Reliability and validity of the PAD questionnaire: a measure to assess pain-related decline in physical activity.

Abstract

To examine the reliability and validity of the physical activity decline (PAD) score: a measure for assessing a decline in the level of physical activity in patients with chronic pain. This study was embedded in a prognostic cohort study based on an inception cohort of patients with sub-acute low back pain. Sixty-two patients who developed chronic pain participated in this study. Internal consistency was expressed by Cronbach's alpha and the test-retest reliability was based on an intraclass coefficient (ICC) score. Construct validity was determined using a Pearson correlation coefficient with disability (Quebec Back Pain Disability Scale), a change in physical activity level (DeltaBPAQ) as external criteria for convergent validity. The level of physical activity (Physical Activity Rating Scale) was used as external criterion for discriminant validity. The internal consistency (Cronbach's alpha = 0.92) and reliability (ICC = 0.93) of PAD were shown to be good. The construct-validity of the PAD questionnaire appeared to be adequate, with Pearson coefficients of r = 0.45 (p < 0.01; a change in BPAQ), r = 0.55 (p < 0.01; disability) and r = 0.03 (p = 0.74; physical activity). Based on the fact that 38.7% of the patients had the lowest score of 0, the presence of a floor-effect in the PAD score must be considered. The reliability and validity of the PAD questionnaire in its original Dutch version appears to be good. Further research is warranted regarding the presence of a floor-effect.