Is This A Good Medical Test?

This project uses integrated history and philosophy for medicine to inform medical practice today. The evaluation of medical tests is today in a state of confusion. Traditionally, the main index used to evaluate diagnostic tests is diagnostic accuracy.

Typically, this is done by evaluating the sensitivity (the proportion of patients with a disease that test positive) and the specificity (the proportion of patients without a disease who test negative) of a test. Sensitivity and specificity are commonly assumed to be constants, or to vary only in a limited rage of circumstances.

In contrast to this, others claim that sensitivity and specificity are highly variable, so much so that the ‘diagnostic accuracy paradigm’ should be abandoned.

Instead of focusing on diagnostic accuracy, some recommend that test evaluation should focus on patient outcomes, whilst others strongly disagree.

This confusion is detrimental to medical practice, as it leaves medicine in a state of not knowing how to tell if a medical test is a good one. This project seeks to understand and address this confusion using historical and philosophical tools.

Philosophical analysis reveals that the differing attitudes to test evaluation have at their root differing philosophical assumptions about how homogeneous or heterogeneous patients with the same disease are.

The project provides and intellectual and social history that traces the development of test evaluation over the twentieth century.

It follows how successive generations of researchers have balanced assumptions of homogeneity and heterogeneity and modified them for use in their particular setting.

The central claim of this project is that understanding medical history is key to balancing assumptions of homogeneity and heterogeneity skillfully.