| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
Departments of
*Anesthesia and
Pharmacy, New England Medical Center, Boston, Massachusetts;
National Institutes of Health History Office, National Institutes of Health, Bethesda, Maryland; and
§Department of Anaesthesia and Pain Management, Royal North Shore Hospital, St. Leonards, New South Wales, Australia
Address correspondence to Scott A. Strassels, PharmD, BCPS, New England Medical Center, 750 Washington St., NEMC Box 345, Boston, MA 02111. Address e-mail to Scott.Strassels{at}es.nemc.org or sstrassels@lifespan.org.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Implications: We performed a citation analysis to identify important contributions and contributors to the biomedical literature. Recent pain and analgesia research has been focused on mechanisms of pain, but evidence suggests the importance of understanding the pain experience from the patient’s perspective and the transition from acute to chronic pain.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Evidence-based medicine (EBM) is defined as the "conscientious, explicit, and judicious use of current best evidence in making decisions about the care of individual patients" (2). This approach to medical care has been applied to evaluating the medical literature, in the form of meta-analyses, to synthesize answers to clinical questions.
Citation analysis is an application of EBM in which the appearance, frequency, and citation of articles published in a given field of study are systematically examined (3). Citation analysis has been most widely applied to the biomedical sciences, but it is also used to describe the literature of other fields of study (4–18). Results of citation analyses are used in many ways, including the quantification of a researcher’s impact, with the assumption that articles describing important findings are more likely to be cited. A citation, therefore, is an objective vote of sorts, cast by other contributors to the biomedical literature. Thus, the more citations, the more important the work or journal is presumed to be (19).
Pain medicine is a recent offshoot of anesthesiology, and there is substantial controversy over what topics comprise this body of knowledge. Description and analysis of the development and direction of pain medicine helps researchers, medical care providers, and patients understand recent and ongoing emphases within the field. This perspective helps to provide insight and stimulates the questioning that is at the heart of scientific discovery.
There were three purposes of this project. First, we wanted to analyze the biomedical literature to identify the authors and articles with the most impact on pain and analgesia research and practice over the past 18 yr. Second, we wanted to identify broad trends that have been important in pain and analgesia in this period. Third, we wanted to ascertain whether the information generated from this type of analysis would be useful in identifying current and future trends in this branch of medical practice.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
We then repeated the analysis for January 1988 to June 1997 to better estimate the most important trends and topics in pain medicine during the most recent decade available to us. This subset of the original database contained more than 75,500 articles. The 20 articles and 20 authors with the most citations were then evaluated by checking each of the most-cited articles and authors for their applicability to pain or analgesia. Individual contributions of each author were evaluated by examining the 20 most-cited articles each had written. If most of the 20 most-cited articles written by a most-cited author were not related to pain or analgesia, the author was excluded. Articles and authors judged not applicable to pain or analgesia were replaced by the next article or author on the ranked list.
To test the robustness of the assumptions used in these analyses, we performed two univariate sensitivity analyses. Our main assumption was that we were able to identify the most important authors and articles in pain and analgesia, using the methodology described. This assumption was tested by comparing the results of the searches before and after the hand check of the most-cited articles and authors. Finally, we examined the numbers of articles written by most-cited authors and the numbers of authors who wrote most-cited articles to examine how closely the two were related.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
|
|
The most-cited articles and authors from January 1988 through June 1997 are shown in Tables 3 and 4. Unlike the 1981–1997 analysis, however, none of these most-cited articles has the same first author.
|
|
The selection of most-cited authors and articles was highly sensitive to whether each author and article had been hand-checked for applicability to pain or analgesia. Four authors in the 1981–1997 analysis and five in the 1988–1997 period who were originally identified as being among the highest cited were excluded after their publications were hand-checked. Six of the most-cited authors were also first authors of most-cited articles from 1981 to 1997, whereas three of the most-cited authors were first authors of most-cited articles from 1988 to 1997. A preliminary examination suggested that our results were also very sensitive to changes in other assumptions, such as the search terms used and the years assessed (17).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Citation analysis is an evidence-based technique that is useful in detecting longitudinal trends of topics present in a body of biomedical literature. We believe that pain medicine is ripe for such an analysis because there is an urgent need to improve pain treatment. To do so, however, we must understand the pathophysiology of pain in its many forms, and we must also translate recent advances in preclinical pain research into the clinical realm.
In our study, several principal themes were present in the pain medicine literature over the interval examined. Basic research on the mechanisms of pain was the most common emphasis of the most-cited articles, whereas work on NSAIDs, opioids, and persistent and chronic pain was also frequently cited.
We believe that the main topics that appeared among the most-cited articles are likely to continue to be important in pain medicine and research, particularly as new drugs, such as the cyclooxygenase-2 inhibitors, become available. Other topics that we believe will be important include preemptive and "balanced" analgesia, epidural and spinal analgesia, and the mechanism and treatment of chronic and persistent pain syndrome (22,23). Missing from the most-cited articles were those concerning assessment of pain or treatment of pain from the patient’s perspective—subjects we believe to be of increasing importance as medical practice shifts to a more patient-centered perspective (24). Notably, however, 1 of the 20 most widely cited authors in each of the periods we assessed (Melzack and Turk, respectively) have devoted their careers to advancing techniques for the assessment of pain.
It is also useful to consider why this key cluster of topics appeared so frequently in our analyses. One reason may be a ripple effect, in which a preclinical discovery in the laboratory or a finding in another discipline triggers other thoughts, ideas, and new findings. Another possibility is that the usual process of scientific discovery is one of incremental progress, in which advances come about with the help of innumerable others (25). Thus, we might expect that articles reporting basic research would be most commonly cited, whereas application of these results occurs through a host of subsequent applied research projects, each of which may be of less impact than the seminal driving discovery. Another possible explanation has to do with "memes." In the context of this analysis, memes are elements of culture, thought, or behavior that seem to self-propagate (26). The word "meme" is patterned after "gene," the fundamental and self-propagating unit of genetic information (27). This property has been described as a kind of survival of the fittest for ideas. Thus, an article or presentation on opioid receptors or NSAIDs may encourage other researchers to explore a new idea from the perspective described in the cited article. It is also possible that similar research or practice ideas may have occurred to numerous people at the same time, resulting in numerous articles on a similar topic in a relatively short period of time.
We examined the number of frequently cited articles written by frequently cited authors to examine the relationship between the two. This is analogous to testing the hypothesis that history is composed of the biographies of great individuals (17). We found some support for this theory, although the number of most-cited authors who were also first authors of a most-cited paper decreased by 50% between the two periods we assessed, despite the overlap of the two periods. Also striking was the absence of women first authors in these analyses.
As with all research, this analysis has several limitations. Citation analyses are highly sensitive to changes in word choice or in the makeup of a list of terms. In other words, history depends on how it is defined. For example, if we had retained terms relating to anesthesia in our search strategy, the resulting picture of pain medicine would have shown an emphasis on molecular studies. Articles published in the early 1980s are heavily represented in the first part of this analysis because the earlier an article is published, the longer it has to accumulate citations.
Citation analyses are controversial for characterizing a body of literature or rating the importance of an article or author. Clearly, the articles and authors we identified as having a great impact on the pain and analgesia literature are not the only important articles or contributors to the field. In addition, citation analyses should not be used in isolation to make decisions about complicated topics such as funding or tenure. We believe that it is grossly inappropriate to use this information in such a way.
In addition, there are several potential biases that may have affected our results. Articles describing positive findings, new techniques, or innovative uses for existing procedures are more likely to be published than are concept articles. Citation counts for authors may be misleading when authorship does not reflect a substantial contribution to the article, although this artifact can be addressed by counting only articles that meet predetermined criteria, such as counting only those articles for which an author was listed first or second. Other factors may contribute to the chances that an article will be cited, including the number of journals publishing in a field, the number of issues, and the number of articles appearing in an issue. Similarly, if an incomplete literature search is performed, or if authors rely on literature with which they are familiar to the exclusion of other articles, it is possible that the citation counts for a given author or article may be artificially inflated. Finally, although citations of scientific research are implicitly assumed to indicate support for the cited work, an article that reports an incorrect result or conclusion may be cited frequently because it is incorrect. The cold fusion literature of the late 1980s is an example of this phenomenon (28).
Citation analyses claiming to identify important publications and authors are sensitive to changes in the assumptions used to construct the body of literature considered. Similarly, the database assembled by the ISI for this purpose is not exhaustive. As described earlier, books, book chapters, proceedings, volumes, guidelines, government reports, and other types of "gray" literature are not included, nor are Internet Web sites (29). No single database includes all types of literature; thus, this incompleteness is a potential pitfall for investigators seeking to identify bodies of literature for meta-analyses and other systematic, evidence-based reviews of the literature.
No previous attempt at an exhaustive citation analysis of the field of pain and analgesia has been published. Analogous to systematic reviews of the literature for a specific therapeutic question, citation analyses have been used to describe the trends, directions, and general literature of other medical specialties, although not extensively. In a search of MEDLINE from 1966 to 1999, relatively few analyses of medical practice fields were found, including speech-language pathology, dermatology, neurosurgery, medical decision making, tropical medicine, physical therapy, psychosomatic medicine, drug information, and veterinary medicine (4–18,30–32). Just as systematic reviews begin with a clear definition of search terms, bodies of literature must be focused and well defined to make citation analysis a useful tool for widespread assessment of the biomedical literature. This observation explains, in part, why few medical specialties have been the subject of citation analyses and why it is a reasonable tool for studying the past and future of pain medicine.
One possible application of citation analysis is as a tool for guiding policy decisions. As with techniques such as expert and consensus conferences and Delphi methods, the present results demonstrate the utility of using citation analyses to generate objective information relevant to medical research and practice.
Indeed, during the last decade, the field of EBM has emerged as a vigorous international movement championed by groups such as the Cochrane Collaboration (33,34). This movement is already manifest in the field of pain relief (35).
We identified frequently cited articles and authors in the pain and analgesia literature. We also identified past and current trends likely to remain important for pain medicine research and practice using a cross-sectional citation analysis of the biomedical literature. This information may provide insight for improving the understanding and response to issues facing researchers, clinicians, patients, funding agencies, and policy decision makers interested in the future of pain.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. Baltussen and C. H. Kindler Citation Classics in Anesthetic Journals Anesth. Analg., February 1, 2004; 98(2): 443 - 451. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. Walker, L. C. Goudas, M. J. Cousins, and D. B. Carr Combination Spinal Analgesic Chemotherapy: A Systematic Review Anesth. Analg., September 1, 2002; 95(3): 674 - 715. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
