Review of statistical methods reported in oral presentations during seven Greek national anesthesiology conf erences

The application of statistical methods in order to extract safer conclusions from samples of medical data has become a key methodology for synthesis and evaluation in any medical research. This study makes a retrospective overview of statistical methods used for oral presentations in the summaries of Greek anesthesia conferences and tries to “capture” the change in the use of statistics in recent years. Nine hundred and twenty five oral presentations from seven Greek anesthesia conferences were included for further analysis. The results, recorded an increase of randomized studies, the majority of which, had the character of single-blind study, using relatively small samples (n=31-51). Nevertheless, there is a trend towards studies with larger samples. References to the software used for processing of data also increase over the years. The majority of statistic analyses are done by using descriptive statistical methods.Tests for normality of the data are presented in the last conferences, but there is no reference to the power of any study. Extrapolations are mainly based on p value, although, over the years, the use of confidence intervals is appearing also as an alternative. Finally, more and more and different methodologies and statistical analysis tests are being selected. Although these findings refer to only 25% of all abstracts of the 28 conferences carried out until the beginning of this study (2009), they give us a first insight look on the use of statistics among Greek anesthesiologists. The last seems to ameliorate as the e-ducation in that field getting better with time.