Assessment of the musculoskeletal discomfort scale for upper limb among workers in inner Brazil

Abstract

BACKGROUND:

Work-related musculoskeletal disorders for upper limbs (UL-WMSDs) form a complex of occupational diseases common to many professions worldwide. UL-WMSDs are manifested in most cases by pain, resulting in musculoskeletal discomfort.

OBJECTIVE:

This research aimed to evaluate the perception of musculoskeletal discomfort in workers from the interior of the Brazilian states of Alagoas and Bahia through the construction of a scale to assess musculoskeletal discomfort for upper limb.

METHODS:

The discomfort assessment scale was constructed from self-reported pain symptoms by 420 workers from the inner regions of the Brazilian states of Alagoas and Bahia. The reliability and dimensionality of the collected data were analyzed by McDonald’s Omega and exploratory factor analysis, respectively. Item Response Theory (IRT) was used to create parameters for the discomfort scale.

RESULTS:

The musculoskeletal discomfort metric was constructed from the workers’ response with six levels (varying from minimum discomfort to maximum discomfort). At the lowest level of the scale, workers indicated symptoms in the shoulders and wrists were rare. At the highest level of the scale, daily pain symptoms are reported in all regions of the upper limbs. The shoulders are the last region to develop extreme pain symptoms.

CONCLUSION:

The metric was created to present satisfactory psychometric properties and capable measurement of the workers’ level of musculoskeletal discomfort based on self-reported pain symptoms. Therefore, the metric can support measuring discomfort, contributing to decisions that improve a healthier occupational environment for the worker.

Keywords

Keywords UL-WMSDs item response theory pain symptoms metrics in ergonomics

1 Introduction

Work-related musculoskeletal disorders (WMSDs) are considered a serious, common, and expensive occupational health problem [1, 2]. The report from the Special Secretariat for Social Security and Labor in Brazil accounted for around 39,000 workers a year being removed from work due to WMSDs, creating a loss of functionality because of the difficulty of movement, impacting both the professional and personal life of the worker [3]. In some cases, WMSDs are associated with insufficient time to recuperate the parts that form the musculoskeletal system of workers [3 –5]. Therefore, when the psychomotor and cognitive demands of work go beyond workers’ limits, a process of illness begins that affects various regions of the body, especially the upper limbs [6].

In countries like Brazil, WMSDs impose the greatest occupational disease burden occupational disease [7]. Therefore, WMSDs become a complex problem because they have a multifactorial origin [8] and because they are influenced by characteristics of the work environment, and the way the activities are performed [9]. Indeed, pain symptoms result in decreased work efficiency, negatively impacting the increase in production costs and reducing the quality of the product or service being offered [10].

According to data from the Brazilian Ministry of Health, WMSDs have a higher incidence among professionals that work in the sectors of industry, commerce, food, transport, and domestic/cleaning services [7]. Between 2016 and 2017, the Northeast Region had the second highest incidence of WMSD in the country, with 27.2%of cases in workers [11].

Despite this, there is an under-notification of cases of musculoskeletal disorders in inner regions of the Northeastern states, especially in Alagoas and Bahia [12]. Thus, the actual WMSD cases may be higher than the official information from the Brazilian government.

However, it is possible to identify economic sectors with a high number of reported accidents at work, like the hospital care sector, with 6%of accidents between 2012 and 2022 in the city of Delmiro Gouveia (inner of the State of Alagoas); and the electrical energy sector (27%of accidents), public administration (7,44%of accidents), preservation and manufacture of fish products (6,82%of accidents) and retail trade of goods in general (6,05%of accidents) in Paulo Afonso (city of the State of Bahia) according to the National Institute of Social Security of Brazil [13].

Because they are considered costly occupational injuries, WMSDs have been constantly studied in different work contexts. In many cases, statistical modeling has been used to analyze the influence of risk factors on the onset of upper limb work-related musculoskeletal disorders (UL-WMSDs) [14 –16]. However, the literature is incomplete in the construction of scales that consider ergonomic issues to guide organizational policies and practices during the analysis of working conditions [17]. In this context, in most cases, musculoskeletal discomfort is assessed either by a single item or based on the sum of symptoms reported by workers. However, expressing a complex phenomenon through a simple item does not seem the most appropriate. Likewise, for musculoskeletal discomfort scores that are generated based on the sum, simple or weighted, it is not known how accurate and reliable they are.

Metrics or scales to quantify musculoskeletal discomfort are still scant. Alternatively, researchers in ergonomics and occupational health have used Item Response Theory (IRT) to pick up the response patterns associated with musculoskeletal symptoms in the body parts of workers to generate a valid and reliable discomfort score [18, 19]. IRT makes use of a mathematical model that represents the relationship between the probability of an individual marking an alternative response (pain intensity, for example) of an item (body region), and the behavior of the latent trait (musculoskeletal discomfort). of individuals (workers) [20 –22]. Thus, conclusions do not depend on the test as a whole or on the entirety of the research instrument used, but on each item that composes it, enabling a statistical analysis focused on the item [23].

It is worth noting that few studies are focusing on the health of workers in the interior regions of Northeast Brazil [12 , 25]. This is the first study with a sample of workers from the Inner States of Alagoas and Bahia that seeks to construct a scale of musculoskeletal discomfort based on symptoms self-reported by workers in the upper body region. Given the above, the article aims to evaluate the perception of musculoskeletal discomfort through the construction of a scale to assess musculoskeletal discomfort in the upper limb with workers in the industry, commerce, health, and education sectors of cities in the hinterland of Alagoas and Bahia, leveraging IRT to improve response quality.

2 Materials and methods

The steps for this research were: (i) Definition of population and sample; (ii) Research instrument; and (iii) Statistical analysis of data. This study was approved by the Research Ethics Committee of the Federal University of Alagoas, with CAAE number 35014720.6.0000.5013. The research was conducted over a two-year period in locations where consent was obtained. The researchers split up to collect data from workers in both Alagoas and Bahia, which began in February 2020 and ended in January 2021. After collection, the data was organized, which took approximately 12 months (period from February 2021 to January 2022) to complete and be applied according to the proposed methodology.

2.1 Definition of the study population and sample

This research accounts for a sample of data from an Inner Northeastern region, more specifically in the cities of Delmiro Gouveia (State of Alagoas) and Paulo Afonso (State of Bahia). Sampling took place by non-probabilistic means, totaling 420 professionals interviewed from the industry, commerce, health, and education segments (Table 1). The pre-established inclusion criteria for participation in the research were: voluntary worker participation, being at least 18 years old, and having an effective contract with the company, regardless of biological sex. The exclusion criteria applied to workers who did not want to participate in the research during data collection, who were not dismissed due to WMSD or occupational accidents in others works and pregnant women.

Table 1
Description of professional areas

Segments Total Professional areas of the respondents

Health 167 Doctors, nurses, nursing technicians, pharmacists, and others

Education 159 Teachers, librarians, technicians, and professionals in the administrative area

Trade private 35 Various functions linked to service work

Industry 59 Workers from the ceramic industries

Segments	Total	Professional areas of the respondents
Health	167	Doctors, nurses, nursing technicians, pharmacists, and others
Education	159	Teachers, librarians, technicians, and professionals in the administrative area
Trade private	35	Various functions linked to service work
Industry	59	Workers from the ceramic industries

As seen in Table 1, 167 professionals from hospitals, clinics, and health centers were interviewed, including doctors, nurses, nursing technicians, pharmacists, and others. For education, there were 159 responses, including teachers, librarians, technicians, and professionals in the administrative area of 9 municipal schools, 3 state schools, and a university. Trade covered private chain stores in the service sector, totaling 35 respondents. For the industry, 59 workers from the ceramic industries answered the questionnaire.

2.2 Research instrument

The data referring to pain symptoms in the upper limbs were collected in loco through a self-administered questionnaire adapted from the Nordic questionnaire [26], from the diagram detailing symptoms in the forearm region [27] and in the hands [28] (Fig. 1).

Fig. 1

Regions analyzed by the research instrument.

The instrument has a total of twelve items referring to the regions highlighted for the upper limbs. A scale with five response alternatives (1- I never felt pain; 2- I rarely feel pain; 3- Sometimes I feel pain; 4- I often feel pain; 5- I feel pain daily) was used to help determine workers’ perception of pain in each region of the upper limbs. Questions were added about sociodemographic factors such as age, BMI, education, marital status, children, physical activity, and biological sex. The complete instrument is shown in Fig. 1.

2.3 Statistical analysis

The reliability of the instrument was assessed via the McDonald’s Omega test (ω) and the ω values greater than 0.70 indicate reliability [29] Bartlett’s test of sphericity and Kaiser-Meyer-Olkin’s (KMO) test were also applied to analyze the suitability of data for full-information item factor analysis (FA). Bartlett’s test of sphericity the hypothesis that the variables are uncorrelated and KMO is used to compare the partial correlations between pairs of variables without the effects of the others [30, 31]. A p-value lower than 0.05 in Bartlett’s sphericity test and a KMO greater than 0.70 indicate good adequacy of the data to the FA technique.

Parallel analysis was used to determine the number of dimensions to be used in FA. The instrument was considered unidimensional when a factorial load (F) greater than 0.300 and commonality (h²) greater than 0.200 for all parts of the body in a single dimension [19]. Items with F less than 0.300 or h² less than 0.200 were considered low-quality items [31].

Equation 1 presents the TRI gradual response model (GMR) [32] which was used to estimate the discrimination parameters (ai) and difficulty (bi) of the items and the latent trait of musculoskeletal discomfort in the upper limbs (⊖j) of workers. These parameters were estimated with a mean of 0.00 and a standard deviation of 1.00 [20]. $P_{ik} (θ_{j}) = \frac{1}{1 + e^{- a_{i} (θ_{j} - b_{ik})}} - \frac{1}{1 + e^{- a_{i} (θ_{j} - b_{ik + 1})}}$ (1)

The parameter a_i expresses the capacity of an item i to discriminate individuals with different levels of the latent trait θ (musculoskeletal discomfort in the upper limbs of workers); and the b_ik parameter represents the position on the scale of the latent trait θ in which category k of item i has the highest probability [33]. Items with ai greater than 0.70 are adequate [20, 34] and b_ik ∈ [–3.3] [35], as they indicate good quality [36]. The θ_j expresses musculoskeletal discomfort in the upper limbs experienced by worker j.

Based on the IRT parameters, the aim was to construct a musculoskeletal discomfort scale for the upper limbs based on the process of anchoring the response alternatives for each item. Initially, Z-anchor levels spaced by 0.55 θ discomfort units were created in the range of –3 to 3. Then, based on the IRT parameters, the cumulative probability was calculated at each Z anchor level. It was considered that a response alternative k of item i was anchored at a given Z anchor level when its cumulative probability was greater than 50%at this level [37].

Among the benefits of IRT, there is the amount of information provided for each item in different regions of the latent traits scale [38]. The instrument’s information curve indicates the region of the scale with greater accuracy, it ensures that the subset of selected items provides adequate precision throughout the continuum of the latent trait scale [36, 39]. Equation 2 presents the calculation to evaluate the information function of item i for the latent trait θ (Ii(θ)). $I_{i} (θ) = \sum_{X = 1}^{ki} \frac{{P^{'}}_{ix} {(θ)}^{2}}{{P^{'}}_{ix} (θ)}$ (2)

The sum of the information of each item i generates the instrument’s information curve, where P’_ix (θ) is the first derivative of the response curve of the category evaluated at a given latent trait level. Thus, the greater the discrimination of an item, the greater the information it provides to the measurement instrument [38].

3 Results

3.1 Descriptive analysis

The sample has the response of 420 individuals to the research instrument applied in the sectors of industry, commerce, health, and education, in the cities of Delmiro Gouveia in the state of Alagoas and Paulo Afonso in Bahia, a region known for being part of the Brazilian hinterland.

Sociodemographic data are shown in Table 2. The sample, for the most part, is made up of female individuals (72.38%), married (52.38%), with children (69.05%), aged between 18 and 44 years (72.14 %), normal weight (51.67%), completed high school (34.05%) and who perform physical activities (52.14%).

Table 2
Sociodemographic data

Variables Industry (n = 59) Commerce (n = 35) Health (n = 167) Education (n = 159) General (n = 420)

Frequency (%) Frequency (%) Frequency (%) Frequency (%) Frequency (%)

Sex

Female 21 (35.59) 27 (77.14) 134 (80.24) 122 (76.73) 304 (72.38)

Male 38 (64.41) 8 (22.86) 33 (19.76) 37 (23.27) 116 (27.62)

Age

18 to 44 years 48 (81.36) 31 (88.57) 126 (75.45) 112 (70.44) 303 (72.14)

Older than 45 years 11 (18.64) 4 (11.43) 41(24.55) 47 (29.56) 117 (27.86)

BMI

Underweight 1 (1.69) 3 (8.57) 3 (1.80) 2 (1.26) 9 (2.14)

Normal weight 33 (55.93) 17 (48.57) 64 (38.32) 103 (64.78) 217 (51.67)

Overweight 20 (33.9) 12 (34.29) 65 (38.92) 38 (23.90) 135 (32.14)

Type I obesity 5 (8.47) 3 (8.57 24 (14.37) 13 (8.18) 45 (10.72)

Type II obesity 0 (0.00) 0 (0.00) 7 (4.19) 3 (1.88) 10 (2.38)

Type III obesity 0 (0.00) 0 (0.00) 4 (2.40) 0 (0.00) 4 (0.95)

Education

Incomplete Elementary 19 (32.2) 4 (11.43) 5 (2.99) 0 (0.00) 28(6.67)

Complete Elementary 11 (18.64) 2 (5.71) 4 (2.40 2 (1.26) 19 (4.52)

Incomplete High School 9 (15.25) 5 (14.29) 1 (0.59) 0 (0.00) 15 (3.57)

Complete High School 18 (30.51) 22(62.86) 84 (50.30) 19 (11.95) 143 (34.05)

Incomplete Higher Education 0 (0.00) 0 (0.00) 9 (5.39 27 (16.98) 36 (8.57)

Complete Higher Education 2 (3.40) 2 (5.71) 60 (35.93) 71 (44.65) 135 (32.14)

Post Graduation 0 (0.00) 0 (0.00) 4 (2.40) 40 (25.16) 44 (10.48)

Marital Status

Single 26 (44.07) 19 (54.29) 81(48.50) 74 (46.54) 200 (47.62)

Married 33 (55.93) 16 (45.71) 86 (51.50) 85 (53.46) 220 (52.38)

Children

Yes 51 (86.44) 24 (68.57) 81(48.50) 58 (36.48) 290 (69.05)

No 8 (13.56) 11 (31.43) 86 (51.50) 101 (63.52) 130 (30.95)

Physical activity

No day 25 (42.37) 20 (57.14) 71(42.51) 85 (53.46) 201 (47.86)

1 day or more days 34 (57.63) 15 (42.86) 96 (57.49) 74 (46.54) 219 (52.14)

Variables	Industry (n = 59)	Commerce (n = 35)	Health (n = 167)	Education (n = 159)	General (n = 420)
Sex
Female	21 (35.59)	27 (77.14)	134 (80.24)	122 (76.73)	304 (72.38)
Male	38 (64.41)	8 (22.86)	33 (19.76)	37 (23.27)	116 (27.62)
Age
18 to 44 years	48 (81.36)	31 (88.57)	126 (75.45)	112 (70.44)	303 (72.14)
Older than 45 years	11 (18.64)	4 (11.43)	41(24.55)	47 (29.56)	117 (27.86)
BMI
Underweight	1 (1.69)	3 (8.57)	3 (1.80)	2 (1.26)	9 (2.14)
Normal weight	33 (55.93)	17 (48.57)	64 (38.32)	103 (64.78)	217 (51.67)
Overweight	20 (33.9)	12 (34.29)	65 (38.92)	38 (23.90)	135 (32.14)
Type I obesity	5 (8.47)	3 (8.57	24 (14.37)	13 (8.18)	45 (10.72)
Type II obesity	0 (0.00)	0 (0.00)	7 (4.19)	3 (1.88)	10 (2.38)
Type III obesity	0 (0.00)	0 (0.00)	4 (2.40)	0 (0.00)	4 (0.95)
Education
Incomplete Elementary	19 (32.2)	4 (11.43)	5 (2.99)	0 (0.00)	28(6.67)
Complete Elementary	11 (18.64)	2 (5.71)	4 (2.40	2 (1.26)	19 (4.52)
Incomplete High School	9 (15.25)	5 (14.29)	1 (0.59)	0 (0.00)	15 (3.57)
Complete High School	18 (30.51)	22(62.86)	84 (50.30)	19 (11.95)	143 (34.05)
Incomplete Higher Education	0 (0.00)	0 (0.00)	9 (5.39	27 (16.98)	36 (8.57)
Complete Higher Education	2 (3.40)	2 (5.71)	60 (35.93)	71 (44.65)	135 (32.14)
Post Graduation	0 (0.00)	0 (0.00)	4 (2.40)	40 (25.16)	44 (10.48)
Marital Status
Single	26 (44.07)	19 (54.29)	81(48.50)	74 (46.54)	200 (47.62)
Married	33 (55.93)	16 (45.71)	86 (51.50)	85 (53.46)	220 (52.38)
Children
Yes	51 (86.44)	24 (68.57)	81(48.50)	58 (36.48)	290 (69.05)
No	8 (13.56)	11 (31.43)	86 (51.50)	101 (63.52)	130 (30.95)
Physical activity
No day	25 (42.37)	20 (57.14)	71(42.51)	85 (53.46)	201 (47.86)
1 day or more days	34 (57.63)	15 (42.86)	96 (57.49)	74 (46.54)	219 (52.14)

As for the frequency of pain, most workers reported never feeling symptoms in the upper limbs, but when workers reported pain, the main regions with severe and maximum symptoms were concentrated in the right fingers, right thumb, right and left wrists, left proximal ulnar, left arm and right and left shoulders (Fig. 2).

Fig. 2

Frequency of pain perception in the affected regions.

3.2 Assessment of dimensionality

The ωt value was equal to 0.96, indicating the presence of reliability in the data collected by the research instrument. In the same way, Bartlett’s test of sphericity and KMO presented p-values equal to 2.2×10^–16 and 0.89, respectively, allowing the use of FA The parallel analysis graph (Fig. 3), when analyzing the explanatory variance of the set of factors of the items evaluated in the research instrument, suggests unidimensionality of the latent trait of musculoskeletal discomfort in the upper limbs, given that a single dimension stands out with a high eigenvalue.

Fig. 3

Parallel analysis scree plots.

Therefore, the FA method considered the presence of a dimension to estimate the values of F and h² (Table 3). All regions of the upper limbs showed F > 0.300 and h² ^> 0.200, reinforcing that musculoskeletal discomfort in the upper limbs is a latent unidimensional trait. Symptoms in the forearm region had an average F of 0.939, indicating that detailing the symptoms in this region of the body increases the correspondence between the symptoms of pain in the forearm and the perception of discomfort in the workers’ upper limbs. Similar results occurred for the hands, in which the detailing of their anatomical regions resulted in F values greater than 0.800.

Table 3

Factorial load, commonality, and TRI parameters in the analyzed body regions

Upper limb regions	F	h²	a_i	b₂	b₃	b₄	b₅
Finger right	0.824	0.678	2.471	0.595	1.141	1.533	2.286
Finger left	0.863	0.745	2.908	0.674	1.157	1.548	1.962
Thumb right	0.827	0.684	2.504	0.696	1.165	1.551	2.296
Thumb left	0.886	0.785	3.253	0.826	1.231	1.420	1.786
Interphalange right	0.901	0.813	3.545	0.79	1.157	1.518	1.897
Interphalange left	0.936	0.876	4.533	0.819	1.196	1.545	1.801
Palm right	0.925	0.856	4.147	0.919	1.180	1.414	1.655
Palm left	0.904	0.818	3.609	0.886	1.250	1.527	1.762
Wrist right	0.735	0.540	1.843	0.354	1.061	1.481	2.266
Wrist left	0.726	0.527	1.797	0.468	1.272	1.894	2.343
Distal radial right	0.933	0.870	4.397	0.849	1.229	1.518	1.746
Distal radial left	0.931	0.866	4.327	0.885	1.214	1.518	1.780
Distal ulnar right	0.928	0.861	4.243	0.955	1.290	1.575	2.016
Distal ulnar left	0.942	0.888	4.795	0.933	1.397	1.781	2.151
Proximal radial right	0.937	0.878	4.562	0.907	1.242	1.576	1.913
Proximal radial left	0.944	0.891	4.864	1.002	1.276	1.552	1.856
Proximal ulnar right	0.955	0.913	5.504	0.931	1.237	1.538	1.806
Proximal ulnar left	0.943	0.890	4.842	0.957	1.345	1.651	2.268
Elbow left	0.867	0.752	2.961	1.018	1.275	1.692	2.198
Elbow right	0.859	0.739	2.861	0.915	1.212	1.654	2.157
Arm left	0.783	0.613	2.141	0.623	1.098	1.465	2.211
Arm right	0.801	0.641	2.275	0.554	1.078	1.461	1.879
Shoulder left	0.640	0.410	1.419	0.301	1.158	1.899	3.014
Shoulder right	0.629	0.396	1.378	0.156	1.07	1.783	3.049

F –Factor loading; h₂ –Commonality; a_i –Discrimination values; b2, b3, b4, b5 –Difficulty parameter of each alternative answer.

Commonality values were also higher in regions of the upper limbs with greater anatomical detail. In addition, Table 3 presents the discrimination (a_i) and difficulty (b_ik) parameters. All body regions demonstrate a high ability to discriminate individuals with different levels of musculoskeletal discomfort in the upper limbs. It was also evident that workers needed to have an above-average level of discomfort to start reporting symptoms. But, in general, all regions had a high value of the difficulty parameter, indicating that the symptoms in these regions are not commonly experienced by most workers in the sample. These results are in line with the results shown in Table 4.

Table 4

Upper Body region discomfort scale

Scale (0±1)	≤1.09	1.10 –1.64	1.65 –2.19	2.20 –2.74	2.75 –3.29	≥3.3
Discomfort classification	Minimal	Low	Moderate	High	Severe	Maximum
	Wrist right (RA)	Finger right (RA)	Finger right (ST or OF)	Finger left (ED)	Finger right (ED)	Shoulder left (ED)
	Wrist left (RA)	Finger left (RA)	Finger left (ST or OF)	Thumb left (ED)	Thumb right (ED)	Shoulder right (ED)
	Shoulder left (RA)	Thumb right (RA)	Thumb right (ST or OF)	Interphalange right (ED)	Wrist right (ED)
	Shoulder right (RA)	Thumb left (RA)	Thumb left (ST or OF)	Interphalange left (ED)	Wrist left (ED)
		Interphalange right (RA)	Interphalange right (ST or OF)	Palm right (ED)	Proximal ulnar left (ED)
		Interphalange left (RA)	Interphalange left (ST or OF)	Palm left (ED)	Arm left (ED)
		Palm right (RA)	Palm right (ST or OF)	Wrist left (OF)
		Palm left (RA)	Palm left (ST or OF)	Distal radial right (ED)
		Wrist right (ST)	Wrist right (OF)	Distal radial left (ED)
		Distal radial right (RA)	Wrist left (ST)	Distal ulnar right (ED)
		Distal radial left (RA)	Distal radial right (ST or OF)	Distal ulnar left (OF)
		Distal ulnar right (RA)	Distal radial left (ST or OF)	Proximal radial right (ED)
		Distal ulnar left (RA)	Distal ulnar right (ST or OF)	Proximal radial left (ED)
		Proximal radial right (RA)	Distal ulnar left (ST)	Proximal ulnar right (ED)
		Proximal radial left (RA)	Proximal radial right (ST or OF)	Proximal ulnar left (OF)
		Proximal ulnar right (RA)	Proximal radial left (ST or OF)	Elbow left (OF or ED)
		Proximal ulnar left (RA)	Proximal ulnar right (ST or OF)	Elbow right (OF or ED)
		Elbow left (RA)	Proximal ulnar left (ST)	Arm right (ED)
		Elbow right (RA)	Elbow left (ST)	Shoulder left (OF)
		Arm left (RA or ST)	Elbow right (ST)	Shoulder right (OF)
		Arm right (RA or ST)	Arm left (OF)
		Shoulder right (ST)	Arm right (OF)
			Shoulder left (ST)

RA: Rarely, ST: Sometimes, OF: Often, ED: Everyday.

Thus, the rare symptoms in the right shoulder region will start to be reported when workers are discomfort level equal to 0.156, estimating the parameters with a mean of 0.00 and standard deviation of 1.00, the value was above the average. In most regions of the upper limbs, the items had a standard deviation above the mean (b_ik≥1.000) to report that response alternative ‘sometimes have symptoms of musculoskeletal pain’, and the standard deviation was 1.5 times and 2.0 times above the mean to report ‘frequent’ and ‘daily’ symptoms, respectively.

Figure 4 shows the instrument information curve. A low amount of information was evidenced in the region of the scale close to one standard deviation below the mean. On the other hand, the error for this same region was also low, indicating that the instrument generates little information for workers with a level of discomfort slightly below average. The greatest amount of information was grouped in the region of the scale close to 2 standard deviations below the mean; and in the region between the mean and up to 2 standard deviations above the mean. Thus, there are indications that the instrument measures the latent trait of workers with very low levels of discomfort, or medium discomfort, or above average discomfort.

Fig. 4

Instrument information and standard error curves.

3.3 Upper body of the discomfort scale

The accumulated probabilities became greater than 50%for the answer alternative, ‘I rarely have symptoms’, from 0.55 units of musculoskeletal discomfort in the upper limbs (⊖). Therefore, the first level of the scale, called minimal discomfort in the upper limbs, started at 0.55 units of ⊖ with mild symptoms in the wrists and shoulders. In general, the musculoskeletal discomfort scale for the upper limbs (Table 3) presented six levels, which were classified as minimal discomfort (⊖= 0.55), mild (⊖= 0.10), moderate (⊖= 1.65), high (⊖= 2.20), severe (⊖= 2.75) and maximum (⊖= 3.30).

Because it is a cumulative IRT model, at the last level of the scale (maximum discomfort) workers tend to respond that they experience daily symptoms in all regions of the upper limbs, with the shoulders being the last region with this frequency of symptoms. Still, in Table 3, it is possible to analyze the response pattern of each level of the scale of musculoskeletal discomfort in the upper limbs, in which the regions with painful symptoms of each level of the scale and the frequency in which these symptoms affect the workers are expressed.

According to Fig. 5, it was possible to observe that each region of the upper limbs anchors at different points as the level of pain increases and this presents the evolution of WMSD symptoms.

Fig. 5

Evolution of upper limb musculoskeletal discomfort for each scale level.

4 Discussion

The construction of musculoskeletal discomfort scales based on the IRT assumptions has demonstrated satisfactory psychometric properties [34], resulting in results with less bias, given that the findings are based on the analysis of response patterns contributed by an individual to a set of items and not the sum of the responses of these same items [40]. To understand the relevance of response patterns defined based on IRT parameters, researchers have devised Computerized Adaptive Testing (CAT). Wherein, a computer selects a few items contained in a bank of items already calibrated by the IRT procedures to be applied, and an individual doesn’t need to respond to all items so that their discomfort score is caused by pain, can be determined. example [42]. Other researchers have put to the test the response patterns generated by the IRT, starting to compare the response patterns of workers who are at the same level of the discomfort scale, but who have different ages, BMI, and length of service [24]. As a result, more than 95%of the analyzed response patterns, at all levels of the scale, showed no significant difference.

In this study, six different response patterns were captured (Table 4), generating the six levels of discomfort on the scale that ranged from minimal musculoskeletal discomfort to maximum musculoskeletal discomfort in the upper limbs. This research seeks response patterns for musculoskeletal symptoms in a sample of workers from the interior of the States of Alagoas and Bahia, who are part of a population with occupational characteristics that are little known, but with serious social characteristics such as low pay, long working hours, adverse weather conditions, illiteracy and improvised work tools [12].

The first response pattern of the scale of musculoskeletal discomfort in the upper limbs, called the region of minimal discomfort ( $\underset{j}{⊖}$ < 1.10), anchored rare symptoms in the wrists and shoulders. It is important to point out that workers who have shoulder pain can modify their movements during work [43], starting to adopt less safe actions, thus contributing to situations of biomechanical overload in other regions of the body. The wrist and shoulder regions were also the first to show musculoskeletal symptoms in the upper limbs in workers in the Brazilian footwear industry [22]. In the interior region of the States of Alagoas and Bahia, there is still excessive use of upper limbs in repetitive movements in manual activities. This scenario contributes to WMSD symptoms, especially in the shoulders and wrists [19].

The next level of the scale was called mild discomfort ( $1.10 \leq \underset{j}{⊖}$ < 1.65). At this scale level, all parts of the hand (fingers, interphalange, thumb, and palm), forearm (distal and proximal radial, distal and proximal ulnar), and elbow may exhibit rare symptoms. As for the symptoms in the arms, and in the right shoulder and wrist, they begin to occur more frequently (sometimes). Thus, the scale points to an aggravation profile generated from the rare symptoms in the shoulders and wrists. With this, as WMSDs cause tissue damage, the worker protects to save the regions in pain, adopting non-standard and inappropriate movements, overloading regions that should be less activated during work activity [44]. The presence of symptoms in these regions of the upper limbs was observed in workers in occupations similar to the one in this study, in which standing work with repetitive operations may indicate a medium and high risk for upper limb disorders [45], presenting pain symptoms in regions of the shoulders and wrists can be either rare or common [22] and shoulder discomfort is common [46, 60] in industry professionals. Regarding education, pain was found in the shoulders, arms, forearms, and wrist regions [47], while in the health sector there were also signs of pain symptoms in the upper limbs of nurses in relation to organizational factors [15] and in different occupational groups there were indications of pain in the shoulder [18 , 54] and in the hands [48, 54].

In the region of the scale called moderate discomfort ( $1.65 \leq \underset{j}{⊖}$ < 2.20), there is a significant increase in the frequency with which workers report symptoms in the parts that form the upper limbs. Thus, symptoms are no longer rare in both elbows, in addition to the shoulder, wrist, and left ulnar region (proximal and distal). Symptoms can be frequent in all regions of the hands, arms, in the left and direct radial (distal and proximal) region, and in the right ulnar (distal and proximal) region. Thus, it is observed that the process of worsening symptoms occurs jointly in the regions of the upper limbs for the analyzed sample. This phenomenon can be explained by the presence of biomechanical links that make movements involving the upper limbs require the joint use of several regions [50]. Therefore, the presence of a painful symptom in a region will tend to affect symptoms in other regions as it worsens, indicating that WMSDs can be modeled as a cumulative characteristic [12 , 36].

In the region of the high discomfort scale ( $2.20 \leq \underset{j}{⊖}$ < 2.75) it is possible to observe a worsening of the symptoms in the shoulders, left wrist, and in the left proximal ulnar region, becoming frequent symptoms. In contrast, daily symptoms can also be observed in the elbows, right arm, all regions of both forearms, all regions of the left hand, interphalangeal, and palm of the right hand. Culturally, Brazilian organizations do not seek to ensure appropriate cycles of work and rest for the recovery of workers [51], resulting in increasingly severe musculoskeletal symptoms that increase cases of absenteeism [52]. In addition, in the interior region of the States of Alagoas and Bahia, it is common to find work activities with high biomechanical work demands (maintenance of the upper limbs in an uncomfortable position for more than 6 hours, for example) and psychosocial demands (dissatisfaction at work and control of the work, for example) [24]. Such a scenario explains a response pattern captured at this level of the scale, in which symptoms are abundant. It should be noted that musculoskeletal symptoms in the upper limbs of workers in the industry [36], education [53], and in the service sector [54] have already been widely observed in previous studies.

The next region of the scale, called severe discomfort in the upper limbs ( $2.75 \leq \underset{j}{⊖}$ < 3.33), except the shoulders, all anatomical areas analyzed started to present daily symptoms. Compared to the previous level of the scale (high discomfort), at the severe discomfort level, both wrists, the fingers (including the thumb) of the right hand, in addition to the left arm and left ulnar region of the forearm have daily symptoms. Findings by Silva et al. [36] also indicate the wrists (right and left) and left arm experience daily symptoms only at higher levels of discomfort. Because it is a sensitive region with little fat, pain in the ulnar nerve is common in activities that compress the region, such as repeated bending with the elbow [54]. Such movements are observed by education professionals when writing on the whiteboard and by industry professionals who support or handle objects in this position. This behavior helps with the issue of inappropriate postures and movements, both due to the lack of appropriate instruments for the work environment and the lack of training. These characteristics mean that the worker needs to improvise the execution of the activity, which can result in postures that are harmful to health.

The finding relates to notifications of work accidents in the period from 2012 to 2022 in the municipalities examined in the study. In Delmiro Gouveia, the most frequently affected body parts include fingers (25.6%), hand (except wrist or fingers –5.79%), arm (between wrist and shoulder –4.96%) and shoulder (4.13%), with emphasis on the hospital care sector. The most reported occupations were retail salespeople and nursing technicians, both representing 4.17%[11]. In the municipality of Paulo Afonso, the areas predominantly affected were the fingers (23.2%), hand (except wrist or fingers –5%), forearm (between wrist and elbow –4.24%) and shoulder (3.9%). Nursing technicians led the notifications in occupations, with a rate of 11.2%, while the retail trade of general merchandise, with a focus on food products (6.05%), was the most reported economic sector [13]. These results are aligned with the body regions identified in the discomfort scale in the upper body region (Table 4), showing significant levels of discomfort, ranging from high to maximum.

At the last level of the scale, called maximum discomfort in the upper limbs ( $\underset{j}{⊖}$ ≥ 3.33), all regions were analyzed, including both shoulders, these being the last regions reached cumulatively by the musculoskeletal wear caused by the demands arising from professional activities. In industry, shoulder flexion or abduction movements are common, which are predictive factors of chronic or recurrent disorders in the upper limbs [56]. Disorders in the lower back and shoulders are also associated with pushing and pulling movements [57], activities commonly observed by healthcare professionals when handling equipment and patients. Furthermore, in the health sector, the worker maintains inappropriate postures, and stands for long periods, in addition to lifting/transferring heavy loads, without the proper break [58, 59]. In addition, the high level of distress, physical and psychosocial factors are predictors of the onset of pain in the neck and/or shoulder [60].

Accordingly, the study by Miranda et al. [61] indicated that the self-report of nonspecific pain in the shoulder was related to burnout, depression, and inability to express feelings. Wixted et al. [62] related the increased incidence of musculoskeletal disorders in the industrial sector to the increased cognitive demand imposed by the automation of manufacturing processes, verifying that increased distress among employees contributes to a higher incidence of self-reported symptoms in the shoulder and lower back. When it comes to studies in the countryside, Silva et al. [24] already pointed out that shoulder symptoms were associated with long exposure to biomechanical factors related to poor posture of the upper limbs and use of vibrating hand tools, in addition to psychosocial stressors such as low control and job satisfaction, in addition to high job insecurity, work and work-family conflict.

Studies by Silva et al. [36] indicated that daily elbow symptoms tend to occur at the highest level, something not observed in this study. Already Silva et al. [24] found that symptoms in the right hand were the only ones located in the upper limbs to be at the highest level of the musculoskeletal discomfort scale. On the other hand, different results were expected, as Brazil has continental proportions, exposing workers from similar professions to totally different working conditions. On the other hand, Bispo et al. [12], when studying WMSDs in this region of the backlands of Alagoas and Bahia, it was be observed that the region has a high degree of informality at work, high labor turnover, and intense competition for unstable jobs, something that can contribute to that workers hide their musculoskeletal symptoms and that they get worse or become more and more frequent. Furthermore, in Brazil it is estimated that diseases and accidents at work produce losses of approximately R$300 billion, in addition to immeasurable costs to the individual and family [63].

5 Conclusion

In this study, a six-level scale was developed to measure musculoskeletal discomfort caused by work-related musculoskeletal disorders among workers in the interior of the states of Alagoas and Bahia in Brazil. The scale’s psychometric properties were satisfactory, indicating validity for the musculoskeletal discomfort metric.

The use of diagrams of the human body to signal WMSD symptoms is a common practice for understanding WMSDs in the workers studied. Therefore, this study aimed to contribute to expanding the possibilities of using these reported symptoms to create a score that reflects the level of musculoskeletal discomfort experienced by workers.

The scale levels were constructed based on the workers’ response pattern, obtained by the anchoring procedure performed from the estimation of the IRT parameters. As they present a cumulative behavior, the response patterns seem to express the progressive process in which the symptoms affect the parts that form the workers’ upper limbs. It was evidenced that the symptoms begin to occur rarely in the shoulders and wrists (left and right). At the maximum level of discomfort, they affect all regions of the upper limbs, with symptoms in the shoulders (left and right) of the daily type, the last to manifest.

Therefore, the use of scales like this one enables research and monitoring of symptom perception through worker self-report. This aids in identifying workplace improvements, reducing medical costs, and minimizing sick leave. Additionally, it contributes to the organization’s profitability, enhances employee performance, and allows early treatment of symptoms, preventing serious complications.

When analyzing the perception of pain in the upper limbs of workers from the interior of Alagoas and Bahia in a global way, there is the benefit of viewing the professional profile according to regional characteristics, highlighting the particularities of the sample. For this study, there were some limitations: Self-reporting involves personal perception, making it a subjective evaluation; therefore, other types of tests should be explored to increase the reliability of the research. Additionally, due to sample characterization, the results may not extrapolate to other regions or professions.

On the other hand, this study has limitations that can be overcome by future studies. Thus, studies carried out a posteriori can make use of IRT parameter equalization methods to generate separate scales based on different characteristics of the sample, such as gender, professional group, and level of experience, among others. Other studies may analyze the differentiated functioning of items (DFI) by verifying whether individual characteristics of the respondents may compromise the estimation process of the IRT parameters and, consequently, the construction of the musculoskeletal discomfort scale for the upper limbs.

Ethical approval

This study was approved by the Research Ethics Committee of the Federal University of Alagoas (protocol number CAAE 31022420.3.0000.5013).

Informed consent

Informed consent was obtained from all respondents.

Conflict of interest

The authors declare that they have no conflict of interest.

Footnotes

Acknowledgments

None to report.

Funding

None to report.

References

Lee

, Kim

, Jung

, Kim

, Lee

. The association between long working hours and work-related musculoskeletal symptoms of Korean wage workers: data from the fourth Korean working conditions survey (a cross-sectional study). Ann Occup Environ Med. 2018;30(67):December.https://doi.org/10.1186/s40557-018-0278-0.

Nambiema

, Bodin

, Fouquet

, Bertrais

, Stock

, Aublet-Cuvelier

, Descatha

, Evanoff

, Roquelaure

Upper-extremity musculoskeletal disorders: How many cases can be prevented? Estimates from the COSALI cohort. Scand J Work Environ Health. 2020. Doi:10.5271/sjweh.3911.

Brasil. Ministério da Saúde. Quase 39 mil trabalhadores são afastados por LER/DORT em2019. https://www.gov.br/fundacentro/pt-br/assuntos/noticias/noticias/2020/3/(2020).Accessed on July 6, 2021.

Ferreira

, Morais

KCS

, Cirqueira

, Macedo

. Sintomasosteomusculares em professores: uma revisão de literatura. InterScientia. 2015;3(1), 147–162.

Brasil, Ministério da Saúde. Secretaria de Políticas de Saúde Protocolo de investigação, diagnóstico, tratamento e prevenção de lesão por esforços repetitivos/disfunções osteomoleculares relacionados ao trabalho. Brasília: Ministério da Saúde; 2000. Available from: https://bvsms.saude.gov.br/bvs/publicacoes/protocolo_ler.pdf. Accessed on December 20, 2023.

Brasil, Ministério da Saúde. Secretaria de Vigilância em Saúde. Departamento de Vigilância em Saúde Ambiental e Saúde do Trabalhador. Dor relacionada ao trabalho: lesões por esforços repetitivos (LER): distúrbios osteomusculares relacionados ao trabalho (Dort) / Ministério da Saúde, Secretaria de Vigilância em Saúde, Departamento de Vigilância em Saúde Ambiental e Saúde do Trabalhador. - Brasília: Editora do Ministério da Saúde. 2012. Available from: https://bvsms.saude.gov.br/bvs/publicacoes/dor_relacionada_trabalho_ler_dort.pdf. Accessed on July 6, 2021.

Brasil. Ministério da Saúde. LER e DORT são as doenças que mais acometem os trabalhadores, aponta estudo [LER e DORT são as doenças que mais afetam os trabalhadores, diz estudo] https://www.saude.gov.br/noticias/agencia-saude/45404-ler-e-dort-sao-as-doencas-que-mais-acomet-em-os-trabalhadores-aponta-estudo.(2019). Accessed on July 6, 2021

Leite

WKS

, Araujo

AJS

, Silva

, Souza

, Pimentel

, Silva

JMN

, Assis

NLP

, Lemos

ELM

, Oliveira Filho

. New Job Stress Scale: factorial, convergent validity and reliability. Rev Psicol Organ Trab. 2021;21:1463–1472. https://dx-doi-org.web.bisu.edu.cn/10.5935/rpot/2021.2.20608.

Nambiema

, Bertrais

, Bodin

, Fouquet

, Aublet-Cuvelier

, Evanoff

, Descatha

, Roquelaure

. Proportion of upper extremity musculoskeletal disorders attributable to personal and occupational factors: results from the French Pays de la Loire study. BMC Public Health. 2020;20(456). https://doi.org/10.1186/s12889-020-08548-1.

10.

Caruso

. Possible broad impacts of long work hours. Ind Health. 2006;44:531–536. Available from: https://www.jstage.jst.go.jp/article/indhealth/44/4/44_4_531/_pdf/-char/en.Accessed on July 17, 2021. https://doi.org/10.2486/indhealth.44.531.

11.

Brasil. Ministério da Saúde. Secretaria de Vigilância em Saúde. Departamento de Vigilância de Doenças e Agravos não Transmissíveis e Promoção da Saúde. Saúde Brasil 2018: uma análise de situação de saúde e das doenças e agravos crônicos: desafios e perspectivas. Brasília: Ministério da Saúde; 2019. 352p. Available from: https://bvsms.saude.gov.br/bvs/publicacoes/saude_brasil_2018_analise_si-tuacao_saude_doencas_agravos_cronicos_desafios_perspec-tivas.pdf. Accessed on July 6, 2021.

12.

Bispo

LGM

, Moreno

, Silva

GHO

, Albuquerque

NLB

, Silva

JMN

. Risk factors for work-related musculoskeletal disorders: A study in the inner regions of Alagoas and Bahia. Safety Science. 2022;153:1–14.https://doi.org/10.1016/j.ssci.2022.105804.

13.

Instituto Nacional do Seguro Social (INSS). Observatório de Segurança e Saúde no Trabalho. Available from: <https://smartlabbr.org/sst/localidade/2702405?dimensao=perfilCa-sosAcidentes>. Accessed on feb 15, 2022.

14.

, Zhang

, Xu

, Qiao

. Relationship Between the Exposure to Occupation-related Psychosocial and Physical Exertion and Upper Body Musculoskeletal Diseases in Hospital Nurses: A Systematic Review and Meta-analysis. Asian Nurs Res. 2021;15(3):163–173.https://doi.org/10.1016/j.anr.2021.03.003.

15.

Herin

, Paris

, Levant

, Vignaud

, Sobaszek

, Soulat

, ORSOSAgroup (2011) Links between nurses’ , Herin

, Paris

, Levant

, Vignaud

, Sobaszek

, Soulat

; ORSOSA groupLinks between nurses’organisational work environment and upper limb musculoskeletalsymptoms: Independently of effort-reward imbalance! the ORSOSA study. Pain. 2015;152(9):2006–2015. doi: https://doi.org/10.1016/j.pain.2011.04.018.

16.

Leite

WKS

, Silva

, Souza

, Fernandes

JGB

, Colaço

. Risk of WMSDs in monofunctional and multifunctional workers in a Brazilian footwear company. Production. 2017;27(2014):1–15. Available at: https://doi.org/10.1590/0103-6513.231817.

17.

Shi

, MacDermid

, Tang

, Sinden

, Walton

, Grewal

. Confirmatory Factor and Rasch Analyses Support a Revised 14-Item Version of the Organizational, Policies, and Practices (OPP) Scale. J Occup Rehabil. 2017;27(2):258–267.https://doi.org/10.1007/s10926-016-9653-2.

18.

Saltychev

, Vastamäki

, Mattie

, McCormick

, Vastamäki

, Laimi

. Psychometric Properties of the Pain Numeric Rating Scale When Applied to Multiple Body Regions among Professional Musicians. PLoS ONE. 2016;11(9):e0161874.https://doi.org/10.1371/journal.pone.0161874.

19.

Silva

, Silva

JMN

, Silva

LKD

, Lima

TDFM

, Bornia

, Souza

LAH

, Leite

WKS

, Vieira

EMA

. Do men and women have different musculoskeletal symptoms at the same musculoskeletal discomfort level?Ergonomics. 2022;65(11):1486–1508. doi: https://doi.org/10.1080/00140139.2022.2088854.

20.

Andrade

, Tavares

, Valle

Teoria de Resposta ao Item:conceitos e aplicações. São Paulo: AssociaçãoBrasileira de Estatística; 2000.

21.

Sartes

LMA

, Souza-formigoni

MLO

. Avanços na psicometria: DaTeoria Clássica dos Testes à Teoria de Resposta ao Item. Psicol Reflex Crit. 2013;26(2), 241–250.

22.

Silva

JMN

, Gontijo

, Bornia

, Silva

, Leite

WKS

, Vieira

EMA

. Evaluation of musculoskeletal discomfort using item response theory: creation of a scale based on the self-reported pain symptoms. Ergonomics. 2021;64(2):241–252.https://doi.org/10.1080/00140139.2020.1825823.

23.

Araújo

EAC

, Andrade

, Bortolotti

SLV

. Teoria da Resposta ao Item. Revista da Escola de Enfermagem USP. 2009;43(Esp):1000–8.

24.

Silva

, Souza

ERM

, Silva

LKD

, Silva

ATC

, Silva

JMN

. Effects of Multiple Risk Factors on Upper Limb Musculoskeletal Disorders Among Workers in Inner Brazil. WORK. 2022;72(3):885–900. doi: 10.3233/wor-210752.

25.

Silva

, Silva

JMN

, Bispo

LGM

, Souza

DSF

, Serafim

, Torres

MGL

, et al. Construction of a Musculoskeletal Discomfort Scale for the Lower Limbs of Workers: An Analysis Using the Multigroup Item Response Theory. International Journal of Environmental Research and Public Health. 2023;20(7):5307. doi: 10.3390/ijerph20075307.

26.

Kuorinka

, Jonsson

, Kilbom

, Vinterberg

, Biering-Sørensen

, Andersson

, et al. Standardized Nordic questionnaire for theanalysis of musculoskeletal symptoms. Applied Ergonomics. 1987;18(3):233–237. doi: 10.1016/0003-6870(87)90010-x.

27.

Kadefors

, Areskoug

, Dahlman

, Kiibom

, Sperling

, Wikstriim

, et al. An approach to ergonomics evaluation of hand tools. Applied Ergonomics. 1993;24(3):203–211. doi: 10.1016/0003-6870(93)90008-W.

28.

Van der Schatte Olivier

, Van’t Hullenaar

, Ruurda

, Broeders

. Ergonomics, user comfort, and performance in standard and robot-assisted laparoscopic surgery. Surg Endosc. 2009;23(6):1365–71. doi: 10.1007/s00464-008-0184-6.

29.

Zinbarg

, Revelle

, Yovel

, et al. Cronbach’s α,Revelle’s β, and Mcdonald’s ω H: their relationswith each other and two alternative conceptualizations ofreliability. Psychometrika. 2005;70:123–133. doi: 10.1007/s11336-003-0974-7.

30.

Dillon

, Goldstein

Multivariate analysis: Methods and applications. New York (NY): Wiley; 1984.

31.

Hair

, William

, Babin

, Anderson

Análise multivariadade dados. 6^a ed. Porto Alegre: Bookman; 2009.

32.

Samejima

. Estimation of latent ability using a response pattern of graded scores. Psychometrika. 1969;34:1–97. doi: 10.1007/BF03372160.

33.

Borges

JWP

, Moreira

TMM

, Schmitt

, Andrade

, Barbetta

, Souza

ACC

, et al. Measuring the quality of life in hypertension according to Item Response Theory. Revista de Saúde Pública. 2017;51(0). doi: 10.1590/s1518-8787.2017051006845.

34.

Tezza

, Bornia

, Andrade

. Measuring web usability using item response theory: Principles, features and opportunities, Interacting with Computers, 2011;23(2):167–175. doi: 10.1016/j.intcom.2011.02.004.

35.

Chernyshenko

, Stark

, Chan

, Drasgow

, Williams

. Fitting item response theory models to two personality inventories: Issues and insights. Multivariate Behavioral Research. 2001;36(4):523–562. doi: 10.1207/S15327906MBR3604_03.

36.

Silva

JMN

, Gontijo

, Bornia

, Silva

, Leite

WKS

, Vieira

EMA

, et al. Construction of an osteomuscular discomfort scale for the upper region of the body of footwear industry workers. International Journal of Industrial Ergonomics. 2020;80:103048. doi: 10.1016/j.ergon.2020.103048.

37.

Alvarenga

, Santos

TSS

, Andrade

. Item Response Theory-based validation of a short form of the Disordered Eating Attitude Scale (DEAS-s) to a Brazilian sample. Cadernos de Saúde Pública. 2020;36(2):e00169919.doi: 10.1590/0102-311x00169919.

38.

Vincenzi

, Possan

, Andrade

, Pituco

, Santos

, Jasse

. Assessment of environmental sustainability perception through item response theory: A case study in Brazil. Journal of Cleaner Production. 2018;170:1369–1386. doi: 10.1016/j.jclepro.2017.09.217.

39.

Edelen

, Reeve

. Applying item response theory (IRT) modeling to questionnaire development, evaluation, and refinement. Quality of life research. 2007;16(1):5–18. doi: 10.1007/s11136-007-9198-0.

40.

O’Connor

. Comparison of two psychometric scaling methods for ratings of acute musculoskeletal pain. Pain. 2004;110(1-2):488–494. doi: 10.1016/j.pain.2004.04.038.

41.

Gorter

, Fox

, Twisk

. Why item response theory should be used for longitudinal questionnaire data analysis in medical research. BMC Med Res Methodol. 2015;15:55. doi: 10.1186/s12874-015-0050-x.

42.

Kopec

, Badii

, McKenna

, Lima

, Sayre

, Dvorak

;Computerized adaptive testing in back pain: validation of the CAT-5D-QOL, Spine (Phila Pa 1976). 2008;33(12):1384–1390. doi: 10.1097/BRS.0b013e3181732a3b.

43.

Descatha

, Leclerc

, Chastang

, Roquelaure

. Incidence of ulnar nerve entrapment at the elbow in repetitive work. Scand J Work Environ Health. 2004;30(3):232–240. doi: 10.5271/sjweh.784.

44.

Swangnetr

, Kaber

, Phimphasak

, Namkorn

, Saenlee

, Zhu

, et al. The influence of rice plow handle design and whole-body posture on grip force and upper-extremity muscle activation. Ergonomics. 2014;57(10):1526–1535. doi: 10.1080/.2014.934301.

45.

Sakthi Nagaraj

, Jeyapaul

, Mathiyazhagan

. Evaluation of ergonomic working conditions among standing sewing machine operators in Sri Lanka. International Journal of Industrial Ergonomics. 2019;70:70–83. doi: 10.1016/j.ergon.2019.01.006.

46.

Karimi

, Dianat

, Barkhordari

, Yusefzade

, Rohani-Rasaf

. A multicomponent ergonomic intervention involving individual and organisational changes for improving musculoskeletal outcomes and exposure risks among dairy workers. Applied Ergonomics. 2020;88:103–159. doi: 10.1016/j.apergo.2020.103159.

47.

Kayabínar

, Kayabínar

, Önal

, Zengin

, Köse

. The musculoskeletal problems and psychosocial status of teachersgiving online education during the COVID-19 pandemic and preventivetelerehabilitation for musculoskeletal problems. Work. 2021;68(1):33–43. doi: 10.3233/WOR-203357.

48.

Nordander

, Ohlsson

, Åkesson

, Arvidsson

, Balogh

, Hansson

GÅ

, et al. Risk of musculoskeletal disorders among females and males in repetitive/constrained work. Ergonomics. 2009;52(10):1226–39. doi: 10.1080/00140130903056071.

49.

Umar

, Kashif

, Zahid

, Sohail

, Arsh

, Raqib

, et al. The Prevalence of Musculoskeletal Disorders and Work-Station Evaluation in Bank Employees. Physikalische Medizin, Rehabilitationsmedizin, Kurortmedizin. 2019;29(02):99–103. doi: 10.1055/a-0756-9782.

50.

Bovenzi

, Prodi

, Mauro

. A longitudinal study of neck and upper limb musculoskeletal disorders and alternative measures of vibration exposure. International Archives of Occupational and Environmental Health. 2016;89(6):923–33. doi: 10.1007/s00420-016-1131-9.

51.

Guimarães de

LBM

, Ribeiro

JLD

, Renner

. Cost-benefit analysisof a socio-technical intervention in a Brazilian footwear company. Applied Ergonomics. 2012;43(5):948–57. doi: 10.1016/j.apergo.2012.01.003.

52.

Leite

WKDS

, Araújo

AJDS

, da Silva

, de Souza

, Silva

JMND

, Bolis

, et al. Sickness absence from work in the footwear industry: A longitudinal study. Work. 2022;72(2):431–451. doi: 10.3233/WOR-205312.

53.

, Voo

, Maakip

. Psychosocial factors, depression, and musculoskeletal disorders among teachers. BMC Public Health. 2019;19(1):1–10. doi: 10.1186/s12889-019-6553-3.

54.

Kozak

, Wirth

, Verhamme

, Nienhaus

. Musculoskeletal health, work-related risk factors and preventive measures in hairdressing: a scoping review. Journal of Occupational Medicine and Toxicology. 2019;14(1):1–14. doi: 10.1186/s12995-019-0244-y.

55.

Helfenstein Júnior

. Uncommon compressive neuropathies of upper limbs. Best Practice & Research Clinical Rheumatology. 2020;34(3):1–12. doi: 10.1016/j.berh.2020.101516.

56.

Punnett

, Fine

, Keyserling

, Herrin

, Chaffin

. Shoulder disorders and postural stress in automobile assembly work. Scandinavian Journal of Work, Environment & Health. 2000;26(4):283–291. doi: 10.5271/sjweh.544.

57.

Lind

. Pushing and pulling: an assessment tool for occupational health and safety practitioners. International Journal of Occupational Safety and Ergonomics. 2016;24(1):14–26. doi: 10.1080/10803548.2016.1258811.

58.

Chiwaridzo

, Makotore

, Dambi

, Munambah

, Mhlanga

. Work-related musculoskeletal disorders among registered general nurses: a case of a large central hospital in Harare, Zimbabwe. BMC Research Notes. 2018;11:1–7. doi: 10.1186/s13104-018-3412-8.

59.

Tinubu

, Mbada

, Oyeyemi

, Fabunmi

. Work-Related Musculoskeletal Disorders among Nurses in Ibadan, South-west Nigeria: a cross-sectional survey. BMC Musculoskeletal Disorders. 2010;11(1):1–8. doi: 10.1186/1471-2474-11-12.

60.

Andersen

. Risk factors in the onset of neck/shoulder pain in a prospective study of workers in industrial and service companies. Occupational and Environmental Medicine. 2003;60(9):649–654. doi: 10.1136/oem.60.9.649.

61.

Miranda

, Viikari-Juntura

, Heistaro

, Heliövaara

, Riihimäki

. A Population Study on Differences in the Determinants of a Specific Shoulder Disorder versus Non specific Shoulder Pain without Clinical Findings. American Journal of Epidemiology. 2005;161(9):847–55. doi: 10.1093/aje/kwi112.

62.

Wixted

, Shevlin

, O’Sullivan

. Distress and worry as mediators in the relationship between psychosocial risks and upper body musculoskeletal complaints in highly automated manufacturing. Ergonomics. 2018;61(8):1079–1093. doi: 10.1080/00140139.2018.1449253.

63.

ORGANIZAÇÃO DAS NAÇÕES UNIDAS. Gastos com doenças e acidentes do trabalho chegam a R$ 100 bi desde 2012. 2021. Available from: https://brasil.un.org/pt-br/125723-gastos-com-doen%C3%A7as-e-acidentes-do-tra-balho-chegam-r-100-bi-desde-2012#:∼:text=%E2%80%9CEstima%2Dse%20que%20doen%C3%A7as%20e,consi-derando%20o%20PIB%20de%202020.Accessed on December 22, 2023.