Ergonomic observational analysis tools commonly use categories to describe work-related postural position, force, and repetition. Their lack of quantitative measurements have led many ergonomists to consider Observational Tools to be less exact and less valuable than Biomechanical Analysis Methods. At the same time, it is difficult to obtain field measurements required of Biomechanical Tools, especially when a job is characterized by multiple, non-routine tasks.
A recent study by New York and Massachusetts researchers applied a Monte-Carlo method to convert category responses of the PATH (posture, activities, tools, handling) observational method to quantitative values. When Monte-Carlo simulated values were compared to outcomes from direct measurements, degrees of trunk flexion and L5/S1 disc compression force were highly correlated. However, a poor relationship was seen between Monte-Carlo simulated values and direct measurements for trunk lateral bending and shoulder flexion.
With appropriate reserve, a Monte-Carlo method can be applied to observational tool categories to provide quantitative values which can be used in Biomechanical Analysis Methods without compromising the accuracy of calculated lumbar compressive forces.
Study Design
Data Collection
PATH (posture, activities, tools, handling) is an analysis method developed by Buchholz et al. (1996) which categorizes an exposure (i.e., body posture – Trunk neutral, mild flexion, severe flexion, lateral bending, or twisted).
Information was used from a prior study (Paquet et al., 2001) whereby a PATH analysis was applied at fixed 30 second intervals to five male subjects who performed three of six construction tasks in an outdoor simulated construction site. The PATH analysis produced 463 observations that categorized body posture, load handling and activities. At the same time, the volunteers were wearing electrogoniometers and electroinclinometers which recorded arm and trunk position directly. Video recording of tasks was also performed.
The weight of items used to complete tasks was recorded.
In the current study, a Monte-Carlo simulation (a class of computational algorithms that rely on repeated random sampling to compute their results) was applied to each joint (trunk, shoulder, legs). For a specific categorical body position, an angle range was assigned (i.e., mild trunk flexion – 20 to 45 degrees for a mean of 32.5 degrees). Also, the frequency of an observed body posture was determined (i.e., from 463 observations, mild trunk flexion was recorded 66 times or 15.6%).
The body postures and the load in each of the 463 observations were entered into the University of Michigan 3-Dimensional Static Strength Prediction Program to estimate the compressive force on the lower lumbar spine.
For the same 463 observed body postion(s), data generated from the direct measurements (electrogoniometers and electroinclinometers) was entered into the Michigan software to calculate estimated lumbar spine compressive forces.
Data contamination reduced the number of comparable observed body positions to 422 pairs.
Key Outcome of Interest
The comparison of lumbar spine compressive force arithmetic mean values estimated by the Michigan 3-D software derived from:
- The Monte-Carlo generated data inputs from the 422 observational positions
versus
- The data inputs from the direct measurements of the 422 observational positions
Other Key Findings
1) Deviation from a neutral trunk posture occurred in 40 percent of all observations
2) A weight greater than 5 kg was held by the hands in 22 percent of all observations
3) The median lumbar spine compressive force at L4/L5 for the PATH and direct measurement calculations were 1673 N and 1564 N, respectively
4) The median lumbar spine compressive force at L5/S1 for the PATH and direct measurement calculations were 1489 N and 1423 N, respectively.
The Bottom Line – How This Applies To Ergonomists
Observational tools such as PATH can have their categorical determinations quantified by using Monte-Carlo methods. The resultant values can be applied to Biomechanical tools with relative confidence. This will allow for quantitative analysis of jobs characterized by multiple, non-routine tasks (i.e., carpentry work) that provide challenges to extracting direct measurements.
Further studies evaluating the application of Monte-Carlo methods to other observational tools would be helpful.
Article Title: Estimation of compressive forces on lumbar spine from categorical posture data
Publication: Ergonomics, 50, 2082-2094, 2007
Authors: S. Tak, L. Punnett, V. Paquet, S. Woskie, and B Buchholz
This article originally appeared in The Ergonomics Report™ on 2008-07-02.