Article Text

Download PDFPDF
Statistics for evidence-based nursing
  1. Trevor Sheldon, Dsc
  1. Department of Health Studies, University of York, York, UK

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Healthcare professionals and policy makers are increasingly aware of the need for their decisions to be informed by the best available research evidence. Although the articles selected for abstraction in Evidence-Based Nursing have already been appraised, and only the highest quality research selected, practitioners still need basic skills to identify and interpret methodologically sound research. This editorial is one of a series that aims to help you to do this, and is the second of 3 editorials focusing on some basic measurement and statistical issues in healthcare research. In this editorial, we look at how the results of intervention or “treatment” studies are summarised and presented for different types of health outcomes. The next editorial will explain how statistical techniques are used to assess the probability that the observed treatment effect occurred by chance, and what is meant by a “statistically significant” result.EBN notebook

Measures of health and disease

Many measures are used to assess the health outcomes of an intervention, ranging from those trying to capture its effect on people's general health (eg, Short Form-36) to measures of a specific dimension relevant to a particular disease (eg, the Beck Depression Inventory). Some are measures of patients' perceptions of their health; more often, however, they are measures that clinicians or researchers think are important. Regardless, these measures are generally either continuous or discrete; this distinction is important because the type of measure used determines the way the results are presented and analysed.


When the outcomes of a study are continuous (eg, temperature, blood pressure, or cholesterol concentrations), the researchers are usually interested in the extent to which these values change after exposure to an intervention. Studies that use continuous outcome measures may compare the average values of the variable (eg, mean or median) after treatment. In a study of treatment for arthritis, researchers may measure …

View Full Text