Development of an audiovisual taxonomy of pain behavior: Reliability and discriminant validity.

Developed a pain behavior taxonomy derived from patients', spouses', and health care providers' nominations of pain behaviors. 71 chronic low back pain patients and 40 controls (Ss' mean age 38.25 yrs) were videotaped while undergoing a sequence of movements involving sitting, standing, walking, bending, and exercising, and a brief standardized interview component. Two observers rated each S's videotape employing a 16-category observational rating system that included verbal and nonverbal pain behaviors. Seven of the pain behavior categories, including guarding, bracing, position shifts, partial movement, grimacing, limitation statements, and sounds, had acceptable reliabilities and frequencies and were selected for inclusion in a discriminant analysis. Ss were randomly divided into 2 groups, and analysis on the 1st subsample indicated that partial movement, limitation statements, sounds, and position shifts accounted for 75% of the variance in group membership and correctly classified 94.4% of the patients and 95.2% of the controls. On cross-validation, the same 4 categories identified in the initial discriminant function correctly classified 88.9% of the Ss in the 2nd subsample. The utility of this pain behavior observational recording system is discussed in reference to both patient assessment and treatment outcome research. (10 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)