Early results from this work were presented at the European Society of Anaesthesiologists' meeting in Nice, April 2002

Introduction

Medical journals keep readers up to date with advances in basic science and clinical practice and stimulate discussion and debate. For authors and researchers, journals allow the legitimation of research findings by peer-review and publication and advance the careers of individuals and institutions.

Although original research is the most useful and prestigious to the scientific community, experience suggests that those in clinical practice tend to read review articles and editorials and these may have a greater impact on readers' thinking and practice.

There has been some research into peer review but this has mostly examined how original research is handled₁. We aimed to document how editorials and review articles are handled in journals in anesthesiology, critical care and pain medicine.

Methods

We conducted a postal questionnaire survey of the editors of international English language journals in the field of anesthesiology, critical care and pain medicine. The questions related to how topics and authors for review articles are chosen and what peer review processes are used (Figure 1). We also asked if the editor would be willing to have the journal's policy identified in this paper. The responses were categorized.

Figure 1

Figure 1: Journals whose editors responded

Results

We sent letters to 21 editors and received responses from 19 (90.5% response).

Question 1: How are the topics for review articles chosen?

The responses are shown in Table 1. Further comments were offered by nine respondents.

Topics for reviews are chosen by the Editor-in-Chief of one journal, by members of the Editorial Board in four journals and by authors and section editors in one journal each.

Two editors highlighted topicality as an influence on their choice of subject.

Figure 2

Table 1: How review topics are chosen (n = 19 )

Question 2: How are the authors for these topics selected?

The responses are shown in Table 2.

Figure 3

Table 2: How review authors are chosen

Question 3: Do review articles undergo peer review?

Question 4: How are the subjects for editorials selected?

The responses are shown in Table 3. Further comments were offered by thirteen respondents.

Topics for reviews are chosen by the Editor-in-Chief, members of the Editorial Board or section editors in five journals and suggested by reviewers in two journals.

Three editors highlighted topicality as an influence on their choice of subject. Three mentioned specifically that editorials are published relating to articles in the same issue.

Figure 4

Table 3: How editorial topics are chosen (n = 19)

Question 5: How are the authors for these editorials selected?

The responses are shown in Table 4.

Figure 5

Table 4: How editorial authors are chosen

Question 6: Do editorials undergo peer review?

Question 7: Are systematic reviews and more traditional review articles handled any differently?

Consent to publish individual journal data

Discussion

Whilst there is variation in editorial policies between leading anesthesia journals in the selection of both topics and authors for review articles and editorials, the majority of journals prefer to select topics and commission work rather than favor unsolicited material. In deference to the wishes of the minority of editors, we have not identified individual journals in our results.

We are not aware of any previous work examining how editorials and review articles are selected in the anesthesiology and related literature. This is not unusual. Little is known about the internal workings of medical journals, including the peer review system, in general₂. Peer review began essentially to share the editorial workload of journals and only latterly has become regarded as a quality control mechanism in biomedical publication₃. This probably explains the diversity of organization of peer review systems in different journals.

The whole peer review mechanism has been termed a ‘black box'₄ and in truth not much is known about the internal workings of journals. Some general medical journals (the British Medical Journal and the Annals of Internal Medicine, for instance) have opened up their manuscript handling processes but the effects of this have yet to be fully felt. Despite the fact that a recent systematic review₅ found that the effects of peer review are uncertain, it is still regarded as vital₆.

In recent years, much attention has been paid to improving the methodological quality of primary research in anesthesiology. Editorial policy and the standards imposed by journal peer review have contributed greatly to this. Latterly, the same methodological rigor has been applied to the preparation of review articles, resulting in so-called ‘systematic reviews'₇. It has been suggested that traditional or ‘narrative' reviews, typically written by experts, are prone to bias throughout the review process₈. Whilst the existence of publication bias is acknowledged, this most commonly refers to the tendency of journals to publish studies with positive, rather than negative findings. However, when journal editors commission reviews, these are usually ‘traditional' rather than systematic. This may for purely practical reasons, as individuals with both the necessary understanding of systematic review methodology as well as expertise in a particular area of clinical anesthesiology are rare. Further, many topics for review are still relatively new and there may be few or no high-quality randomized controlled trials to draw on.

Despite the recent and continuing trend towards ‘evidence-based' practice, expert opinion is essential both to interpret published work and to point the way when evidence is incomplete or absent. Editorials help fulfill this vital function₉ and can be especially recommended as it is clear that what they contain is opinion, rather than evidence. Both are necessary in medical practice but should not be confused. Jefferson has challenged editors to be explicit about how editorials are produced₁₀ and we have gone some way towards addressing this. Although editorials are more likely to be commissioned, and less likely to be peer-reviewed than review articles, this may again be for essentially practical reasons: if editorial comment is linked to an article in the same issue of the journal, publication might be delayed if a lengthier review process is used₉.

One less well appreciated aspect of biomedical peer review relates to its social significance within the professional communities in which it operates. In general, the more specialized the journal and the smaller its readership, the more closely defined is the ‘community of practice' it encompasses₁₁. It is unclear what effect, if any, this has on the dynamics of peer review. Collins' studies in the physical sciences have shown how scientific knowledge is shaped by social consensus amongst scientists as well as factual inquiry₁₂. In the context of biomedical peer review, Davidoff has noted that ‘Approval by peer review is perhaps the most single most powerful expression' of professional recognition in science, which is in turn the ‘principal instrument of social control within the scientific community'₆. Our data are compatible with the notion that what is chosen for publication reflects the normative influence of a small social group of ‘experts'. Whilst this ‘gatekeeping' function may be appropriate to maintain scientific quality, does it also constrain the breadth of material that appears in print?

This study is simple and modest in its aims, which we believe is appropriate for an initial enquiry into this subject. Future work could usefully examine the relationship between editorial policies and the quality and diversity of what is published.