Skip navigation

Out-takes


New EB Journal on the Block

Evidence-based Health Policy & Management is a sibling publication to Evidence-based Medicine.

The purpose of the new journal is to provide health managers and policy-makers with the best evidence available about the financing, organisation and management of health care. The format is the same as that for Evidence-based Medicine. More than 30 journals are systematically scanned for appropriate articles - reviews or trials - and summarised concisely so that the bones of the results and the quality of evidence can be picked. Most useful are the accompanying commentaries by someone experienced in the field, which help us with the `spin' and allow us to place the results in the proper context.

Again, as with EBM, there are short articles of classic papers, on skills, or educational or evidence resources to help us use the information better. This new journal is inexpensive ([sterling]55 for a years personal subscription), and though the first issue is a bit thin (on pages rather than content, but a journal on evidence based policy and management is never likely to be thick), it will fill out as the editors and their helpers get into their stride.

Evidence-based Health Policy & Management is available from Pearson Professional Ltd, PO Box 77, Harlow, Essex CM19 5BQ, UK. Fax +44 (0) 1279 639609.

Mental Health Information on the Web

Many professionals working in mental health units do not enjoy easy access to library and information services. In order to support evidence-based practice in mental health, the Director of R&D has provided resources for connection to the Internet and the development of OXAMWEB, a collection of mental health information on the Web.

OXAMWEB (website http://strauss.ihs.ox.ac.uk/oxamweb.html ) is the THREE-CLICK SITE, designed to provide useful information within "three clicks", and with "EB" benchmarks soon.

Users will thus go straight to the stuff, rather than waste hours chasing round the Web.

There is both an evidence-based section, and traditional textbook-style information. It is designed for users and their families, as well as professionals.

Do please check it out and send your comments to david.gill@psych.ox.ac.uk .

Book Reviews

Critical Appraisal

The Pocket Guide to Critical Appraisal. IK Crombie. BMJ Publishing Group 1996, 66pp, ISBN 0-7279-1099-X, price £8.95.


A number of people have asked Bandolier 's advice on evaluating papers. The answer is to get some critical appraisal skills. There are a number of ways of doing this, and the best is probably attending a course of the type arranged by the Critical Appraisal Skills Programme (Claire or Katie on 01865 226968). But if you can't do that, or want an introduction, and want something readable, and particularly if you want something simple to keep on your shelf, then Iain Crombie's book is for you.

It covers all the bases, and tells you what to look out for when appraising clinical trials, cohort studies, case-control studies, and review articles. It has some stuff on statistics - but never heavy and always accessible. A useful addition to the shelves that will break neither the bank nor shelf!

Statistics for the frightened

Clinical Investigations and Statistics in Laboratory Medicine
RG Jones & RB Payne. ACB Venture Publications 1997, 192pp, ISBN 0-902429-21-3, price £21.00.


Most of us feel an overwhelming sense of dread if asked to think about statistics. So if you read a book review which told you that there was a statistics book that might make good reading on the beach, you would probably conclude that the reviewer was either deranged or some kind of anorak!

Richard Jones and Brian Payne have managed to write such a book, though. While the title suggests it is for laboratory medicine types, actually this book is a gentle, hand-holding, informative and innovative entrance to healthcare research for anyone. They give some enchanting one-page backgrounds to famous statisticians, which makes them, and their subject, more human. They lead the new researcher into important areas often ignored by the cognoscenti - like how to put together a grant application, and how to make a really good poster presentation.

But most importantly there is nothing in this book to frighten the horses. In their introduction they tell us that statistics needs more symbols than the Greek and Roman alphabets combined - but they use precious few, meaning that those of us who aren't classicists don't feel alienated.

Bandolier was fortunate to have more than one copy. One remains firmly on the shelf for moments of panic (what is a negative predictive value??), while the other went to a new medical student. But the book will be useful for most people of any background who want to set out to do research themselves, or who want to understand research done by others. It looks like a very useful teaching text.

Correspondence: IYDT vs NNT

Some interesting thoughts on data framing from Jon Brassey, TRIP (Turning Research Into Practice), Gwent Health Authority in Pontypool. The continual battle to get research into evidence has to be given as much help as possible. To that end I refer to a psychological phenomena known as resonance. According to Tversky and Kahneman [1] humans are motivated to retain a current freedom rather than to obtain a freedom they could have, but don't.

Now what bearing has that on the real world? Gonzales, Aronson and Costanzo [2] carried out a study looking at resonance and the uptake of home insulation. With half of the group they described how much money they could save each year on heating bills. With the other half they gave the same information but framed it in terms of the amount of money lost. The result - those in the latter group had a 50% higher uptake rate.

Parallel this with NNT. In the above example NNT would be the amount of money saved group. Perhaps it would be better if you turn it around and have an 'If You Don't Treat' (IYDT). So if an intervention has an NNT of three, instead of saying "If you treat three people one will benefit" use "If you don't treat three people one will come to harm". Although I haven't come across any application in the health service I feel it may be of some benefit. If the evidence from outside the health service is applicable, one should see IYDT having a greater impact that NNTs.

References:

  1. A Tversky, D Kahneman. The framing of decisions and the psychology of choice. Science, 1981; 211:453-8.
  2. D Gonzales, D Aronson, K Costanyo. Using social cognition and persuasion to promote energy conservation : a quasi experiment. Journal of Applied Social Psychology, 1988; Vol 18 (12, pt2): 1049-66.



previous or next story in this issue