Publication Patterns of Ophthalmology Residency Dissertations in Turkey

To investigate the frequency of converting a residency dissertation to a journal article in Turkey. To compare the different residency programs and their publications rates; to sub classify of dissertations according to sub-specialties. Ophthalmologists who graduated before 2010 conducted surveys. The surveys questions included graduation year, residency program, topic of the dissertations, type of the study (prospective, retrospective, or experimental), publication rate from the dissertations, and type of journal for published dissertations. Three-hundred-eight ophthalmologists conducted surveys. Fifty seven (18.5%) dissertations were published in SCI-Expanded Indexed journals (39 of them (23%) from the university hospital programs and 17 of them (13%) from state hospital programs (p=0.036)) and 97 of dissertations were published in national journals. 92 of 170 (54%) dissertations from the university based residency programs and 59 of 131 (45%) dissertations from the state hospital based residency programs were published in national journals or international journals indexed in the SCI-Expanded Index (p=0.17). A publication rate from dissertations was similar between big city university programs and small city university programs (23.3%; 22.6%). The most frequently studied topics of dissertations were cataract-refractive surgery (21.5%), followed by retina (21%), glaucoma (21%), cornea-anterior segment (19%), and strabismus (7%). Of the study design, 57% of them were prospective, 32% were retrospective, and 10.5% were experimental study. In this study, it seems to be that lesser than one fifth of Ophthalmology Residency dissertations in Turkey converted to journal articles in SCI-Expanded journals. It is obvious that the overall quality of the dissertations was inadequate to be published in indexed journals and it is necessary to take some measures to increase the scientific quality.


INTRODUCTION
Today, the most important indicator of scientific production for individuals, academic centers, and countries is the number of publications in indexed journals.Residency dissertations are one of the most important sources of scientific publications.The completion of a dissertation is obligatory for ophthalmology training at both university and state hospital programs in Turkey.However, the scientific quality of dissertation is controversial.Converting a residency dissertation to a journal article is a reliable indicator of the scientific quality of dissertation.The data on publication rate from residency dissertation is limited (1)(2)(3)(4)(5)(6).In this article, we investigated the frequency of converting the ophthalmology residency dissertations to journal articles with comprehensive surveys.

MATERIAL AND METHOD
The study was conducted among ophthalmologists who graduated in 2010 or before.Most of the ophthalmologists participated surveys at 2012 Turkish Ophthalmology Society meetings.Rests of the ophthalmologists were contacted directly or with cell phone.The questions of surveys included graduation year, residency program, study design (prospective, retrospective or experimental), type of the journal (whether in SCI-Expanded Index or not) for published paper, and topic of dissertation.If a physician published some part of the dissertation, it was accepted as a published article.Topics of dissertations were classified into 8 groups: 1) Cataract and refractive surgery, 2) Retina, 3) Glaucoma, 4) Corneal-anterior segment, 5) Strabismus, 6) Neuro-ophthalmology, 7) Uvea, and 8) Oculoplastic surgery.
Physician name or initials were used when entering data to avoid counting the same person twice.Additional columns were added to differentiate between university or state hospital based programs, and between big city and small city universities.Answered questions in the surveys were only analyzed.
We included SCI-Expanded Indexed journals in the study since SCI-Expanded Index has been widely accepted by most of the scientists and contains of peer-review, international and respected journals.Furthermore, a publication in SCI-Expanded journals is an important criterion for academic promotions in Turkey.The impact factors were not evaluated in this study.
In this study, "SSPS 20.0 for Windows" (SPSS Inc., Chicago, Illinois) statistical program was used for statistical analyses.Non-parametric data were analyzed with non-parametric tests.Percentage (%) was used for a descriptive statistic analysis of nonparametric data.Mann-Whitney-U test was used for comperative groups analyses.A p value less than 0.05 was accepted as statistically signifincat.

RESULTS
Three hundred and eight ophthalmologists participated in completing the surveys.This number is 11.8% (308/2606) of all ophthalmologists practiced in Turkey in 2010.Half of the dissertations (%50; 154 dissertations) were converted to the journal articles.Of the articles, 57 of them were published in journals indexed in SCI-Expanded Index (39 of 57 were from the university based residency programs and 17 of 57 were from the state hospital based residency programs (p=0.036)).A type of residency program was not detected in one of the published article.Of the dissertations, 97 of them (31.5%) were published in the national journals not indexed in the SCI-Expanded Index.Ninety two of 170 dissertations from the university based residency programs and 59 of 131 dissertations from state hospital based residency programs were published in national or international journals (p=0.17).The publication rate in the journals indexed in SCI-Expanded Index from dissertations was similar between the residency programs of the older universities of three big cities and newer universities of the other cities (23.3%; 22.6%) (Table 1).

When a distribution of dissertations according to years
Note: The reason of discrepency between total number and total number of institutions was due to incomplete answers of questions in surveys by some participants.2, 3).

DISCUSSION
Since the study data was collected from surveys, our results may be subjective.This condition can be seen as a deficit, but it still can be a healthy way to access accurate data.Currently, patient's satisfaction studies and quality of life assessment studies are usually conducted based on subjective surveys.We previously reported that total ophthalmic article count had been significantly increased during the period of 1990 -2013 (7).In addition,

Table 4. Distribution of the Dissertations According to Topics
the previous related studies generally collected their data by searching from the central data base (1,4,6).
Although more objective data can be collected by those ways, it may cause high false negative results that may be related to several factors including 1) searching with wrong name, last name, or article name, 2) differentiation of key words and titles, and 3) presence of similar names.
Our results were similar to the previous studies involving different branches of medicine.In Salmi et al.'s study investigating 300 medicine dissertations between 1990 and 1993, 17% of them were published as an article in indexed journals (1).Twenty four percent of dissertations from two medical schools in Croatia were published in indexed journal between 1990 and 1999 (2).In Peru, 17.6% of 482 dissertations were published in indexed journals between 2000 and 2003 (5).The rate of converting a dissertation to a journal article was reported 24% in one study from Finland and was 30% (highest rate in the literature) in one study from India (4).In Özgen et al study from Turkey, the publication rate was 6.2% in journals indexed in SCI-Expanded Index among 22625 registered dissertations (6).We believe this rate did not reflect the actual rate due to high false negative results related factors that we discussed above.Our study is the first article related to ophthalmology residency dissertations.In Ozgen et al.'s study, the publication rate in journals indexed in SCI-Expanded Index from ophthalmology dissertations was 3.8% (6).The factors that causing high false negative results while searching data base may cause this low rate, we believe our rate 18.5% would be more accurate rate.
In our study, the publication rate from residency dissertations in journals indexed in SCI-Expanded Index was similar between the big city university programs and the small city university programs.However, the publication rate was higher in the university hospital residency programs than the state hospital residency programs.The publication rate of ophthalmology dissertations was increased until 2000, and then it became stable.However, the publication rate per year in peer-review journals did not show remarkable change over time (around 50-60%).The rate of publication in journals indexed in SCI-Expanded Index was increased between 1995 and 2000, and then it became stable after 2000.In addition, most of the studies were prospective and only one-tenth of the studies were experimental.We believe that if experiment opportunities increase and more grants are provided for experiments, the ratio of experimental studies will increase.
The publication rate from dissertations in indexed journals is an important indicator of scientific achievements of individuals, academic centers, and countries.Required steps should be taken in order to use this source more effectively.When the publication rate in any journals (~50%) and in SCI-Expanded indexed journals (~20%) is considered, significant time and money have been wasted during residency training.Publications from residency dissertations are not considered during academic promotions in some countries such as Turkey and India that has caused remarkable decrease in publication rate (8).

Özgen et al. suggested residency option without disserta-
tion should be offered to physicians who do not consider academic career that may help to reserve sources such as time and money (6).We do not agree their suggestion because it may damage established dissertation tradition and development of scientific and analytic thinking skills in residents.Incentive awards from government and private institutions given to published articles may increase the rate of publication from residency dissertations.

Table 1 .
Distribution of the dissertations according to the academic institutions

Table 3 .
Distribution of the dissertation/publication ratio according to years was analyzed, the oldest dissertation was performed in 1971.After 1990s, the participants' number was dramatically increased.The publication rate per year in peerreview journals did not show remarkable change over time (around 50-60%).The rate of publication in journals indexed in SCI-Expanded Index was increased between 1995 and 2000, and then it became stable after 2000 (Tables

Table 2 .
Distribution of the Publications in SCI-Expanded Journals According to Years

Table 5 .
Distributions of the Dissertations According to Study Design