Defining the clinically important difference in pain outcome measures

The purpose of this study was to determine the levels of change on standard pain scales that represent clinically important differences to patients. D ata from analgesic studies are often difficult to interpret because the cli nical importance of the results is not obvious. Differences between groups, as summarized by a change in mean values over time, can be difficult to ap ply to clinical care. Baseline scores vary widely and group mean difference s could reflect large changes in a few patients, small changes in many pati ents, or any combination of these outcomes. Determination of the proportion of patients who have a clinically important improvement in their pain woul d provide a more interpretable result with direct clinical implications. Ho wever, determining a clinically important outcome requires information abou t the degree of change over time that is clinically important. Data from th e titration phase of a multiple cross-over randomized clinical trial of ora l transmucosal fentanyl citrate (OTFC) for the treatment of cancer-related breakthrough pain were re-analyzed to examine the differences in pain score s between treatment episodes that did and did not yield adequate pain relie f. The scales evaluated were absolute pain intensity difference (PID, 0-10 scale), percentage pain intensity difference (PID%, 0-100% scale), pain rel ief (PR, 0 (none), 1 (slight), 2 (moderate), 3 (lots), 4 (complete)), sum o f the pain intensity difference (SPID over 60 min), percentage of maximum t otal pain relief (% Max TOTPAR over 60 min), and global medication performa nce (0 (poor), 1 (fair), 2 (good), 3 (very good), 4 (excellent)). Adequate relief was defined by the patient's decision not to use another dose of opi oid medication as a rescue, in addition to the study medication, to treat e ach painful episode. One hundred thirty OTFC naive patients contributed dat a on 1268 episodes of breakthrough pain. The scales that were converted to a percentage change yielded the best accuracy in predicting adequate relief , with balanced sensitivity and specificity. The best cut-off point for bot h the % Max TOTPAR and the PID% was 33%. The best cut-off points for the ab solute scales were absolute pain intensity difference of 2, pain relief of 2 (moderate), and SPID of 2. The global medication performance of 2 (good) had excellent values as well. This study presents data-derived cut-off poin ts for the changes in several pain scales, each reflecting the clinically i mportant improvement for patients treating breakthrough cancer pain episode s with OTFC. Confirmation in other patient populations and different pain s yndromes will be needed. The use of consistent clinically important cut-off points as the primary outcome in future pain therapy clinical trials will enhance their validity, comparability, and clinical applicability. (C) 2000 International Association for the Study of Pain. Published by Elsevier Sci ence B.V. All rights reserved.