A Brief Overview of Base Rates

Abstract

The term “base rate” refers to the prevalence of an event, such as a symptom, sign or disorder, within a given population. For example, the base rate of dementia in the general population of individuals over age 85 is approximately 20% (American Psychiatric Association, 1994). The base rate of self-reported memory problems at seven years post-injury in a sample of hospital patients admitted following closed head injury was 38% (Oddy, Coughlin, Tyerman, & Jenkins, 1985). The importance of base rates in the clinical sciences is fundamental. Variations and covariations in symptom base rates among various populations (including the lack of symptoms, or a base rate of 0%) inform the definitions and the diagnosis of diseases and disorders. This aspect of the application of base rate information in clinical decision-making is commonly recognized. Even if a clinician is not thinking “base rates” per se, he or she is aware that certain symptoms are more likely to be observed in certain disorders, whereas other symptoms typically are not associated with the disorder. There are, however, additional ways in which base rates can and should inform clinical decision-making that are less generally recognized or applied. These applications of base rate information are related to the interpretation of the significance of diagnostic test findings, such as the differentiation between findings of statistical versus clinical significance and the determination of the predictive value of diagnostic tests. This chapter will provide a brief overview of some of these considerations, but will begin with a brief discussion of symptom base rates, classification of diseases and disorders, and the process of clinical decision-making.