kind of Evidence Is It That Evidence-Based Medicine Advocates Want Health Care Providers and Consumers to Pay Attention To?” BioMed CentralHealth Services Research 2(1) (2002): 3.
Haynes, R.B., P.J. Devereaux, and G.H. Guyatt. “Physicians’ and Patients’ Choices in Evidence Based Practice.” British Medical Journal 324(7350) (2002): 1350.
Horton, R. “The Grammar of Interpretive Medicine.” Canadian Medical Association Journal 159(3) (1998): 245–249.
Leeder, S.R. and L. Rychetnik. “Ethics and Evidence-Based Medicine.” Medical Journal of Australia 175(3) (2001): 161–164.
Longino, H. Science as Social Knowledge. Princeton: Princeton University Press, 1990.
Malterud, K. “The Legitimacy of Clinical Knowledge: Towards a Medical Epistemology Embracing the Art of Medicine.” Theoretical Medicine 16(2) (1995): 183–198.
Miettinen, O.S. “Evidence in Medicine: Invited Commentary.” Canadian Medical Associ- ation Journal 158(2) (1998): 215–221.
Miles, A., P. Bentley, A. Polychronis, and J. Grey. “Evidence-Based Medicine: Why All the Fuss? This Is Why.” Journal of Evaluation in Clinical Practice 3(2) (1997): 83–86.
Miller, S. and L. Safer. “Evidence, Ethics and Social Policy Dilemmas.” Education Policy Analysis Archives 1 (1993): 1–14.
Naylor, C.D. “Grey Zones of Clinical Practice: Some Limits to Evidence-Based Medicine.” Lancet 345(8953) (1995): 840–842.
Norman, G.R. “Examining the Assumptions of Evidence-Based Medicine.” Journal of Evaluation in Clinical Practice 5(2) (1999): 139–147.
Pellegrino, E.D. “The Ethical Use of Evidence in Biomedicine.” Evaluation and the Health Professions 22(1) (1999): 33–43.
Polychronis, A., A. Miles, and P. Bentley. “The Protagonists of ‘Evidence-Based Medi- cine’: Arrogant, Seductive and Controversial.” Journal of Evaluation in Clinical Practice2(1) (1996): 9–12.
Poynard, T., M. Munteanu, V. Ratziu, Y. Benhamou, V. Di Martino, J. Taieb, and P. Opolon.
“Truth Survival in Clinical Research: An Evidence-Based Requiem?” Annals of InternalMedicine 136(12) (2002): 888–895.
Reid, M.C., D.A. Lane, and A.R. Feinstein. “Academic Calculations Versus Clinical Judg- ments: Practicing Physicians’ Use of Quantitative Measures of Test Accuracy.” AmericanJournal of Medicine 104(4) (1998): 374–380.
Round, A. “Introduction to Clinical Reasoning.” Journal of Evaluation in Clinical Practice Straus, S.E. and F.A. McAlister. “Evidence-Based Medicine: A Commentary on Common Criticisms.” Canadian Medical Association Journal 163(7) (2000): 837–841.
Tanenbaum, S.J. “What Physicians Know.” New England Journal of Medicine 329(17) Tonelli, M.R. “The Philosophical Limits of Evidence-Based Medicine.” Academic Medi- cine 73(12) (1998): 1234–1240.
Toulmin, S. The Uses of Argument. Cambridge: Cambridge University Press, 1958.
Upshur, R. “Certainty, Probability and Abduction: Why We Should Look to C.S. Peirce Rather than Gödel for a Theory of Clinical Reasoning.” Journal of Evaluation in ClinicalPractice 3(3) (1997): 201–206.
Upshur, R. “The Ethics of Alpha: Reflections on Statistics, Evidence and Values in Medicine.” Theoretical Medicine and Bioethics 22(6) (2001): 565–576.
Upshur, R. “Priors and Prejudice.” Theoretical Medicine and Bioethics 20(4) (1999): 319– Upshur, R. “Seven Characteristics of Medical Evidence.” Journal of Evaluation in Clinical Practice 6(2) (2000): 93–97.
Upshur, R.E., E.G. VanDenKerkhof, and V. Goel. “Meaning and Measurement: An Inclusive Model of Evidence in Health Care.” Journal of Evaluation in Clinical Practice7(2) (2001): 91–96.
Walton, D. “Argumentation and Theory of Evidence.” In New Trends in Criminal Investi- gation and Evidence Vol. Ii, 711–732. Antwerp: Intersentia, 2000.
Walton, D. The New Dialectic: Conversational Contexts of Argument. Toronto: University Walton, D. “What Is Reasoning? What Is an Argument?” Journal of Philosophy 87 (1990): Zarkovich, E. and R.E. Upshur. “The Virtues of Evidence.” Theoretical Medicine and Bioethics 23(4–5) (2002): 403–412.
Sunnybrook and Women’s HealthSciences CentreJoint Centre for BioethicsUniversity of TorontoRoom E349B2075 Bayview AvenueToronto, OntarioCanada M4N 3M5E-mail: rupshur@idirect.com

Source: http://cfpb.nl/fileadmin/cfpb/images/publicaties/artikelen/2003_Upshur_and_Colak.pdf