Medicine encompasses both art and science: Evidence-based guidelines are no exception

In this issue of Paediatrics & Child Health, the Fetus and Newborn Committee (FNC) of the Canadian Paediatric Society presents ‘evidence-based’ guidelines for the management of babies at risk for newborn infection (1). How should these guidelines affect our practices? How much, if at all, should we modify them for our patients and institutions? How absolute are the recommendations? All worthy endeavours have searched for perfection and truth – why should health care be the exception? Historically, this quest was completed through experience, anecdote and tradition – this type of ‘evidence’ served us, our professions and our patients for centuries. We are now, however, in a new age in which the truth is fluid and what we believe to be ‘written in stone’ can quickly transform to fading sketches from a previous era. The new language to help us understand how we know what we know is that of ‘evidence-based health care’. Now we can categorize evidence as ‘internal’, consisting of personal experiences and the art of medicine, and as ‘external’, the science of our profession. Sackett et al (2) recognized the dualism of evidence in their definition of evidence-based medicine: Good doctors use both individual clinical expertise and the best available external evidence, and neither alone is enough. Without clinical expertise, practice risks becoming tyrannised by evidence, for even excellent external evidence may be inapplicable to or inappropriate for an individual patient. Without current best evidence, practice risks becoming rapidly out of date, to the detriment of patients. On this occasion, the FNC adapted the Centre for Evidence-Based Medicine’s framework for its evaluation of evidence (3). To appreciate this tool, one requires knowledge of the language of evidence-based health care. Of particular importance are the terms ‘level of evidence’ and ‘class’ or ‘grade’ of recommendation. One could argue that all experiences, whether personal history, expert opinion or scientific documentation, constitute ‘evidence’ – in reality, all these sources are important to evidence-based health care. The initial task, however, is to rank these sources of evidence according to the process by which they are derived – these are our ‘levels’ of evidence. Evidence from clinical trials, particularly those in which two similar but randomly selected samples are compared using a single intervention versus placebo, is assigned a relatively high ranking. At the other end of the spectrum, particularly when no studies exist that answer a specific question, consensus by a group of experts is acceptable as ‘evidence’. The latter, however, ranks lower than the clinical trial. Thus, we have different ‘levels’ of evidence ranging from randomized controlled trials that concur (level 1a) to expert opinion (level 5) (Table 1). All varieties of evidence are subject to potential bias – the ranking of levels of evidence is an attempt to explicitly recognize how likely it is that a particular type of evidence is free from bias. TABLE 1 Levels of evidence One must also appreciate that the ‘level’ of evidence does not necessarily represent ‘quality’ of evidence. Randomized, controlled trials may be poorly designed or may be too small, and, therefore, give misleading results. Similarly, experts may have biases or conflicts of interest that weaken the value of their opinions. Another dimension is the direction that evidence points (whether supporting, refuting or indeterminate with respect to the question). In this respect, it is not uncommon for two similar studies, or two expert committees, to point in opposing directions. Sometimes complex statistical analyses, including meta-analyses, are required to combine similar studies to determine the direction of evidence – experts may still disagree on the direction of evidence. Many readers are aware of the controversy regarding resuscitation of newly born babies with air or 100% oxygen, despite numerous studies on the subject (4). Conflicting interpretations are in the nature of evidence-based appraisal and should not necessarily be considered a failing of the process. Once evidence is interpreted, levels determined, quality appraised and direction charted, one can move on to the next stage: How should I change my practice? At this stage, one should look at the ‘class of recommendation’. When level quality and direction of evidence are unequivocal, for example, when a number of convincing, well-designed randomized controlled trials concur (ie, consistent level 1 studies), the class of recommendation is A. Grade B recommendations reflect studies with largely level 2 or 3 evidence (mostly observations of populations [or ‘cohorts’] or multiple cases, sometimes using historical comparisons or comparisons with other cohorts). With level 4 or 5 evidence, the class of recommendation would be C or D, respectively (Table 2). TABLE 2 Grades of recommendations It should be recognized that level 1 evidence may not always be available (or even possible to obtain) – for example, we will never test antibiotics against placebo in neonatal meningitis. On the other hand, although a grade D recommendation has potential for bias (level 4 or 5), it can be a very important and very strong recommendation. For example, a recommendation that an individual trained and experienced in neonatal intubation should always perform or supervise the procedure is not based on randomized, controlled trials, but is a common sense recommendation designed to promote patient safety. If one returns to the FNC statement, one can now see that most of the recommendations are grade B. We clearly do not have the best evidence, and consequently, clinical acumen – the art of medicine – should play an increased part in decision making. The FNC reinforces this concept by encouraging education of parents, as well as clinical observation. It is, therefore, vitally important to accept that such recommendations are not an exclusive course of action, but guidelines that will be refined as more evidence becomes available. At the least, we know what we do not know and how unsure we are. The Centre for Evidence-Based Medicine process is not unique. A similar methodology was used to develop the 2006 international neonatal resuscitation guidelines (5), and is being introduced into the Canadian neonatal practice with the Evidence-Based Practice for Improving Quality Initiative (6). As we become increasingly aware of these tools, we should also recognize the importance of experience and judgment when we cannot answer clinical questions with evidence. The evidence-based health care language brings clarity and honesty to practice, and in its own way, respects the art of medicine while refining the processes of science.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.