textbook of clinical trials machin d, day s, green s (editors) (2004) isbn 0471987875; 428 pages;...

1
in general. The examples guide the reader through each method and are very much applicable to the relevant chapter. I particularly felt that the PROACTIVE approach was very well described and could be used by any statistician. There are also a number of very good references for each chapter, which would help any reader develop a good understanding of the decision-making area. Also included is a disc to accompany the text that contains supplementary information. This book would be a useful acquisition for a statistician who needs the basics of decision-making, and it would certainly be a good book for a non-statistician. My only criticism would be that it may be too basic for statisticians, who may benefit more by using it solely as a pointer to other publications. Alun Bedding Eli Lilly (DOI: 10.1002/pst.157) Textbook of Clinical Trials Machin D, Day S, Green S (editors) (2004) ISBN 0471987875; 428 pages; £150; h225; $278 Wiley; http://eu.wiley.com/WileyCDA/WileyTitle/productCd- 0471987875.html A woman’s preaching is like a dog’s walking on his hinder legs. It is not done well; but you are surprised to find it done at all. (Dr Samuel Johnson) I have very low expectations of the genre. Edited textbooks, in my view, are nearly always a disaster. The process of getting the authors to deliver is like one of those fictional clinical trials in which you have equal numbers of patients per centre and the trial proceeds at the pace of the slowest recruiter. Unlike such a trial, each author seems to operate to a different protocol and in producing their overview of the whole the editors have to struggle against almost insuperable odds to make sure that the homogeneity test is not failed and that a meaningful overall message can be synthesized. The limitations of the genre mean that this book is not entirely satisfactory, despite having exceeded my expectations. However, although it is only a qualified success it is interesting enough to deserve that any review of it should end on a high note, so I shall save my praise for later and give you my major criticism now. This is simply that the coverage is so eccentric and incomplete that the qualifying phrase ‘some aspects of’ is needed before ‘clinical’ in the title. For instance, we have a whole section on cancer trials, with six chapters for breast, childhood, gastrointestinal, haematologic, melanoma and respiratory but no place for prostate. Other sections cover cardiovascular, dentistry, dermatology and reproductive health. I would have thought that a section on brain and neurological diseases would have been appropriate, with chapters on Parkinsonism, multiple sclerosis, motor- neurone disease, migraine, and perhaps pain more generally, as well as trials in psychiatry and Alzheimer’s. Instead the first five topics on this list are completely missing from the book and we have a whole section on psychiatry, in which, admittedly, there is a chapter on ‘Alzheimer’s disease’, but with not only separate chapters for ‘Anxiety disorders’, ‘Cognitive behaviour’ and ‘Depression’ but a separate introduction as well. This introduction by Brian Everitt is excellent, but it isn’t really relevant to Alzheimer’s and could have replaced the three other chapters in the section (also all very good in their way) to make way for other topics. There are even more serious omissions, however. The preface of the book makes a pitch for topicality in mentioning SARS. However, any reader who takes this as an indication that infectious diseases will be covered will be sorely disappointed. There is nothing on AIDS, nor on malaria, nor on TB, apart from the discussion of the famous MRC trial, and there is nothing on vaccines either and certainly nothing on SARS or on anything that could have the remotest relevance to it. In short, the shape of the book is a complete mess. It fails to cover what it should and covers much that is less important. I suspect that the editors had a different plan originally and have had to abandon it in response to dilatory would-be authors. The stopping rule has used calendar date rather than information accrual or significance. However, this does not mean that there is not much to enjoy in the book. The introductory chapters are all excellent. In particular, I liked the chapters entitled ‘Brief history of clinical trials’ by Day and Ederer and ‘General issues’ by Machin. Many other chapters are also impressive. I particularly admired those on ‘Breast cancer’ by Berry, Smith and Buzdar, on ‘Depression’ by Dunn, and on ‘Respiratory’ by K. all ! en. The first of these gives many insightful discussions of ethics, survival analysis and adaptive designs, the second of causality, and the third is an excellent illustration of how the challenge of a specific disease requires careful thought rather than standard recipes by way of intellectual response. There are many other fine chapters in addition to these. In fact, if the book leaves many conditions uncovered and hence many aspects unaddressed, it nevertheless does succeed in covering very many different areas raising quite distinct methodological problems. Although a little pricy for the private purchaser, it is the sort of book that I would expect any university department with courses in clinical trials methodology to purchase several copies of, as it will provide useful and stimulating reading material for students who desire to progress beyond that which is offered in introductory texts. In short, although the book can be criticized for its sins of omission, there are none of commission. What there is in it is good and the student of clinical trials and the reader of Pharmaceutical Statistics will find much of interest. Stephen Senn Department of Statistics, University of Glasgow, UK (DOI: 10.1002/pst.158) Copyright # 2005 John Wiley & Sons, Ltd. Pharmaceut. Statist. 2005; 4: 81–82 Book reviews 82

Upload: stephen-senn

Post on 06-Jul-2016

215 views

Category:

Documents


0 download

TRANSCRIPT

in general. The examples guide the reader through each

method and are very much applicable to the relevant chapter.

I particularly felt that the PROACTIVE approach was

very well described and could be used by any statistician.

There are also a number of very good references for each

chapter, which would help any reader develop a good

understanding of the decision-making area. Also included is a

disc to accompany the text that contains supplementary

information.

This book would be a useful acquisition for a statistician who

needs the basics of decision-making, and it would certainly be a

good book for a non-statistician. My only criticism would be

that it may be too basic for statisticians, who may benefit more

by using it solely as a pointer to other publications.

Alun Bedding

Eli Lilly

(DOI: 10.1002/pst.157)

Textbook of Clinical Trials

Machin D, Day S, Green S (editors) (2004)

ISBN 0471987875; 428 pages; £150; h225; $278

Wiley; http://eu.wiley.com/WileyCDA/WileyTitle/productCd-

0471987875.html

A woman’s preaching is like a dog’s walking on his hinder

legs. It is not done well; but you are surprised to find it done

at all. (Dr Samuel Johnson)

I have very low expectations of the genre. Edited textbooks, in

my view, are nearly always a disaster. The process of getting the

authors to deliver is like one of those fictional clinical trials in

which you have equal numbers of patients per centre and the

trial proceeds at the pace of the slowest recruiter. Unlike such a

trial, each author seems to operate to a different protocol and in

producing their overview of the whole the editors have to

struggle against almost insuperable odds to make sure that the

homogeneity test is not failed and that a meaningful overall

message can be synthesized.

The limitations of the genre mean that this book is not

entirely satisfactory, despite having exceeded my expectations.

However, although it is only a qualified success it is interesting

enough to deserve that any review of it should end on a high

note, so I shall save my praise for later and give you my major

criticism now. This is simply that the coverage is so eccentric

and incomplete that the qualifying phrase ‘some aspects of’ is

needed before ‘clinical’ in the title.

For instance, we have a whole section on cancer trials, with six

chapters for breast, childhood, gastrointestinal, haematologic,

melanoma and respiratory but no place for prostate. Other

sections cover cardiovascular, dentistry, dermatology and

reproductive health. I would have thought that a section on

brain and neurological diseases would have been appropriate,

with chapters on Parkinsonism, multiple sclerosis, motor-

neurone disease, migraine, and perhaps pain more generally, as

well as trials in psychiatry and Alzheimer’s. Instead the first

five topics on this list are completely missing from the book and

we have a whole section on psychiatry, in which, admittedly,

there is a chapter on ‘Alzheimer’s disease’, but with not only

separate chapters for ‘Anxiety disorders’, ‘Cognitive behaviour’

and ‘Depression’ but a separate introduction as well. This

introduction by Brian Everitt is excellent, but it isn’t really

relevant to Alzheimer’s and could have replaced the three other

chapters in the section (also all very good in their way) to make

way for other topics.

There are even more serious omissions, however. The preface

of the book makes a pitch for topicality in mentioning SARS.

However, any reader who takes this as an indication that

infectious diseases will be covered will be sorely disappointed.

There is nothing on AIDS, nor on malaria, nor on TB, apart from

the discussion of the famous MRC trial, and there is nothing on

vaccines either and certainly nothing on SARS or on anything

that could have the remotest relevance to it. In short, the shape of

the book is a complete mess. It fails to cover what it should and

covers much that is less important. I suspect that the editors had a

different plan originally and have had to abandon it in response to

dilatory would-be authors. The stopping rule has used calendar

date rather than information accrual or significance.

However, this does not mean that there is not much to enjoy

in the book. The introductory chapters are all excellent. In

particular, I liked the chapters entitled ‘Brief history of clinical

trials’ by Day and Ederer and ‘General issues’ by Machin. Many

other chapters are also impressive. I particularly admired those

on ‘Breast cancer’ by Berry, Smith and Buzdar, on ‘Depression’

by Dunn, and on ‘Respiratory’ by K.aall!een. The first of these gives

many insightful discussions of ethics, survival analysis and

adaptive designs, the second of causality, and the third is an

excellent illustration of how the challenge of a specific disease

requires careful thought rather than standard recipes by way of

intellectual response. There are many other fine chapters in

addition to these. In fact, if the book leaves many conditions

uncovered and hence many aspects unaddressed, it nevertheless

does succeed in covering very many different areas raising quite

distinct methodological problems. Although a little pricy for the

private purchaser, it is the sort of book that I would expect any

university department with courses in clinical trials methodology

to purchase several copies of, as it will provide useful and

stimulating reading material for students who desire to progress

beyond that which is offered in introductory texts.

In short, although the book can be criticized for its sins of

omission, there are none of commission. What there is in it is

good and the student of clinical trials and the reader of

Pharmaceutical Statistics will find much of interest.

Stephen Senn

Department of Statistics, University of Glasgow, UK

(DOI: 10.1002/pst.158)

Copyright # 2005 John Wiley & Sons, Ltd. Pharmaceut. Statist. 2005; 4: 81–82

Book reviews82