Comparisons of the accuracies of two methods for assessing trunk posture associated with low back disorders.
Lu-M-L; Waters-T; Krieg-E Jr.; Werren-D
Proceedings of the 17th World Congress on Ergonomics (IEA2009), Beijing, China, August 9-14, 2009. Madison, WI: International Ergonomics Association, 2009 Aug; :1-8
The purpose of the present study was to determine the accuracies of two methods (human posture simulation and subjective posture rating) for assessing trunk posture for use in studies of prevention of low back disorders due to manual lifting. Subjective rating methods for analyzing a large number of lifts are typically time-consuming and prone to errors. The human posture simulation, developed by researchers at National Institute for Occupational Safety and Health (NIOSH) may expedite the process of the assessment. The simulation method involved acquisition of body postural data with an electro-magnetic motion capture system (Ascension Flock of Birds MotionStar with Motion Monitor software, Innovative Sport Training Inc., Chicago, 2003). The system was capable of measuring the 15 body joint angles defined by the University of Michigan's Three Dimensional Static Strength Prediction Program (3DSSPP) and the geometric parameters used for calculating the Lifting Index (LI) of the NIOSH Revised Lifting Equation (RNLE). Among the variables measured by the system, trunk flexion defined by 3DSSPP and lift asymmetry defined by RNLE were used for comparing the accuracies of the two methods. Eight subjects (5 males and 3 females) participated in this study. Six common lifting postures of different degrees of trunk flexion (0 degrees, 30 degrees, >60 degrees), and asymmetry (0 degrees, 45 degrees) posed by the subjects were measured with the motion capture system and used as the gold standard. The accuracy, precision and reliability of the simulation method were evaluated by comparing the results of the gold standard to simulating the photographs of the posed postures performed by the same subjects. The absolute difference between the simulation data and the gold standard was used as the error. The photographs were also sent to eight professional ergonomists for rating the postures. The results of the subjective posture rating were compared to the gold standard. The average errors in simulating and rating trunk flexion were found to be 6 degrees and 13 degrees (significantly different, p<0.05), respectively. The average errors in simulating and rating lift asymmetry were approximately 14 degrees and 16 degrees, respectively. The average precision measures of the rating method for assessing both trunk flexion and asymmetry were slightly better than the simulation method, but the average precision measures of both methods were reasonably small (<6 degrees). The average intraclass correlation coefficients of the simulation and rating methods for assessing trunk flexion were 0.82 and 0.65, respectively, indicating that the simulation method offered a better within-subject reliability for determining trunk flexion than the rating method. The findings suggest that the human postural simulation method is more accurate and reliable than the observation method for assessing general trunk posture associated with the development of low back disorders.
Ergonomics; Workers; Worker-health; Occupational-health; Occupational-exposure; Manual-lifting; Manual-materials-handling; Computer-models; Simulation-methods; Humans; Men; Women; Musculoskeletal-system-disorders; Back-injuries
Proceedings of the 17th World Congress on Ergonomics (IEA2009), Beijing, China, August 9-14, 2009