Introduction

With the best of intentions, medical interventions have the capacity to do harm as well as good, and before any new treatment is accepted into clinical practice it should pass through a series of pre-clinical and clinical evaluations to ensure that it is both safe and has some efficacy. Safety is usually first evaluated in laboratory studies followed by observations of either healthy volunteers or patients. Clinical effectiveness must be evaluated through some kinds of comparisons against no treatment or other competing therapies - on either a formal or an informal basis. Only rarely will an intervention be so obviously effective that formal comparison is not required: for example, defibrillation following cardiac arrest. The effects of most interventions are much more subtle and require careful evaluation to ensure that we are indeed measuring the effect of treatment rather than differences in the prognosis and potential responsiveness to treatment between the individuals being compared [1].

Random allocation in a clinical trial ensures that assignment to therapeutic groups is unbiased. There is no other way of creating comparison groups that has this property, because it can never be assumed that all factors relevant to prognosis and responsiveness to treatment have been distributed between the groups in an unbiased way. Although other methods of assigning treatment can attempt to match for factors that are known, there are always likely to be many unknown, and therefore unmatchable, factors that may influence outcome. There is good empirical evidence that the results of non-randomized studies give results different to those of randomized controlled trials addressing the same question [2-4].

The randomized controlled trial is therefore widely accepted as the gold standard of treatment evaluation, and has been an essential tool of medical research since the 1940s [5]. The first clinical trial in which the treatments were assigned by a genuinely random process was probably a British Medical Research Council trial comparing bed-rest with or without streptomycin in the treatment of tuberculosis [6]. The first randomized trial in cancer was initiated in 1954 and compared two regimens of chemotherapy in the treatment of acute leukaemia [7]. Current emphasis on evidence-based medicine has increased the international awareness and acknowledgement of the importance of randomized trials in reliably comparing the advantages and disadvantages of different treatment policies and for accurately assessing their relative impact on the outcome measures of interest.

While the central role of randomization was established many years ago, the environment in which trials are now conducted has changed enormously since those times, and those responsible for running trials have had to and must continue to adapt. The next chapter after this considers one of the areas of the greatest change, namely the public perspective. When designing trials, we need to be certain that we are addressing questions that are clinically relevant, measuring outcomes that are important to patients and their families, and presenting information concerning participation in trials in an accessible and readily understandable format. It is incumbent upon those who conduct trials to help promote the public understanding of the principles involved and to educate and encourage not only patients but also the healthy public to agree to take part in trials as occasions present. In improving communication with patients and increasing their involvement in trials, we need, perhaps, to question the terms we use. 'Randomized controlled trial', for example, means little to most patients. Much thought needs to be given to explaining randomization in ways that are both accurate and reassuring. There is scope for much wider involvement of the medical and scientific communities, patients and patient advocacy groups in promoting together a high standard of collaborative research to which all groups contribute constructively. Our experience suggests that there is an enormous resource of goodwill among patients that can be harnessed.

The first step in initiating a clinical trial is formulating an idea which typically arises from an individual or small group. Much work is then needed to filter out research questions that have already been answered and to translate ideas into realistic and feasible protocols. In Chapter 3 we discuss the importance of deciding, on the basis of information already available and the type of cancer under study, whether a trial is needed and, if so, what type of trial will best address the question posed. We describe the standard stages in developing a new treatment, from establishing its safety and acceptability, through assessing its potential benefits and hazards compared with other treatments already in use, to monitoring its use in routine practice. Understanding these stages, the need for randomization - including its role in phase II trials, and the advantages and disadvantages of all types of comparative study is essential.

Most new treatments will be evaluated in a randomized trial before they can be accepted into clinical practice. In Chapter 4 we describe how to design a randomized trial, drawing particular attention to the nature and number of treatment groups, the timing and unit of randomization and the practical methodology of randomization.

The results of many randomized trials are inconclusive, often because inadequate numbers of patients were included. In Chapter 5 we describe the need to estimate sample sizes realistically, emphasizing the importance of deciding what difference in primary outcome measure should be targeted and the information that needs to be considered in determining this. Reasons why unequal allocation is sometimes desirable, and why and how sample size may need to be adjusted during an intake in the light of event rates, are also discussed.

The assessment of health-related quality of life is becoming increasingly important in cancer clinical trials; indeed, many of the principles and much of the methodology were developed and continue to be developed in the context of such trials. We review the current status in Chapter 6, and emphasize the importance of formulating hypotheses, deciding what outcome measures to study, and how they should be measured.

In Chapters 7 and 8, we describe how plans are put into practice. Once a trial has been designed, much has to be done before it can be launched. We describe what needs to be done, who does it, and what resources are required. To attract funding from a sponsor and support from the clinical community and patient advocacy groups, the proposal must make a persuasive case for conducting the trial and show that it stands a good chance of answering its questions reliably. For a trial to be run successfully and ethically it generally requires independent data monitoring and supervision, and must conform to the general principles of good clinical practice. It should usuallybe conducted by a multi-disciplinary group, competent to organize the trial, to elicit and maintain the trust and enthusiastic collaboration of clinicians, to analyse the data, and to present and publish the results.

This book is not the place for presenting analytical and statistical methodology in great detail, or for reviewing statistical packages, for which specialist texts should be consulted. Nevertheless, all those involved in conducting and collaborating in clinical trials need to be aware of the important principles of analysis, which we cover in Chapter 9. This is essential if they are to appreciate the reliability of the findings of trials in which they are involved and to appraise the literature critically. It is of fundamental importance to have a feel for the nature and variety of the data and hence the methods appropriate for analysing them, to appreciate what is required for making informative and reliable comparisons, and to be able to present, communicate and interpret results.

We discuss reporting and interpreting the results of trials in Chapter 10: what information to present, how and where to report it, and when; and how to interpret it. All trials should be reported, with due attention to CONSORT guidelines, including trials that have to be terminated prematurely; and reports must assure readers that the trial was conducted correctly, that the data were fully and relevantly analysed, and that the findings and their interpretation are reliable and unbiased. We also suggest sensible approaches to communicating with the media and preparing press releases.

A single trial, even a large randomized trial impeccably conducted, will rarely change clinical practice internationally. Systematic reviews and meta-analyses, which we discuss in Chapter 11, are therefore essential tools in objectively evaluating all the available evidence on which treatment decisions are based. The requirements for conducting meta-analyses reliably are substantial. We discuss the practical issues involved, from formulating the question, through writing a protocol, collecting or abstracting data, analysing the data, to reporting the results, drawing particular attention to the value of individual patient data analyses compared with those based on published or summary data. We also highlight the need for international registration of all randomized trials.

Finally, in Chapter 12, we describe the characteristics of, and benefits that accrue from, an established trials centre and research group. The potential benefits are great and stem largely from the ability to build up an experienced staff and a group of experienced collaborators, to plan and conduct coherent programmes of trials in collaboration with other groups over many years, and to conduct trials-associated research based on large accumulations of data.

In summary, we hope that this book will not only prove helpful to other researchers but also play its part in encouraging a positive and collaborative attitude to trials, and particularly randomized trials, in the treatment of cancer, in all sections of the community.

10 Ways To Fight Off Cancer

10 Ways To Fight Off Cancer

Learning About 10 Ways Fight Off Cancer Can Have Amazing Benefits For Your Life The Best Tips On How To Keep This Killer At Bay Discovering that you or a loved one has cancer can be utterly terrifying. All the same, once you comprehend the causes of cancer and learn how to reverse those causes, you or your loved one may have more than a fighting chance of beating out cancer.

Get My Free Ebook


Post a comment