Abstract
Introduction
Musculoskeletal disorders (MSDs) are often referred to as physical and biochemical insults to individual or groups of muscles, nerves, tendons and ligaments, or to any other body parts that support the back, neck or limbs [1]. Swelling and varying levels of discomfort are the major consequences of MSDs that profoundly affect one’s normal routine. MSDs are relatively widespread among individuals who experience substantial bio-mechanical loads, like those in factory jobs and manufacturing, but also among those who do not experience such loads, like office workers [1–3]. Despite the lack of biomechanical forces, the development of MSDs is frequently noticed among individuals with a sedentary working and personal life style, particularly if they have inadequate amounts of physical activity as compared to ‘active’ individuals. In line with this, the incidence of MSDs among office workers has recently become higher in this large section of the workforce [4].
Detecting and monitoring MSDs within a specific population demands accurate and reliable methods of measurement. To date, qualitative and semi-quantitative measurements, namely questionnaires and utilization of instruments like the goniometer to gauge the range of motion, are commonly used to assess MSDs [5, 6]. The questionnaire method is preferred in epidemiological research [7], where large numbers of participants are involved, and it is also cost effective [8]. However, lack of measurement reliability with regard to the Cornell questionnaire, Borg scale and goniometer [9, 10] have been cited. In the present study, the reliability of the Cornell questionnaire (musculoskeletal disorder questionnaire), the Borg scale (perceived exertion) and goniometer (measure the range of motion) were evaluated and compared, since these are the tools most commonly utilized by occupational therapists. We hypothesize that there is a high reliability for the Borg scale, the Cornell questionnaire and the goniometer and that these are three reliable instruments to measure the severity of pain among office workers.
Methods
Sample size
The sample sizes were based on “Sample Size Tables for Clinical Studies” [11], in order to provide a supposed ICC = 0.75 (I), a desired CI = 95% (confidence interval) and 25% for drop-out rate. Accordingly, 120 subjects were investigated.
Participants
One hundred twenty healthy office workers, body mass: 87.1 ± 10.3 (kg), age: 27 ± 5.1 (years); height: 1.78 ± 0.16 (m) (mean ± SD), who had at least 1 year of experience in office working were chosen randomly.
All participants were employees who worked in governmental offices and their native tongue was Malay. They were working in Malaysia during January and February, 2015. In total, 1205 volunteers participated in this research. To take part, they must possess the following characteristics: Be adults (male/female) who [1] were 22–32 years old, [2] were without any disabilities or problems that would limit the performance of simple physical actions, and [3] possessed a level of knowledge/understanding that would enable them to answer the questionnaires. Before any physical tests were performed by a physician, participants completed the questionnaires, and those who had any long-term medical problems were omitted from the study, in order to prevent risk during exercise testing [12]. All procedures were performed according to the Helsinki Declaration of 1964. In addition, all subjects provided written, informed consent before participating in the research. The study was approved by Department of Occupational Health’s Ethic Committee, University Putra Malaysia, Malaysia.
Data collection
Participants were approached before work time. Training was given by the researchers regarding how to answer the questionnaires correctly.
On the first day, the Borg questionnaire was given to the participants, and they were asked to not do anything other than their daily routines during the data collection process. The time allotted for answering the questionnaire was 15 minutes. The subjects were then recalled after 8 days, in order to fill out the Borg questionnaire a second time, this second result being used to measure the questionnaire’s reliability [13]. The same process was adopted for the Cornell questionnaire. On all occasions, a researcher was available to answer any questions. In addition, the ranges of motion of the neck (bending to right and left), shoulder (abduction and adduction, extension and flexion), hip (extension and flexion) and knee (extension and flexion) were measured by a goniometer (12-inch plastic goniometer, Lafayette Instrument Company, USA), this taking about one hour. The same examiner was used on both occasions, in order to evaluate the method’s accuracy using a test-retest experimental design [14].
Reliability
To measure the internal reliability and consistency of the Borg questionnaire, subjects filled out the questionnaire twice under the supervision of a certificated physiotherapist. The interval between the two tests was 7–10 days, as recommended in previous research. The Pearson correlation coefficient was used to measure the reliability of the test, with separate evaluations for the neck (rotation and bending), shoulder (abduction, elevation and flexion) and lower back (bending, rotation, flexion and extension) [15]. To measure the internal reliability and consistency of the Cornell and Borg questionnaires, a Pearson correlation coefficient was used, with separate assessments of frequency, severity and interference. To measure the internal reliability and consistency of the goniometer, subjects laid prone on an examination table (Winco 8570, SW MED-Source, USA) with 30" height. Values were re-evaluated after one hour, to assess the method’s accuracy in a test-retest experimental design. The ICC coefficient was used to assess the reliability of the test, with separate evaluations of the range of motion in the neck (bending to right and left), shoulder (abduction and adduction, extension and flexion), hip (extension and flexion) and knee (extension and flexion) [14, 16]. All of the measurements were performed by a certificated physiotherapy and under the supervision of a specialized orthopedist.
Statistical analysis
Statistical analysis was with SPSS (Statistical Package for Social Sciences) software for Windows (Version of 21.0), with the level of significance being set at P <0.05. The ICC (inter-class correlation) and Pearson correlation coefficient were used to analyze test–retest reliability in measured variables, and additional assessments used the single-measure ICC (one-way random effects model), which signifies the percentage of total variance that can be accounted for by variation between the two times ofmeasurement.
Results
Borg scale reliability (8 days Test–Retest)
The comparison between the results of the Borg scale (with 8 days between measurements) showed a high level of correlation for all items: Flexion, extension, bending and rotation of the lower back; elevation, abduction and flexion of the shoulder; and bending and rotation of the neck (Table 1).
Goniometer reliability (One-hour Test–Retest)
Goniometry was done for the hip, hamstring, shoulder and neck, when the participants were lying prone on an examination table. All the measurements were performed by a certificated physiotherapist under the supervision of a researcher. The first measurement was done in the morning around 10 am and the second measurement about 1 hour later, the subjects being in the same position (Table 2).
Cornell questionnaire (8 days Test–Retest)
The questionnaire was filled out by the participants under the supervision of a researcher. The first time it was answered around 10 am and, on the second occasion, at the same time but after a delay of 8 days (Table 3).
Discussion
All participants were volunteers, and they were representative of the general population with regard to their occupational and educational backgrounds [17]. Unlike in the present study, in most previous measurements of musculoskeletal disorders, the reliabilities of the Borg scale, goniometer and Cornell questionnaire – the most commonly used instruments - were not measured. In this study, the items of the Borg questionnaire focused on the lower back (contains 6 elements), shoulder (3 elements) and neck section (4 elements). In similar research done in 2014 by Markus et al., the reliability of these scales was measured among blue collar workers [15].
The results of the present study using the Borg scale showed that it reliably measures perceived exertion among office workers. The ICC range coefficients was (0.805–0.954, p<0.001). This result supports our hypothesis and is also in line with previous research findings [18, 19]. It means that this scale can be used to measure the severity of pain among this part of the population. This scale has previously been used mostly among athletes, but here we used it to measure similar variables in office workers. We used an 8 day delay between the test and re-test in order to measure the reliability of the questionnaire. Previous research has used a 2-week delay only between test and re test, but our preference was for a delay of 8 days [20]. Use of less than a one-week delay might mean that subjects remember their previous answers, and this is undesirable in such an experimental protocol. A long delay (more than 10 days) is not suitable, since individuals might be experiencing new and different types of pain.
The results from this study relating to the Cornell questionnaire support our hypothesis and show that it is a reliable instrument for measuring the severity of pain. The ICC range of coefficients was (0.883–0.975, p <0.001), which is a high and acceptable range. This questionnaire focused on the neck, hip, wrist, arm, lower back, shoulder, thigh and knee, and contains 17 elements. The last modifier of this questionnaire, Professor Allen Hedge, used this questionnaire mostly for measuring musculoskeletal disorders among office workers, and also measured the reliability and validity of this questionnaire [21]. However, he measured the reliability of the questionnaire only with regard to pain severity. The delay time between the test and re-test was the same as that used for the Borg scale (8 days). The results of the present study are in line with those found previously [4].
The goniometer results showed a high reliability in measuring the range of motion in the hamstrings (2 elements), hip (2 elements), shoulder (4 elements) and neck (2 elements) among office workers. The ICC range coefficients were (0.785–0.978, p <0.001), acceptable values. The measurements for the test and re-test components were performed with a 1-hour delay, and by a physiotherapist under the supervision of specialized orthopedists. The results showed that a manual goniometer is a reliable instrument to measure ranges of motion, findings that were in line with previous research [14, 16].
Having a sound means for data collecting is extremely important to the performance of valid assessments in ergonomics studies. The present results indicate that the methods used give reliable results, and so the methods can be utilized effectively to study office workers who suffer from pain.
Conclusion
The implication of this research is that the CMDQ is a sufficiently reliable and valid method for evaluating different areas of musculoskeletal discomfort with different severities of pain. This test is suitable for use in medical studies where the initial focus of the research relates to the levels, frequency and duration of pain and musculoskeletal discomfort. Nonetheless, the potential for over-estimating problems remains high in a self-assessment context, so specific attention should be paid to the methods used to categorize people as active or inactive in a clinical context. Additional analysis and validation may be needed in the future for the CMDQ to be considered as a totally validated instrument for data collection in other studies.
Limitations
The findings of this research are limited by the small number of subjects and may not have enough generalizability, as the assessments were only done among office workers in Malaysia. So, for the future researches, it is suggested to do the same process among a larger population of office workers in different locations.
Conflict of interest
The authors declared no potential conflicts of interests with respect to the research, authorship, and/or publication of this article.
Funding
There was no grant or funding for this research and it was done by the personal budget of researchers.
Footnotes
Acknowledgments
The authors thank the staff at the Telecom MalaysiaTM for their participation and cooperation.