Introduction: Determining the amount of changes produced in the treatment of chronic pain patients is important. Statistical tests of significance are usually used for this purpose. Although important, the use of statistical tests of significance in determining treatment improvements is associated with a number of limitations. To overcome these limitations, investigators have recommended using clinical significance to determine whether or not the treatment outcome is clinically significant.
Method: Chronic pain patients from a pain clinic (57) were selected, using a convenience sampling method. These patients completed twice, a series of questionnaires with a 4-week interval: Pain-Self-Efficacy Questionnaire, Depression Anxiety Stress Scales, Multidimensional Pain Inventory, Physical Disability Questionnaire and Short form Health Survey. Pearson correlations were calculated between the scores of the patients obtained over the 4-week interval. These test-retest coefficients were used to establish a reliable change index for the two measures of pain self-efficacy and depression of the 11 chronic pain patients treated with cognitive behaviour therapy. Statistically significant changes of the pain self-efficacy and depression were examined, using paired sample t-tests.
Results: The results of tests of statistical significance provide evidence of statistically significant improvements at post-treatment in both measures of pain self-efficacy and depression. The results of a reliable change index (as one of the measures of clinical significance) documented that 9 of the 11 patients on the pain self-efficacy variable and 8 of the 11 patients on the depression variable achieved clinically significant gains.
Conclusion: While statistical tests of significance provide little or no information regarding the variability in treatment response from person to person, tests of clinical significance go beyond this limitation and indicate who achieved clinically significant change.