Secondary Logo

Journal Logo

QUANTIFYING LOW BACK INJURY RISK ASSOCIATED WITH MANUAL LIFTING JOBS IN DISTRIBUTION CENTERS: GP54.

Lavender, Steven A.1; Marras, William S.2; Ferguson, Sue A.2; Splittstoesser, Riley E.2; Yang, Gang2

Spine Journal Meeting Abstracts: October 2011 - Volume - Issue - [no page #]
GENERAL POSTERS
Free

1The Ohio State University, Integrated Systems Eng. & Orthopaedics, Columbus, OH, US; 2The Ohio State University, Biodynamics Laboratory, Columbus, OH, US

INTRODUCTION: Quantifying low back disorder (LBD) risk associated with manual lifting tasks in distribution centers has been difficult in large part due the varying weights of the lifted items, the varying locations relative to the body, and the variable work rates. The objective of the current study was to develop LBD risk models applicable to distribution centers based upon biomechanically relevant exposure metrics.

METHODS: Biomechanical exposures were measured on 195 workers working 50 jobs in 21 distribution centers using a wearable, sonic‐based, load moment and movement data capture system. In most of the jobs, 4 to 5 people were sampled for approximately four hours. These data were pooled to create a sample for each job which was the unit of analysis. Low back injury rates from each job were collected retrospectively from the company's OSHA 300 records over the prior 3‐year period. Jobs with no LBD incidence over the prior three years were considered "low" risk (n=15). Jobs with a LBD incidence rate greater than 12 per 200,000 hours were considered "high" risk (n=15). Jobs with I200k values between these low and high risk criteria were removed.

RESULTS AND DISCUSSION: The multivariate logistic regression identified an injury risk model that included: (1) The average of the job's peak dynamic forward load (box) moments; (2) the average of the maximum dynamic slide forces; and (3) the average duration of the non‐load exposure period (seconds) during work periods (does not include breaks). This model had sensitivity and specificity values of 87 and 73 percent, respectively. It should be recognized that many variables were significantly related to injury risk and that different models could be developed using other biomechanical exposure variables. We selected this particular model based on its sensitivity, specificity and explanatory power.

© 2011 Lippincott Williams & Wilkins, Inc.