CONTENTS

EDITORIAL

COLUMNS

REVIEWS

SHORT PAPERS

MEETING REPORTS

ANNOUNCEMENTS

Issue 3, December 1997

ISSN 1368-1591

Health on the Internet

Anthony, Denis
Health on the Internet
Blackwell Science 1996 158p. 063040726
URL: http://www.blackwell-science.com/PRODUCTS/BOOKS/4062.htm

This book gives a very broad and often quite technical overview of the Internet using some examples relevant to Health Care Professionals. It includes useful and sometimes thought provoking chapters on Connecting to the Network , Commercial Uses of the Internet, Electronic Publishing, Distributed Teaching Resources, Setting up an On-Line Service, Telnet, Downloading Files, Electronic Mail, Electronic Mail Lists, and USENET Newsgroups.

Like any book on the rapidly developing Internet some of the information it contains is already out of date. The most obvious example is Denis Anthony's own site NURSE which is referred to several times. When the reader visits the site he finds that it is no longer updated. Denis Anthony has, however, made some useful hot links from the original page. For contacts the link is to Nursing Connection on http://www.nurse-dk.com and for other services Nursing and Health Care Resources on the Net at the University of Sheffield (http://www.shef.ac.uk/~nhcon).

I was surprised that OMNI, the Organising Medical Networked Information evaluated UK resource site (http://www.omni.ac.uk) is not mentioned and that not all those sites which are recommended have their full URLs clearly listed. The world wide web searching chapter does not discuss 'free text' searching or hierarchical subject listing nor does it give examples of the commercial search engines.

Health on the Internet was not, however, intended to be merely a guide to Internet resources but rather to inform health care professionals about its potential role in learning and communication. In this it succeeds.

Anne H. Weist BA RGN MSc ALA MIInfSci
Trust Librarian
Whipps Cross Hospital
London, UK

Medicine and the Internet

Luis G. Pareras, MD
Medicine and the Internet
Little Brown and Company, London, c1996
ISBN 0316690597

This new book about the Internet is aimed at medical practitioners and 'claims to be 'the best independent reference guide to the most medical information on the Internet'. Weighing in at a door-stopping six-hundred and fifty-three pages it certainly seems to be one of the biggest.

Written by an expert in medical informatics, this rather hefty tome seems to be trying to be a basic introduction to the Internet as well as an essential reference guide. Unfortunately it really only succeeds on one of these counts.

Dividing the book into five parts, the author starts from scratch with a brief history of the Internet, the basic protocols behind it and the requirements of connection. There then follows a brief introduction to some of the main Internet tools and a description of some of the current and future medical applications of the Internet. The second part deals with communications and includes exhaustive sections on e-mail, listservs, IRC and newsgroups. Part three covers the major Internet tools (including the WWW, Gopher, telnet and FTP). The fourth part covers searching the Internet and finding individuals on the Internet. Mention is made of many of the usual suspects such as Medical Matrix, Yahoo and Webcrawler (although OMNI is ignored). I was pleased to see that he follows my own standard maxim about getting someone's e-mail address - try phoning them before embarking on anything more complicated. The final part is a descriptive guide to selected medical resources on the Web. At the end of each chapter there are lists of appropriate Internet resources and addresses.

Every chapter in this book is extremely detailed and contains step-by-step instructions. Indeed at times the author seems to go into too much detail, especially when dealing with some of the more elderly tools such as GOPHER, ARCHIE and VERONICA (which he strangely includes in a section entitled Tools of the Future).

This concentration on some of the older Internet tools could in part reflect the fact that this book was published in 1996 and like many books on this topic is already slightly out of date. PUBMED, for example, is not mentioned and a sizable minority of the addresses and resources cited were no longer available when I tried to access them. There are also a number of curious omissions in a book of this size. Otherwise comprehensive chapters on e-mail and listservs contain only vague references to netiquette and there are only the briefest mentions of issues such as the quality of Internet resources or confidentiality. The book itself is well laid out and has a detailed table of contents but this does not compensate for a lack of an index which is surely a major flaw in a work this size.

Overall I would not recommend this work to those seeking a quick and basic introduction to the Internet - its sheer size militates against this. The author himself acknowledges that, 'the lack of time available to doctors for activities other than the practice of their profession is well known', and there are more manageable texts on the market. However, this book may well prove useful to those seeking a basic reference work to dip into now and again when looking for detailed information on a specific tool or Internet application.

Donald M. Mackay MA(Hons) DipLIS ALA
Librarian
Health Education Board for Scotland
UK

Medicine and the Internet

McKenzie, Bruce C
Medicine and the Internet, 2nd Edition
Oxford University Press, 1997, 353 pp
ISBN: 0-19-2628526
£16.95
URL: http://www.oup.co.uk/scimed

Editor's note: 'In recognition that the Editor of SIMQ may have a conflict of interest in the publication of this review, editorial responsibility for this review has been assumed solely by the Reviews Editor, Mr Frank Norman, and the decision to publish the review made by confidential voting of the Editorial Board Members, with the Editor abstaining.'

Now in its second edition, the excellent Medicine and the Internet is an extremely comprehensive guide to the business of connecting to, using and getting the most out of, the Internet. Although aimed at healthcare professionals and doctors in particular, the first part of the book, which discusses setting up a computer and modem and establishing a connection, would be as useful for a layperson.

The subtitle, Introducing Online Resources and Terminology, neatly sums up the author's approach to his subject. In his preface he firmly states his intention not to avoid jargon as 'Learning the language (like the Latin of gross anatomy) makes universal communication possible', a concept which most doctors will readily appreciate. This is therefore a book that does not talk down to its readers, it expects and demands a lot but the rewards are generous. The clear detailed exposition of all the technical and ethical issues surrounding the use of the Internet as well as the valuable reference aspect of the work should inspire any new or seasoned Internet user with the confidence to explore further. Because of this approach though, the work is probably less suitable for the complete computer novice but more useful for the doctor, healthcare professional or researcher who has some knowledge of computers and the terminology perhaps through use of their practice, hospital or university computer and who is prepared to put time and effort into expanding that knowledge.

It is well written with chapters ranging from the basics of setting up a computer and modem and getting signed up with a service provider, through accessing health related resources on the Internet with a look at specific examples, to setting up personal Web pages and becoming an information provider. Along the way are chapters on e-mail and Usenet as well as some of the less well-known Internet tools such as Archie, Gopher and Veronica, which tend to get overlooked. The chapter on ethics will be appreciated by those concerned about online medical ethics, netiquette and security issues. The work has been enhanced by material written by outside contributors covering specialised fields such as the setting up of a medical mailing list or the pitfalls of communication with patients by e-mail, a problem which the UK reader has yet to encounter but which is relevant elsewhere particularly in the USA.

The layout is good with the graphics being mainly screen shots with a few line drawings and photographs. Technical points and useful hints and tips are highlighted. There is an excellent glossary, index and an extensive list of well-chosen World-Wide Web links. These sorts of lists usually suffer from being out of date before the book is published but the problem is solved in this case by having an associated Web site which contains both updates to the book (which may be incorporated into future editions) and an updated list of hypertext links. This is a brilliant idea and safeguards the investment made in the printed text.

Overall then if you are a healthcare professional with some familiarity with computers and you need to get online, then buy this book, follow the links, keep it updated and you will probably have no need to refer to anything else.

Anne A. P. Hicks MBChB MRCGP BA
General Practitioner
Blackbrook Surgery
Taunton, UK

Free Medline: PubMed and Internet Grateful Med

URL: http://www.nlm.nih.gov/databases/freemedl.html

Medline is one of the most successful databases in the world. It contains nearly 9 million records dating from 1966 to present and has a highly effective indexing system that has made it the information source of choice for medical information seekers across the world.

Medline or its predecessor, the MEDical Literature Analysis and Retrieval System (MEDLARS), was one of the first literature indexing services to use the emerging computer technology of the early 1960s for the 'more rapid production of bibliographic publications and for conducting individualized searches of the literature for health professionals via offline processing'. Over the years MEDLARS has changed, first from an off-line service to an online service; then becoming available on CDROM and now finally into a WWW service. As computer technology has changed, so too has the sophistication of the database interfaces, such that now anyone with an Internet connection can use what was once an arcane and difficult database. Medline has probably been made available in more different versions than any other database. While these (usually) have all contained the same basic bibliographic data, it has been clear that different search interfaces can give radically different search results. The US National Library of Medicine (NLM) have made two different WWW interfaces to Medline available free of charge to the world at large. The NLM Press Release puts emphasis on Medline as a source of information for patients - 'good information [is] the best medicine', but it is more properly seen as a major link in the chain of information sources for evidence-based practice.

PubMed

The origins of the PubMed project lie in the Entrez service, which started in 1992. This was a combined interface to gene sequence and protein sequence databanks, with links between related records. A subset of Medline records dealing with molecular biology was added and was gradually expanded . By late 1996 the whole Medline file was incorporated into the Entrez interface. Soon after this the name 'PubMed' appeared and in May 1997 the NLM anounced that PubMed was to remain free of charge as a major route to searching Medline.

There are several components to the search interface in PubMed. The default is the 'Basic search' screen, but the 'Advanced search' screen is only slightly more complex and much more useful as it allows you to modify your search query as you proceed. Both of these screens are easy to use for a very simple search, but difficult to use for slightly more complex searches. There are some clever features, such as automatically recognising that a term like 'Baldock C' should be searched as an author, and automatically recognising many phrases as terms, so that 'health service' is searched as a phrase rather than as two separate terms. Sometimes this cleverness fails though. If you search for 'Internet', the system automatically changes your search to include the much more general term 'Computer communication networks', thereby retrieving large number of references mostly having nothing to do with the Internet. The step-by-step way in which the Advanced search screen allows you to build up a search statement is reasonably clear if you are familiar with the terminology of Boolean searching, but a little laborious. You can type in complex search statements directly but only if you can master the PubMed command language, which has a number of quirks and is generally not user-friendly. Searching is made easier by the 'Journal browser', which makes it easy to locate and search for a journal title without needing to know the correct abbreviation, and the 'Clinical queries' screen. The latter is intended for clinicians and has built-in search filters with the objective of reducing the retrieval to 'articles that report research conducted with specific methodologies'. Whether an easier search is always a better search is perhaps questionable, and there are those who will worry that searchers may rely excessively on these search filters, rather than attempting to think for themselves.

Searching on Pubmed is easy but not always effective. However, the interesting part is yet to come. When your search is completed, the results display on the screen in a brief format, with a feature for selecting articles which are of interest. Clicking on the author generates the complete Medline record, including the abstract if it is available, or you can display the complete record for all articles you have selected (N.B 'complete' in this case does not mean the full text of the article). The complete Medline record may also include a number of hyper-links and these are what mark out PubMed from other interfaces. Firstly, there are links when appropriate to related gene sequence and protein sequence records in the Entrez service, also to chromosome maps and 3D structures of proteins. These links are a reflection of PubMed's origins as a tool for researchers. It also has links to OMIM records (Online Mendelian Inheritance in Man), a useful compilation of information on genetic diseases and syndromes. A record may also have a link to the full-text of the article, where a publisher makes this available in a form that Pubmed can work with (however, you should be aware that you may not beable to view the full-text if you do not have a subscription to the title). Finally, and most importantly, there is a link to 'related articles'. For every article in Medline a list of related articles (or 'neighbors') is generated by using an automatic algorithm that compares the title, abstract and MeSH terms of each article. If you click on the 'Related records' button you can view these articles. If you prefer, you can select articles of interest in your search results, and then view the related records for the selected articles all together. You can continue following links to further related records indefinitely.

Internet Grateful Med

If Pubmed is ideal for browsers, then Internet Grateful Med (IGM) is an interface for searchers. IGM developed from an NLM program called 'Grateful Med', that allowed PC users to dial into Medline via phone lines and submit a search interactively. In April 1996 IGM was launched as a service to registered users only, then as a freely available service for all in May 1997. The main IGM search screen is more complex than PubMed's, but it usefully includes easy ways to limit results by date, language, publication type, age groups and gender. It also has a very powerful feature: 'Find meta/MeSH terms' that can suggest indexing and other terms to search with. A full examination of this feature is beyond the scope of this review, but suffice to say that it can guide novices and experts alike through the maze of medical terminology, and help to ensure you do not overlook relevant terms when you are building up your search strategy. Another feature 'Analyze search' can suggest ways to modify your chosen search terms to make them fit Medline's indexing structure better. Both of these features will probably be clearly understood only after reading the excellent survival guide. With the aid of these two features, and a little thought, IGM opens up the intricacies of effective Medline searching to all. It does not allow any of the linking that PubMed features however.

Database content

Both PubMed and IGM include Medline back to 1966, as well as 'OldMedline' (1964-65) and 'PreMedline' (articles in process, perhaps just one week after original publication). IGM in addition allows you to search several other databases, including some AIDS databases, Health Services databases, a History of Medicine database and a directory of health organisations. There are plans eventually to make further NLM databases available via IGM. Both interfaces allow results to be printed or saved to a file (using the basic WWW browser features). Records thus saved can be reused in personal bibliographic databases.

Searching or linking?

For the researcher, with genetic or molecular biology interests, PubMed's Medline is very attractive because it sits at the centre of a large mass of heavily interconnected, inter-related data. Those with more clinical interests should not assume that PubMed is the best system for them, just because it is available on the WWW, but it is a system with some valuable features. If the functionality of Internet Grateful Med could be built into it, PubMed would become a serious rival to the established interfaces.

Traditionally the way to use Medline has been to search for 'keywords', authors' names or indexing terms, and if that failed then to think of a different term and try a new search. PubMed introduces a new idea: linking to related articles and records in other databases. It does not have a particularly powerful or flexible search system - all but the simplest of searches require a clunky search language that is difficult to remember. However, using the hyperlinks that are integral to the WWW (where you click on a word to retrieve another document), PubMed lets you start with one search, select relevant records and then at the click of a mouse view records that are on similar subjects. Hence linking rather than searching is the key to using PubMed. In some ways it is an experiment - it is very useful and is quite an intuitive way of browsing through relevant articles, but can you be sure of finding everything on a topic just by following links unless you begin with a good, well-thought-out and comprehensive search? Until this question is answered , access to systems with better search interfaces (e.g. Ovid, SilverPlatter or Knowledge Finder) continues to be necessary.

Frank Norman BSc Dip.Lib
Deputy Librarian
National Institute for Medical Research, UK

The Virtual Hospital

URL: http://www.vh.org

Hospital webmasters across the globe must envy the University of Iowa College of Medicine. Iowa took the initiative in establishing their Internet project in 1993, when the WWW was still relatively novel to most people, and thereby earned the right to what must be one of the most resonant titles on the web - the Virtual Hospital. It is a project of the Electric Differential Multimedia Laboratory, in the Department of Radiology at Iowa, and grew out of earlier work there in developing multimedia teaching materials. It aims to provide 'patient care support and distance learning to practising physicians and other healthcare professionals', as well as information to the general public on common health topics. Now 4 years old, the Virtual Hospital has earned an enviable number of awards and commendations. It is a medical WWW site to be reckoned with.

Quality medical information

Looking at the site as a whole, it is clear that there is a commitment to presenting high quality medical information. The site subscribes to the HON code principles, (although this fact is not displayed on the front page or any other prominent page). The HON code is a code of conduct for medical Web sites, developed by the Health on the Net Foundation and intended to encourage Web sites to provide medical/health information that meets specific quality standards. In addition, every page of the Virtual Hospital is scrupulously labelled with full authorship details and a statement of its peer review status. The presentation and organization of the information across the site is also exemplary - attractive and readable. Unfortunately there are one or two 'dead' links, apparently resulting from recent reorganisation of the site.

A large portion of the content of the Virtual Hospital has been provided by the faculty of the University of Iowa College of Medicine, and staff of the University of Iowa Hospitals & Clinics. In addition, it re-publishes some patient information that is in the public domain, particularly from US government sources.

Because of its large size, it is difficult at first to get a clear impression of how much material is in the Virtual Hospital. It is extensive, though not entirely comprehensive. There are about 20 'multimedia textbooks', 16 teaching files, 9 patient simulations, 8 grand rounds/lectures, 9 clinical practice guidelines, 4 continuing medical education courses and 2 clinical reference handbooks. These cover a wide range of topics but, especially with the textbooks and teaching materials, there is a bias towards imaging, the respiratory system and paediatrics, presumably reflecting the interests of the site's developers. Most major areas of medicine are represented, but some by only 1 or 2 resources. This is partly made up for by the presence of the University of Iowa Family Practice Handbook, an electronic version of the 700 page book of the same name (published by Mosby, 1990). Separate pages list resources helpful to medical students, resources for naval medicine, and resources for child health. The medical students pages also usefully point towards other comprehensive resources (such as the Merck Manual) and some major guides to resources available elsewhere on the Internet.

Patient information

A substantial number of patient information resources are included in Virtual Hospital, collected together as the Iowa Health Book. The full-text of about 400 items are provided, each dealing with a specific health problem. In some cases these are just hyperlinks to resources on external sites but in each case the authorship, date of publication, MeSH indexing terms and peer revew status are given (but in many cases the latter is just 'Under Review'). Sources used include University of Iowa, the US Centers for Disease Control, US National Cancer Institute, US medical Academies and major hospitals.

Weaknesses

Although I would have no hesitation in recommending the Virtual Hospital as a source of medical information, there is room for improvement in some areas. The search facility (essential in such a large and varied site) generates rather confusing results, showing large amounts of text on the screen out of context. The search syntax is simple but non-standard and there is no apparent way to use controlled vocabulary (although some parts of the site are indexed with MeSH terms). As the site grows it will be important to pay more attention to this. Where links off-site are included alongside internal links this should be made more explicit, particularly in the patient information part of the site. It is also a little frustrating to discover that some of the textbooks are 'works in progress', comprising an extensive list of chapter headings with only a few completed chapters. A warning icon by these titles would be useful.

Conclusion

The Virtual Hospital is well worth investigating and bookmarking for future reference. Each reader will have to make their own decision about the individual specialist textbooks contained within, but no-one should ignore them. The Iowa Health Book is a very strong collection of patient information resources. It does, understandably, have a strong US flavour and it would be interesting to see this balanced.

Frank Norman BSc Dip.Lib
Deputy Librarian
National Institute for Medical Research, UK

Society for the Internet in Medicine

Comments to: simq@cybertas.demon.co.uk

Copyright © 1997 Society for the Internet in Medicine. All rights reserved.
Date: December 12, 1997
Document URL: http://www.cybertas.demon.co.uk/simq/issue3/reviews.html