Evidence-based medicine – perspectives of a community-based paediatrician

Have we succumbed to a new tyranny in medicine–evidence-based medicine (EBM)? With apologies to my former professor, Dr David Sackett, but has the randomized controlled trial (RCT) become a new idol? Indeed, EBM has been satirized as a new religion (1). At the risk of not being ‘politically correct’ (2), I would argue that EBM can be abused and substituted for sound clinical judgement. The exponents of the new paradigm we call EBM recognized this: Thus, knowing the tools of evidence-based practice is necessary but not sufficient for delivering the highest quality of patient care. In addition to clinical expertise, the clinician requires compassion, sensitive listening skills, and broad perspectives from the humanities and social sciences (3). EBM, when properly understood and applied, improves patient outcomes but it is often misunderstood and used inappropriately (4). It is important to recognize the limitations of EBM, as well as its benefits in the care of our patients. Strauss and McAlister (4) point out some of the limitations of EBM: Shortage of coherent, consistent scientific evidence, difficulties in applying evidence to the care of individual patients, barriers to the practice of high-quality medicine, limited time and resources, paucity of evidence that evidence-based medicine works. One common example that many community generalists would share around the concept of ‘barriers to the practice of high-quality medicine’ is the issue of a lack of resources for the appropriate care of one's patient. For example, there is an unacceptable waiting time for a child you suspect of having autism to see a child psychiatrist to confirm the diagnosis and enable the family to obtain publicly funded resources for early intervention. This is equally true for other psychiatric disorders. Indeed, the wait time for many subspecialties in paediatrics is often in the order of many months. EBM is an offshoot of the philosophical movement known as scientific positivism, which states that only that which is scientifically verifiable is true. The critique of scientific positivism is that its central premise is itself not scientifically verifiable, but must be accepted as a sort of first principle or axiom. Similarly, the principle of EBM must be itself accepted as an axiom of the system. This can be seen if we frame the question, ‘What is the evidence that EBM improves patient outcomes?’ In fact, there is no such evidence because such a trial would be extremely difficult and unethical to perform. Evidence-based medicine, like other models of care, has limitations, …efforts need to be directed toward …conducting studies to test whether and how evidence-based medicine affects processes of care and patient outcomes (4). Let me clearly state that my comments are not to be taken as an excuse to return to magical thinking or charlatanism. The outstanding achievements of modern medicine have been accomplished through the rigorous application of the scientific method in developing evidence-based guidelines. Patient outcomes have certainly improved with the application of these treatments. We are also aware of therapeutic disasters when the principles of controlled observations and rigorous research designs were not observed (eg, tying the internal mammary arteries in the treatment for angina pectoris [5]). Having said this, there is an a priori reason why we should not substitute EBM for a more holistic approach, and that is, quite simply, that the evidence is never complete; there is always the possibility of newer evidence that will invalidate a particular diagnosis, therapeutic choice or diagnostic test. Newer diagnostic tools inevitably result in a greater understanding of pathophysiology (eg, the explosion of molecular genetics). Often, there is no definitive RCT in the literature to help one with a clinical decision. Sometimes the results of a Cochrane summary (ie, the evaluation of available RCTs regarding a particular question) are contradictory or inconclusive (6). One could also argue that the ‘gold standard’ RCT is best used when a clearly defined therapy (ie, drug A versus drug B) is being tested in a well-defined group of patients. This may be the case in a secondary or tertiary care centre, but is not the case in primary care in which multiple pathologies may be at work on a broader group of patients. An RCT may not be the most useful tool when evaluating problems in primary care where there is a less highly selected group of patients with a single defined etiology (7). The outcome measure used in an RCT may be inappropriate, as in the case of rosiglitazone, in which the primary outcome measure was the lowering of glycated hemoglobin. The clinical trials leading up to the release of this drug on the markets focused on an outcome that may not have been relevant (glycemic control) and ignored important side effects, assuming that a reduction in glycated hemoglobin translates into improved patient outcome. Indeed the side effects (weight gain, edema and changes in lipid levels) detected during the clinical trials may well be more important with regard to outcome than glycemic control. In fact, Nissen and Wolski (8) showed that rosiglitazone increases the risks of myocardial infarction, despite improved glycemic control. Recently, Ioannidis (9) argued cogently that much published research is eventually shown to be false. This results from attempts to achieve a conclusive result based on a single study assessed statistically using P<0.05. A finding from a well-conducted, adequately powered randomized controlled trial starting with a 50% pre-study chance that the intervention is effective is eventually true about 85% of the time....Conversely, a meta-analytic finding from inconclusive studies where pooling is used to ‘correct’ the low power of single studies, is probably false if R=<1:3. Research findings from underpowered, early-phase clinical trials would be true about one in four times, or even less frequently if bias is present (9). There is also a darker side to the quest for evidence-based guidelines when vital evidence is withheld from the public domain, or falsified, invalidating the ‘evidence’ on which a particular therapy is recommended. An example is the COX-2 inhibitors. In the enthusiasm to release a new drug, rofecoxib, a COX-2 inhibitor, Bombardier et al (10) overlooked an important cardiovascular side effect in a subgroup of patients, even though the drug showed promise in the group of patients for which the clinical trial was designed (ie, those with rheumatoid arthritis). Subsequent analysis of the data indicated a substantial risk for cardiovascular events in patients taking the drug, and resulted in the withdrawal of rofecoxib (Vioxx, Merck Frosst Canada Ltd) from the market (11). This has recently led to the institution of a registry for all RCTs; any researcher who wishes to publish in any of the major medical journals will now have to ensure that their RCT is registered in the public domain, with all data (both positive and negative) available (12,13). A further question is one of using data from trials that were conducted unethically, as in the infamous Tuskegee syphilis study (14), in which 399 poor black sharecroppers were denied effective treatment so that researchers could follow the natural history of syphilis. Similar comments can be made about the Nazi doctors and the use of eponyms such as Hallervorden-Spatz syndrome, honouring a pathologist, Hallervorden, who actively participated in the Nazi euthanasia program to obtain specimens of human brains for his research (15). A further issue is that conclusions drawn from RCTs are based on a statistical assessment of the participants in the clinical trial, but you must apply the conclusion to your particular patient. The danger is that your particular patient may not be similar to the patients in the clinical trial. This is especially a problem when trying to apply the results of a study which has been performed in a secondary or tertiary care centre. The patients enrolled in such a trial may be quite different from the patients seen in a community practice. At any given clinical encounter, the physician does not have all the information needed to make an evidence-based decision. For example, every individual patient has a particular pharmacogenetic profile, which is largely unknown. Another major problem in paediatrics is that most drugs are ‘orphans’ (ie, there is no good information derived from studies in paediatrics and so we extrapolate from adult studies). It is good to remind oneself that extrapolation means outside the principle of EBM – we are extrapolating or going beyond the evidence on an assumption. The assumption is that what holds true for adult patients will hold true for paediatric patients. Of course, this assumption is not evidence based! Moreover, we now know that this assumption is often untrue. Consequently, sometimes evidence-based principles ‘break down’ in the concrete world in which individual physicians must act. The renowned medieval philosopher and theologian, Thomas Aquinas, made a distinction derived from Aristotle between the speculative and practical intellect. The speculative intellect is the locus of analytical, theoretical thinking, and this is the home of EBM. The practical intellect, however, is the locus of applying EBM to the practice of medicine – the specific judgement (having taken into account the evidence and the particulars of this patient before me) that the physician makes in determining the appropriateness of a particular treatment. The practical intellect is the final common pathway leading to a particular judgement for a particular patient. For example, a nine-year-old girl is brought to your office at 17:00 on a Friday afternoon complaining of an earache. You correctly diagnose an otitis media. EBM would suggest that the correct course is not to prescribe an antibiotic but rather to prescribe analgesics and re-examine the patient 48 h later; if there is no resolution, then it is appropriate to prescribe an antibiotic. But other considerations may override the principles of EBM – the mother may not be able to come back for a re-examination or the family is about to catch a plane for a holiday, etc. Thus, a more holistic approach is mandated and the principle of EBM breaks down. There is always a certain unpredictability in our encounters with patients. Let us take another example. There are many excellent paediatric hospitals that care for patients with cystic fibrosis. A recent article (16) in The New Yorker describes how all of them follow the same evidence-based guidelines for the treatment of cystic fibrosis and yet, if one examines their results, they are distributed along a bell curve; most of the hospitals were only average in their results but one hospital stood out – Rainbow Babies & Children's Hospital in Cleveland (USA) – where a respirologist, LeRoy Matthews, established a program for the treatment of cystic fibrosis in 1957. When national mortality rates for cystic fibrosis in the best centres was around 20%, Matthews claimed that his mortality rate was less than 2%. What was the difference? The difference turned out to be innovation and constantly challenging the evidence-based guidelines which record past experience (16). We must remain focused on our patients and challenge ourselves to do better, to go beyond the evidence-based guidelines, to excel, thereby constantly improving the guidelines themselves. There have always been two streams in medicine – the Hippocratic and Aesculapian. EBM and guidelines belong to the Hippocratic stream. The Aesculapian stream emphasizes healing, and the importance of the psychological and spiritual aspects involved in healing. Because physicians do not always take this into account, many patients turn to complementary medicine for relief of their symptoms. Modern 21st century medicine needs to be reminded of both the Hippocratic and the Aesculapian heritage in its long and renowned history. A good physician must take a holistic approach in any clinical encounter with a suffering human being. EBM must never be abandoned, but it must be incorporated into the practical intellect along with the particular concrete aspects of this patient to make an appropriate and efficacious therapeutic decision. The cornerstone of caring for our patients and practicing good medicine is the encounter with the patient through our history and physical examination. We then combine this with the principles of EBM, recognizing its hermeneutic weakness, to arrive at the best possible outcome for our patient. It is then that EBM becomes “‘ebullience-based medicine,’ a lively, enthusiastic, continually joyful expression of our good fortune at having the privilege to be able to care for and advocate for children” (17).

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.