Making Sense of Statistics for Family Practitioners: "Understanding the score"

Abstract

A plethora of scoring systems have evolved in medicine Ranks or scores are a common feature of quantitative research conducted by family practitioners. In addition ranking is prevalent in the laboratory, e.g. malaria parasitaemia ranked as 0, +, ++, or +++, and at the patient bedside, that is, grading the severity of stroke or leprosy lesions as mild, moderate or severe (+, ++, +++) or of a cancer as Stage l, ll, lll or lV.A convenient way of combining the ranks of a group of related criteria is the creation of an "index". Basic indices result from simply adding the individual ranks of all components, while more sophisticated indices provide individual weightings to each component before they are summed. lt is useful to create indices but they should be used with circumspection.We will use the Apgar scoring system, which is used for the evaluation of newborn babies, to highlight features commonly shared by indices that should always be explicitly considered before use.