Publication bias in post-graduate medical education dissertations in India

Article type
Authors
Kumar A1, Reljic T1, Charan J2, Chaudhari M3, Moitra M4, Mhaskar R1, Djulbegovic B1, Kosambiya J5
1Division of EBM, Morsani College of Medicine, Internal Medicine, USA
2Deparment of Pharmacology, GMERS Medical College, Patan (Gujrat), India
3Department of Pharmacology, Government Medical College, Surat , India
4Department of PSM, Government Medical College, Surat , India
5Department of PSM, Government Medical College, Surat, India
Abstract
Background:
Successful completion of a post-graduate degree in medicine in India requires a thorough investigation of a scientific issue in the form of dissertation, which is an important aspect of training future physicians in research. While dissertations are a mandatory requirement for graduation, publication of dissertation is not. Publishing findings, hypotheses, theories, lines of reasoning and evidence is critical to the progress of science. A formal determination of publication bias in post-graduate medical education dissertations has not been done.

Objective:
Assessment of publication bias in post-graduate medical education dissertations.

Methods:
We conducted a quantitative cross-sectional e-survey and followed the Checklist for Reporting Results of Internet E-Surveys. The survey covered three areas of demographics, publication status and publication details. We contacted three medical associations and colleges in India. We used descriptive statistics to summarize data; we assessed differences in publication rates using the Mann-Whitney test.

Results:
The survey completion rate was 88% (212/241). Most responders were males (67%; 141/212) and completed the MD program (82%; 174/212) followed by MS program (15%; 32/212); 17% (35/212) of responders were from the specialty of preventive and social medicine and 12% (25/212) from pharmacology. The median year of post-graduation degree was 2009 (range 1982-2014). The overall publication rate was 36% (76/212); the adjusted publication rate with matched citations was 25% (52/212). Pharmacology graduates reported significantly higher publication rates than other specialties (P value < 0.001). The most common reasons cited for non-publication were time constraints (46%) followed by lack of awareness about the importance of publication (32%) and lack of qualified mentors (18%).

Conclusions:
Publication bias exists in dissertation studies associated with post-graduate medical education in India. Despite a mandatory publication policy by Medical Council of India a significant number of dissertations remain unpublished. There has been a positive publication trend (Figure) in recent years.