2.5.1. Description of work activities
The major work activities commonly observed in small and medium-scale cashew nut processing mills are boiling, shelling, peeling, and grading (Azam-Ali & Judge, 2001;
Mohod et al., 2010). All of these tasks are manually performed by workers. Figure 2.1 shows the work activities assessed in the present study. Boiling is about heating the raw cashew nut using a steam cooker where workers carry and lift cashew bags. This task is commonly carried out on raised shoulders and with bending trunk. The primary purpose of shelling is to crack the roasted cashew nuts, which is performed using conventional cashew nut shelling equipment. This activity involves prolonged standing, bending, and twisting the trunk for long hours. Peeling activity is about removing the thin layer on the cashew kernels, and grading is to sort the kernels by size. In these two activities, workers commonly adopt different floor-sitting postures, including sitting with cross-legs, squatting, and legs folded at knees.
Figure 2.1 Typical work activities of small and medium-scale cashew nut processing mills: (a) boiling, (b) shelling, (c) peeling, (d) grading.
2.5.2. Study design
This study was conducted across four major cashew processing states (i.e., Andhra Pradesh, Assam, Meghalaya, and Orissa) of eastern India. According to the (CEPCI,
(a) (b)
(c) (d)
2019), these states nearly produce 30.07 % of total cashew in India. In these states, a total of sixteen mills (i.e., Andhra Pradesh (6), Assam (1), Meghalaya (1), and Orissa (8)) were visited. As mentioned earlier, four major processing activities, namely boiling, shelling, peeling, and grading, were examined during the study of WMSD risk.
The cross-sectional study design was carried out with 290 randomly selected cashew workers from (eastern) India. The inclusion criteria were: workers with work experience of 12 months in the current job, age greater than 18 years, and giving their individuals’ consent to participate in the study. The workers reporting pain due to injuries, accidents, or any other diseases were excluded. All the participants were informed about the purpose of the study in layman's terms. Participation was voluntary.
2.5.3. Instruments
A questionnaire survey was carried out to collect details related to cashew workers in small and medium-scale cashew nut processing mills. The questionnaire consisted of three parts, as shown in Appendix I. The first part of the questionnaire collected data on demographic details (gender, age, height, weight, marital status, education level, work activity, and smoking habit). In the second part, work-related details such as work experience, daily working hours, the perceived speed of work, perceived work fatigue, job satisfaction, current health status, and stress level were collected. The third part of the questionnaire included the standard Nordic Musculoskeletal Questionnaire (NMQ) (Kuorinka et al., 1987). Using NMQ, participants were interviewed about the prevalence of WMSDs during the past 12 months. A body map indicating the nine different body regions (i.e., neck, shoulders, elbows, hands/wrists, upper back, lower back, hips/thighs, knees, and ankles/feet) was shown to the participants (Appendix I). Subsequently, they were asked to identify those body parts that they had experienced any pain or discomfort in the past 12 months. A 5-point scale (1= very mild pain to 5 = very severe pain) pain rating scale was used to record the severity of pain. Additionally, any disruption of normal activities due to WMSDs was also recorded.
The working postures of cashew workers were assessed, and it was observed that the whole body of workers is involved in most of the tasks. Therefore, the
analysis of different working postures of cashew workers was performed using the Rapid Entire Body Assessment (REBA) method (Hignett & McAtamney, 2000) for quantifying the risk of WMSDs. According to this method, a special scoring method is employed based on the range of motion of body parts. The body parts are divided into two groups as group A (neck – trunk – legs) and group B (shoulders – elbows – wrists). Load coupling and static/dynamic nature of activity parameters are also included. All the scores are combined to give a final or grand REBA scores and five action levels, as shown in Figure 2.2. Based on the deviation of body posture from the neutral position, the final REBA score increases, and the checklist can be seen in Appendix II.
Figure 2.2 REBA scores and corresponding action levels.
During the study, working posture of cashew workers was photographed and video recorded. The most engaging and frequent work postures observed in all the activities were taken into consideration. Later on, stick diagrams were prepared, and the REBA employee assessment sheet (Appendix II) was used to determine scores, risk levels, and action levels for each work activity. REBA scores and risk levels were categorized activity-wise: boiling, shelling, peeling, and grading.
2.5.4. Statistical analysis
Statistical analysis of the data was performed using SPSS v20.0 (IBM Corp., Armonk, NY, USA). Kolmogorov-Smirnov test was used to check for normality of data.
Demographic and work-related characteristics of the study population were reported as Mean (Standard Deviation), frequencies, and percentages (%). Multivariate logistic regression analysis (backward stepwise) was performed to predict the influence of risk
factors on occurrence of WMSDs among different body regions. According to Hair, Tatham, Anderson, & Black (2006), multivariate logistic regression is a well-known statistical modelling method for understanding relationships between independent and dependent variables. In the present study, the dependent variables were neck, shoulders, elbows, hands/wrists, upper back, lower back, knees, and ankles/feet disorders. Independent variables include demographic and work-related characteristics. The odds ratios (ODs), p-value, and 95 % confidence intervals (95 % CIs) were used for the description of association between the prevalence of WMSDs and study variables. The assumptions of logistic regression models (presence of outliers and collinearity) were checked, and the model’s fit was checked by the Hosmer-Lemeshow goodness-of-fit test. P <0.05 was considered statistically significant.