The value of participating in clinical trials: the whole is greater than the sum of its parts

Engaging clinicians in the process of trial design, and the ensuing dialogue about best practice, can improve routine care

Evidence-based medicine remains the Holy Grail for clinicians and policy makers, and is dependent on the results of large randomised phase 3 clinical trials and on meta-analyses of all related phase 3 trials of treatment in a specific patient population.1 Irrespective of whether clinical trials test a new intervention or a drug or are comparative effectiveness studies, the findings have the potential to not only affect routine clinical practice but to inform the evidence base to lay a solid foundation for optimal care in the future as well as raise new research questions. The cost of these large trials and the resources required to mount such studies are considerable, which inevitably raises questions about whether the study is essential to answer a clinical question. Indeed, it is not uncommon to be asked by grant reviewers “do we need to do these studies in Australia?”

However, it is worth reflecting on the impact of randomised controlled trials on clinical practice to appreciate the associated ripple effects associated with participating in clinical trials. For example, consider a specific trial designed to test the role of pre-operative radiotherapy added to standard pre-operative chemotherapy before total gastrectomy for localised gastric cancer (TOPGEAR trial; ClinicalTrials.gov identifier: NCT01924819). This randomised study compares the experimental treatment (ie, the addition of pre-operative radiation therapy) with the standard of care (in this example, pre-operative chemotherapy alone).2 If the patients randomly allocated to the experimental arm have a statistically significant and clinically meaningful improvement in survival and quality of life over the control group, with acceptable toxicity, the trial findings will lead to a new standard of care. But for such a study to be relevant for day-to-day clinical practice, investigators must first agree on what is the optimal standard of care against which the new intervention is to be tested. If the standard arm is not considered “state of the art” evidence-based treatment, then any benefit associated with the experimental intervention is unlikely to influence clinical practice.

For complex diseases, such as locally advanced gastric cancers, that require multidisciplinary input, defining the optimal standard of care is not always simple, even among experts, as there are many variables to consider. Standard care (based on the results of previous trials) needs to include consensus on what constitutes standard pre-operative treatment, the type and extent of baseline investigations, the type and extent of surgery, the extent of lymph node dissection, the use and type of post-operative treatment, the nature and duration of post-operative tests as well as pre- and post-operative supportive care — in other words, every aspect of diagnosis, treatment and follow-up must be standardised after considering the best available evidence before randomisation to the experimental intervention occurs.

At the end of this process, but well before the grant application has even been written, well before the protocol has received approval from any human research ethics committee and, in this case, years before the results of the randomised study are known, many clinicians involved (even peripherally) in this process of trial design have gained new insights and knowledge about the condition to be studied. While it is acknowledged that a systematic review or similar review of the literature might equally be capable of defining best practice, the intrinsic nature of a large multicentre clinical trial is such that many clinicians across a range of disciplines need to engage and agree on what constitutes best practice for the control arm for the study to proceed. Hence, our hypothesis is that engaging clinicians in the process of the design of clinical trials, and the ensuing dialogue of what is best practice based on an analysis of previous studies, can itself lead to a change in routine care, potentially one of the most important yet intangible advantages of conducting clinical trials through Australian clinical trials networks. Indeed, were TOPGEAR to have been conducted in Europe or the United States without Australian involvement, the stimulus for the changes in routine practice illustrated above, arguably, would not have occurred until well after the results of the study were published.

Does this argument stand the test of real world observations? That trial participants do better than their non-trial participants is widely accepted,3 but the explanation for this phenomenon may be multifactorial, including stricter adherence to protocol-based interventions and detailed inclusion and exclusion criteria that select fitter patients. However, although limited, there are data supporting the hypothesis that institutions involved in research have better outcomes, independent of the actual research findings. In one study of over 170 000 patients enrolled into trials of acute coronary syndromes, centres were divided into tertiles (ranging from medians of 0% to 4.9% of patients) based on trial accrual.4 Adherence to evidence-based guidelines increased from 76.9% to 81.1% (lowest to highest tertile; P = 0.008) and in-hospital mortality was significantly reduced from 5.9% to 3.5% (P = 0.003; adjusted for all hospital- and patient-level variables) in high tertile hospitals, despite the fact that the majority of patients were not enrolled into these studies.4 In the United Kingdom, the National Health Service Trusts most active in research, based on extent of research funding or recruitment, demonstrated better risk-adjusted inpatient survival in funding tertiles (lowest v highest; odds ratio [OR], 1.05; 95% CI, 1.033–1.068; P < 0.0001) and patient recruitment tertiles (lowest v highest; OR, 1.069; 95% CI, 1.052–1.086; P < 0.0001).5 Similar benefits of research participation, which increased with sustained participation from 2001–2008, were seen in post-operative mortality and 5-year survival for over 200 000 patients with colorectal cancer across all National Health Service hospitals caring for patients with colorectal cancer.6

We are not suggesting that these data are conclusive, as these compelling but nonetheless observational findings (the specific explanations for which are also likely to be multifactorial) are only supportive of the central hypothesis. Nor do we advocate that every proposed, large phase 3 trial should be automatically funded for these reasons alone. Careful consideration needs to be given to the importance of the question being addressed and how this could affect patient outcomes. In addition, due consideration needs to be given to the extra effort required by volunteers on human research ethics committees; the additional time commitments and effort required from research nurses and data monitors, biostatisticians and relevant clinicians; and the vital contributions patients make towards improving the health care of future generations. Indeed, as the workload required to fund, initiate, complete, analyse and report a clinical study continues to escalate, many will question whether there are sufficient resources to justify the implications of such a conceptual approach.

However, when considered against the enormous resources committed to so many existing clinical activities for which the evidence base is unknown,7 perhaps a readjustment of our thinking is warranted. At least some of our so-called standard clinical activities are likely to add substantial cost to the system but are potentially of minimal clinical benefit, an observation confirmed by the data underpinning national disinvestment approaches such as the EVOLVE program.8

Indeed, we may ultimately find that engaging a broad group of clinicians through existing and emerging trials networks in a trials-based approach to health care represents a critical and cost-effective investment that not only improves quality of care and patient outcomes, but somewhat paradoxically improves standards of care well before the results of clinical trials are known. Funding and resources to conduct clinical trials will undoubtedly remain an increasing challenge; however, to the extent that the Medical Research Future Fund strategy has identified the need for new investment in the infrastructure of clinical trials networks9 and large scale clinical trials, there is good reason for optimism.