PDF
ePub
Citation
Print
I. Introduction
Recently, the number of SCI (Scientific Citation Index) articles has been widely used for evaluation of medical school's productivity so that all the medical schools are trying to increase the number of SCI articles. According to the 2007 Report of Korean Ministry of Education, Science and Technology, the number of articles published through the journals indexed in Science Citation Index Database by Korean researchers has been increased to 25,494, which occupies 2.71% of SCI articles of the world1). According to the data, the number of articles of clinical medical science was 4,277, which occupies 16.77% of the SCI articles of 2007 in Korea. The SCI publication status of Korean medical science field has been reported by National Medical Academy in 20062). In the report, the accumulated number of SCI articles is 19,695 for biomedical research and 26,023 for clinical medicine. Thus, the number of articles of medical schools has shown huge expansion in recent years. For these sparkling advances in medical scientific publications, many factors are related with the productivity. These days, the medical schools are competitively investing the human and financial resources on the research. Especially the expanding number of professors and research faculties are implicated. But the investment plans were not based on their resources and capabilities so that the resources management is not efficiently functioning for their goals. Though those reports mentioned above listed the number of SCI articles, there is no report which analyzed the relationship between the resources and productivity of medical publications of each medical school. The results of such study may be of interest to a number of stakeholders, such as potential students, employers of graduates, university administrators, potential donors, funding agencies, and governments.
Thus the study was conducted to compare a quantitative and qualitative scientific publication status of the Korean medical schools during the past 12 years in relation with the resources of the medical schools. The interests are oriented to: how many papers have been published by Korean medical schools; where have these papers been published; how many times have these articles been cited; how are the output and impact measures distributed among the medical schools; and finally how the resources of the medical schools are affecting the productivity. The current research is expected to serve a basic data to set a strategy to increase the SCIE articles for the medical schools.
II. Materials and Methods
Scientific Citation Index Expanded (SCIE) Articles: The question frame is 'How many is the papers?" and "How often have the papers, published between 1997 and 2008 by full time faculties in academic year 1998 to 2005, been cited as of January 31, 2009?" The original articles and reviews are included, but bibliographies, meeting abstracts, editorials, introductions to special issues, letters, discussion pieces, corrections, and book reviews were excluded. Twelve years of archives of the SCIE articles of each medical school from the year 1997 to 2008 were extracted from SCOPUS database (Elsevier B.V., Amsterdam, Netherlands), which is one of the largest abstract and citation database of research literature and quality web sources. In the SCOPUS database, the articles were sorted by using the command designed to extract only full papers (original article and reviews) and to confine the affiliation as "medical school or medical college".: (AF-ID (########) AND AFFIL (medicine) OR AFFIL (medical) AND NOT AFFIL (oriental medicine) AND NOT AFFIL (veterinary medicine)) AND SUBJAREA (mult OR agri OR bioc OR immu OR neur OR phar OR medi OR nurs OR vete OR dent OR heal) AND PUBYEAR AFT 1996 AND PUBYEAR BEF 2009 AND (EXCLUDE (DOCTYPE, "cp") OR EXCLUDE (DOCTYPE, "le") OR EXCLUDE (DOCTYPE, "er") OR EXCLUDE (DOCTYPE, "sh") OR EXCLUDE (DOCTYPE, "ed") OR EXCLUDE (DOCTYPE, "ip") OR EXCLUDE (DOCTYPE, "no")) (Fig. 1). AF-ID is designated identification number of each institution. The AF-ID, the name of institutions and name variants of the institutions are listed in Appendix 1. The command describes list articles where 'the database field affiliation ID is ########' and 'the affiliation includes medicine or medical' and 'the affiliation exclude oriental medicine or veterinary medicine' and 'subject of the articles contains multidisciplinary, agriculture, biochemistry genetics and molecular biology, immunology, neuroscience, pharmacology, medicine, nursing, veterinary medicine, dentistry and health professions' and 'the publication year is from 1997 to 2008' and 'the types of articles are limited to article and reviews'. The affiliation name and name variants of medical school is listed in appendix 1. Among the articles, the SCIE listed journal was extracted. The criteria of SCIE, impact factors and citation numbers were classified according to the database of Thomson Scientific Company. Resources of the Medical Schools: The statuses of resources of medical schools were collected from biannual report of educational status of medical schools in Korea by medical dean's association. The reports of 1998/9, 2000/1, 2002/3, and 2004/5 were used. The generated data fields were 1) graduate students; basic medical science (master, doctor course), clinical medical science (master, doctor course), 2) hospital; number of hospitals, number of beds, number of interns, number of residents, number of fellows, 3) professors; basic science (full professor, associate professor, assistant professor, instructor), clinical science (full professor, associate professor, assistant professor, instructor), 4) research faculties; basic science (assistant, assistant researcher), clinical science (assistant, assistant researcher). The status of the human resources of the medical schools is listed in appendix 2. SPSS 15.0 and Pearson correlation test was done was used for the statistical analysis.
III. Results
1. SCIE articles
The numbers of SCIE articles of medical schools from 1997 to 2008 and the number of citations of each medical school are listed in Table 1. The total number of searched SCIE articles of the 40 medical schools during the period was 35,469 but the sum of the number of SCIE of the medical schools was 51,480 due to the multi-institution, multi co-authoring articles. The number of SCIE articles per year (SCIE/yr) has increased from 1,024 in 1997 to 8,462 in 2008 (Fig. 2). The total number of SCIE article may cause optical illusion because the number was multiplied according to the affiliation of the authors. Also the number does not differentiate whether the authors are corresponding or co-authoring. Thus, the number of SCIE articles of corresponding author (CA-SCIE) of designated school was searched to estimate the productivity of each medical school. The total number of CA-SCIE of 40 medical schools was 29,048, which is 56.4% of the number of SCIE. Annual increase is 8.6% for the number of SCIE articles and 7.6% for CA articles. Among the 40 medical schools, the top four medical schools (Seoul National, Yonsei, Seong Kyun Kwan, Ulsan) occupies only 23% of total number of professors but they occupies 40% of SCIE articles and 50% of CA articles. Most of the medical schools are having less than 10,000 citations and 1,000 CA-SCIE (Fig. 3). 27.5% of medical schools publish less than 50 SCIE articles per year and 57.5% of medical schools publish less than 50 corresponding articles per year. Among the SCIE articles, 4,272 articles have foreign institutions as corresponding author's affiliation and 813 articles have national research institutions such as KRIBB (Korea Research institute of Bioscience and Biotechnology) or KAIST (Korea Advanced Institute of Science and Technology) as corresponding author's affiliation.
2. SCIE per capita
The SCIE/yr during the period ranges from 10.3 to 652.8 and median was 55.2 (Table 2). The number of SCIE per professor per year (SCIE/Prof./yr) ranges from 0.12 to 1.76 and median was 0.43. But the use of absolute number of SCIE is error prone since it does not discriminate multi-author article over single-author article. The CA-SCIE per year (CA-SCIE/yr) has increased from 608 in 1997 to 4,607 in 2008. The CA-SCIE/yr during the period ranges from 2.8 to 432.3 and median was 25.3. The number of CA-SCIE per professor per year (CA-SCIE/Prof./yr) ranges from 0.03 to 1.17 and median was 0.2. The average number of SCIE article per professor per year (SCIE/Prof./yr) in Korea is 0.5 and the average number of corresponding SCIE article per professor per year (CA-SCIE/Prof./yr) is 0.25. Only one medical school (Seoul National University) exceeded 1.0 in both factors.
3. Citations
The citation number of each article was obtained from the SCOPUS Database. The score provides the total citation number of each individual articles and it does not consider longevity of the article. The number of citations reflects past several years accumulated citations of the articles. Also, the number is multiplicated by the number of affiliations of the authors. So the citation number of CA-SCIE can be considered as true citation number. The accumulated citation number of CA-SCIE was 226,277. Among the SCIE articles published, 75.7% of articles have less than 10 citations and only 10.4% of the articles have citations of more than 20. The average number of citation per CA-SCIE was 7.8. Meanwhile, for the CA-SCIE of foreign correspondences, 52.3% of articles have less than 10 citations and 19.6% of articles have citations 30 or more.
4. Impact factor
The current status of impact factors of CA-SCIE are shown in Table 3. Among the 29,048 CA-SCIE articles, 28,370 articles were found to have impact factors. 90.8% of the articles have impact factors less than 5 and only 9.2% of the articles have impact factors of 5 or more. Meanwhile, the percentage of articles with impact factor 5 or more is 33.3% for the articles which have correspondence as foreign institute. Also, it is 20.3% for the articles which have correspondence as National Research Center such as KRIBB or KAIST.
5. H-Index
H index is easier to demonstrate the productivity and citation status of the authors or institutions. Table 3 shows H index of the medical schools. The H index ranges from 3 to 30 and median is 12. The Seoul National University and Yonsei University had H index of 30. Meanwhile the H index of foreign institute is 33.
6. Relationship with resources of the professor
The number of SCIE articles, the number of SCIE articles per professor (SCIE/Prof.), the number of SCIE articles cited 30 or more (highly cited), the ratio of SCIE article cited 30 or more in total number of SCIE articles were compared. Total number of SCIE article is closely related to the number of SCIE articles cited 30 or more. But the ratio of highly cited articles does not related with the total number of SCIE articles. The number of SCIE articles per professor was related with the number of highly cited articles. The ratio of highly cited articles did not have relationship with other factors. The correlations between research product and resources are listed in Table 4. Total number of SCIE articles and highly cited articles are related with the total number of professors, the number of basic science professors and the number of clinical science professors. The number of basic science professors did not related with the total number of SCIE articles and highly cited articles but it was significantly related with the number of SCIE per professor. The number of highly cited articles is increasing as the number of full professors increasing but the ratio of full professors did not show any relationship with the factors. The most strongly related professor factor was the number of associate professors of basic medical science (r=.899, p<0.01). The average number of article per professor per year is 0.5 and the average number of CA-SCIE per professor per year is 0.2. The highest number of SCIE article per professor per year is 1.76 in medical school of Seoul National University. In the meantime, the number of articles per professor of Gwangju Institute of Science Technology (GIST) was 5.07. Seoul National University record was 1,403 in 2008 to make SCIE/Prof. about 3.8. The number of full professors and the number of professors of basic science showed high correlation coefficient for the number of SCIE articles and the number of CA- SCIE. But for the SCIE/Prof., the total number of professors showed weak correlations. And the number of professor of clinical science showed no correlation. The clinical professors did not contribute the number of SCIE article per professor. The number of full professor and the number of basic science professor showed relationship with CA-SCIE per professor. The number of full professors and the number of associate professor are highly related with the proportion of CA-SCIE. The only factor affected both the total number of SCIE article and the number of SCIE article per professor was the number of researchers per professor.
7. Relationship with resources of the graduate students
The total number of students of master course and the total number of doctor course were related with the total number of SCIE articles and the number of highly cited articles. But the number of graduate students did not related to the number of SCIE articles per professor and the ratio of highly cited articles. The number of graduate student per professor did not show any relationship with the factors.
8. Relationship with resources of the clinical faculties
The number of total SCIE articles and highly cited articles are increased as the number of bed is increased. Also, the number of residents was related with the total number of SCIE articles and highly cited articles. The number of resident per professor was related with the total number of SCIE articles and the number of highly cited articles. Clinical fellows had no relationship with any of the research product related factors.
9. Journals
The number of submitted journals was 2,202. The Journal of Korean Medical Science was ranked to top to have 1339 articles occupying 3.78% of the SCIE publications followed by Yonsei Medical Journal to have 919 articles occupying 2.59%. Top 30 journals occupied 24.3% of the articles and 11.9% of citations (Table 5).
IV. Discussion
SCI is scientific citation index database used since 1963, which is now served as on-line service named as Web of Science. The expanded version of SCI is SCIE, which is not different in the aspect of quality of the journals, so expanded version was used. As the competitions among the universities get higher in the research area, the numbers of SCIE articles become a representative of medical schools' research productivity. With these efforts, the number of SCIE articles of Korean medical schools has shown huge expansion during last 12 years. The Korean Ministry of education science and technology (MIST) reported that the number of SCI articles of Korea in 2007 was 25,494 ranking 12th among 180 countries. The number of SCIE articles of clinical medical science field was 4,277 occupying 2.03% of Korean SCIE articles1). The result of current analysis is discordant to the report by National Medical Academy2) and by Korean Ministry of Education1). The discordant comes from the difference of method of analysis. The current study strictly handled using SCOPUS DB and SCIE list of Thomson Scientific Company because the Pubmed or Medline do not discriminate thorough information about affiliation. And further, they do not provide citation information and differentiate multi-institution, multi-author articles. Because the type of the articles was limited to 'original article' and 'reviews' to exclude other type of articles such as short communications, letters to the editor, comments and editorials, the number of SCIE article is smaller than previous report. Moreover, the number of corresponding article is much smaller than expected because only one medical school can be counted to be correspondence.
Those previous reports did not give information of relationship of the medical schools and their resources though the total number of SCIE articles is largely affected by the number of professors and the resources of the university. Therefore, the understanding how the resources of medical school influence on the productivity of SCIE publications of medical school is needed to plan the strategy to promote research for the medical schools. So the author tried to analysis the relationship of medical school's resource and productivity.
The result showed that the number of SCIE articles has been increased from 1,400 in 1998 to 8,462 in 2008 which is 8.3 fold large. Also the number of corresponding SCIE articles of Korean medical schools is increased from 820 in 1998 to 4,607 in 2008 which is 7.6 fold large. The current research demonstrated that the quantitative factors such as total number of SCIE articles is significantly related with qualitative factors such as number of SCIE articles per professor and the total number of highly cited articles.
1. Professors
Most of the professors resources are related with the quantity factor such as total number of SCIE articles and quality factor such as total number of highly cited articles (cited 30 or more). The number of instructors of basic medical science was negatively related with the total number of SCIE articles. Possible explanation for this phenomenon is that to publish a SCIE article, it took more than one year after they were hired so that they negatively affect the total number of SCIE articles. The only factor affect the number of SCIE articles per professor was the ratio of professors of basic medical science (r=.444, p<0.01). The associate professors, especially in basic medical science professors are most highly pressed for tenure. And unlike the professors of the clinical science who have to share much of their activity to manage the patients, the professors of basic medical sciences can concentrate for the research.
2. Graduate school
All of the factors of graduate school except the number of graduate school student per professor were related with the number of SCIE publications. It is quite unreasonable that as the number of graduate school students, especially those of basic medical science, increases, the SCIE article number should be increased because the graduate school student should contribute to the research and submit scientific articles before they graduate. Also, no resource factors of graduate school student were associated with SCIE articles per professor. These phenomenon may suggest that many of graduate school students are not properly involved to research activity and not contributing to the professor's research.
3. Clinical faculties
The number of research faculties (assistant researcher or research assistant) per professor was only factor to affect the articles per professor. The number of residents or clinical fellows did not relate with SCIE per professor. Of the note, the number of hospital bed per clinical professor was significantly affected to the number of SCIE articles per professor. The possible explanation of this is that the increased number of hospital bed per professor means they can have larger clinical patient cases to make SCIE publications increase.
To increase SCIE articles per professor, the ratio of professors of basic medical science and the number of research faculties per professor need to be increased. The number of associate professors of basic medical science and total amount of research fund are strongly associated with the total number of SCIE articles. These two factors are also strongly associated with total number of articles cited 30 or more. For the clinical professors, the number of hospital bed was related with SCIE per professor. The clinicians with more clinical cases can produce more SCIE articles.
4. Citation
In the report by MIST, the citation number per SCIE article was 3.72. But according to current analysis, the average number of citation is 7.8. The difference is that the report of MIST used only 5 years of citation. But our report reflects total accumulated number of citation regardless of year. Table 6 is top four universities' citation status which shows that the citations of the articles published in from 2000 to 2004 are higher than recent 5 years. And the current analysis did not discriminate self-citations. Although it is possible that an author may be citing his or her work, there is no way to judge self citations are meaningless and moral- hazard. Moreover, it is becoming standard practice to include self-citations in meta-review analysis since there is ample evidence that this practice is common3).
5. Impact factor
The impact factor, often abbreviated IF, is defined as the mean number of citations a journal receives over a 2-year period. Impact factors are calculated each year by Thomson Scientific for those journals which it indexes, and the factors and indices are published in Journal Citation Reports. It is derived from citations to all articles in a journal so the number is not giving information about the quality nor the quality of the work. By merely counting the amount of citations and disregarding the prestige of the citing journals, the ISI IF is often regarded as a metric of popularity, not of prestige. But even with the debates about the pros4)5) and cons6), the IF is being used to decide whether or not the professors get promoted or to have tenure or are offered a position in a department, or are awarded a grant. It may be used for simple measurement to compare the quality of work. The proportion of CA-SCIE having IF more than 5 is 9.2%. On the other hand, those of national research center and foreign institutions are 20.3% and 33.3% respectively. The proportion of CA-SCIE cited 30 or more was 5.3% for Korean medical schools, while that of foreign institutions and national research center were 19.6% and 9.2%. With these simple comparisons, the articles of Korean medical schools can be considered to have lesser citations than these institutions.
6. H-Index
H-index has developed by physicist Jorge Hirsch7) to quantify both the impact and productivity. The index is based on the distribution of citations received by a given researcher's publications. H-index is highest number of papers a scientist has published that have each received that number of citations. For example: If A has 7 articles cited 5 times, 5 articles cited 5 times and 3 articles cited 6 times, the H-index of A is 5. In other words, a scholar with an index of h has published h papers each of which has been cited by others at least h times. Thus, the h-index reflects both the number of publications and the number of citations per publication. The h-index serves as an alternative to more traditional journal impact factor metrics in the evaluation of the impact of the work of a particular researcher8). In the current study, Seoul National University and Yonsei University were highest to have H-index of 30 followed by Ulsan University of 25. For the H-index, the most highly cited articles are contributing, it determination is a relatively simple. But the H-index seems to be useful method to compare the productivity of the medical schools.
7. Limitations of the study
Every citation study has limitations in the design because the database contains a number of imperfections. The limitations listed below should be kept in mind when interpreting the results. These factors significantly affected to the total number of SCIE articles and caused significant gap in comparison with the numbers of SCIE which the medical schools are scoring as their product. For the number of medical schools, the exact number of medical schools in Korea is 41 because the Yonsei University has two medical schools. But the authors of two medical schools of Yonsei University are frequently using same affiliations so that they were counted as one university. For the resource of publications database, the study only focused on SCIE journals though the articles indexed in Pubmed or Index Medicus Database may be the intellectual impact of academic research. By focusing on SCIE journals, the analysis can provide a valid measure of academic research. Also, only the original articles and reviews are included. The bibliographies, meeting abstracts, editorials, introductions to special issues, letters, discussion pieces, corrections, and book reviews were excluded. The study was closed by January 31 and conducted from February 1, 2009. The articles of 2008 published after January 31, 2009 were not included. And there were miscellaneous problems such as incorrect spelling of affiliation, for example some authors listed their nation as North Korea. There were dozens of articles suspected to have duplicated publication which were discarded before analysis. For the resources data, the data of resources were obsolete to the biannual report of medical dean's association, of which is not complete. Some data is missing so that some important information other than human resources such as the amount of research fund and the number of research centers was not able to be analyzed. For the citations, the citation number did not count the longevity of the articles. The older articles may have higher number of citations in relation with the recent articles which means recent articles may be counted to have lesser citation numbers. And the self citation was not considered to count the citation number.
In conclusion, the current study showed there are huge variations across the medical schools in the aspect of scientific publications. Most of the medical schools of Korea showed limited capacities in relation with foreign or national research institutes. One of the reasons is that the graduate school students, even with the high numbers, are not contributing for the research of the professors. Instead, research faculties other than graduate school students such as research assistant or assistant researchers affect significantly to the SCIE publications. The research strategy should be planned with the consideration of the medical school's resources. With larger hospital and higher number of hospital bed, the clinical professors may play major role to produce SCIE articles. For the medical schools with limited resources, the basic science professors with assistant researchers may play key role to increase SCIE production.
XML Download