This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Browman, G. P. Search for Related Content PubMed PubMed Citation Articles by Browman, G. P.

Essence of Evidence-Based Medicine: A Case Report

From the Program in Evidence-Based Care, Cancer Care Ontario, and Health Information Research Unit, Department of Clinical Epidemiology and Biostatistics, Faculty of Health Sciences, McMaster University, Hamilton, Ontario, Canada.

Address reprint requests to George P. Browman, MD, Health Information Research Unit, McMaster University Health Sciences Centre, 1200 Main St West, Room 3H7, Hamilton, Ontario, Canada L8N 3Z5.

ABSTRACT

PURPOSE: To illustrate the complexities of the evidence-based approach in clinical oncology practice and the implications for guidelines and evaluation of processes of care.

PATIENT AND METHODS: A case report is presented in which a limited systematic review of the literature was used to address a specific clinical problem in an individual patient. Experts' opinions were also sought.

RESULTS: A reasonable clinical decision was made by a participating patient based on indirect evidence of benefit that would be insufficient to support the same decision as a health policy in some jurisdictions.

CONCLUSION: The practice of evidence-based oncology requires clinical judgment about the validity and applicability of research evidence. The factors that influence an evidence-based decision in the clinical context differ from those in the broader policy context, which could lead to legitimate differences in recommendations based on the same information. Used properly, the individual case report can be a powerful tool to illustrate complex clinical decision phenomena.

EVIDENCE-BASED MEDICINE has been defined as the conscientious, explicit, and judicious use of the best available evidence from health care research in the management of individual patients.¹ The evidence-based approach relegates the case report to a category of lowest quality of evidence to inform clinical decisions about therapeutic alternatives,² mainly because isolated observations collected in an unsystematic way cannot be relied on as a valid source of generalizable information. However, case reports represent stories, or anecdotes, that can be framed in a way that is meaningful to convey ideas in a compelling way.^3,4 Here I present a case report of a clinical decision that serves to highlight the complexity of the evidence-based decision approach. Through this method I hope to broaden understanding of the evidence-based approach and point out the nuances that often are ignored by critics in theoretical discussions.^5-7 Recent publications have begun to examine the fundamental principles behind the evidence-based approach.^8-10 This case report complements such discussions.

PATIENT AND METHODS

An otherwise healthy and active 61-year-old male psychologist was investigated for prostatic symptoms. He was eventually diagnosed with locally advanced prostate cancer. He had a high prostate-specific antigen level (400 IU) and a Gleason score of 9. The bone scan was negative and he had no bone pain.

The patient's professional background allowed him to understand the issues in clinical decision-making, and he was an active participant. After extensive discussions with several medical colleagues, he decided to follow the treatment regimen recommended by his oncologist, which consisted of the combination of hormonal therapy with goserelin, an analog of gonadotrophin-releasing hormone, and radiotherapy. Radiotherapy was planned to begin several weeks after the start of hormonal therapy. The patient's choice of treatment was heavily influenced by the results of a randomized controlled trial that reported a survival benefit for this approach compared with radiation alone.¹¹ He rejected advice from a medical colleague to consider chemotherapy because of a lack of compelling evidence of benefit and the known side effects of treatment. The patient responded well to hormonal therapy, with a dramatic two-log decrease in prostate-specific antigen level before the start of radiotherapy.

Approximately 1 week before the start of radiation, the patient read the results of a randomized trial that he interpreted as strong evidence that treatment with bisphosphonates reduces the probability of bone metastases and prolongs survival in women with a high risk of recurrence from breast cancer.¹² He asked a medical colleague with training in clinical epidemiology and who is an oncologist (but who does not specialize in genitourinary malignancies) to review the literature and advise on whether bisphosphonates should be recommended for him. He explained that his main aim of treatment was to reduce the chance of bone metastases and improve his chances of longer survival. He also asked whether bisphosphonates might be indicated to prevent osteoporosis produced by androgen blockade. The patient requested that his colleague give him his professional advice within a week.

The colleague searched his own files and Medline for the years 1980 to 1998, restricting the search to the issue of use of bisphosphonates and leaving the decision regarding other treatment options to the patient's oncologist. He used the terms prostate neoplasms, bisphosphonates, bone disease, osteoporosis; clinical trials, phase III, randomized; and meta-analysis, systematic review, systematic overview, and practice guideline. The Cochrane library was also consulted. He decided to include all studies in humans, with a preference for randomized controlled trials. All or a majority of patients in the trials had to have prostate cancer. He was interested mainly in trials of bisphosphonates in prostate cancer patients for the purpose of preventing bone disease.

In addition, the clinical epidemiologist consulted three clinical experts for their informal opinions, any additional material in their files, and unpublished information, including ongoing trials of which they might be aware. The experts who were consulted included a medical oncologist with a clinical and research interest in prostate cancer, a medical oncologist with a special interest in bisphosphonates but who did not treat prostate cancer, and a rheumatologist with a research interest and clinical experience in the management of osteoporosis.

RESULTS

Literature Search
The literature search yielded 67 articles, six of which were randomized controlled trials of systemic bisphosphonates for prostate cancer.^13-18 There was one comparative cohort study.¹⁹ One article was a recently published systematic review,²⁰ one article reported on histomorphometric results of bisphosphonate treatment²¹ from a reported randomized controlled trial,¹⁵ and one was a dose-finding study.¹⁸

In addition, one of the clinical experts provided references from his personal file. The references included one narrative review,²² a comparative trial for which the method of patient allocation was not clear,²³ a preliminary report of the published trial by Lipton et al,¹⁸ one noncomparative cohort study,²⁴ and three additional references, all of which were also identified in the formal literature search.

State of the Evidence
There were no randomized trials that tested bisphosphonates in patients with early-stage or localized prostate cancer with the intention of influencing subsequent bone metastases, patient survival, or risk of clinically significant osteoporosis from androgen blockade.

All trials were conducted in patients with prostate cancer who had already-established bone metastases, with the main outcome of pain control. Secondary outcomes were related to effects on bone histomorphometry, biochemical markers of bone turnover, and radiographic changes in metastases.

The results of the randomized trials with respect to pain control were inconsistent. A pooled quantitative analysis was not performed because the published studies were not relevant to the clinical question of interest. Thus, there was no direct evidence to inform the clinical questions that prompted the review. Interested readers who want further information are referred to the systematic review by Bloomfield.²⁰

Opinions of Experts
The medical oncologist who was experienced in the treatment of prostate cancer believed that there was no evidence to support the use of bisphosphonates in this clinical situation, but "....it wouldn't hurt." He would not offer it as part of routine management in his practice. The medical oncologist with a special interest in bisphosphonates but without experience in the management of genitourinary malignancies interpreted the results of trials of bisphosphonates for pain related to bone metastases as less impressive for prostate compared with breast cancer. He also expressed the opinion that "it wouldn't hurt to try it" for a patient with prostate cancer to reduce the probability of future bone metastases. The osteoporosis expert was the most positive, stating that he would "definitely consider the use of bisphosphonates" to prevent osteoporosis from treatment with androgen blockade, although he admitted he was not aware of the prevalence of clinically significant osteoporosis in this patient population.

Discussion With the Patient
Having collected this information, the clinical epidemiologist prepared a written report in a format designed to inform a discussion between the patient and his main health care provider. Before sharing the report with the patient, he asked for clarification of the patient's primary reason for wanting bisphosphonates and his rationale. The patient was most interested in the potential for survival benefit and quoted studies in breast cancer¹² and multiple myeloma.^25,26 He also stated that "millions of people with osteoporosis were receiving bisphosphonates, and we would be aware if there was a high probability of harm."

In his written report, the clinical epidemiologist concluded that "there is insufficient evidence to support the use of bisphosphonates for the treatment of early prostate cancer with the intention of delaying or preventing bone metastases [and]....we may not be able to generalize the effects of bisphosphonates from breast to prostate cancer. If a well-informed patient [with localized prostate cancer] insisted on being treated with bisphosphonates, I would comply with the request... I would not support a general policy that offers this treatment to [all] patients with early-stage prostate cancer."

This was presented to the patient as part of a face-to-face interview in preparation for a more detailed discussion with his oncologist.

All of the clinical experts agreed that based on the evidence, a general policy to make bisphosphonates available in this clinical circumstance could not be supported. They agreed it would be inappropriate to recommend this treatment in a practice guideline on this topic. Using the aforementioned rationale, the patient chose to be treated with bisphosphonates and was supported in his decision. The patient assumed the responsibility of payment for the drug.

After the main discussion, the patient was asked whether, instead of taking bisphosphonates, he would agree to participate in a clinical trial of bisphosphonates in which there was a 50% chance of being randomized to a placebo. He said he would not agree to enter such a trial.

LESSONS LEARNED ABOUT EVIDENCE-BASED MEDICINE

The Problem of Generalizing Evidence
There was no direct evidence from rigorous studies to inform the clinical decision. The patient's decision was based on an implicit generalization from results in other disease sites. The readiness of the clinical experts to be supportive of the decision indicates the difficulty even evidence-based clinicians have in dealing with complex clinical issues for which evidence is weak and points out the flexibility required in knowing how to apply it. The case report brings up the issue of whether the burden of proof ought to be placed on those who would generalize from indirect evidence (the current practice in evidence-based medicine) or on those who would not generalize from similar clinical scenarios. Thus, clinical judgment is part of an evidence-based decision about whether the condition from which the evidence is being generalized is similar enough to support the decision. In this case, the apparent small risk of adverse effects of the treatment played a role in the patient's decision, and the clinicians' readiness to support it ("...it wouldn't hurt"). But the implications are different from a health policy perspective for cases in which side effects in terms of cost are not trivial.

Evidence Is Interpreted and Acted on in a Context
Evidence does not exist in a vacuum. This case report illustrates, or at least brings to light, differences in how the same evidence can be legitimately used to come to apparently conflicting conclusions when applied to clinical policy versus health policy and when applied to clinical policy from different practitioner perspectives. Clinical recommendations also will be affected by more than the efficacy data gleaned from the evidence, but by the inherent tradeoffs in terms of side effects.

Because of the lack of direct evidence for benefit of bisphosphonates in this clinical situation, it would be irresponsible to recommend its routine use as a matter of policy to prevent bone metastases from prostate cancer, despite the reasonableness of using it in a specific clinical case. (There is laboratory evidence to suggest that bisphosphonates can lead to osteopenia when used in patients with established bone metastases from prostate cancer, which was part of the information provided to the patient.²¹) Because of the lack of direct evidence, it seems unlikely that a practice guideline on this topic would even include a reference to the role of bisphosphonates to prevent prostate-related bone metastases.

From a policy perspective, there also may be different legitimate conclusions from the same evidence depending on who is paying for the health service. The expense to society in a publicly funded health care system, such as that in Canada, may not be justifiable by the quality of the available information. However, the same society respects the principle of full disclosure of information to patients, their participation in clinical decisions, and their right to make choices. Thus, the dissonance between the clinical and policy decisions can be understood. This does not invalidate the evidence-based approach, but it does illustrate its complexity and the need for further theoretical and empirical studies in this area.

Patient Preferences and Participation in Clinical Decisions
The case report clearly illustrates the value of a well-informed patient's participation in a clinical decision of this type. This particular patient was able to analyze the clinical situation and consider the evidence. The patient's decision to reject the use of chemotherapy as an adjunct to hormonal and radiotherapy as part of initial management because of the lack of evidence, and the influence of a randomized trial on his decision to accept the initial management strategy, illustrates his respect for the importance of study design as one aspect of an evidence-based approach. However, even this patient was selective about what value to place on different pieces of information. Furthermore, his interest in using evidence as the basis of a decision did not extend to an agreement to participate in a trial that might produce the information that is currently lacking.

Other patients who are less sophisticated in evaluating research evidence may have come to the same conclusion based on other factors. More importantly, consumers who are less sophisticated interpreters of the evidence may come to conclusions that are clearly inconsistent the evidence.

Divergence of Expert Opinions
Although all of the clinical experts were supportive of the patient's ultimate decision, their threshold for "trying" this unproven treatment in the context of prostate cancer differed. Support from the methodologist (an oncologist) and the prostate cancer experts were more measured than that of the oncologist with an interest in bisphosphonates. The opinion of the osteoporosis expert was the most enthusiastic. Clearly, all were influenced by their estimate of a low probability of harm, despite lack of direct evidence for efficacy.

Systematic Reviews Need To Be Practical To Inform Decisions
The patient requested that he be able to make a decision informed by the evidence within 1 week. The methodologist did his best to gather as much relevant literature as he could in an unbiased fashion. He did not do the following: (1) write a protocol defining the systematic review process; (2) rigorously predefine the study eligibility criteria; (3) have an independent reviewer select appropriate studies; (4) hand-search any journals, except for the latest publications of three prestigious cancer journals; or (5) make an effort to locate studies in languages other than English.²⁷ Because there were no relevant trials found, assessment of trial quality was not necessary, nor was a quantitative summary of the evidence.

When an important clinical decision needs to be made within a time frame of 1 week, compromises are needed. In this example, shortcuts were taken so that a reasonable effort was made to collect relevant literature. This distinguishes the utility of the systematic review as a decision tool from its role as an object of scientific inquiry. Of course, the more valid the systematic review, the more confident we can be in its value as a decision tool. However, when such reviews do not exist, we need to study how to make compromises between the "perfect review," which may be unattainable within realistic constraints, and scientific nihilism, which could lead to greater biases in how decisions are made.

Implications for Evaluating Patterns of Practice According to Guidelines
Clinical practice guidelines have been defined as systematically developed statements to assist practitioner and patient decisions about appropriate health care for specific clinical circumstances.²⁸ The definition clearly respects the role of the patient in the decision process and recognizes guidelines as tools, not rules, for informing clinical judgments, not replacing them. This case report highlights how the values and complexities inherent in practice can legitimately lead to defensible decisions that are inconsistent with a guideline.

This has important implications for programs designed to evaluate the impact of guidelines based on processes of care.²⁹ The main message is that we should not assume that practice decisions that are inconsistent with a guideline are necessarily inappropriate. Although guidelines should influence the patterns of practice, on average, we need to become less judgmental in our criticisms of individual practices that seem at variance with guidelines. At a maximum, such examples should be viewed as "flags" for investigation or opportunities to learn about how to make guidelines more responsive to clinical needs.

The example also demonstrates how the "evidence-based" and "patient empowerment" movements may be on a collision course, especially as we think about how to approach the evaluation of guidelines. The primacy of patient choice and the increasing prevalence of the use of scientifically unproven therapies, such as use of alternative therapies among cancer patients,³⁰ are hints that we need to come to grips with the complexities of the evidence-based approach if we are to use it to maximum advantage.

Value of the Case Report
The weakness of the clinical anecdote, which causes clinical methodologists to relegate the information to the lowest rung of the quality-of-evidence ladder, is that it does not provide representative information for broad generalizations in choosing clinical interventions or making causal inferences.

Here the strengths of the case report are used to illustrate the complexities involved in applying the evidence-based approach. Enkin and Jadad³ discussed the strength of the case report, and Jadad⁴ previously demonstrated its power even within the evidence-based paradigm. Individual stories are a powerful means for relaying messages in a concrete way. In this instance, the case report, or anecdote, serves better than an abstract discussion to highlight the value and limitations of evidence-based medicine. It demonstrates the need for flexibility in applying evidence-based principles and the complementary value of the clinical expert and the input of the patient.

This case report is an illustration of the evolution of current thinking of evidence-based practice: that evidence is only one of different aspects involved in making a clinical decision.³¹ As the rhetoric of proponents and antagonists of evidence-based medicine is published in our journals, we need to keep in mind the single most important feature of the evidence-based approach: to be accountable for our decisions through an explicit approach that documents how our decisions are made. This anecdote is presented in that spirit.

ACKNOWLEDGMENTS

Supported by Cancer Care Ontario, The Ontario Ministry of Health, and the Health Evidence Application and Linkage Network, a Network of Centres of Excellence funded through the Medical Research Council of Canada and the Social Science and Humanities Research Council.

I thank the HEALNet leadership for stimulating discussions on the meaning of evidence-based decision making, and Drs Alex Jadad and Murray Enkin for reviewing the manuscript and contributing improvements to it.

REFERENCES

1. Sackett DL, Rosenberg WMC, Gray JAM, et al: Evidence-based medicine: What it is and what it isn't. BMJ 312:71-72, 1996[Free Full Text]

2. Cook DJ, Guyatt GH, Laupacis A, et al: Rules of evidence and clinical recommendations on the use of antithrombotic agents. Chest 102:305-308, 1992

3. Enkin M, Jadad AR: Using anecdotal information in evidence-based decision making: Heresy or necessity? Ann Oncol 9:963-966, 1998[Abstract/Free Full Text]

4. Jadad AR: Are you playing evidence-based games with our daughter? Lancet 347:274, 1996

5. Anonymous: Evidence-based medicine, in its place. Lancet 346:785, 1995[Medline]

6. Grahame-Smith D: Evidence-based medicine: Socratic dissent. BMJ 310:1126-1127, 1995[Free Full Text]

7. Miles A, Bentley P, Polychronis A, et al: Evidence-based medicine: Why all the fuss? This is why. J Evaluation Clin Pract 3:83-86, 1997[Medline]

8. Miettinen OS: Evidence in medicine: Invited commentary. Can Med Assoc J 158:215-221, 1998[Medline]

9. Davidoff F: The future of scientific medicine. Can Med Assoc J 159:243-244, 1998[Medline]

10. Horton R: The grammar of interpretive medicine. Can Med Assoc J 159:245-249, 1998[Medline]

11. Bolla M, Gonzalez D, Warde, et al: Improved survival in patients with locally advanced prostate cancer treated with radiotherapy and goserelin. N Engl J Med 337:295-300, 1997[Abstract/Free Full Text]

12. Diel IJ, Solomayer E-F, Costa SD, et al: Reduction in new metastases in breast cancer with adjuvant clodronate treatment. N Engl J Med 339:357-363, 1998[Abstract/Free Full Text]

13. Strang P, Nilsson S, Brandstedt S, J et al: The analgesic efficacy of clodronate compared with placebo in patients with painful bone metastases from prostatic cancer. Anticancer Res 17:4717-4731, 1997[Medline]

14. Kylmala T, Taube T, Tammela TL, et al: Concomitant i.v. and oral clodronate in the relief of bone pain: A double-blind placebo-controlled study in patients with prostate cancer. Br J Cancer 76:939-942, 1997[Medline]

15. Elomaa I, Kylmala T, Tammela T, et al: Effect of oral clodronate on bone pain: A controlled study in patients with metastatic prostate cancer. Int J Urol Nephrol 24:159-166, 1992

16. Adami S, Mian M: Clodronate therapy of metastatic bone disease in patients with prostatic carcinoma. Recent Results Cancer Res 116:67-72, 1989[Medline]

17. Smith JA Jr: Palliation of painful bone metastases from prostate cancer using sodium etidronate: Results of a randomized, prospective double-blind, placebo-controlled trial. J Urol 141:85-87, 1989[Medline]

18. Lipton A, Glover D, Harvey H, et al: Pamidronate in the treatment of bone metastases: Results of 2 dose-ranging trials in patients with breast or prostate cancer. Ann Oncol 5:S31-S35, 1994 (suppl 7)

19. Fernandez-Conde M, Alcover J, Aaron JE, et al: Skeletal response to clodronate in prostate cancer with bone metastases. Am J Clin Oncol 20:471-476, 1997[Medline]

20. Bloomfield DJ: Should bisphosphonates be part of the standard treatment of patients with multiple myeloma or bone metastases from other cancers? J Clin Oncol 16:1218-1225, 1998[Abstract]

21. Taub T, Kylmala T, Lamberg-Allardt C, et al: The effect of clodronate on bone in metastatic prostate cancer: Histomorphometric report of a double-blind randomized placebo-controlled study. Eur J Cancer 30A:751-758, 1994

22. Adami S: Bisphosphonates in prostate carcinoma. Cancer 80:1674-1679, 1997 (suppl) [Medline]

23. Clarke NW, McLure J, George NJR: Disodium pamidronate identifies differential osteoclastic bone resorption in metastatic prostate cancer. Br J Urol 69:64-70, 1992[Medline]

24. Murad AM, Andrade Filho ACC, Santos M, et al: Multicentric clinical trial on the use of disodium clodronate by intravenous infusion at fifteen day intervals in patients bone osteolysis. S Am J Cancer 1:80-86, 1997

25. Berenson JR, Lichtenstein A, Porter L, et al: Efficacy of pamidronate in reducing skeletal events in patients with advanced multiple myeloma. N Engl J Med 334:488-493, 1996[Abstract/Free Full Text]

26. Berenson JR, Lichtenstein A, Porter L, et al: Long-term pamidronate treatment of advanced multiple myeloma patients reduces skeletal events. J Clin Oncol 16:593-602, 1998[Abstract]

27. Cook DJ, Mulrow CD, Haynes RB: Systematic reviews: Synthesis of best evidence for clinical decisions. Ann Intern Med 126:376-380, 1997[Abstract/Free Full Text]

28. Field MJ, Lohr KM (eds): Institute of Medicine. Guidelines for Medical Practice: From Development to Use. Washington, DC, National Academic Press, 1992

29. Browman GP, Levine MN, Graham I, et al: The clinical practice guideline: An evolving health care technology. Cancer Prev Control 1:7-8, 1997[Medline]

30. Risberg T, Lund E, Wist E, et al: Cancer patients use of unproven therapy: A 5-year follow-up study. J Clin Oncol 16:6-12, 1998[Abstract/Free Full Text]

31. Haynes RB, Sackett DL, Gray JRM, et al: Transferring evidence from research into practice: 1. The role of clinical care research evidence in clinical decisions. ACP J Club 125:A-14, 1996 (editorial)

Submitted October 22, 1998; accepted March 19, 1999.

This article has been cited by other articles:

				G. P. Browman Development and Aftercare of Clinical Guidelines: The Balance Between Rigor and Pragmatism JAMA, September 26, 2001; 286(12): 1509 - 1511. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Browman, G. P. Search for Related Content PubMed PubMed Citation Articles by Browman, G. P.