Commentary: 'The end of clinical freedom': relevance in the era of evidence-based medicine

‘Clinical freedom is dead, and no one need regret its passing’, admonishes John Hampton, in his classic essay in which he urges individual physicians to set aside their opinions, dogma and biases and, instead, base their clinical practice on evidence from rigorous randomized controlled trials (RCTs).1 Using clinical examples in cardiology for which there was uncertainty in the early 1980s, Hampton posed two challenges.1 First, the implementation of unproven management and diagnostic strategies has potential negative implications for both patients and healthcare systems. Second, where evidence exists (either supporting the efficacy of a particular intervention or lack thereof), health-care providers and practitioners fail to implement the evidence in practice, as they are ‘seduced’ with theoretical ‘benefits’ of new and unproven management options. Quality of healthcare has been defined as ‘the degree to which health services for individuals and populations increase the likelihood of desired health outcomes and are consistent with current professional knowledge’.2 Simply stated, poor quality healthcare can involve too much care, too little care, the wrong care or the right care delivered poorly.3 Despite a changing culture towards the implementation of evidence in practice, why are these issues still relevant today? In the three decades since Hampton’s call for more evidence to support medical practice, the number of published RCTs (considered to be the touchstone of excellent evidence) has grown exponentially.4 Furthermore, the number of publications related to the field of cardiology has increased 3-fold (4912 in 1995 vs 11 298 in 2005) in the past decade alone.5 However, there are still a vast number of interventions widely used for which there is little good-quality evidence to support their use. As highlighted in a review by Tricoci et al.,6 only a mere 314 (1.2%) of the 27 111 guideline recommendations put forth by the American College of Cardiology/American Heart Association (ACC/AHA) between 1984 and 2008 were supported by level of evidence A (data derived from multiple RCTs or meta-analyses). Furthermore, only 19% of class I recommendations (i.e. procedure or treatment is beneficial, useful and effective) have level of evidence A.6 Not every guideline recommendation can be, or needs to be, evaluated with an RCT, as persuasive evidence supporting a recommendation can be occasionally obtained from studies other than RCTs, especially when the harms or benefits are large (i.e. smoking cessation). Even with this proviso, it is of concern that the majority of clinical practice, even in cardiology—which has a stronger tradition of conducting randomized trials—is still based on weak evidence. Implementing therapies or diagnostic modalities without appropriate clinical research poses two potential problems, as highlighted by Hampton.1 First, patients may be exposed to potential harm if new technologies are applied to the population without undergoing proper investigation. Furthermore, this practice leads to inappropriate expenditures by the healthcare systems that are increasingly constrained and have limited resources. Use of potentially useless or harmful interventions persists today, as it has been demonstrated that ∼20% of patients with a chronic condition and 30% of patients with an acute condition received ‘contraindicated’ care.3 While the appeal of novel therapies and new diagnostic strategies are enticing to healthcare professionals, their widespread implementation prior to their reliable evaluation in a population can have detrimental effects. For instance, a significant body of observational data suggested that hormone replacement therapy (HRT) may prevent coronary artery disease and osteoporosis in post-menopausal women. HRT was widely adopted and prescribed to ∼40% of women between the age of 50 and 74 years old in the USA.7 However, following the rigorous evaluation of this therapy in a large definitive trial that randomized over 16 000 women to receive combined oestrogen-progestin therapy vs placebo, HRT was proven to be harmful, exposing individuals to an increased risk of breast cancer, coronary artery disease, stroke and venous thromboembolism.8 Other notable interventions in cardiology that were initially considered promising but were subsequently proven to be harmful, include the use of anti-arrhythmic medications for the suppression of ventricular ectopy post-myocardial infarction and the use of the phosphodiesterase inhibitor, milrinone, as therapy for patients with decompensated heart failure.9,10 Novel interventions in healthcare are often implemented prior to quality evidence supporting their benefits are available. The withdrawal of such interventions from routine care, when evidence against their value emerges, is an even more challenging problem. Despite sound evidence refuting their benefit or even demonstrating harm, such interventions may be so strongly ingrained in clinical practice that change seems to be an insurmountable task. Such barriers to change have been identified at the patient, physician and healthcare system level. Significant barriers at the physician level include issues relating to knowledge management (i.e. overwhelming volume of literature to review, access to current research and lack of critical appraisal skills), financial disincentives, organizational barriers (i.e. lack of facilities or equipment), peer group barriers in which local standards of care are not in line with current guidelines, issues relating to the healthcare professional’s knowledge, attitudes and skills, and professional–patient interaction barriers.11 The institution of evidence-based for the appropriate integration of novel interventions into the healthcare system, including a priori strategies for their potential removal, is required. In addition to the overuse of ineffective and inadequately evaluated treatments, there is also an underuse of treatments in clinical practice that are supported by quality evidence. Unfortunately, even when robust evidence strongly supports clinical practice, healthcare professionals often fail to deliver the expected level of care. This evidence–practice gap persists today as studies suggest that 30–40% of patients with acute and chronic care conditions do not receive care according to current scientific evidence.12 Furthermore, preventative care studies show that 50% of patients do not receive recommended care.3 This situation is likely to be far worse in low- and middle-income countries. For instance, high-quality evidence with multiple RCTs and class I A recommendations from national guidelines support statin use for the secondary prevention of cardiovascular events.13,14,15,16,17 However, despite this evidence and recommendations, multiple studies have demonstrated dismal rates of statin adherence in cardiac patients 6-month post-discharge from hospital in developed countries, and far lower rates in developing countries.18,19,20 Thus, while there is a need to continue to produce high-quality evidence, there is just as importantly a need to translate that knowledge into clinical practice in order to minimize the underuse of proven therapies and the overuse of potentially ineffective or harmful therapies. Hampton’s appeal for the ‘end of clinical freedom’ and the need for clinical trials to guide medical decision-making continues to resonate today. Healthcare professionals and healthcare systems must strive to reduce the evidence–practice gap in order to limit the exposure of patients to unnecessary risks of iatrogenic harms and reduce wasteful healthcare systems’ expenditure. As Hampton puts it, ‘One man’s provision is another man’s deprivation’.1 The funds that are potentially wasted in inappropriate healthcare can be put to better use to provide much needed and proven services and treatments, with likely improved health outcomes. While the ultimate goal of evidence-based medicine is to improve the under-use of beneficial interventions (clear evidence that benefits outweigh the harms), and reduce the over-use of harmful interventions (clear evidence that harms outweigh the benefits), significant practice variations persist most notably in the grey zone, where there is often a lack of evidence or the best choice depends on how patients value the benefits vs harms.21 While we should strive to reduce the grey zone by generating high-quality evidence, healthcare professionals will continue to be faced with tailoring management to meet the needs of their individual patients (particularization of management). Regardless of the quality of large, multi-centre RCTs, the adoption of study findings into clinical practice always requires a combination of clinical and scientific judgement. Furthermore, as outlined above, often a mismatch persists today even in relatively well-researched fields, such as cardiology, between the recommendations contained in clinical practice guidelines and the evidence that supports such recommendations. Therefore, until we have quality evidence supporting every clinical recommendation, a degree of clinical freedom is inevitable. Even when the evidence for or against a treatment emerges, the particularization of evidence-based medicine must continue to combine individual clinical expertise (which integrates patient presentation, co-morbidities, preferences, costs and setting) with the best available evidence.22 Conflict of interest: None declared.

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.