Simplifying Outcome Measurement

Abstract

A comparative evaluation of outcome instruments and global assessment was performed. To test patient global assessment as a substitute for the use of more comprehensive outcome instruments in treatment trials of chronic low back pain. Treatment outcome can be measured with pain scales and functional instruments. In the absence of a gold standard, the patient him- or herself is the basic reference for outcome, for which the instruments give a more or less exact measurement. Global assessment, which is a retrospective recording, may overestimate improvement as a result of recall or motivational bias. In this study, 294 patients treated for chronic low back pain were evaluated with a visual analog scale for back pain, the Oswestry Disability Index, the Million Score and general function score for disease-specific disability, and the Zung Depression Scale for depressive symptoms. The correlation between the pretreatment and posttreatment scores for the outcome instruments (Δ scores) and the global assessment scores was calculated; effect sizes were compared; sensitivity and specificity with receiver operating characteristics (ROC) curves were estimated; and associations of global assessment with pretreatment and posttreatment scores were determined. All the Δ scores showed significant correlations with patient global assessment and with each other. The effect size of global assessment tended to be greater than that of the outcome instruments. The specificity and sensitivity of the disability instruments and pain scale were approximately 75%, whereas they were lower for depression. The associations between global assessment and outcome instrument scores did not produce evidence that global assessment was biased. Patient global assessment is a valid and responsive descriptor of overall effect in randomized controlled trials of treatment for chronic low back pain.