Civil Engineering Reference
In-Depth Information
group membership. This combination of workplace and trunk motion factors formed the basis of the
LBD risk model. The model was selected for its statistical importance of the predictors and for biome-
chanical plausibility. The model variables remained consistent when tested with the various intermediate
data sets. The empirical stability of the model was checked by predicting the classification of 100 jobs,
based on the preliminary model. This model resulted in an odds ratio of 10.6.
By averaging individual probability values for moment, lift rate, sagittal flexion, twisting velocity, and
lateral velocity, the LBD risk model is able to predict the probability of high-risk group membership
(LBD risk) for any repetitive job. A chart depicting this information is shown in Figure 49.3. It is import-
ant to understand that the predictive power of this model is a result of the interaction of these five vari-
ables. Individually, each of these five factors is unable to reliably distinguish between the injury rate
groups, but when they are considered in combination, the predictive power increased tenfold.
49.3.5 Validation and Predictive Ability
The iLMM's LBD risk model is one of a few tools for assessing injury risk that have been tested for their
predictive abilities. It was validated in a prospective study (Marras et al., 2000) to determine how well it
reflects a job's LBD risk. A total of 36 jobs from 16 different companies were tracked in a prospective
cohort study, in which 142 employees performing these jobs were studied both before and after work-
place interventions were made. Assessments of LBD risk were made using the iLMM and associated
model, and the jobs' LBD incidence rates also were computed pre- and postintervention. The results
from this study indicated that a statistically significant correlation existed between changes in the
jobs' estimated LBD risk assessment values (using the iLMM) and changes in their actual LBD injury
rates during the observation period. This relationship is graphically depicted in Figure 49.4. As this
figure shows, jobs whose interventions produced LBD risk values still considered “high” resulted in
small changes, on average, in their actual low-back injury rates. For those jobs where the intervention
produced a “medium” level of LBD risk, the average injury rate drop was over 50%. Finally, the
FIGURE 49.3 The LBD risk model, showing the five factors scaled relative to risk. The vertical line indicates the
overall probability of high-risk group membership, or LBD risk, for a particular job.
Search WWH ::
Custom Search