Trends in the quality of work presented at the society of british neurological surgeons meetings: 1975 to 2010.

Background: The quality of scientific publications in clinical journals is well studied but the quality of work presented at medical conferences less so. Aims: To describe trends in the quality of presentations at the Society of British Neurological Surgeons [SBNS] conference between 1975 and 2010 and the factors associated with higher quality work in order to consider what might improve publication rates. Methods: Analysis was conducted in 5-year time periods (i.e. 1975-1979, 1985-1989, 1995-1999, 2005-2009). Published abstracts were used to identify conference presentations. Quality metrics included level of evidence of the presentation and eventual publication within 5 years. Publication 5-year citation count and destination journal impact factor were further used to assess publication quality. Statistical analysis was carried out using SPSS. Results: Of the 1711 presentations in total, 479 (28%) were published. The British Journal of Neurosurgery (93, 19%) was the favoured destination. Although the total number of publications has increased, given the increase in the number of presentations, the proportion of work published has decreased (80/179; 45% in the 1970s to 113/721; 16% in the 2000s). The growth in the impact factor of published work was better than that found in leading neurosurgical journals, but lower than for leading medical journals. In a multivariate model, presentations using a higher level of evidence increased the likelihood of publication (AOR 6.7 95% CI 3.7, 12.1), whilst presenting at conferences after the 1970s reduced the likelihood of publication; 1985-1989 (AOR 0.3, 95% CI 0.2, 0.4), 1995-1999 (0.4, 95% CI 0.3, 0.7) and 2005-2009 (0.1, 95% CI 0.1, 0.2). Conclusion: SBNS conferences today contain more presentations and yield more publications than ever before. However, the increased volume may dilute the quality of work presented.