What does it mean?: Anchoring psychosocial quality of life scale score changes with reference to concurrent changes in reported symptom distress

As customarily presented the results of quality of life (QOL) studies tend more to puzzle than to enlighten their intended audience of physicians, policy makers, and the general public. They ask quite legitimately what a difference of three "points" on the Rand Mental Health Index means. Or, a fortiori, a difference of 0.17 "standard deviations" or "responsiveness units." How large need a difference be in order to justify action or decision? At the level of the individual question, the numbers have plain, intuitive content. But once they are combined into statistical abstractions, interpretation becomes much more difficult. It is intrinsically easier to understand findings couched in symptom-distress terms than in psychosocial scales. People know the difference between hurting a little and hurting a lot. We have aimed here to transfer via empirical analysis, some of this intuitive advantage into the psychosocial realm. In two parallel, multi-center clinical studies comparing the effects of drug treatments in hypertension and in angina, respectively, 1003 patients completed, up to five times at intervals of several weeks, a battery of psychosocial quality of life (PQOL) measures and concurrently an extensive symptom distress check list. The resulting databases yielded, for each combination of a PQOL subscale and a symptom distress (SD) index, over 200000 pairs of change scores, each associating some manifestation of PQOL change, between one visit and the next, with a concurrent SD change. We have documented a clear and surprisingly consistent association between changes in the level of reported symptom distress and concurrent psychosocial changes as measured by the Rand Mental Health Index. With further refinement and consensus-building, QOL researchers should eventually be able to explain psychosocial findings quantitatively with reference to the difference between "no distress" and "some distress," or that between "some distress" and "very much distress."