Interpreting Output of Statistical Studies

This review provides an overview of the elements of statistical analysis so that the reader may better understand the output of statistical studies, which is essential to the modern practice of critical care, in a rapidly evolving field. Beginning with hypothesis testing, the review progresses through an explanation of variable types and demonstrates how to quantify and categorize variables, with examples; it then goes on to explain the principles of basic comparative analysis, which helps identify simple differences between cohorts, and then highlights the importance of potential confounders to help readers understand simple strategies used to control for confounding, such as using different types of study designs and different methods of statistical analysis. By using contemporary, influential articles from the critical care literature to illustrate these principles, we hope to illuminate the importance of interpreting output from statistical studies to better inform evidence-based practice. Key words: confounding, hypothesis testing, statistical analysis