Health care interventions are intended to increase longevity, prevent morbidity, and make patients feel better. The first 2 of these end points are relatively simple to measure. However, measuring how people are feeling is not as easy. Physiological or laboratory tests commonly are used for the direct measurement of the patients’ well-being, assuming that a strong link exists between the 2 parameters. However, when investigated empirically, the relation between physiological or clinical measures and patients’ symptoms and well being is, at best, modest and often highly variable.