Professional Documents
Culture Documents
BUKU QEC A Practical Method For The Assessment of Work-Rela
BUKU QEC A Practical Method For The Assessment of Work-Rela
net/publication/267241521
Article in Proceedings of the Human Factors and Ergonomics Society Annual Meeting · October 1998
DOI: 10.1177/154193129804201905
CITATIONS READS
90 2,772
2 authors, including:
Peter W Buckle
Imperial College London
176 PUBLICATIONS 8,106 CITATIONS
SEE PROFILE
All content following this page was uploaded by Peter W Buckle on 20 November 2014.
A newer exposure tool has been developed for health and safety practitioners to assessthe exposure to risks for
work-related musculoskeletal disorders. The tool is based on the practitioners’ needs for such a tool and “state
of the art” research findings. QEC has been tested, modified and validated based upon various simulated and
practical tasks, with the help of up to 150 practitioners. The studies have shown that the tool has a high level of
sensitivity and usability, and exhibits largely acceptable inter/intra-observer reliability. Field studies also
indicate that the tool is, in practice, reliable and applicable for a wide range of tasks. With a short training
period and some practice, assessment can normally be completed within 10 minutes for each task.
indicated that such duration was adequate for most observers to and frequency of movement). Two types of manual handling
complete the assessment,The order of the recorded tasks was tasks and two types of repetitive manual assembly tasks were
arranged such that the task characteristics were different performed (controlling weight/heights/work distance etc.). The
between adjacent tasks. frequency of arm or back movement was controlled using
24 practitioners participated in the test and they all came audible signals presented only to the subject. The tasks were
from health and safety occupations. Their mean age was 41.2 randomly performed by one male subject (age: 34 years,
years (SD=9.16, Range=26-59). The average time they had stature: 1.72 m), during which the tasks were assessedby 18
worked in occupational health and safety was 8.8 years practitioners (5 males and 13 females, mean age=41.3 years
(SD=7.33, Range=2-33), and the average experience with risk (SD=9.31, Range=27-58), average time worked in health and
assessmentwas 3.0 years (SD=2.59, Range=O-9). safety=5.2 years (SD=3.88, Range=l-18)). The observers were
The practitioners (observers) were divided into five divided into five groups with 3 or 4 people in each group.
groups, and each group assessedthe tasks independently. Validity tests were also conducted in field studies by
Before watching the videos, the observers spent 5-10 minutes comparing 6 practitioners’ assessment on a wide range of
going through the “Guide to the use of the exposure tool”, practical tasks (60 in total) with more detailed video analysis of
during which they could discuss issues about the use of the tool. these tasks performed by an analyst.
Then the tape was played and the practitioners made their
assessmentson each task. If anyone could not complete the Results
assessmentwithin the time given, the tape was re-played to For the test of inter-observer reliability, both Cohen’s
make sure all observers could finish their assessment. kappa and percentage agreement were calculated using the data
Intra-observer reliability of 24 observers and of those who had l-7 and 4-7 year’s
A test-retest was conducted with 8 observers assessing the experience in making risk assessment (Table 1). The
same set of 18 recorded tasks twice in 3-week interval. Their individual’s test-retest results are given in Table 2.
average age was 41.0 years (SD=12.81, Range=26-59), average The need to re-wind the tape due to unfinished assessment
experience in health and safety work was 7.63 years (SD=6.30, was experienced only with the first and second tasks. During
Range=2-20), and average experience in risk assessmentwas this early stage, some observers were hesitant with the
2.5 years (SD=3.25, Range=O-7). assessment but after observing up to two tasks, they became
Validitv test familiar with the method and could complete the assessments
The measurement validity was tested by comparing the within the recording time. It was thus suggested that the results
practitioners’ assessmenton simulated tasks with computer- from the first two tasks could be biased and they were therefore
aided 3D motion analysis using the SIMI system. The tasks excluded from further analysis. Table 3 shows some
were designed so that they could be performed at “known” preliminary results for the validity test of the tool in both
levels of physical exposure to certain parts of the body (posture laboratory and practical work environment.
PROCEEDINGS of the HUMAN FACTORS AND ERGONOMICS SOCIETY 42nd ANNUAL MEETING-1998 1353
Discussion
Inter-observer reliabilitv items. It is anticipated that an improved training process may
The agreement between 24 observers on most assessment increase the assessment reliability further.
items which were tested indicated a ‘fair agreement’ as Assessment validity
evaluated by kappa analysis (Landis and Koch, 1977)(Table 1). There is no ‘correct’ answer as to what the actual
The agreement of ‘experienced’ practitioners showed some exposure is for a practical task. For tasks with ‘known’ body
improvement, particularly for back posture, shoulder/arm and postures and movement (either controlled or measured by a
wrist/hand movement, suggesting a ‘moderate agreement’. sophisticated method), most assessment items were ‘correctly’
Agreement on back movement was ‘slight to fair’, suggesting measured at an acceptable accuracy level (Table 3). It was
that further improvements may be needed for this assessment found that, in the field studies, one practitioner regarded back
item. Inter-observer agreement on neck posture was not high flexion of less than 90” as ‘moderately flexed’ rather than
(K=0.25). Similar problems have been encountered by other ‘extremely flexed’, resulting in a high level of disagreement
researchers (Kilbom, 1994). For wrist/hand posture, kappa with the detailed video analysis. Further analysis with the
analysis was found to be unsuitable due to low data variation results of five observers showed that the agreement on back
between categories. posture was 64.5%. Another reason for this ‘low level’
With percentage agreement, most assessment items were agreement was possibly that the posture was assessedas the
either close to or above 70%. According to Baty et al, (1986), ‘worst event’ which may happen in a short time, during which
inter-observer agreement of 75% can be regarded as the observer may not be looking at the worker and therefore
acceptable. Present studies suggest that QEC has the potential missed the recording.
to meet the basic requirement of its inter-observer reliability for The ‘score system’ is, at this time, largely hypothetical,
most of its assessmentitems. It should be emphasised that which considers the interaction/combination of risk factors.
assessing recorded tasks can be different from assessing real Epidemiologic evidence is still not sufficient to support such a
ones, especially with the tasks in the present studies - the pattern, but some evidence can be found in the literature. It is
recordings’ quality varied among tasks (eg. lighting), and they difficult to determine whether this system is ‘true’ or ‘correct’,
were recorded from either a fixed camera position or from a far and equally difficult to determine whether it is ‘untrue’ or
distance. Better agreement is expected if the assessment is ‘incorrect’ (it is perhaps a significant achievement if one can do
conducted using real tasks or better video recordings. so in either way). At this stage the system serves as a
Intra-observer reliabilitv compromise between the ‘known’ and ‘unknown’ concerning
Table 2 shows that intra-observer reliabilities for almost the ‘weighting’ and ‘interaction’ of ‘risk factors’, and should
all assessmentitems reached ‘moderate agreement’ level, and only be taken as a reference.
the test-retest agreements were all statistically significant. The
kappa results also suggested that people with/without previous Conclusion
experience in risk assessment are able to reach an agreement at QEC is a new method which has been developed for
a similar level. This is encouraging because the tool is aimed at practitioners to assessexposure to the risks of WMSDs. Based
non-skilled users and is intended to be used by a person who on test results obtained so far, this tool is found to be sensitive
assessesa job before and after an ergonomic intervention. The for assessing the change in exposure before and after an
‘experienced’ observers did not achieve a higher level of ergonomic intervention, and for exposure comparison either
agreement for all items as compared to the whole group. One between two or more workers performing the same task, or
explanation is that time-based experience (years in job) may not between people performing different tasks. The tool is also
necessarily represent the ability of using the tool. People with shown to be largely reliable and applicable to a wide range of
the same experience level may not have the same skill level in jobs. With brief training (self-learning) and some practice,
making such an assessment,and they may be better at judging assessmentcan normally be completed within 10 minutes for
some types of exposure, but may not be good at assessing other each task. However, studies also suggested that improvements
1354 PROCEEDINGS of the HUMAN FACTORS AND ERGONOMICS SOCIETY 42nd ANNUAL MEETING-1998
are needed, particularly for the assessment of frequent body Landis, J.R. and Koch, G.G., 1977, The measurement of observer agreement
for categorical data. Biometrics, 33, 159-174.
movements. It is anticipated that measurement reliability will
Li, G. and Buckle, P., 1997, The development of a practical tool for
improve with regular use of the tool and further experience of musculoskeletal risk assessment. In: Contemporary Ergonomics 1997, (ed.
making risk assessments. S.A. Robertson), London: Taylor & Francis, 442-447.
Acknowledgement NIOSH, 1997, Musculoskeletal Disorders and Workplace Factors: A critical
This research was funded by the HSE. Review of Epidemiologic Evidence for Work-Related Musculoskeletal
Disorders of the Neck, Upper Extremity, and Low Back. (ed. B. Bernard)
References
Buckle, P. and Li, G., 1997, A practical approach to musculoskeletal risk
assessment in the real workplace. In: From Experience to Innovation. Vol. 4,
(eds. P. Seppbla, T. Luopaj;irvi et al.), 138-140.
Appendix 1: Quick Exposure Checkfor work-related musculoskeletal risks - QEC
Job tide: Task: Assessment conducted by: Worker’s name: Date: Time:
Pa-rt A: Observer’s assessment
j;; ke;.$p)
d3: High
l Is the visual demand of this task -
el: Low? ( There is almost no need to view fine details )
e2: High? ( There is a need to view some fine details )
l Do you have difficulty keeping up with this work?
Never
f-ii Sometimes
f3: Often
l HOW sty&f 2 you find this work?
Low
;:I Medium
g3:
g4: High
PROCEEDINGS of the HUMAN FACTORS AND ERGONOMlCS SOCIETY 42nd ANNUAL MEETING-1998 1355
b3 6 8 10 I ”L 00 rn II
1” ”c; I ”I)
7
Exposure to the Shoulder/arm
Cl c2 c3 Score 1 Dl D2 D3 i &ore2 bl b2 b3 swre 3
al 2 4 6 4 6 2 4 6
a2 4 6 8 6 8 4 6 8
a3 6 8 10 8 10 6 8 10
a4 8 10 12 10 12 8 10 12
-
Tot&sc,mre Rw shoulder/arm
-
6 2 4 6
4 6 8 4 6 8
8 I/l
1”
r ” IA
b3 6
_.
Exposure to the Wrist/hand
‘*
J2 4 6 4 6 8
1” 6 8 ‘,_ 6 8 10
Score4 Total sMImfar the w&t/hand
>
,\,,sy& 5_,,
,:
:
1 =Sumdscofes1t05
2 4 6 2 4 “’ \-
4 6 8 4 6
8 10 6 8 ‘_
b3 6
I I I I
Exposure to the Neck
I I I I I Tcitsl scare for the neck
el e2 ,2&@2. .1--‘
^ i. / =swms1+2
2 4 ,.”
JCL
6
b3 6 8 10 6 8