Overview of statistical methods usage in Indian anaesthesia publications

Background and Aims: Despite the importance of statistics being well established for medical research, it remains a neglected area of understanding and learning. The present survey aimed to examine the use of various statistical methods in a two-year sample (2019–2020) of representative Indian anaesthesia journals and compare it with an international top-ranked journal. Methods: The literature survey included analysis of 748 original articles from 'Indian Journal of Anaesthesia' (179), 'Journal of Anaesthesiology Clinical Pharmacology' (125) and 'Anesthesia & Analgesia' (444) published over the period. Original research articles were identified from the table of contents of each issue. Articles were assessed for statistical methods, categorised as being descriptive, elementary, multivariable, advanced multivariate or diagnostic/classification. Results: Compared to Anesthesia & Analgesia, the Indian journals (considered together) had a significantly greater use of mean (standard deviation) (91.2% versus 70%) and percentages (79.5% versus 67.6%) (P = 0.000 each); and lesser for Wilcoxon (5.4% versus 14.6%) and Pearson/Spearman (5.1% versus 13.5%) correlation tests (P = 0.000 each), multivariable tests including various regression methods (P < 0.001), classification/diagnostic tests [Receiver operating characteristic (ROC) curve analysis, P = 0.022; sensitivity/specificity, P = 0.000; precision, P = 0.006; and relative risk/risk ratio, P = 0.010] and a virtual absence of complex multivariate tests. Conclusion: The findings show limited use of advanced complex statistical methods in Indian anaesthesia journals, usually being restricted to descriptive or elementary. There was a strong bias towards using randomised controlled designs. The findings suggest an urgent and focussed need on training in research methodology, including statistical methods, during postgraduation and continued medical training.