How Do I Find .......?
How do I find out if the intervention I want to use is effective?
Over the years most clinicians develop a portfolio of interventions for the conditions that present most commonly. Some of these may be ineffective or not cost-effective or both. Now every journal you pick up tells you that improved patient care depends on selecting only those interventions for which the evidence for clinical effectiveness is unequivocal.
I want it now!!It is hard to find the time to check the literature. Often you want the answer immediately. It is tempting to rely on one or two sources, because any venture further afield into the literature shows that a lot of the evidence is contradictory. So what can you do?
Unfortunately, you are not going to find the evidence in one, easily obtainable package, nor will it be in one place. You will need time and information-searching skills. Libraries are an invaluable source of help. The Librarian will be able to search a wide range of databases which will provide pointers to the evidence in the published and semi-published, report-type literature.
SearchingSome of you will want to do the searching yourself. Index Medicus, the main paper-based source for bio-medical and clinical references, is now widely available in compact disk format - CD-ROM. Unfortunately the electronic versions only go back to 1966 so if you need older material the paper version is still important.
You may have bought MEDLINE on CD-ROM for the practice or department. Two companies are marketing CD-ROM versions currently - Silver Platter and CD-Plus. The main difference between the two is the type of search software that they provide; Silver Platter uses SPIRS and CD-Plus uses OVID. Whoever does the search, it is important to understand that there is a hierarchy in the published journal literature, from primary reports of research to reviews and overviews.
Systematic reviews and RCTsWhat you want to look for are systematic reviews of randomised controlled trials (RCTs). These will aggregate the results of several or many RCTs. With aggregation many of contradictions are dissipated. To search MEDLINE for RCTs is more difficult than it sounds, mainly because of the way in which the National Library of Medicine indexes articles for inclusion in MEDLINE. The UK Cochrane Centre in Oxford has made a special study of this problem.
Efficient searching in MEDLINE can also be difficult if you do searches infrequently. Your local librarian will be able to help you work out the best search strategy to find randomised controlled trials and will also be able to give you tips on how best to search MEDLINE in general. It is important to realise that for some questions, MEDLINE may not be the best source of evidence. Other databases exist which focus on different aspects of health care and which may be more appropriate.
You probably subscribe to the BMJ and Lancet. You are certainly deluged with `freebies' and unsolicited magazines. You may have the time to be self-indulgent and browse. If not it is worth while adopting a simple elimination checklist for what you do and don't read. This will mean that you have to set aside some time to determine your criteria and to decide which titles are worth concentrating on and which should be immediately `binned'.
David Sackett and his colleagues in Clinical Epidemiology (pp 362-378) give some useful guides on this process. The ACP Journal Club which comes out every 2 months provides succinct summaries of papers which present sound evidence (see Desert Island Texts ).
New sources of informationIn the last few years important new sources of information about clinical effectiveness have emerged in the UK. The NHS has encouraged and supported the development of three new centres for information - the UK Cochrane Centre in Oxford which produces the database of randomised controlled trials on pregnancy and childbirth, the Centre for Reviews and Dissemination in York which produces the Effective Health Care Bulletin and the Nuffield Institute for Health in Leeds which produces the Outcomes Briefing . It is worth making sure that you are on the mailing list for these publications by writing to one of the addresses below.
It is also worth getting a copy of Executive Letter EL(94)74 dated 28 September 1994 on Improving Effectiveness, either from your local District Health Authority or direct from the Department of Health. It contains lots of useful addresses for information.
We all rely heavily on people as a source of information. On the whole people are more user-friendly than databases. Sometimes the quickest way to find out whether an intervention is effective is to telephone an expert in the field (preferably one that understands the importance of evidence!) or if you don't know who is working in the field you can try using the PHISH (Public Health Information Sharing) database available from Steve Ashworth at Buckinghamshire Health Authority (0296 394022).
InternetFinally, if you are a 'techy' you will know all about the Internet and will need no conversion. For the rest of you it may be worth considering a subscription to the Internet. But be warned! Useful sources of health information may be hard to find, you will run up huge telephone bills and you will need oceans or stratospheres of time to meander about the superhighways of cyberspace. Belonging to one of the international, subject-based mailing lists - newsnets - will give you access to clinicians all over the world with similar interests.
In a future issue of Bandolier we'll tell you how to get hold of a hard copy of the document you want to read. In the meantime if you need help on finding the evidence contact your local postgraduate medical centre librarian or if you are not sure who that is or where the library is, get in touch with Judy Palmer or David Stewart in the Health Care Libraries Unit in the John Radcliffe Hospital and we will point you in the right direction.
previous or next story in this issue