Cancer has been at the forefront of developments in clinical trials methodology for the past half-century. Some of the earliest randomised controlled trials (RCTs) were in the field of oncology and clinical research has routinely used RCTs to assess new therapies. However, before we congratulate ourselves, perhaps prematurely, we need to look to see how the evidence from RCTs and other clinical studies has been used.

The movement called evidence-based medicine has not suddenly invented the concept of using evidence from clinical experiments – Western medicine has been predicated on this concept for many hundreds of years. However, the tool of the RCT has been honed and we now have the means to synthesise information in a systematic manner that reduces the risk of bias. These developments have come at a time when electronic communication has allowed us for the first time to keep an effective track of clinical research, scientific publications and to bring all of the information together using the methodology of systematic reviews.

In the past, the need for reviews of current knowledge was met by "narrative reviews", usually written by an expert in the field. Such individuals were held to have a thorough grasp of the literature and the ability to interpret it. Readers of reviews need unbiased information and there is consistent evidence (including from cancer) that narrative reviews of healthcare interventions rarely use methods

designed to reduce the risk of bias. The fact that narrative reviews are written by experts in the field who have often carried out some of the research that they are reviewing compounds the risk of bias.

Systematic reviews, the bedrock of evidence-based medicine, are designed to reduce the risk of bias. They also bring together all of the pertinent evidence from trials judged to be of good quality. Where appropriate, evidence from these trials can be pooled using the technique of metaanalysis. The result of such a process is to reduce the risk of bias and to maximise the chances of finding a useful outcome, because the intervention is being examined in the largest population possible and not in a few discreet RCTs.

This book uses an evidence-based approach to look at the strength of the underlying evidence used to support some of the key decisions in cancer care. The authors have not been commissioned to carry out systematic reviews to answer each of these questions. Each systematic review is a complex and time-consuming exercise and it would be impractical in a book of this nature. Authors were asked to use systematic searches of the medical literature and to summarise their findings. Where there are systematic

reviews these were presented and discussed in the light of the rest of the literature. Where no systematic reviews were available reviewers summarised the available literature with a particular emphasis on RCTs. However, new systematic reviews were usually not carried out.

The conclusions for each review question have been graded according to the strength of the evidence underlying that conclusion. Readers may wish to use these grades in thinking about the believability of the conclusions, but should bear in mind that such grades are a crude approximation. While they provide a summary of the strength of evidence, they are also included to stimulate readers to automatically think about how believable the evidence really is.

Inevitably, in a book of this type there will be many questions that were not included. The limited list of key questions were selected by the authors with the section editor. Often this was driven by those areas where there was known to be RCTs and sometimes systematic reviews. There is an emphasis on questions deemed to be of significance by clinicians because this is where the research has been carried out. Questions of particular importance to patients and their families have often been much less well researched and because of this are even harder to systematically review.

Users of this book should be able to read about the evidence underlying many of the decisions that underpin our current approach to treating the common cancers. Often clear conclusions elude us because there is a paucity of data and some of it is of doubtful quality. In these circumstances systematic reviews are often an essential starting point of new trials.

I hope that readers will find that this book is a useful starting place when looking for evidence for how we currently treat cancer.