Q J Med 2003; 96: 1-5
© 2003 Association of Physicians
Editorial |
Research evidence and the individual patient
Hadassah Medical School, Kaplan Medical Center, Rehovot, Israel e-mail: amiMD{at}clalit.org.il Department of Ambulatory Care and Prevention, Harvard Medical School, Boston
Clinical decision-making used to be based on physicians' experience and authority. Now this is no longer enough. The dissemination of the practice of evidence-based medicine (EBM) has closely appended science to art.1 Clinicians worldwide have access to powerful, precise and up-to-date information sources on which to base both diagnosis and treatment. Medical textbooks, which until recently, referred to tests as being ‘positive’ or ‘negative’ and to treatments as being ‘useful’ or ‘ineffective’ for a given condition, are being transformed. They are less likely to contain these vague adjectives, whose meanings are uncertain,2 or the tables that cite extremely rare causes on the same footing as highly common ones. An analytical, quantitative and critical approach to medical data is on the rise. Factual, numeric information, often derived from the scientific examination of large numbers of patients, is now widely available, through modern textbooks, journal articles and electronic databases. Many of these databases can be accessed and searched in real time, and they are often continuously updated. As a result, EBM, which is the use of this type of information as the basis of decision-making, is becoming a paradigm of correct medical practice.3 More and more, it underlies clinical thinking, and it affects and improves many medical decisions.
The potential benefits of EBM are numerous and well described.4,5 Nevertheless, concerns have been published,6–9 and enthusiasm for this recently unearthed treasure trove should not be undiscerning. Although there is wide agreement as to the immense value of the practice of EBM in the science and art of clinical decision-making, its clinical usefulness could be enhanced if clinicians understood the gaps that exist between the research evidence and the care of the individual patient, and dealt with them effectively.
First, the characteristics of patients studied in randomized controlled trials (RCT) and the individual patient under care in the physician's office or on hospital ward, are often different. As a result, even data collected from carefully conducted and controlled studies may not be directly applicable to an individual patient. Applying average results derived from groups of patients to a unique patient is bound to be problematic. As Mant10 says: ‘a clinical trial is the best way to assess whether an intervention works, but it is arguably the worst way to assess who will benefit from it’. In addition, differences related to the population (genetic, cultural, environmental, healthcare facilities) and individual differences (age, co-morbidity, past or current treatments, non-biological variables) may affect the translation of evidence from study sample to individual patient.10,11 Moreover, patients studied in large trials are selected according to strict criteria, meticulously investigated, and carefully followed, by physicians who are not only highly experienced, but also committed to their research project. Research studies are often conducted at outstanding teaching facilities with personnel, equipment and laboratory capabilities exceeding those available for most patient care. As a result, patients in controlled trials may get care and follow-up that is considerably better than ordinary. Their compliance is often above average, in part because willingness to follow medical advice is part of the trial's selection criteria.
The potential gaps between research evidence and the individual patient can be illustrated throughout by looking at one well-studied example: warfarin treatment for stroke prevention in non-valvular atrial fibrillation. Most trials excluded many patients, and often less than 10% of eligible patients were actually randomized.12 In actual practice, patients are older, have more co-morbid conditions, have much longer intervals between coagulation checks, and are less often within the target ratios.13–15 Factors related to different physician's expertise and time constraints in the community are also likely to contribute, and the reported rate of serious bleeding may be at least twice as high for patients in the community (3% vs. 1.3% per year).16,17 Perhaps treatment should not be offered to low-risk patients, identified through studies that examine risk stratification in atrial fibrillation.18 Treatment failures in ordinary practice (i.e. ischaemic strokes) may also be more frequent than reported in randomized trials based in referral centers.15 As one critic of EBM put it ‘the conditions of trials are often far removed from clinical practice, as are the patients .^.^. or patients ’.9
These differences should be carefully and critically taken into account when translating the evidence. Some differences may be avoided by finding studies that were done in ordinary practice settings,13–15 and more such studies are needed. Searching published studies for information on risk, benefit and harm in subgroups of patients that more closely approximate those in practice, is also useful when available. However, most trials are not large enough to allow precise estimates in subgroups, and pooled data from many randomized controlled trials may be necessary to obtain adequate information on patients in clinically important subgroups.16 At the same time, carefully performed analyses of observational data (such as the Duke cardiovascular diseases database) can complement and extend RCT, often providing important information for individual patients and longer-term follow-up.19,20
Once information that is as relevant as possible to the patient in hand has been identified, a relatively simple approach can be adopted to tailor it to the patient. As suggested by the EBM Working Group, the clinician can estimate the patient's risk of both a clinical event and intervention-related adverse events, relative to that of the patients in the study.21 Such estimation of relative differences based on clinical judgment, may be quite accurate, and used to generate patient-specific number-needed-to-treat or number-needed-to-harm data.
In extrapolating from large trials to an individual patient and in reconciling EBM with patients' values, the criteria proposed by Glasziou et al. could be used.11 Their model requires four steps: deciding whether the patient is essentially similar to those in the study and stratifying findings according to individual patient characteristics; assessing whether the intervention can be approximated in a non-study setting; quantifying benefits and harm; and incorporating individual preferences. Alternatively, clinical decision analysis may be helpful.22 Familiarity with this method allows the clinician to derive a unique, patient-tailored decision that takes into account multiple health outcomes, even when the evidence is incomplete, complex and probabilistic, such as anticoagulation in a patient with atrial fibrillation.23,24 Recent studies clearly demonstrate the importance of the individual patient's risk profile, and how it could be used in individualizing treatment choices.25 Nevertheless, review of decision analysis in patient care reveals that significant limitations still exist.26 Clearly, the effect of individual variables on patient outcomes must be skilfully assessed, and then incorporated into the controlled trial data, before it is adopted for the care of the individual patient.
Second, practicing physicians, pressured by work overload and shortened time for seeing each patient, may be tempted to prematurely narrow the clinical question and look for the best evidence related to it. Unlike physicians in trials, whose carefully planned protocol demands meticulous evaluation, their examination might overlook one of many important secondary causes of atrial fibrillation which require different management, such as an occult murmur of mitral stenosis;27,28 or their history taking may fail to discover that the patient is taking quinine for night cramps, which interacts with warfarin, increasing haemorrhagic risk.29 Thus, careful history and physical examination remains the cornerstone of medical decision-making, and should precede the application of EBM, to avoid serious gaps in knowledge and adverse patient consequences. The physical examination should itself be guided by evidence-based medicine.30,31
A third gap between research evidence and practice is of particular importance in the era of escalating medical costs and managing care. The physician's recommendation may be well founded in research evidence, but even wealthy countries encounter local barriers to implementation because of scarce resources. The contention that interventions driven by research evidence not only increase effectiveness but ultimately decrease costs, remains so far unproven.32
Fourth, the patient's own wishes and preferences should determine how evidence is applied to them. The patient's autonomy, rights and point of view should be respected and taken into consideration.33 This can be done best only after the evidence has been consulted. In that way, the pros and cons of the situation can be accurately transmitted to the patient and family, enabling them to judge, give an informed consent and be compliant. Providing an accurate basis for the patient's own decisions is another important advantage of EBM.34,35
In the atrial fibrillation example, the patient might find it too difficult to submit to frequent blood tests, or decide that the risk of serious bleeding is too much to accept. A recent meta-analysis of six RCTs showed a relative risk reduction of 62% with warfarin compared with placebo, and 36% compared with aspirin.36 Unfortunately, despite its well-proven efficacy in preventing stroke, warfarin therapy of these patients is still markedly underused: up to two-thirds of eligible patients do not receive warfarin.37,38 Part of this may be due to patient reluctance,39 but recent studies consistently found that patients placed a high value on avoiding stroke and a much lower value on avoiding a bleed.40 Physicians' thresholds were significantly different, emphasizing that clinicians cannot confidently substitute their own values for those of their patients. They must take care to avoid inadvertently skewed presentation of information, and strive at a truly balanced shared decision making between them and the patient.40,41 This point is well illustrated by the abovementioned study, which shows just how substantially different individual responses are. Responses of the 61 patients interviewed, regarding the minimum number of strokes that need to prevented before warfarin is acceptable, ranged from 1 to 11(!) strokes in 100 patients over two years, and the patient's view could not be predicted by other associated factors.40,42
Finally, in every patient's illness and reaction to illness there is a strong emotional component. The so-called ‘psychosocial’ element is often hidden behind the organic presentation, but affects it in a complex interplay, and has an important bearing on the patient's prognosis. Elements such as the patient's social support, economic status, level of satisfaction, stress and perhaps most important, feelings of hope and optimism, all affect the patient's course and prognosis.43–47 Physicians relying on the evidence alone, tend to disregard this information, which is relevant to the individual's patient's outcome but does not show on randomized controlled trials. Yet it is there, and may even be managed to improve outcomes.48
With the current dominance of the biomedical approach, the patient's preferences and emotions are likely to be given only fleeting attention unless a patient-centered paradigm is intimately incorporated into the physician's thinking and consideration.49 This is especially important, because patients differ widely in their values and needs; even individual patients' circumstances and preferences may change with time and should be re-evaluated frequently.50 An effective and sensitive communication between the physician, the patient and the family is an essential tool for incorporating and adopting individual circumstances to the evidence gathered from the analysis of populations.51 Furthermore, active patient participation in consultations may actually be associated with a better control of their diseases.52 Thus EBM may not fulfil its full promise if the patient's unique personal concerns and attitudes are not fully appreciated and allowed to affect evidence-derived clinical decisions.33,53 This is as true in secondary and tertiary care settings as it is in primary care.
Medical decisions should certainly be based on the best available research evidence. However, the importance of the physician's traditional skills is not diminished in the process. Clinical expertise is nowadays more crucial than ever. It enables the physician to carefully collect all relevant information from the patient, assess it to form a central question, then collect the best available research information and carefully apply it to the individual patient. In doing so, the physician should bear in mind the differences between the setting of the randomized controlled trial and the current situation, and the patient's own wishes and feelings. Incorporating clinical skills and above all clinical judgment7,54 with a thorough appraisal of research evidence and matching the results to the unique individual patient should be today's standard of good care. In this way (Table 1
), the inherent gap between clinical research and practice will become as narrow as possible, and the myriad benefits of evidence-based medicine will become even more impressive.
|
References
1. Davidoff F. In the teeth of the evidence: the curious case of evidence-based medicine. Mt Sinai J Med 1999; 66:75–83.[Medline]
2. Timmermans DRM. The roles of experience and domain of experience in using numerical and verbal probability terms in medical decisions. Med Decis Making 1994; 14:146–56.
3. Haynes BR, Sackett DL, Gray JMA, Cook DJ, Guyatt GH. Transferring evidence from research into practice: 1. The role of clinical care research evidence in clinical decisions. ACP Journal Club 1996; 125:A14–16.[Medline]
4. The Evidence-Based Medicine Working Group. Evidence-based medicine: a new approach to teaching the practice of medicine. JAMA 1992; 268:2420–5.
5. Rosenberg W, Donald A. Evidence-based medicine: an approach to clinical problem-solving. Br Med J 1995; 310:1122–6.
6. Feinstein AR. Clinical judgment revisited: the distraction of quantitative models. Ann Intern Med 1994; 20:799–805.
7. Tonelli MR. The philosophical limits of evidence-based medicine. Acad Med 1998; 73:1234–40.[ISI][Medline]
8. Rosser WW. Application of evidence from randomised controlled trials to general practice. Lancet 1999; 353:661–4.[CrossRef][ISI][Medline]
9. Goodman NW. Criticizing evidence-based medicine. Thyroid 2000; 10:157–60.[ISI][Medline]
10. Mant D. Evidence and primary care: can randomized trials inform clinical decisions about individual patients? Lancet 1999; 353:743–6.[CrossRef][ISI][Medline]
11. Glasziou P, Guyatt GH, Dans AL, Dans LF, Straus S, Sackett DL. Applying the results of trials and systematic reviews to individual patients. ACP J Club 1998; 129:A15–16.[Medline]
12. Lip GYH. Thromboprophylaxis for atrial fibrillation. Lancet 1999; 353:1620.[CrossRef][ISI][Medline]
13. Gottlieb LK, Salem-Schatz S. Anticoagulation in atrial fibrillation: does efficacy in clinical trials translate into effectiveness in practice? Arch Intern Med 1994; 154:1945–53.[Abstract]
14. Evans A, Kalra L. Are the results of randomized controlled trials on anticoagulation in patients with atrial fibrillation generalizable to clinical practice? Arch Intern Med 2001; 161:1443–7.
15. Kalra L, Yu G, Perez I, Lakhani A, Donaldson N. Prospective cohort study to determine if trial efficacy of anticoagulation for stroke prevention in atrial fibrillation translates into clinical effectiveness. Br Med J 2000; 320:1236–9.
16. McMahan DA, Smith DM, Carey MA, Zhou XH. Risk of major hemorrhage for outpatients treated with warfarin. J Gen Intern Med 1998; 13:311–16.[CrossRef][ISI][Medline]
17. Risk factors for stroke and efficacy of antithrombotic therapy in atrial fibrillation: analysis of pooled data from five randomized controlled trials. Arch Intern Med 1994; 154:1449–57.[Abstract]
18. The Stroke Prevention in Atrial Fibrillation Investigators. Predictors of thromboembolism in atrial fibrillation: I. Clinical features of patients at risk. Ann Intern Med 1992; 116:1–5.[ISI][Medline]
19. DeLong ER, Nelson CL, Wong JB, Pryor DB, Peterson ED, Lee KL, Mark DB, Califf RM, Pauker SG. Using observational data to estimate prognosis: an example using a coronary artery disease registry. Stat Med 2001; 20:2505–32.[CrossRef][ISI][Medline]
20. Alexander KP, Newby LK, Hellkamp AS, Harrington RA, Peterson ED, Kopecky S, Langer A, O'Gara P, O'Connor CM, Daly RN, Califf RM, Khan S, Fuster V. Initiation of hormone replacement therapy after acute myocardial infarction is associated with more cardiac events during follow-up. J Am Coll Cardiol 2001; 38:1–7.
21. McAlister FA, Straus SE, Guayatt GH, Haynes RB. Users' guide to the medical literature. XX. Integrating research evidence with the care of the individual patient. JAMA 2000; 283:2829–36.
22. Lilford RJ, Pauker SG, Braunholtz DA, Chard J. Decision analysis and the implementation of research findings. Br Med J 1998; 317:405–9.
23. Sarasin FP. Decision analysis and the implementation of evidence-based medicine. Q J Med 1999; 92:669–71.[ISI]
24. Thomson R, Parkin D, Eccles M, Sudlow M, Robinson A. Decision analysis and guidelines for anticoagulant therapy to prevent stroke in patients with atrial fibrillation. Lancet 2000; 355:956–62.[CrossRef][ISI][Medline]
25. Col NF, Pauker SG, Goldberg RJ, Eckman MH, Orr RK, Ross EM, Wong JB. Individualizing therapy to prevent long-term consequences of estrogen deficiency in postmenopausal women. Arch Intern Med 1999; 159:1458–66.
26. Elwyn G, Edwards A, Eccles M, Rovner D. Decision analysis in patient care. Lancet 2001; 358:571–4.[CrossRef][ISI][Medline]
27. Arnsdorf MF. Causes of atrial fibrillation. UpToDate, 2001 [http://www.uptodate.com].
28. St Clair EW, Oddone EZ, Waugh RA, Corey GR, Feussner JR. Assessing housestaff diagnostic skills using a cardiology patient simulator. Ann Intern Med 1992; 17:751–6.
29. Chan TY. Drug interactions as a cause of overanticoagulation and bleedings in Chinese patients receiving warfarin. Int J Clin Pharmacol Ther 1998; 36:403–5.[ISI][Medline]
30. Simel DL, Rennie D. The clinical examination: an agenda to make it more rational. JAMA 1997; 277:572–4.
31. McGee S. Evidence-based Physical Diagnosis. Saunders, Philadelphia, 2001.
32. Brown GC, Brown MM, Sharma S. Health care in the 21st century: evidence-based medicine, patient preference-based quality, and cost effectiveness. Qual Manag Health Care 2000; 9:23–31.[CrossRef][Medline]
33. Kassirer JP. Incorporating patient's preferences into medical decisions. N Engl J Med 1994; 330:1895–6.
34. Laupacis A, Sackett DL, Roberts DS. An assessment of the clinically useful measures of the consequences of treatment. N Engl J Med 1988; 318:1728–33.[ISI][Medline]
35. Straus SE, McAlister FA. Evidence-based medicine: a commentary on common criticisms. CMAJ 2000; 163:837–41.
36. Hart RG, Benavente O, McBride R, Pearce LA. Antithrombotic therapy to prevent stroke in patients with atrial fibrillation: a meta-analysis. Ann Intern Med 1999; 131:492–501.
37. The Clinical Quality Improvement Network (CQIN) Investigators. Thromboembolic prophylaxis in 3575 hospitalized patients with atrial fibrillation. Can J Cardiol 1998; 14:695–702.[ISI][Medline]
38. Gage BF, Boechler M, Doggette AL, Fortune G, Flaker GC, Rich MW, Radford MJ. Adverse outcomes and predictors of underuse of antithrombotic therapy in medicare beneficiaries with chronic atrial fibrillation. Stroke 2000; 31:822–7.
39. Protheroe J, Fahey T, Montgomery AA, Peters TJ. The impact of patients' preferences on the treatment of atrial fibrillation: observational study of patient based decision analysis. Br Med J 2000; 320:1380–4.
40. Devereaux PJ, Anderson Dr, Gardner MJ, Putnam W, Flowerdew GJ, Brownell BF, Nagpal S, Cox JL. Differences between perspectives of physicians and patients on anticoagulation in patients with atrial fibrillation: observational study. Br Med J 2001; 323:1218–21.
41. Ross JM. Commentary on applying the results of trials and systematic reviews to individual patients. ACP Journal Club 1998; 129:A17–18.[Medline]
42. Montgomery AA, Fahey T. How do patient's treatment preferences compare with those of clinicians? Qual Health Care 2001; 10(Suppl. 1):139–43.
43. Frasure-Smith N, Lesperance F, Talajic M. Depression following myocardial infarction: impact on 6 months survival. JAMA 1993; 270:1819–25.[Abstract]
44. Greer S. Psychological responses to cancer and survival. Psychol Med 1991; 21:43–9.[ISI][Medline]
45. Berkman LF, Leo-Summers L, Horwitz RI. Emotional support and survival after myocardial infarction: a prospective, population-based study of the elderly. Ann Intern Med 1992; 117:1003–9.[ISI][Medline]
46. Philips DP, Ruth TE, Wagner LM. Psychology and survival. Lancet 1993; 342:1142–5.[CrossRef][ISI][Medline]
47. Sobel DS. Mind matters, money matters: the cost-effectiveness of mind/body medicine. JAMA 2000; 284:1705–6.
48. O'Boyle CA. Diseases with passion. Lancet 1993; 342:1126–7.[CrossRef][ISI][Medline]
49. Sweeney KG, MacAuley D, Gray DP. Personal significance: the third dimension. Lancet 1998; 351:134–6.[CrossRef][ISI][Medline]
50. Peters RM. Matching physician practice style to patient informational issues and decision-making preferences. Arch Fam Med 1994; 3:760–3.[Abstract]
51. Bensing JM, Verhaak PF, van Dulmen AM, Visser AP. Communication: the royal pathway to patient-centered medicine. Patient Educ Couns 2000; 39:1–3.[CrossRef][ISI][Medline]
52. Kaplan SH, Greenfield S, Ware JE. Assessing the effects of physicians-patient interactions on the outcomes of chronic disease. Med Care 1989; 27(Suppl. 3):S110–27.
53. Jacobson LD, Edwards AGK, Granier SK, Butler CC. Evidence-based medicine and general practice. Br J Gen Pract 1997; 47:449–52.[ISI][Medline]
54. Guyatt GH, Haynes RB, Jaeschke RZ, Cook DJ, Green L, Naylor CD, Wilson MC, Richardson WS. Users' guide to the medical literature. XXV. Evidence-based medicine: principles for applying the users' guides to patient care. JAMA 2000; 284:1290–6.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M. Links Analogies between reading of medical and religious texts BMJ, November 18, 2006; 333(7577): 1068 - 1070. [Full Text] [PDF]
|
|
|
|
 |
|
 S. C. Durso Clinical Guidelines for Older Adults With Diabetes Mellitus--Reply JAMA, October 18, 2006; 296(15): 1840 - 1840. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||