Low Back Injury Risk Assessment Tools - Fundamentals and Assessment Tools for Occupational Ergonomics - page 687

the Lumbar Motion Monitor (LMM) (Figure 35.1) that was developed to capture the instantaneous pos-
ition, velocity, and acceleration of the lumbar spine in the three cardinal planes of human motion. Their
approach to developing an assessment tool was to use this tri-axial goniometric device to capture the
trunk kinematic profiles of workers performing their normal work tasks and then to relate these kin-
ematic characteristics (along with a cadre of other task descriptors such as lifting frequency, moments
about the spine created by the load, job satisfaction, the static workplace variables from the NIOSH
Lifting Equation, etc.) to the historical incidence of low back injuries. They sampled 403 industrial
jobs and then used multiple logistic regression techniques to form a relationship between historical
injury data and the task parameters. Their results showed that five parameters were adequate to dis-
tinguish between the high- and low-risk jobs in this data set: lift rate, maximum sagittal angle,
average twisting velocity, maximum lateral velocity, and maximum moment (model odds ratio of
10.7). The result of their work is a low back injury risk assessment model that takes as inputs these
five task variables and the output is a single value that describes the probability of high-risk group
membership (PHRGM) for that job.
The principal strength of this model is that it is based on the empirical relationship between outcome
measures (injury and job turnover rates) with quantifiable job characteristics, including human perform-
ance-related variables. With this approach comes the ability to begin to consider the role that individual
differences (i.e., lifting and MMH techniques) may play in the etiology of low back injury. While this
model was able to overcome the static biomechanical modeling limitations of the NIOSH Lifting
Guides, a limitation to the generalizabilty of this model is that it was developed using data collected
from a sample of jobs where workers performed “repetitive jobs without job rotation.” Since this was
an empirical model, the specific job dataset that was used to develop the relationship between work
characteristics and risk will have a great influence on the model output. Since nonrepetitive jobs were
not included in the dataset, certain characteristics of these types of jobs may not be represented in
this model's predictions. Further, because of the special emphasis placed on the variables describing
trunk dynamics that resulted from this sample of jobs, some high-risk activities, such as lifting heavy
loads in awkward, static postures will often escape identification.
FIGURE 35.1 The Lumbar Motion Monitor.