Article Text

This article has a correction. Please see:

Download PDFPDF

Sports medicine clinical trial research publications in academic medical journals between 1996 and 2005: an audit of the PubMed MEDLINE database
  1. A W Nichols
  1. Division of Sports Medicine, Department of Family Medicine and Community Health, John A. Burns School of Medicine, University of Hawaii at Manoa, Honolulu, Hawaii, USA
  1. Dr A W Nichols, Associate Professor and Chief, Division of Sports Medicine, Department of Family Medicine and Community Health, John A. Burns School of Medicine, 651 Ilalo Street, Medical Education Building, Honolulu, Hawaii 96813-5534; nicholsa{at}hawai.edu

Abstract

Objective: To identify sports medicine-related clinical trial research articles in the PubMed MEDLINE database published between 1996 and 2005 and conduct a review and analysis of topics of research, experimental designs, journals of publication and the internationality of authorships.

Hypothesis: Sports medicine research is international in scope with improving study methodology and an evolution of topics.

Design: Structured review of articles identified in a search of a large electronic medical database.

Setting: PubMed MEDLINE database.

Participants: Sports medicine-related clinical research trials published between 1996 and 2005.

Interventions: Review and analysis of articles that meet inclusion criteria.

Main outcome measurements: Articles were examined for study topics, research methods, experimental subject characteristics, journal of publication, lead authors and journal countries of origin and language of publication.

Results: The search retrieved 414 articles, of which 379 (345 English language and 34 non-English language) met the inclusion criteria. The number of publications increased steadily during the study period. Randomised clinical trials were the most common study type and the “diagnosis, management and treatment of sports-related injuries and conditions” was the most popular study topic. The knee, ankle/foot and shoulder were the most frequent anatomical sites of study. Soccer players and runners were the favourite study subjects. The American Journal of Sports Medicine had the highest number of publications and shared the greatest international diversity of authorships with the British Journal of Sports Medicine. The USA, Australia, Germany and the UK produced a good number of the lead authorships. In all, 91% of articles and 88% of journals were published in English.

Conclusions: Sports medicine-related research is internationally diverse, clinical trial publications are increasing and the sophistication of research design may be improving.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Clinical sports medicine fellowship training programme opportunities have flourished in the USA over the past 25 years. Physicians with primary medical specialty residency training in orthopaedic surgery, family medicine, internal medicine, paediatrics, emergency medicine and physical medicine and rehabilitation are eligible for various types of sports medicine fellowships. Scholarly activities and clinical research efforts within the field of academic sports medicine has grown with the rise in the number of formally trained sports medicine physicians.

Sports medicine research efforts aim to identify ways to maximise human performance and to optimise athletic healthcare. Bleakeley et al. examined the quality and evidence of sports medicine research by sampling four major sport and exercise medicine journals. They discovered that randomised controlled trials comprised no more than 10% of all original research articles, the most commonly published studies were observational or descriptive in nature, the most frequent topic of study was sport science and few studies investigated the treatment of injuries and medical conditions in athletes.1 2

The aim of this study was to identify a large sample of sports medicine research clinical trials articles published between 1996 and 2005 through an extensive search of the PubMed MEDLINE electronic database. The PubMed MEDLINE electronic database originates from the United States National Library of Medicine of the National Institutes of Health and includes over 17 million citations from medical and life science journals.3 Among the data extracted and analysed from the publications that met the inclusion criteria for the study were research topics of study and methods, journals of publication, characteristics of experimental human subjects, countries of origin of lead authors and journals, and languages of publication.

To meet the requirements of a systematic review, a study must include an exhaustive search of the medical literature, focus on a particular topic and explicitly state the methods used for the search and data synthesis. Several authors have shown that a topic cannot be exhaustively searched from within a single database; instead, searches conducted within different databases complement each other.48 A recent systematic review of strategies to prevent injury in adolescent sport included a computer-based literature search of 7 databases, in which only 12 of the 154 retrieved papers were eligible for study inclusion.9 This study is best considered to be a “literature audit” as it does not meet the criteria for a systematic review. The literature search involved only a single (though very large and internationally recognised) database and the data extracted and synthesised is intentionally broad in scope so as to provide an overview of the state of sports medicine clinical trial research during the 10 year study period.

METHODS

Literature search

This PubMed MEDLINE database literature search of sports medicine clinical research trial publications between 1996 and 2005 was conducted on 13 July 2007. The search strategy, optimised in consultation with a US medical school senior librarian, is summarised in table 1. Abstracts of identified articles were skimmed and print articles of clinical trials that addressed subjects relevant to human sports medicine, exercise science, sports injuries, and human performance were included. Articles published electronically only, letters, reviews, case reports, studies that were not clinical trials, and animal studies were excluded.

Table 1 Literature search strategy using PubMed MEDLINE database on July 13, 2007

Data collection procedure

Abstracts of all selected articles were read. If the abstract contained incomplete information, the entire article was obtained and read with special attention to research-study design methods. Using the matrix method,10 each paper was evaluated using a structured abstract form. Information collected from each article included: lead author’s name and country of origin, year of publication, journal name, journal country of publication, journal language of publication, research topic category and subcategory, clinical trial study design, number of experimental subjects, gender(s) of study subjects and specific sport(s) of subjects studied. Each English-language study was classified into one of the following clinical trial study types: (1) prospective cohort study, (2) retrospective cohort study, (3) cross-sectional study, (4) case–control study, (5) randomised clinical trial (RCT), and (6) non-randomised clinical trial/miscellaneous (non-RCT/Misc). Articles that were published in a language other than English were not considered for classification into clinical trial study types because of the difficulties in article retrieval and language translation. Research study categories and subcategories were based upon an outline of the curriculum requirements for primary care sports medicine fellowship accreditation. The four main study subject categories wre: (1) diagnosis, management and treatment (Dx/Mx/Tx) of sports-related injuries; (2) diagnosis, management and treatment (Dx/Mx/Tx) of sports-related injuries; (3) basic sports science; and (4) musculoskeletal rehabilitation. Subcategories are listed in table 2. Each journal’s country of origin was determined by accessing the PubMed MEDLINE Journals Database.3

Table 2 Clinical trial research study topic categories and subcategories numbers of publications (%) overall and by 5 year periods from 1966 to 2005

Data analysis

The data were analysed using Excel spreadsheet and statistical software. Descriptive statistics were obtained and significance was tested using the χ2 test, with the level of significance set at p<0.05.

RESULTS

The search strategy identified 414 articles (table 1) Of these, 377 were published in English and 37 were published in various non-English languages. In total, 379 (91.5%) articles, including 345 (91.0%) of the English-language articles and 34 (91.9%) of the non-English-language articles met the study inclusion criteria and were entered into the study matrix. To facilitate data analysis, studies were grouped into the following 5 year publication year ranges: 2001–2005 English,11252 2001–2005 non-English,253272 1996–2000 English,273375 and 1996–2000 non-English.376389 The numbers of publications in English and non-English languages for each individual year, each 5 year period and overall are shown in table 3. Figure 1 offers a trendline of the numbers of publications occurring during each individual year of the study (R2 = 0.8681). The numbers of “English-language only publications” and “total publications” was greater during the 2001–2005 than the 1996–2000 time period.

Figure 1 Numbers of sports medicine clinical research trial publications by year between 1996 and 2005 with trendline and R2 value.
Table 3 Numbers of sports medicine clinical research trial publications in professional journals in English and in non-English languages between 1996 and 2005

Table 4 lists the numbers of English-language publications by different study design types overall and for each 5 year period, including numbers of publications per study type, average number of publications per study type and percentage of all publications per study type. The numbers of publications were significantly greater during the 2001–2005 year period than the 1996–2000 year period for each of the six study types except for retrospective cohort study types. Overall, RCTs were the most common study type during the study period, accounting for 36.8% of publications.

Table 4 Numbers and percentages (%) of English-language clinical trial professional journal publications by research study type from 1996 to 2005

Table 2 shows the numbers of publications overall and during each 5 year time period by research study subject topic category and subcategory. The study topic category, “diagnosis, management and treatment (Dx/Mx/Tx) of sports-related injuries and conditions”, yielded the highest number of publications (n = 162, 42.7%), followed by the “health promotion and preventative aspects” (n = 110, 29%) and “basic sports science” categories (n = 78, 20.1%). The most frequently studied subcategories were the “knee” and “conditioning and training techniques”, with 48 (12.7%) publications each, followed by “exercise physiology”, “the lower leg/ankle/foot” and “epidemiology”.

Table 5 Gender of subjects who participated in clinical trials between 1996 and 2005

Table 2 shows that a total of 75 487 subjects participated in English-language published sports medicine clinical trial studies between 1996 and 2005. The numbers of subjects increased between the 1996–2000 and 2001–2005 5 year periods. Overall, 27.4% of the clinical trials studied male subjects only, 10.6% involved female subjects exclusively and 62% investigated both male and female subjects.

In total 32% (132/414) of published clinical trials studied only participants from a single sport. Soccer players and runners (15 studies each) were the most frequently studied groups of athletes. Participants in volleyball, American football, team handball, bicycling, rugby, Australian football and rowing were also commonly studied. Overall, athletes from 33 different sports were studied exclusively in at least one clinical trial.

The search identified 94 different professional journals that published sports medicine clinical trial studies between 1996 and 2005. English was the language of publication for 83 (88.3%) journals. Table 6 lists the numbers of publications, countries, languages and first years of publication for the 21 journals with ⩾3 sports medicine clinical trial articles during the study period. In total 11 (52.4%) of these originates from the USA, 3 (14.3%) each from the UK and Germany and 1 (4.8%) each from Australia, Belgium, Denmark, France and Italy. The Journal of Bone and Joint Surgery publishes both American and British versions. The American Journal of Sports Medicine had the most publications overall (n = 36), followed by the British Journal of Sports Medicine (n = 33), International Journal of Sports Medicine (n = 26), Clinical Journal of Sports Medicine (n = 24), Journal of Strength and Conditioning Research (n = 23) and Sportverletzung Sportschaden (n = 19). The American Journal of Sports Medicine also had the highest number of publications between 2001 and 2005, whereas the Clinical Journal of Sports Medicine had the most between 1996 and 2000. Lead authorships originated from 29 different countries (table 7). Overall, the USA produced the highest number (n = 128) of lead authorships, followed by Australia (n = 37), Germany (n = 36), the UK (n = 28), Canada (n = 24), Sweden (n = 20) and Denmark (n = 12).

Table 6 Journals that published ⩾3 or more sports medicine clinical trial articles between 1996 and 2005
Table 7 Lead authors’ countries of origin among sports medicine clinical trial publications from 1996 to 2005

A predominance of papers (n = 345, 91%) were published in English. Among the other languages in which studies were published, 26 (76.5%) articles were published in German, 4 (11.8%) in French, 2 (5.9%) in Russian and 1 (2.9%) each in Finnish and Danish. German authors published 24 (66.7%) articles and Swiss authors published 2 (33.3%) articles in their native German language. French authors published four (36.4%) articles in French, Russian authors published two (100%) in Russian, Finnish authors published one (25%) in Finnish and Danish authors published one (8.3%) in Danish.

Of the 15 journals with the most (⩾6) sports medicine clinical trial publications between 1996 and 2005, authors from the USA contributed articles to 11 (73%) different journals. Australia (8), Canada and the UK (7 each), Sweden (6), Germany and Norway (5 each) and the Netherlands (4) also contributed authorships to substantial numbers of these journals. Table 8 ranks the journals by numbers of lead authorships from different countryies and also shows the percentage of lead authorships that originated from a journal’s country of publication and the ratio of the number of lead authorships from different countries divided by the total number of publications for a journal. The American Journal of Sports Medicine (USA) had lead authors from 11 different countries. Medicine and Science in Sports and Exercise (USA) and the International Journal of Sports Medicine (Germany) also published the works of authors from various countries. Arthroscopy (USA) and Knee Surgery Sports Traumatology and Arthroscopy (Germany) had high international publication diversity ratios, indicating that in spite of having relatively few publications overall, the authors’ countries of origin were diverse relative to the total numbers of publications. Only the Journal of Science and Medicine in Sports (Australia) and Sportsverletzung Sportschaden (Germany) were found to have publications only from authors within each journal’s home country.

Table 8 Diversity of different country lead authorships for the 15 journals with ⩾6 sports medicine clinical trial publications between 1996 and 2005

DISCUSSION

During the study period, the number of sports medicine clinical trial research publications showed an increasing trend year on year and a 127% rise between the 1996–2000 and 2001–2005 year periods. An informal extension of the literature search to the 30 years before the study period (1966–1995) confirmed that the observed increasing trend in publication numbers also predated the study period. Sports medicine research publications were notably uncommon before 1975. The data identified substantially more sports medicine-related RCTs as a percentage of total clinical trials than that observed by previous investigators.1 2 This could be due to a true increase in the number of published RCTs during the recent years, different search strategies and databases and perhaps, different study time periods.

The study results found studies categorised as “diagnosis, management and treatment of sports-related injuries and conditions” to be the most common research topics, which conflicts with an earlier review study that found “sports science” to be the most common topic of research study and research on “injury prevention and treatment” to be less common.1 2 Such disparities in research topics could be due to a shift in research topic focus toward more clinically based issues in recent years or to differences in research topic classification schemes.

Knee injury research was found to be a popular subject of study, accounting for 12.7% of all publications during the period. In addition, the substantial rise in the number of “neurological” related studies between 1996–2000 and 2001–2005 probably reflects recent efforts to better assess and treat sports-related head injuries using tools such as neuropsychiatric testing. “Basic sports science” topics accounted for the second largest number of research topics. Within this category, “exercise physiology” and biomechanics” showed rising popularity, as did “conditioning and training techniques” from within the “health promotion and preventative aspects” category. The enhanced interest in these research topics correlates with the increased sophistication of sports training and conditioning techniques used by athletes today. Notably, the number of “musculoskeletal rehabilitation” related clinical trial articles increased by more than 3-fold between 1996–2000 and 2001–2005, indicating improvements in the application of evidence-based rehabilitation practices. The results revealed surprisingly few clinical trials that studied “nutrition”, “injury protection/safety equipment”, “exercise prescription”, and the “preparticipation examination”. The paucity of research efforts in these crucial categories of study offer great opportunities for young investigators.

As expected, the growth in sports medicine clinical trial publication numbers parallel increases in the numbers of experimental subjects who participated in such studies. The number of research subjects increased by 118% between 2001–2005 and 1996–2000. The majority of studies included both male and female subjects (62%) and of the remaining articles that studied subjects of only one gender, the number of those that studied men exceeded those that studied women by nearly threefold. Perhaps the large increases in female sports participation over the past few decades will lead to similar increases in clinical trial research on female athletes.

Runners were favourite experimental subjects of researchers, facilitated by the ease of performing exercise physiology, conditioning and training technique research in the laboratory using treadmills and cycle ergometers. Studies of soccer players were also common, reflecting the sport’s enormous global popularity. Much research involved sports that are popular and specific to a particular region and that often appear in region-specific journals, such as Australian-rules football, skiing and team handball.

The 22 journals with the most sports medicine clinical trial publications originate from eight different countries. All but one of these journals began publication before the 10 year study period and nearly all specialise in sports medicine, exercise physiology, orthopaedic surgery or other sports medicine-related fields. The earliest sports medicine journals were Medicine and Science in Sports and Exercise (formerly Medicine and Science in Sports) and the British Journal of Sports Medicine, both of which started publication in 1969. Several newer sports medicine journals, including the American Journal of Sports Medicine (1976), the Clinical Journal of Sports Medicine (1991), the Scandinavian Journal of Medicine and Science of Sports (1991) and the Journal of Strength and Conditioning (1993) followed.

Authors from the USA contribute the largest number of sports medicine-related clinical trial publications to the PubMed MEDLINE indexed medical literature. Although this number is perhaps biased in that the PubMed MEDLINE database originates in the USA, authors from the USA maintained high scholarly productivity throughout the study period, as did authors from Australia, Germany, the UK and Canada. Researchers from Japan and Belgium demonstrated dramatic recent gains in publication numbers during the 2001–2005 period. The data clearly confirms that sports medicine research efforts are global in nature, as publication authorships originated from 29 different countries, including 18 countries in Europe alone.

English is the predominant language of publication for clinical trials in sports medicine indexed in the PubMed MEDLINE database. Even countries in which English is not the primary language often publish English-language journals, including International Journal of Sports Medicine and Knee Surgery Sports Traumatology and Arthroscopy, both published in Germany, Scandinavian Journal of Medicine and Science in Sports, published in Denmark and the Journal of Sports Medicine and Physical Fitness, based in Italy. Notably, these particular journals were found to have more international diversity in authorships than those published in a country’s native, non-English language. In contrast, non-English-language journals published in the journal country’s native language, such as the German-language Sportsverletzung Sportschaden from Germany and the French-language journal, Revue de chiurgie orthopédique et réparatrice de l’appareil moteur from France, have little international diversity of authorships, generally publishing only the works of authors from their home countries. German researchers in particular publish large numbers of their articles in the German language. Authors from several non-English-speaking countries, including Italy, Sweden, Norway, Belgium, Japan, Denmark, Netherlands, Finland and Norway, published exclusively or near-exclusively in English.

Of the 15 top publishers of sports medicine research, only Sportsverletzung Sportschaden and Journal of Science and Medicine in Sports limited publications to the works of authors from each journal’s home country. Authors from the USA were published in each of these 15 journals except for the Scandinavian Journal of Medicine and Science of Sports and Revue de chiurgie orthopédique et réparatrice de l’appareil moteu. Authors from Canada, Italy, Sweden, Australia, Norway, New Zealand and the UK also published articles in a variety of different journals, compared with German authors who mostly published in journals based in Germany. Journals with high international diversity of journal authorships, such as the American Journal of Sports Medicine, British Journal of Sports Medicine, Medicine and Science in Sports and Exercise and International Journal of Sports Medicine have perhaps each earned strong international academic reputations, global recognition and prestige, which serve to attract manuscript submissions from throughout the world.

A potential limitation of this PubMed MEDLINE database literature review of sports medicine research studies is that the search strategy was restricted to publications classified as “clinical trials”. Thus, for the search to identify all appropriate clinical trials, articles must have been correctly classified. Further, the search was limited to a single, though very large and internationally recognised database. Although this study did not attempt to meet the criteria for a systematic review, if the search is extended to include other databases, it is likely that more articles would be retrieved that meet inclusion criteria. Although widely recognised as an internationally inclusive medical research electronic database, the USA-based PubMed MEDLINE database might perhaps have a bias toward journals of American origin and other English-language medical journals. The addition of other databases, such as the European-based EMBASE, could potentially alter the international composition of sports medicine clinical trial articles. An interesting future project would be to compare the various electronic medical databases for content of sports medicine-related topics. Such comparisons have been performed in several areas of medical specialisation.48 An additional limitation is that there is no way to determine how many manuscripts were submitted for publication and to compare this with the number of manuscripts that reach publication stage. It would be interesting to know the publication rate as a percentage of manuscripts submitted and determine whether this has risen or declined during the study period.

The review may also have unintentionally presented an inflated number of experimental subjects who participated in sports medicine clinical trials, as certain subject study groups could have been used repeatedly in more than one published clinical trial. An additional potential limitation of the study is that certain journals publish in more than one language and yet are classified by the primary language only. The option to publish in more than one language may also influence the manuscript submission decisions of authors.

This PubMed MEDLINE literature review of sports medicine clinical trial research literature publications over a recent 10 year period confirms that the numbers of publications overall and by specific research study design are increasing. Topics of research study have evolved in parallel with new discoveries and current issues in sports medicine. The review identified several understudied research topics that would provide excellent clinical trial research opportunities for young investigators.

The review also highlights the widespread global contributions to academic sports medicine research. The USA is the largest contributor to the sports medicine research literature by numbers of authorships and numbers of journals. Australia, Germany, the UK and Canada are showing rapid increases in their relative contributions to the sports medicine research literature. The majority of sports medicine clinical research trials are published among relatively few sports medicine journals. These periodicals originate from several different countries and are primarily published in English, and many are internationally diverse in authorships.

The escalating number of formally trained sports medicine physicians should continue to add to the rapidly expanding body of scholarly activity and medical literature in the specialty. Further, as world communication continues to improve, the international scope of sports medicine-related research should thrive and add to the diversity of research ideas, collaborations and information dissemination.

What is already known on this topic

There is little published information about the current state of sports medicine clinical trial research.

What this study adds

  • This review of sports medicine clinical trial research publications identified in the MEDLINE database between 1996–2005 provides a snapshot of the current state of sports medicine research.

  • Data collected and analysed included publication numbers, research methods, authorships, journals, research topics and the internationality of research.

Acknowledgments

I thank V Tanji, Director of the John A Burns School of Medicine Health Sciences Library, for her assistance in conducting the literature search.

REFERENCES