Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Book reviews


Introducing palliative care. Robert Twycross. Radcliffe Medical Press 2003, Fourth Edition. ISBN 1-85775-915-X, 190 pp, £not known.

Palliative care is an interesting topic for evidence-based medicine. First of all, there is little evidence in terms of randomised controlled trials, and therefore few systematic reviews, though the Pain, Palliative, and Supportive Care Cochrane review group and their reviewers are beginning to help us understand what evidence there is. Then there is the question of the difficulty of conducting clinical research in palliative care.

It is complicated by palliative care being a complex package of care tailored to each individual patient, and the fact that, by its very nature, patients die, making conventional trials difficult. After a while one wonders about whether this makes clinical trials themselves the best way to understand evidence, and though that is forced by the difficulties, satisfactory alternatives are themselves hard to find.

In these circumstances, the insights from a lifetime of experience help, and Robert Twycross provides this in the most recent edition of his book. There is lots of good stuff here about things that might not always be in the forefront of one's mind, like the special requirements of different religious beliefs. But perhaps what sets it out as special is the section on symptom management, with evaluation, explanation, management, monitoring and attention to detail (EEMMA as an acronym), and particularly the guidelines. These include guidelines for starting a patient on oral morphine, or transdermal fentanyl, or management of opioid constipation.

All in all a useful book for professional carers, and an opportunity for the younger and less experienced to gain some valuable insights into palliative care.


Systematic reviews to support evidence-based medicine. Khalid Khan, Regina Kunz, Jos Kleijnen, Gerd Antes. Royal Society of Medicine Press 2003. ISBN1-85315-525-X. 135 pp, £not known.

If you want an unashamedly Cochrane view of systematic reviews, then this is the book for you. It sets out the steps of framing questions, identifying literature, assessing quality, summarising evidence and interpreting findings in exactly the way it is done by the Cochrane Collaboration. It is one way of looking at the world of reviews, but not everyone will agree with it.

Bandolier readers will find it unfamiliar territory, because they will find few L'Abbé plots or NNTs, and instead be immersed in Forrest plots, odds ratios and relative risk. There is much talk of heterogeneity, and the material on clinical heterogeneity is good. But there is also stuff on funnel plots for publication bias that may just be plain wrong.

Validity of trials is a difficult, but enormously important topic. Validity can mean many things, but the single most important meaning is to that of whether a trial has the ability to answer a question. The dictionary definition is that which is sound or defensible. Validity is situation dependent, and criteria for validity might include the severity of a condition, the dose or intensity of intervention, the duration of an intervention or the duration of observation. Validity is crucial to what trials go into a review. Yet validity is found only as a glossary entry in the book.

This book will undoubtedly help some people to do their own reviews, and some of those reviews will be right. The worry is whether the book will stop them doing reviews which are wrong. That may be too harsh a judgement, but it is a worry.


Medical statistics made clear. Ashis Bannerjee. Royal Society of Medicine Press 2003. ISBN 1-85315-544-6. 137 pp, £not known.

So what is a z score? Or a risk ratio? Or a sensitivity analysis? Most of us have problems with some or all of these. We sort of know what they mean, but would be hard pressed to write an essay on them, and we hope we understand them enough to make sense of other people's use of them.

This neat little book is not a statistical text book, but rather a series of definitions and explanations, mostly in layman's language, and without requiring any great mathematical knowledge or expertise. The section on epidemiological concepts has interesting sections listing things like evidence supporting causality, and evaluation of randomised trials, and that old bugbear, confounding. There is even a nice definition of the Hawthorne effect. Later chapters even have simple things well explained, like a Latin square.

This is a useful little book for the shelf, and you will soon find yourself thumbing through it.


The menopause. What you need to know. Margaret Rees, David Purdie, Sally Hope. BMS Publications 2003. ISBN 0-9536228-2-7. 105 pp £6.99.

The updated 2002 version of this useful monograph, which has been around since 1994. Bandolier reviewed its predecessor in 1999 and found it useful then. It is even better now, with some nice treatment of risks. As before, it is a useful book for healthcare professionals, but will, notwithstanding, be of interest to informed consumers, so accessible is the writing.
previous story in this issue