The measurement of symptoms and side effects in clinical trials of chronic pain

Clinical trials in chronic pain have used various methods to assess the effect of medications. In addition to measuring symptom relief, researchers must also address adverse events (AEs) associated with the medication to evaluate overall therapeutic results. This paper reviews methods of measuring symptoms and AEs, including passive capture, scripted prompting, prospective assessments of side effects of interest, and prospective comprehensive symptom checklists. Methods of measuring therapeutic results have advantages and disadvantages. Although passive AE capture (unscripted, open-ended questions about symptoms) is not constrained by preconceptions of potential AEs, it sometimes fails to capture clinically significant AEs. Scripted prompting (prespecified, scripted, open-ended questions about symptoms) is likely to address the latter problem but preclude consistency across trials. Prospective assessments (prespecified symptom inventories) can offer greater sensitivity and consistency in detecting side effects of specific treatments. Comprehensive symptom distress inventories can be more sensitive measures of overall treatment benefit, which may be important when efficacy, passive AE capture, and comprehensive quality of life (QOL) batteries fail to differentiate between treatments. In cancer populations these inventories have been statistically correlated with survival and other important clinical outcomes, even after controlling for disease status, global QOL, psychological state, and performance status. Other important considerations are patient perceptions of the importance of symptoms, the correlations between QOL and symptom distress, and the usefulness of global ratings, in which patients integrate any perceived benefits of the medication with tolerability and other factors, such as convenience and cost.