Download Comparisons of ergonomic evaluation tools ALLA RULA REBA and OWAS for farm work

International Journal of Occupational Safety and
Ergonomics
ISSN: 1080-3548 (Print) 2376-9130 (Online) Journal homepage: https://www.tandfonline.com/loi/tose20
Comparisons of ergonomic evaluation tools (ALLA,
RULA, REBA and OWAS) for farm work
Yong-Ku Kong, Sung-yong Lee, Kyung-Suk Lee & Dae-Min Kim
To cite this article: Yong-Ku Kong, Sung-yong Lee, Kyung-Suk Lee & Dae-Min Kim (2018)
Comparisons of ergonomic evaluation tools (ALLA, RULA, REBA and OWAS) for farm
work, International Journal of Occupational Safety and Ergonomics, 24:2, 218-223, DOI:
10.1080/10803548.2017.1306960
To link to this article: https://doi.org/10.1080/10803548.2017.1306960
Accepted author version posted online: 17
Mar 2017.
Published online: 02 May 2017.
Submit your article to this journal
Article views: 955
View Crossmark data
Citing articles: 3 View citing articles
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=tose20
International Journal of Occupational Safety and Ergonomics (JOSE), 2018
Vol. 24, No. 2, 218–223, https://doi.org/10.1080/10803548.2017.1306960
Comparisons of ergonomic evaluation tools (ALLA, RULA, REBA and OWAS) for farm work
Yong-Ku Konga , Sung-yong Leea , Kyung-Suk Leeb and Dae-Min Kimc∗
a Department
of Industrial Engineering, Sungkyunkwan University, Republic of Korea; b Rural Resources Development Institute, NIAST,
RDA, Republic of Korea; c Division of Mechatronics Engineering, Dongseo University, Republic of Korea
Introduction. The purpose of this study was to validate the agricultural lower limb assessment (ALLA) ergonomic checklist,
which was developed for various agricultural tasks in Korea. Methods. One hundred and ninety-six working postures were
selected from the real agricultural tasks to verify ALLA, a lower limb body posture assessment tool, and then evaluated by
16 ergonomic experts. Hit rate, quadratic weighted κ, one-way analysis of variance and t-test analyses were applied to compare ALLA with other assessment tools. Results. ALLA analysis had a superior hit rate with ergonomic expert assessment
compared with other assessment tools. Quadratic weighted κ analysis also showed that ALLA provided superior estimates
of risk levels for farm working postures. Discussion. ALLA would be an appropriate assessment tool to estimate risk factors
for various lower limb body postures which frequently occur in agricultural tasks in Korea. ALLA is a simple and accurate
risk assessment tool that could be usefully applied to identify and mitigate risk factors and work-related musculoskeletal
disorders in agricultural tasks, and also to evaluate the effects of control and intervention for working conditions.
Keywords: ergonomic risk assessment tools; agricultural lower limb assessment; rapid entire body assessment; rapid upper
limb assessment; Ovako working posture analysis system
1. Introduction
Many studies have investigated work risks using
ergonomic risk assessment tools to prevent musculoskeletal disorders. The most common assessment tools include
rapid entire body assessment (REBA) [1], rapid upper
limb assessment (RULA) [2], the Ovako working posture
analysis system (OWAS) [3], posture, activity, tools and
handling (PATH) [4] and occupational repetitive action
(OCRA) [5]. However, these ergonomic assessment tools
focus on the posture of the upper rather than lower limbs,
and do not apply well to the agricultural, manufacturing
or construction sectors. These tools are therefore unable to
address the continued increase in skeletal disease in Korea,
particularly in agricultural workers.
Musculoskeletal disorders have steadily increased in
Korea since 2003. The decrease in economically active
farm workers, aging of farm workers and reduction of
population per farmhouse has brought about increased cultivated acreage per farmhouse. Thus, with a decreasing
labor force, significance increases in labor intensity have
eventuated. Consequently, the incidence of musculoskeletal disorders for agricultural, forestry and fishery workers
(61.5%) was 2.5 times that of other sector workers (25.1%)
[6]. The many studies produce various results according
to the specific country, race, and age. Even statistically
significant results from one study may transfer to another
*Corresponding author. Email: [email protected]
© 2017 Central Institute for Labour Protection – National Research Institute (CIOP-PIB)
from a different country, due to anthropometric and cultural
differences. Kirby et al. [7] showed that lower limb postures affect the stability and mobility of the whole body,
and, thus, affect the whole working posture load. Gallagher
et al. [8] showed that maximum achievable lifting load is
affected by lower limb posture.
Kong et al. [9] developed the agricultural lower
limb assessment (ALLA) tool, a diverse and segmented
ergonomic lower limb assessment tool suitable for farm
work for Koreans. ALLA is especially useful for lower
limb burdening work, and is based on results acquired
through human body physiological experiments targeting
Koreans using lower limb postures not often observed in
farm work.
This study measured the electromyography (EMG) of
muscles, heart rate and subjective discomfort to maintain
each of 13 lower limb postures. The risk level on working
posture, exposed time and considering both working posture and exposed time can be evaluated through four risk
levels (1 = medium, 2 = moderate, 3 = high, 4 = very
high). The 13 ALLA lower limb postures can be categorized into postures resulting from the change of knee
angles, squatting and kneeling. Rick assessment for each
posture can be performed by analyzing the risk inherent to
the posture and the risk derived from retaining the posture
(Figure 1). Kong et al. [9] proposed ALLA and validated
International Journal of Occupational Safety and Ergonomics (JOSE)
219
Figure 1. Agricultural lower limb assessment (ALLA).
Note: The full color version of this figure is available online.
it through a comparative analysis with other evaluation
tools. However, statistical significance was not explained,
because no evaluation of various postures was conducted.
This study aims to compare ALLA for farm work,
targeting various lower limb postures, against other
ergonomic assessment tools (OWAS, RULA and REBA)
used for musculoskeletal factor surveys.
2. Methods
2.1. Posture selection
This study selected 196 postures occurring in Korea’s
agricultural tasks (farm work) to verify the validity of
lower limb assessment tools, including ALLA, RULA,
REBA and OWAS. An evaluation tool should be able
to evaluate various postures precisely, irrelevant of high
or low risk. Therefore, this study selected the working
postures as follows.
The risk of each crop was scored using a 100-point
scale, based on the general health status and fatigue, disease rate per crop, labor intensity and hard body part per
crop and on survey results on subjective symptoms of
musculoskeletal disorders per crop from Kim [10].
This study then classified risks into four levels according to the score, and selected crops at different working
heights – under knee (bottom–knee), near waist (knee–
breast) and above shoulder – for each crop in each risk
level. A total of 10 crops were selected, as presented in
Table 1, because there is no crop corresponding to the near
waist working height for the very high and medium risk
levels.
To identify the working postures for the 10 selected
crops, researchers visited farms and took photographs or
referred to farm work reports [11–18].
2.2. Evaluation by ergonomic experts
Sixteen ergonomic experts (10 ergonomists, 4 industrial
medical experts and 2 agricultural experts) evaluated each
crop’s working postures using a 10-point scale, where
1 = very safe posture and 10 = very risky posture, excluding factors such as weights or vibration on the given
posture. According to the mean scores measured by the
ergonomic experts for each posture, the risk was classified
into four levels according to the mean scores: 1 = very low
(1.00–3.25), 2 = low (3.25–5.50), 3 = high (5.50–7.25)
and 4 = very high (7.25–10.00).
This study considered the suitability of each evaluation
tool by analyzing whether the tool-evaluated results were
consistent with the expert assessment.
2.3. Evaluation of ergonomic assessment tools
To evaluate each assessment tool, the risk level was estimated for lower limb postures following the ALLA system.
However, risk levels cannot be estimated using only lower
limb postures for RULA, REBA and OWAS. Therefore,
for those systems, final risk levels were estimated using
220
Y.-K. Kong et al.
Table 1. Characteristics of the selected 10 crops.
Risk level
Working height
Very high
Above the shoulder
Near waist (knee–breast)
Under knee (bottom–knee)
Grape
Table 2.
κ
<0.20
0.21–0.40
0.41–0.60
0.61–0.80
>0.80
Strawberry
Criteria of κ analysis.
Strength of agreement
Poor
Fair
Moderate
Good
Very good
the upper limb postures corresponding to the lower limb
postures.
ALLA, RULA and OWAS all use four action categories levels, whereas REBA uses five action levels. Therefore, to align the systems for the purposes of the current
study, REBA action level 0 (no action undertaken) was
amalgamated with action level 1.
2.4. Statistical analysis
Hit rate, quadratic weighted κ, one-way analysis of variance (ANOVA) and one-way t-test analyses were used
to examine consistency between expert assessment and
ALLA, OWAS, REBA and RULA.
First, hit rate analysis classifies expert assessment
results by risk level, and then evaluates the postures corresponding to each risk level from each assessment tool,
producing a rate of common outcomes or hit rate. For
example, suppose the expert assessment produced 30 postures with risk level 1. Each assessment tool evaluates these
30 postures, producing various risk levels (1–4) which are
divided by the total number of postures, namely 30, and
then the rate is demonstrated.
Second, quadratic weighted κ analysis was performed
using the κ value calculated through the expert assessment
results and the evaluation result of each assessment tool.
The κ value was between 0 and 1, and the κ criteria are
presented in Table 2.
Third, one-way ANOVA was conducted to check the
effects of the assessment tools (RULA, REBA, OWAS and
ALLA) against the final risk level from expert evaluation,
by group (risk level 1–4). To identify differences between
the assessment tools, when significant differences were
observed for a given final risk level, due to an assessment
tool, the Tukey test was performed post analysis.
High
A little high
Peach
Rice
Cucumber
Tangerine
Chrysanthemum
Tomato
Medium
Pear
Oriental melon
Lastly, the one-way t test was performed to probe
whether the assessment tools provided significantly different outcomes from expert assessment.
3. Results
3.1. Hit rate analysis
Table 3 presents the hit rates for the various risk levels. The
number of each working posture from 1 to 4 risk levels of
expert assessment results on 196 postures was 12, 72, 96
and 14, respectively.
The hit rate of expert assessment results and the ALLA
evaluation results was 15.1% in the working postures with
risk level 1 of the expert assessment results. The working
postures with risk levels 2 and 3 were over 50%, namely
52.1 and 57.1%, respectively, and the working postures
with risk level 4 showed 37.1% of hit rate.
The hit rate of expert assessment and REBA assessment
was 100% for working postures with expert assessed risk
level 1. However, the risk level of all working postures was
assessed as risk level 1 by REBA, irrelevant of the expert
assessment.
The hit rate of expert assessment and RULA assessment was 100% for working postures with expert assessed
risk level 1, but only 12.3% for working postures with
assessed risk level 2, and postures with risk levels 3
and 4 were RULA assessed as risk levels 1 and 2,
respectively.
The hit rate of expert assessment and OWAS assessment was similar to that for RULA. The hit rate for
working postures with expert assessed risk level 1 was
98.1%, with 28.8% for risk level 2 postures. Working postures with expert assessed risk levels 3 and 4 were OWAS
assessed with risk level 1 and risk level 2, respectively.
3.2. Quadratic weighted κ analysis
Quadratic weighted κ analysis was used to examine the hit
rate between expert assessment and the assessment tools,
as presented in Table 4.
The κ value for expert assessment and ALLA assessment was 0.803, which shows good consistency. κ for
expert assessment and RULA assessment was 0.627, which
shows relatively small consistency, whereas for REBA and
OWAS the κ value was 0.490 and 0.501, respectively,
showing only moderate consistency.
International Journal of Occupational Safety and Ergonomics (JOSE)
221
Table 3. Results of hit rates (%) for all ergonomic risk evaluation tools for the expert assessments
ALLA
Risk level
Expert
assessment
1
2
3
4
12
72
98
14
1
2
15.1
35.7
20.0
84.9
52.1
REBA
3
4
1
20.5
57.1
62.9
2.7
22.9
37.1
100
100
100
100
2
3
RULA
4
1
2
100
87.7
57.1
80.0
12.3
42.9
20.0
OWAS
3
4
1
2
98.1
71.2
60.0
11.4
1.9
28.8
40.0
88.6
3
4
Note: ALLA = agricultural lower limb assessment; OWAS = Ovako working posture analysis system; REBA = rapid entire body
assessment; RULA = rapid upper limb assessment.
Shading represents accuracy (%) of expert assessment results and each assessment’s results. Number represents hit rate, no number
means 0 hit rate. For example, 12 postures were evaluated as a ‘risk level 1’ by experts’ evaluation and then these were evaluated as
‘risk level 1’ – 15.1% or ‘risk level 2’ – 84.9% by ALLA in this study.
Table 4. Results of quadratic weighted κ analysis for all
ergonomic risk evaluation tools.
Checklist
κ
Strength of
agreement
ALLA
REBA
RULA
OWAS
0.803
0.490
0.627
0.501
Very good Moderate Good Moderate
Note: ALLA = agricultural lower limb assessment;
OWAS = Ovako working posture analysis system;
REBA = rapid entire body assessment; RULA = rapid
upper limb assessment.
3.3. One-way ANOVA and t-test analysis
All assessment tools were significantly different from
expert assessment for all expert assessment groups
(α = 0.05).
Figure 2 shows each assessment tool’s evaluation by
the expert assessment group. ALLA showed the highest
assessed risk for all expert groups, with REBA always
showing the lowest.
For the expert assessed risk level 1 group, ALLA was
significantly higher at 1.9, whereas REBA and RULA were
the lowest assessed risk (1.0) and not significant different
from the expert assessment.
For the expert assessed risk level 2 group, ALLA was
again the highest assessment at 2.1, with no significant difference from the expert assessment. OWAS, RULA and
REBA mean assessment was 1.3, 1.1 and 1.0, respectively, all of which are significantly different from expert
assessment.
The expert assessed risk level 3 group showed similar outcomes to the expert assessed risk level 2 group.
ALLA assessment was the highest at 2.6, and was not significantly different from the expert assessment, whereas
OWAS, RULA and REBA assessment was 1.5, 1.3 and 1.0,
respectively, which were all significantly different from
expert assessment.
Finally, for the expert assessed risk level 4 group, mean
ALLA, OWAS, RULA and REBA assessments (3.1, 1.7,
1.3 and 1.0, respectively) were all significantly different
from expert assessment.
In Figure 2 the assessments have also been grouped
by letter (A, B, C, D), where the same letter indicates the
groups are not significantly different.
4. Discussion
This study aimed to verify ALLA, a lower limb posture
assessment tool for agriculture, by comparing ALLA and
existing widely used ergonomic assessment tools (OWAS,
RULA and REBA) with direct expert assessment using
identified working postures occurring in agricultural tasks.
ALLA reflects the expert assessed risk level better
than the other ergonomic assessment tools considered. The
other assessment tools largely evaluated the risk level of
most agricultural tasks as 1 or 2, regardless of the expert
assessment. Hence they tended to produce a high hit rate
for expert assessed risk level 1 or 2 groups, but poor hit
rates for expert assessed risk level 2 or higher groups. This
outcome is a result of lower limb posture assessment being
largely underestimated in the OWAS, RULA and REBA
systems, whereas ALLA was specifically developed to
address this limitation [9]. Quadratic weighted κ analysis
was used to examine consistency between each assessment
tool and expert assessment. ALLA was quite consistent
with expert assessment, and significantly superior to the
other assessment tools.
The one-way ANOVA also showed that ALLA had
the highest assessed risk level across all expert assessed
groups, whereas REBA had the lowest. Although REBA
does include load assessment on lower limb postures, compared with RULA, REBA’s hit rate was the lowest because
only four risk levels (1, 1, 2 or 3) were achieved upon
evaluation of all postures, setting them as the basis except
for lower limb postures. On the other hand, RULA provided only two risk levels (1 or 3). Therefore, REBA mean
assessment was smaller than RULA, even though there was
no significant difference.
This study also analyzed differences between expert
assessment and each assessment tool. For the expert
222
Y.-K. Kong et al.
Figure 2. Results of one-way analysis of variance (ANOVA).
Note: *significant difference in statistics; A, B, C, D = significant grouping in statistics; ALLA = agricultural lower limb assessment;
OWAS = Ovako working posture analysis system; REBA = rapid entire body assessment; RULA = rapid upper limb assessment.
The full color version of this figure is available online.
assessed risk level 1 group, no significant differences
were shown except for ALLA. For expert assessed risk
levels 2 and 3 groups, only ALLA assessment showed
no significant difference from expert assessment. Finally,
for the expert assessed risk level 4 group, all assessment tools were significantly different from expert assessment, although ALLA was the closest outcome to expert
assessment.
Lee et al. [19] evaluated load change according to the
height of a support and duration of the squatting posture
using a psychophysical method. Lee et al. [19] assessed
31 lower limb loads using subjective discomfort, similar
to ALLA. This study analyzed and selected lower limb
postures frequently occurring in farming work in Korea,
based on postures from the studies discussed that showed
significant differences in subjective discomfort.
Previous studies compared tasks using various assessment tools. Lee et al. [19] compared the most widely used
three tools, OWAS, RULA and REBA, and analyzed the
features from the work posture loads. They determined that
waist postures played a pivotal role in deciding overall load
level in OWAS, compared with shoulder postures. Because
the current study evaluated various lower limb postures,
and unified upper limb postures as basic postures, OWAS
assessment risk levels 1 and 2 were mainly shown.
Lee et al. [19] also reported differences in cognitive
discomfort, because RULA did not discriminate lower
limb postures sufficiently, which is similar to the current
study that has shown consistency is low between expert
assessment and RULA assessment for lower limb postures.
Lee et al. [19] showed that REBA was more suitable
for evaluating whole body working postures than other
assessment tools. However, the current study shows that
the consistency between REBA and expert assessment was
relatively lower than the other assessment tools. This is
because OWAS, RULA and REBA do not reflect various
lower limb postures occurring in farming work.
Lee et al. [20] reported that the OCRA assessment
tool correlation with labor intensity in the automobile
industry was higher than that of RULA, and suggested
the reason for this was that OCRA’s evaluation ratio on
repetitiveness, a characteristic of the car assembly process, was relatively higher. They argued that the use of a
proper tool suitable for specific work characteristics was
required. Similarly, the current study shows that ALLA,
which focuses on lower limb postures frequently occurring
in farming work, is more suitable than assessment tools that
largely ignore lower limb posture.
Lee et al. [21] analyzed specific body parts for specific
work using OWAS, RULA, REBA and quick exposure
check (QEC). They compared waist bending when working with weights and waist bending, squatting, neck extension and neck bending for assembly work, and acquired
different results for each situation. They also concluded
that an appropriate assessment tool should be used according to posture. Therefore, the current study recommends
using ALLA, because this lower limb assessment tool has
shown superior agreement with expert assessment of risk
levels for lower limb postures.
Because experts commonly use REBA, RULA and
OWAS, these existing assessment tools become the standard for posture evaluation. Although existing assessment
tools consider upper and lower limb postures, they cannot
independently evaluate upper or lower limb postures.
5.
Conclusion
ALLA risk level assessments show high consistency with
expert risk level assessment for farming work postures.
International Journal of Occupational Safety and Ergonomics (JOSE)
Therefore, ALLA is suitable for evaluation of lower limb
postures for farming work.
The current study shows that ALLA can conveniently
and objectively evaluate work-related musculoskeletal disorder (WMSD) risk levels for lower limb postures commonly observed in farming work in Korea, such as squatting, going down on one’s knees and sitting on the floor.
Therefore, ALLA can be used as an assessment tool to
evaluate the effects of disease prevention, management
and working environment improvement by objectively
evaluating WMSD risk levels in agricultural tasks.
Although ALLA is suitable for measuring risk level by
body part, it is insufficient to evaluate whole body postures.
Extending the current study, an upper limb assessment
tool – agricultural upper limb assessment (AULA) [22] –
has been developed, and combined with ALLA, providing
a whole body assessment tool – agricultural whole body
assessment (AWBA) [23]. The application of AWBA will
be assessed through further verification studies.
[9]
[10]
[11]
[12]
[13]
[14]
Disclosure statement
No potential conflict of interest was reported by the authors.
[15]
Funding
This work was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF)
funded by the Ministry of Education [NRF-2016R1A1A09
918189].
[16]
[17]
References
[1] Hignett S, McAtamney L. Rapid entire body assessment
(REBA). Appl Ergon. 2000;31:201–205. doi:10.1016/S00
03-6870(99)00039-3
[2] McAtamney L, Nigel Corlett E. RULA: a survey method for
the investigation of work-related upper limb disorders. Appl
Ergon. 1993;24(2):91–99. doi:10.1016/0003-6870(93)900
80-S
[3] Karhu O, Kansi P, Kuorinka I. Correcting working postures
in industry: a practical method for analysis. Appl Ergon.
1977;8(4):199–201. doi:10.1016/0003-6870(77)90164-8
[4] Buchholz B, Paquet V, Punnett L, et al. PATH: a work
sampling-based approach to ergonomic job analysis for
construction and other non-repetitive work. Appl Ergon.
1996;27(3):177–187. doi:10.1016/0003-6870(95)00078-X
[5] Occhipinti E, Colombini D. [Proposal of a concise index
for the evaluation of the exposure to repetitive movements of the upper extremity (OCRA index)]. Med Lav.
1996;87:526–548. Italian.
[6] Rural Development Administration. [Comparison analysis
of main disorder and health behavior of farmers]. Suwon:
Farmers Health & Safety Information Center; 2004. Korean.
[7] Kirby RL, Price NA, MacLeod DA. The influence of foot
position on standing balance. J Biomech. 1987;20(4):423–
427. doi:10.1016/0021-9290(87)90049-2
[8] Gallagher S, Marras WS, Bobick TG. Lifting in stooped
and kneeling postures: effects on lifting capacity, metabolic
[18]
[19]
[20]
[21]
[22]
[23]
223
costs, and electromyography of eight trunk muscles. Int
J Ind Ergon. 1988;3:65–76. doi:10.1016/0169-8141(88)90
007-8
Kong Y-K, Han J-G, Kim D-M. [Development of an
ergonomic checklist for the investigation of work-related
lower limb disorders in farming – ALLA: agricultural lower
limb assessment]. J Ergon Soc Korea. 2010;29(6):933–942.
Korean. doi:10.5143/JESK.2010.29.6.933
Kim K. [A study on the farmers’ health status and musculoskeletal workload] [doctoral dissertation]. Seoul: Seoul
University; 2008. Korean.
Rural Development Administration (RDA). [Research
report on risk assessment of agricultural tasks for the development of safety and health management system in rural
villages]. Suwon: RDA; 2006. Korean.
Rural Development Administration (RDA). [Research
report on risk assessment of agricultural tasks for the development of safety and health management system in rural
villages]. Suwon: RDA; 2007. Korean.
Rural Development Administration (RDA). [Research
report on risk assessment of agricultural tasks for the development of safety and health management system in rural
villages]. Suwon: RDA; 2008. Korean.
Rural Development Administration (RDA). [Research
report on risk assessment of agricultural tasks for the development of safety and health management system in rural
villages]. Suwon: RDA; 2009. Korean.
Rural Development Administration (RDA). [Research
report on risk assessment of agricultural tasks for the development of safety and health management system in rural
villages]. Suwon: RDA; 2010. Korean.
Rural Development Administration (RDA). [Development
of agricultural safety management experience program].
Suwon: RDA; 2011. Korean.
Rural Development Administration (RDA). [Development
of agricultural safety management experience program].
Suwon: RDA; 2012. Korean.
Rural Development Administration (RDA). [Development
of agricultural safety management experience program].
Suwon: RDA; 2014. Korean.
Lee IS, Jung MK, Choi KI. [Comparison of observational posture evaluation methods based on perceived discomfort]. J Ergon Soc Korea. 2003;22(1):43–56. Korean.
doi:10.5143/JESK.2003.22.1.043
Lee K-S, Kim J-H, Jung M-S, et al. [Comparison of assessment by OCRA Checklist and RULA at an auto manufacturing plant]. J Ergon Soc Korea. 2007;26(4):153–160.
Korean. doi:10.5143/JESK.2007.26.4.153
Lee KS, Yeom JW, Hur HM, et al. [Comparisons of different
evaluation tools for musculoskeletal disorders]. 2009 Fall
Conference of the Ergonomics Society of Korea; 2009 Dec
06–07; Daegu, Korea. Korean.
Kong Y-K, Lee S-J, Lee K-S, et al. [Development of
an ergonomic checklist for the investigation of workrelated upper limb disorders in farming – AULA: agricultural upper-limb assessment]. J Ergon Soc Korea.
2011;30(4):481–489. Korean. doi:10.5143/JESK.2011.30.
4.481
Kong YK, Lee SJ, Lee KS, et al. Development of
an ergonomics checklist for investigation of workrelated whole-body disorders in farming – AWBA: agricultural whole-body assessment. J Agric Saf Health.
2015;21(4):207–215. doi:10.13031/jash.21.10647