Introduction

Scientific theses are academic texts that systematically present data and analyses obtained through study. These provide experience for assistants in specialized education to assess their research skills, critical thinking abilities, analytical skills, and ability to use scientific methods (1).When examining the publication rates of specialty theses in various medical fields in Turkey, the rates range from 6.2% to 48.3% (2-6). Indeed, more than half of these are not published. Failure to publish theses in national or international peer-reviewed journals implies a waste of effort, time, resources, and intelligence, as well as a missed opportunity to disseminate knowledge. When a thesis is viewed solely as a means to complete specialized education, the quality of the study may decrease (7,8). The publication of a thesis in a peer-reviewed journal signifies that the generated knowledge is deemed acceptable within the scientific community, indicating the quality of the study (9,10).

In our country, the subspecialty program of endocrinology and metabolism requires an additional 3-year training period, which has been entered through an examination since 2007, after completing the main specialization program in internal medicine. For the specialization in internal medicine, it is mandatory to conduct a thesis; however, presenting the thesis is sufficient, and it is not mandatory to publish the thesis as a scientific publication. In our country, the requirement for a thesis in the subspecialty of endocrinology and metabolism was abolished in 2014. However, there are still some specialists who choose to conduct a thesis after this date. The additional training period required for subspecialization, the desire of physicians to specialize in a more in-depth manner, and their focus on a more specific patient population or medical condition may increase the academic expectations of subspecialists. There has been no study conducted in our country regarding the factors influencing the publication of endocrinology fellowship theses.

In this study, we evaluated the publication rate of endocrinology theses in international peer-reviewed journals and the factors influencing theses’ likelihood of publication through a bibliometric analysis of theses published between 1980 and 2023.

Methods

Compliance with Ethical Standards

This study is a bibliometric analysis. Because the study was conducted by researching public databases on open websites and did not involve animals or humans, it does not require official ethical committee approval or informed consent. This study was conducted in accordance with the Helsinki Declaration and the registration and conditions of the National Thesis Center of the Higher Education Council.

Study Design

The Endocrinology and Metabolism subspecialty theses between January 1980 and April 2023 were scanned from the online archive of the National Thesis Center of the Higher Education Council from April 13 to April 30, 2023. The detailed search terms and fields used are shown in Figure 1. A total of 115 theses were identified after selecting “Minor Specialization in Medicine” as the search criterion. Pediatric endocrinology subspecialty theses were excluded from the study. The remaining 93 adult endocrinology subspecialty theses were included in the study. All of them were from 2000 onwards. To access theses that could not be identified through detailed searches, the membership list of the Turkish Society of Endocrinology and Metabolism on their official website was also scanned again in the National Thesis Center. An additional 184 adult endocrinology thesis data were obtained and included in the study. In total, 277 thesis data points were recorded (Figure 2).

Theses were searched on the Web of Science, Scopus, TR-Dizin, and Google Scholar. The published theses were divided into three categories: those published in international journals indexed in Science Citation Index-Expanded (SCIE) and Scopus; those published in other international journals; and those published only in national journals indexed in TR-Dizin. Publications that fit into multiple categories were hierarchically included in the upper category. The province and institution where the theses were conducted, thesis year, thesis topics, study design, publication dates of published theses, index of the journal where they were published, the author’s number of first-author publications in journals indexed in SCIE and Scopus, the author’s Scopus H-index, number of publications by thesis advisors in journals indexed in SCIE and Scopus, H-index of the thesis advisor, academic status of the thesis advisor, and gender of the author and thesis advisor were recorded. These parameters were compared between theses with publications in SCIE- and Scopus-indexed journals and those with publications in other indices (other international journals and national journals indexed only in TR-Dizin). The publication rates of theses and those published in journals indexed in SCIE and Scopus were examined. Since theses started to be scored for academic promotion after 2016, the publication rate of theses before and after 2016 was analyzed.

Statistical Analysis

The data analysis was performed using SPSS 25 (Statistical Package for Social Sciences). Descriptive statistics are presented as median (minimum-maximum) and mean ± standard deviation for numerical variables and as observation count and percentage (%) for nominal variables. The normality of the distribution of numerical variables was tested using the Kolmogorov-Smirnov test. The presence of a statistically significant difference in numerical variables between the two groups was evaluated using the Mann-Whitney U test. Nominal variables were assessed using the chi-square test, Fisher’s exact test, and Fisher-Freeman-Halton exact test. A logistic regression analysis was conducted to determine the criteria influencing the publication of theses. The odds ratios (OR), 95% confidence intervals (95% CI), and p-values for each variable were reported. The results were considered statistically significant at p<0.05.

Results

A total of 277 theses were included in this study. The descriptive data are presented in Table 1. The distribution of theses by province and year is shown in Figures 3 and 4, respectively. The most common topics for theses were thyroid and diabetes (Table 1). Twenty-one (90.6%) theses were conducted at universities, while 26 (9.4%) were conducted at training and research hospitals.

Of the 277 theses, 142 (51.3%) had been published in international or national journals. Among these, 117 (42.2%) were published in SCIE/Scopus-indexed journals, 107 (38.6%) were published in SCIE-indexed journals, 8 (2.9%) were published in other international indexed journals, and 17 (6.1%) were published in national journals indexed only in TR-Dizin (Table 2, Figure 5). The time to the publication of theses was a mean of 3.8±3.3 years, with a median of 3 years (range: 0-21).

When looking at theses in 21-year intervals, between 1980 and 2000, 19 theses (33.3%) were published, while between 2001 and 2023, 123 theses (55.9%) were published. When looking at theses in 10-year intervals, the highest number of publications occurred between 2010 and 2019 (n=67, 47.2%) (Figure 6). When considering 21-year intervals, specifically between 2001 and 2023, it had the highest number of publications (n=123, 86.6%). Of the published theses (n=142), 90 (49.2%) were conducted in the three major cities (Ankara, Istanbul, and Izmir), while 52 (55.3%) were conducted in other cities.

The authors’ H-index had a mean of 11.5±7.2, with a median of 11 (range: 0-51). The authors’ mean number of first-author publications was 4.6±5.1, with a median of 3 (range: 0-38).

The thesis advisors’ H-index had a mean of 19.1±7.4, with a median of 19 (range: 1-51). The thesis advisors’ average number of publications was 76.7±49.1, with a median of 67 (range: 1-313).

When the published and unpublished theses were compared, there were statistically significant differences in the number of first author publications, thesis year, 10-year interval, 21-year interval, and institution status (p<0.001, p=0.024, p=0.001, p=0.002, p=0.011, respectively) (Table 3). In the regression analysis, a relationship was found between the thesis having a publication and the thesis being conducted in a training and research hospital, a higher number of first-author publications, and a more recent year of the thesis (Table 4).

The H-index of thesis advisors for theses published in SCIE/Scopus-indexed journals was statistically significantly higher than that of thesis advisors for theses published in other international indexes or only in TR-Dizin (p=0.029). Additionally, the majority of these publications occurred after 2000 (p=0.045) (Table 5).

Discussion

In our study, the publication rate of endocrinology theses was 51.3%, with a publication rate of 42.2% in SCIE/Scopus-indexed journals. The publication of endocrinology theses was found to be associated with the thesis being conducted in a training and research hospital, a higher number of first-author publications by the author, and a more recent year of the thesis. Thesis advisors of theses published in SCIE/Scopus-indexed journals had higher H-index values, and a significant portion of these publications were from theses conducted after 2000.

In our study, we found that the overall publication rate of endocrinology theses was higher than the overall average in Turkey. de Nonneville et al. (11) demonstrated that 70% of medical oncology theses in France resulted in publications. The higher publication rate of minor specialization theses may be due to higher academic expectations among minor specialization experts.

In 2012, Sipahi et al. (7) showed that the rate of publication of theses in national journals is higher than that in international journals. However, in our country, the publication rate of medical theses in SCIE and other international indexed journals has increased over the years. Researchers may have realized that publishing their work in international journals can enhance scientific collaboration, expand their readership, and gain greater international recognition, and they may have developed a preference for publishing in international journals as well as local journals. In addition, this can be attributed to the selection of thesis topics that contribute to the scientific literature and are internationally recognized, resulting in higher-quality publications. In addition, the influence of academic promotion criteria, which assign higher scores to studies published in international peer-reviewed journals, may have affected journal selection.

Most endocrinology theses were published after 2000, indicating a relationship between thesis publication and more recent years. Several factors may have contributed to this trend, such as the recent promotion of original research in medical disciplines, the advancement of technical capabilities in training institutions for assistants, easier access to reference articles through widespread internet use, and simplified submission processes to journals, which may have encouraged authors to pursue thesis publication (12). Indeed, an increase in the usage of English and the availability of support from organizations providing foreign language assistance may have led to an increase in authors’ submissions to international journals.

In different specialization fields in our country, the average time for publications derived from theses ranges from 2.8 to 3.8 years (2,6,12-15). This is a longer period of time than the two years in international settings (10). After completing specialization training in our country, mandatory working in state hospitals with a high workload and departure from academic environments may have delayed the conversion of theses into publications.

In our study, the average publication time of endocrinology theses was found to be 3.8 years, which is longer than the average published in Turkey and the global literature. Although studies derived from endocrinology theses are published more in journals indexed in databases such as SCIE and Scopus, the longer publication time compared with the literature may be attributed to incorrect journal selection and delays in peer review processes due to the focus on journals that do not charge fees due to high application and publication fees.

In terms of publication rates and the databases in which theses are published, no significant differences were found among state universities, training and research hospitals, and private universities (14,16). In another study, it has been shown that the publication rate of theses completed in university hospitals is five times higher compared to those conducted in state hospitals (3). This difference may be due to more than a 10-year difference between the research years of the studies. In our study, the majority of theses were conducted at state universities. However, it was found that the institution associated with the publication of theses was primarily a training and research hospital. With a larger patient population and better financial resources, training and research hospitals may have an advantage in accessing the materials and technical facilities required for thesis research.

In our study, the publication rates were similar between clinical and experimental studies. In the literature, the study design has been associated with the publication status of theses (2,17). It has been shown that experimental studies are published more frequently in SCIE-indexed journals (4,13,16,18). There are also studies that found similar publication rates for clinical prospective studies and animal experiments (12,15). Both are valuable types of scientific studies. In our study, no significant difference was observed between the publication rates of clinical and experimental studies. However, it is important to note that the number of experimental studies in our study was much lower than that in clinical studies.

In our study, a decrease in the publication rate of theses after 2014 was observed (Figures 4 and 6). This decline can be attributed to the removal of the subspecialty training thesis in our country. Similar to our study, in a study where the thesis was removed from the postgraduate curriculum in India, a significant decrease in publications from those departments was observed during that period (19). When evaluating the impact of the changes in academic promotion criteria in our country in 2016 and the inclusion of publications derived from theses in the scoring system, it can be observed that most publications were from theses completed before 2016. Although it was not mandatory before 2016, the high rate of thesis publications indicates the high level of academic activity and interest among endocrinologists.

In the literature, factors such as having another article in which the author is the first author, being the first author in the thesis publication, and having a higher number of author publications have been identified as important factors in the conversion of theses into publications (2,14). The number of first-author publications, as well as metrics like the H-index, are important indicators of academic activity. In our study, the number of first-author publications was associated with thesis publication. However, the H-index of authors was similar between theses that were published and those that were not. In studies, the first authors assume significant responsibilities at various stages, from study design to publication. In multicenter or national data studies, even if the experience and knowledge of scientific publishing may not be sufficient, co-authors can have a high H-index because of the citations received. Despite the recognition of the H-index as an indicator of academic activity, authors may have few or no first-author publications. In addition, presenting the thesis as an oral presentation, which is sufficient for academic promotion, may have a negative impact on authors’ motivation to publish their theses. Ultimately, in our study, the number of first-author publications, which enhances the experience of scientific publishing, was found to have a greater impact on thesis publication than the H-index of authors.

There are studies investigating the factors related to the thesis advisor in the publication of the thesis. In our study, it was observed that the thesis advisors of authors with published theses in SCIE/Scopus had higher H-index scores. This indicates the importance of the academic activity of thesis advisors in the publication of theses. However, in another study, the impact of the advisor H-index was not found to be significant (2). There are studies that show the influence of the academic title of thesis advisors (such as assistant professors) as an effective factor in the publication of theses, as well as studies that demonstrate no significant impact (2,5,14). Theses are important sources in terms of increasing the publication status of both the advisor and the author (10). Therefore, it is possible for assistant professors who serve as thesis advisors at the early stages of their academic careers to provide more support to the authors. In our study, the advisor’s academic title was not found to be a significant factor influencing publication. However, most of the thesis advisors were professors, and there were very few assistant professors. Another factor, the publication count of the thesis advisor, did not have a significant effect on our study. While the H-index of thesis advisors is a significant factor, the number of publications does not appear to have a significant effect, highlighting the importance of quality publications. However, there are studies in the literature that find the advisor’s publication count to be an important factor in the publication of theses (2).

Study Limitations

The possibility of certain theses going undetected is a potential limitation of our study due to the registration of some theses uploaded to the national thesis center being recorded as “specialization in medicine” instead of “minor specialization in medicine”. Despite these limitations, this is the first study that sheds light on the factors affecting the publication of adult endocrinology theses and analyzes adult endocrinology theses bibliometrically by scanning both national and international journal indexes.

Conclusion

Our study is the first to be conducted on subspecialty theses in our country. In our study, the rate of publication in SCIE/Scopus indexed journals for endocrinology theses was higher than the national average. However, there are still many theses waiting to be published. The academic expectation is higher for subspecialists who have received additional training, which is entered by an exam to specialize in more specific subjects. The publication of theses contributes to the scientific literature, disseminates knowledge and experiences, and enhances the academic recognition of healthcare professionals in our country. It is important to provide specialists with the necessary support, such as time, financial assistance, and training, to ensure the production of high-quality theses and facilitate their publication, particularly in internationally indexed journals.

Ethics

Ethics Committee Approval: This study is a bibliometric analysis. Because the study was conducted by researching public databases on open websites and did not involve animals or humans, it does not require official ethical committee approval.

Informed Consent: This study is a bibliometric analysis. Because the study was conducted by researching public databases on open websites and did not involve animals or humans, it does not require official informed consent.

Authorship Contributions

Concept: C.T.B., M.Y., Design: C.T.B., M.Y., Data Collection or Processing: C.T.B., Analysis or Interpretation: C.T.B., M.Y., Literature Search: C.T.B., Writing: C.T.B., M.Y.

Conflict of Interest: No conflicts of interest were declared by the authors.

Financial Disclosure: This study received no financial support.