The Cochrane Anaesthesia Review Group

The Cochrane Anaesthesia Review Group At a meeting held during the European Society of Anaesthesiologists' meeting in Amsterdam at the end of May 1999, anaesthesiologists from around the world met to found the Anaesthesia Group within the Cochrane Collaboration. Yet, for all the rapid growth of the Cochrane Collaboration, there will still be many anaesthesiologists who have not heard of it, or who may not be familiar with its aims. First, though, by way of introduction, let us look a little at the life of the man whose ideas first inspired it. Cochrane's life and work Archie Cochrane was born in Scotland in 1909. He studied natural sciences at Cambridge, then went on to study medicine in London, graduating in 1938. One of his student years was spent in the International Brigade during the Spanish Civil War. He returned to medical research but joined the Royal Army Medical Corps in 1940. He was captured in Crete in 1941 and spent some years as a prisoner of war in various parts of Europe. After the war, he conducted meticulous and painstaking epidemiological fieldwork on pneumoconiosis among the mining communities of South Wales, and became Professor of Chest Diseases at the Welsh National School of Medicine. In 1969, he became full-time director of the British Medical Research Council's epidemiology unit, focusing particularly on randomized controlled trials aimed at discovering the optimal place of treatment and the optimal length of hospital stay. He became more interested in these and other applications of epidemiology to clinical practice. In 1971, he became the first President of the Faculty of Community Medicine and died in 1988. Cochrane's most famous book, Effectiveness and Efficiency: Random Reflections on Health Services appeared in 1972 (though it has been re-issued this year in a new edition [1]). Few could have foreseen the effect the slim volume was to have on healthcare throughout the world. Cochrane himself remained vaguely surprised and somewhat dismayed by the attention, feeling that it overshadowed what he considered his more important work on the epidemiology of chest diseases in the 1950s. Cochrane defined his terms carefully. He observed that, if we are to get the best from our health services, we must be able to measure benefit as well as cost. He saw that, to do this, we must be able to measure the effect of a particular medical action in altering the natural history of disease. This he called effectiveness, and promoted the randomized controlled trial as probably the best way of measuring it. We need also to determine the optimum use of personnel and materials in achieving these results. This he called efficiency. These approaches are familiar to us now but this was the first time that such ideas had been so widely aired. Effectiveness and Efficiency is often acknowledged to be the start of the whole 'evidence-based medicine' movement. However, it would be quite wrong to say that the unfeeling application of science was Cochrane's aim. In fact, he recognised that there were problems in applying the effectiveness/efficiency approach to most health service activity, because although these two indices could be used for assessment of therapy (attempts at cure) they could yield no information on the health service's other main function, that of providing, in his words, 'board and lodging and tender, loving care'. He proposed a third index, equality, referring mainly to equality in these nontherapeutic aspects of hospitals, but for some reason this did not feature in the book's title, though it pervades the work nonetheless. Cochrane's challenge: the genesis of the Cochrane Collaboration In a further publication in 1979 [2], Cochrane wrote: It is surely a great criticism of our profession that we have not organised a critical summary, by specialty or subspecialty, adapted periodically, of all relevant randomised controlled trials In the same essay, he singled out obstetrics (whether rightly or wrongly) as the medical specialty which had made least effort to seek good evidence on which to base its practices. This rebuke encouraged an international collaborative effort which had been established to prepare systematic reviews, using the statistical technique of meta-analysis when appropriate, of all the randomized controlled trials relevant to care during pregnancy and childbirth. This began to yield results in the late 1980s, the findings being published as the book Effective Care in Pregnancy and Childbirth[3] and electronically as the Oxford Database of Perinatal Trials. Such systematic reviews began to be used in healthcare decision-making and clinical practice, and were also taken up enthusiastically by lay people. The encouraging reception given to this first wave of collated evidence on effectiveness in healthcare coincided with the introduction of the United Kingdom National Health Service Research and Development Programme in 1991, and Cochrane's idea found favour within it. In October 1992, the world's first Cochrane Centre opened in Oxford. Within a year it was clear that this was destined to become a world-wide venture, and its name had to be changed to the UK Cochrane Centre. In October 1993, 77 people from nine countries founded the Cochrane Collaboration. Since then, 14 other Cochrane Centres have been opened throughout the world, to support local researchers and act as points of contact. Now the Cochrane Collaboration is a truly international organization whose aims are simple: to help people make well-informed decisions about health by preparing, maintaining and disseminating systematic reviews of the effects of healthcare interventions. Its nine guiding values are shown in Table 1. The eventual goals are to create a register of all completed and continuing randomized controlled trials, to combine the results of trials that meet certain quality standards into systematic reviews, and to make these widely available to healthcare practitioners and consumers. Preparation and maintenance of the reviews is coordinated by Collaborative Review Groups. Currently there are over 50 Groups registered, covering areas as diverse as stroke, injuries, infectious diseases, ENT and schizophrenia. Not all areas of healthcare are represented - yet - and clearly the output of the Collaboration reflects the interests of the Groups that have come forward.Table 1: The principles of the Cochrane Collaboration Cochrane reviews are published electronically on the Cochrane Library. This is available on CD-ROM, floppy disk and on-line. As well as the reviews, the Library also contains the Controlled Clinical Trials Register (a database of over 220000 trials, some either unpublished or not previously identified on Medline), a large collection of references on review methodology, and the Database of Abstracts on Effectiveness (material not prepared by Cochrane groups but still meeting certain quality standards). The CD-ROM is published quarterly, and the last issue contains 628 complete reviews and 593 protocols for future reviews [4]. Review articles - traditional and modern Busy doctors need efficient methods of keeping up to date, and review articles have long been popular as time- and effort-saving summaries of much larger amounts of primary research. However, Mulrow's 1987 paper pointed out that traditional reviews (often now called narrative reviews to distinguish them from systematic reviews) rarely used scientific, systematic methods [5] and this exposes them to a number of potential biases, analogous to those in primary research. Ten years later, Smith drew the same conclusions after examining reviews in four major anaesthesia journals [6]. Researchers and journal editors now take great pains to ensure that primary studies are carried out in such a way as to keep possible biases to a minimum. The randomized, blinded controlled trial has been favoured because its design tends to reduce bias in selection, allocation and performance, and more attention is also paid to ensuring completeness of follow-up and analysis. Only recently have the same scientific methods been applied to review articles. Review articles can also be prone to a number of biases. Biases operating during the identification of studies include language bias (searches limited to the English language may miss valuable material in other tongues [7]) and publication bias (studies that report positive finding are more likely to be published than those whose results are negative [8]). The latter may be helped in due course by registers of planned and current trials. Bias in conducting the review can be reduced by making explicit the criteria used to decide which studies are included and which left out, and having more than one person making the decision independently, comparing their actions as necessary [9]. Some assessment of the articles' methodological quality should be made, and poor quality articles rejected, with reasons for rejection stated. This contrasts with the traditional review process where the reviewer (usually an expert in the field of study) prepares the review without making the process explicit and declaring any biases, if applicable. In fact, Oxman and Guyatt (themselves known for their expertise in this area) have produced research which purports to show that, the more expert the reviewer, the less time is spent in the review's preparation and the poorer the quality of the review [10]. This may simply be due to the fact that experts know more about the subject and so can write more efficiently, but also implies that they will include their own work most readily because it is closest to hand. Systematic reviews offer a potential solution to these problems, and further benefits have been outlined by Mulrow [11]. Systematic reviews may be time-consuming but a review is usually quicker and cheaper than embarking on a new study. Because they include a diversity of studies, systematic reviews allow some assessment of the wider applicability of findings, and also allow consistency of treatment effect to be explored. Conversely, data inconsistencies and conflicts come to light also. An often-cited advantage is increased power. If the numerical results of different trials are combined using the statistical techniques of meta-analysis, an overall treatment effect may become apparent which was not evident in the results of the studies before they were pooled. Furthermore, for a given treatment effect, the precision of the estimate increases because, as more trials are added to the analysis, the confidence limits become progressively narrower. Evidence-based medicine The underlying assumption of the Collaboration is that health care interventions will be more effective if they are based on complete and up-to-date evidence instead of out of date research, anecdote and conjecture [12]. This view is shared by the whole evidence-based medicine movement, whose enthusiasts argue that the traditional basis of medical practice - the accumulated experience of a seasoned clinician - is open to all kinds of bias and exposes patients to care that is at best inconsistent and ineffective and at worst potentially harmful. Critics have argued that the whole enterprise is no more than 'cookbook' medicine, or fear that the process has been hijacked by Governments and health policy-makers as a means of controlling costs. In response to these views, Sackett (perhaps evidence-based medicine's most famous exponent) clarified his position in a placatory editorial: 'Evidence based medicine is the conscientious, explicit and judicious use of the current best evidence in making decisions about the care of individual patients. The practice of evidence based medicine means integrating individual clinical expertise with the best available external clinical evidence from systematic research' [13]. Goodman, in an excellent overview of evidence-based medicine and its problems [14], has pointed out that it is frankly insulting to doctors to suggest that they have not been basing their practice on evidence. The evidence they have traditionally used may not have been derived from careful scrutiny and synthesis of published research, but that is partly because there was very little research to draw on. In fact, for much of medical practice, including anaesthesia, this is still true. To quote Naylor [15]: 'Clinical practice seems to consist of a few things we know, a few things we think we know (but probably don't) and lots of things we don't know at all'. Perhaps if the activity had been christened 'literature-based medicine' or collation-based medicine' it would have provoked less outrage and been less vigorously opposed. Anaesthesiology is in an unusual position in that much of what we do is not innately therapeutic in itself, but is designed to facilitate other interventions. Few outcomes with which we may be associated are directly attributable to anaesthesia; most are the net result of a package of influences including surgical technique and peri-operative care as well. Intensive care is in a similarly difficult position [16]. The heterogeneity of this group of patients, with their many complex problems, makes it difficult to study one intervention in isolation. Continuous technological advance brings new techniques often without time for adequate evaluation of the old, whilst the cost of research means that work is more likely to be carried out in certain areas where specific funding is available (e.g. drug trial) than on more important clinical questions. One way round the difficulties with recruitment into trials is the approach adopted by the Intensive Care Audit and Research Centre in the United Kingdom. The alternative to studying small, selected subgroups of a particular population is to collect data on the entire population and for some years now the Centre has been starting to do this with a view to establishing a database of therapies and outcomes, and this comprehensive approach must surely complement attempts to synthesise such research literature as is available. If there is a danger with evidence-based medicine, it is perhaps that the slavish adherence to epidemiological principles in the care of our patients, whilst it may make us more efficient, may make us too inhuman. Although Cochrane's name is now for ever associated with the systematic review, we should not forget his kindness and compassion. The Cochrane Anaesthesia Review Group What, then, will be the scope of the Cochrane Anaesthesia Review Group? Broadly speaking, we expect to concentrate on anaesthesia, peri-operative medicine and intensive care medicine, though some aspects of prehospital, resuscitation and emergency medicine might also fall within the scope of the Group. Our main topic areas are therefore: 1 Drugs in anaesthesia and intensive care medicine; 2 Anaesthesia and medical disease; 3 Regional anaesthesia; 4 Ambulatory anaesthesia; 5 Peri-operative medicine; 6 Patients in the post anaesthesia and intensive care units; 7 Technology; and 8 Prehospital medicine, resuscitation and emergency medicine. Already there are Collaborative Review Groups dedicated to Pain and Palliative Care, Injuries, Heart Diseases, Airways and Colorectal Cancer, and careful coordination will be needed to make sure that there is no duplication of reviewing where the interests of Groups overlap. How can anaesthetists help? This major enterprise can only succeed with the support of the anaesthesiological community in general. There are many ways to get involved with the work of the Collaboration. We need volunteers to hand-search journals for randomized controlled trials for inclusion in the register; we need translators, and experts who can offer their services for peer-reviewing protocols and reviews prior to publication. We need those who can help with analyses and methodology, and Collaborative Review Groups will need help obtaining financial support. Most of all, though, we need people to take on the commitment of writing reviews, for these are the foundation on which the Group will be built. Finally, though it will take some time before the Group's reviews appear on the Cochrane Library, we would encourage all anaesthesiologists to read them when they appear, comment upon them, introduce others to them and, of course, use them in practice so that our patients everywhere can benefit. Further information can be found on our website at: HTTP://www.cochrane-anaesthesia.suite.dk. A. Smith (Editor in Cochrane Anaesthesia Review Group) Department of Anaesthesia, Royal Lancaster Infirmary, Lancaster LA1 4RP, United Kingdom T. Pedersen (Co-ordinating editor) A. Møller (Deputy co-ordinating editor)