Incident Analysis and the Scientific Method

Abstract

It is unclear why the scientific method has been systematically ignored in incident – analysis. As example the case of a fatal hemorrhage in a hospital during an observational period of 1 year was used. In case of a fatal hemorrhage the physician in charge of the analysis will first make an inventory of how many fatal hemorrhages of the same kind have occurred in the period of 1 year. The number seems to be no less than ten. The question is, is this number larger than could happen by chance. This chapter assesses the use of z-tests and chi-square tests for answering the question of chance findings. The scientific method is often defined as an evaluation of clinical data based on appropriate statistical tests, rather than a description of the cases and their summaries. The above example explains that the scientific method can be helpful in giving a clue to which incidents are based on randomness and which are not so. The scientific method is in a nutshell: reformulate your question into a hypothesis and try and test this hypothesis against control observations.