Musculoskeletal Risks in Washington State Apple Packing Companies

3.0 STUDY DESIGN AND METHODS

3.1 Background

The Field Group worked with the Washington State Growers League, fruit packing house industry representatives, labor groups, the medical community, equipment vendors, and researchers from other universities starting in 1997 to develop cooperative relationships and to gain support for an occupational health pilot study of the fruit packing industry. In 1997 and 1998 the Field Group staff conducted four walkthroughs of Yakima valley packing houses to learn more about the various jobs and tasks performed during fruit packing. Although the basic packing processes were similar at the companies, several differences between small and large packing houses were recognized during the walkthroughs. Company representatives expressed a need for more information on musculoskeletal hazards, assistance with job task analysis, feedback on appropriate control strategies, and assistance with developing ergonomic programs. In the spring of 1999, packing house industry representatives agreed to support a pilot study of the musculoskeletal hazards in this industry.

3.2 Study Design

The purpose of the pilot study was to obtain a broad overview of musculoskeletal hazards in Washington State apple packing houses. A cross sectional, descriptive survey design was used to characterize musculoskeletal risk factors and reports of WMSD that might be related to selected apple packing house tasks, and the degree to which automation might influence the hazards and symptom reporting. A triangulation strategy was used, which allowed comparison between three different points of reference. The three sources used were: 1) management and workers compensation records of musculoskeletal injuries; 2) on-site observations of job tasks; 3) worker self-reported symptoms and perceptions of risk. One unique aspect of this project was the collaboration between industrial hygienists and an anthropologist to explore worker symptoms and risk perception. All study procedures, including subject consent, met the requirements of University of Washington Human Subjects Review Committee.

3.3 Recruitment

3.3.1 Company Selection

The Washington Growers League (League) distributed a recruitment flyer (Appendix A) describing the pilot study to its members and asked interested companies to contact the League directly. The League forwarded the names of interested companies and their contact people to the Field Group. Five companies contacted the League within a two-week period. The Field Group conducted a brief telephone survey with the company contact person to collect information on the company’s packing operation, number of workers, types of work related injuries, and availability to participate in the study. Four companies operated single packing houses and one company operated two packing houses in the Yakima area. All five companies engaged in packing apples; three packing houses also packed pears and one also packed cherries. Two companies operated one shift per day and three companies operated with at least two shifts during part of the year. All five companies packaged fruit throughout the year and all companies participated in the Washington State Workers Compensation Program. Not all companies were available to participate in a study during the spring of 1999. Three companies were invited to participate in the study based on the following eligibility criteria:

1. the company used either manual or automated packing processes or both
2. the company was located in the Yakima valley
3. the primary language of the workers was either English or Spanish
4. the company was available to participate in a study during April and May 1999.

Participating companies were offered as an inducement a company-specific final report identifying and discussing musculoskeletal hazards observed in their company and possible solutions for identified concerns.

3.3.2 Worker Selection

Six different job titles within each company were targeted for evaluation based on the following criteria:

1. repetitive tasks
2. job was performed regularly during most of a work shift
3. large number of workers had the job title.

The job titles selected for inclusion in the study were sorter, manual tray packer, semiautomatic tray packer, manual bag packer, semi-automatic bag packer, and segregator. Workers in these six job titles were invited to participate in the study at an informational meeting held at their place of work; the meeting was endorsed by company management and held during work time. The informational meeting was conducted in English and Spanish. Workers interested in participating approached the Field Group researchers after the informational meeting. All workers in the designated job titles were eligible to participate; however, due to study resource constraints participation was limited to 10 subjects per job title per company. Companies 1 and 2 were small, so all workers in the designated job titles were encouraged to participate. Company 3 had a large workforce consequently 10 full time workers were randomly selected from each of the six job titles. Workers completed an informed consent form prior to data collection. All worker participants received a small monetary remuneration acknowledging their participation in the study.

3.4 Exposure Measures

3.4.1 Workstation Design Evaluation and Videotaping

The layout of each workstation was sketched and videotaped and the dimensions of the process equipment were measured. The height and width of each conveyor system and the weight of a full box of apples, a tray, and a bag were measured in each job category. The ambient room temperature and lighting levels were also measured and recorded. Nonstationary equipment at each workstation, such as packing horse carts, was also assessed for height and adjustability for range of worker heights. Sorting, grading, and packing lines of each company were videotaped on the day of on-site assessment to assist in analyzing the work processes and to review job tasks. Permission was obtained from workers prior to videotaping.

3.4.2 Job Task and Task Cycle Time Observations

Observational methods are commonly used in ergonomic field assessments to quantify the number and type of awkward postures, repetitive movements, and other musculoskeletal hazards (Pinzke, 1997). For the purposes of observation each of the six-selected job titles was broken down into one to four specific tasks that defined a job cycle. Designated tasks were used to modify a version of the Ergonomic Surveillance Checklist developed by the Safety and Health Assessment Research for Prevention (SHARP, 1999) program, a division of Washington State Department of Labor and Industries (Appendix B). The checklist enumerated 38 potential risk factors related to forceful hand exertions, awkward postures, contact stress, and general musculoskeletal hazards. Twenty of the 38 risk factors had left and right components (e.g. hand, wrist, forearm, and elbow factors) that were listed separately. An operating protocol for the use of the assessment tool was also developed. Four analysts (three industrial hygienists and one ergonomist) received one day of training on the use of the assessment tool, including definition of all risk factors on the checklist and a review of tasks associated with each fruit packing job title. Analysts also reviewed videotape of the workers performing tasks in the six job titles prior to using the assessment tool.

Time-motion study of the jobs was conducted during on-site observation. During the on-site observations a minimum of 10 and a maximum of 20 job cycles were recorded by each analyst as time allowed. Each job cycle was observed to identify and note checklist risk factors and to time the job cycle. The average cycle time of each job and the percentage of the work-shift the observed risk factors were present were calculated by multiplying the percent of time the risk factor was observed by the percent of time a given task occurred in a cycle. The findings for all observed risk factors are found in Appendix E. Two assumptions were made with regard to the calculations: 1) the risk factor observed during the sample time was present the entire time, and 2) a work shift was 480 minutes. It was not feasible to perform an interanalyst statistical test for reliability because analysts were not observing the same workers at the same time.

The four analysts observed four to six job titles at each company. At Company 1, workers were engaged in all six jobs, while at Companies 2 and 3, only four jobs were done. Each job cycle contained one to four tasks. Each analyst worked to observe each job up to 20 times. This target was not achieved for all observed jobs due to low production levels on the day of observation and inadequate time to complete observations during one shift. Table 1 presents the targeted and obtained observations for each company.

Table 1 Analyst Observations

Job Title	Number of Job Tasks	Targeted # of Observations Per Company			Obtained # of Observations Per Company
		1	2	3	1	2	3
Sorter	1	80	80	80	80	80	80
Manual Tray Packer	4	320	320	*	140	260	*
Semi-Automatic Tray Packer	3	240	*	240	115	*	240
Manual Bagger	4	320	320	*	156	137	*
Semi-Automatic Bag Packer	2	160	*	160	160	*	160
Segregator	2	160	160	160	148	144	160

3.4.3 Lifting Hazard Analysis

The segregators who move filled boxes to pallets for shipping or storage engage in tasks that require considerable lifting. To assess the degree of risk to the segregators, two computer programs that predict lifting requirements were used: the NIOSH lifting model (NIOSH, 1991) and the University of Michigan 3D Static Strength Model (University of Michigan, 1993). The NIOSH lifting model was used to evaluate the segregator job for the potential for injury to the lower back. This model considers the frequency and distance of lifts and weight of the load. The University of Michigan 3D Static Strength Model was used to evaluate the forces and stresses on the upper body during lifts that occur above shoulder height. Despite the fact that the 3D model may underestimate risks from frequent lifting, it was deemed appropriate for analysis of segregating because the job included awkward back and upper body postures. Analyses were conducted for a 95th percentile male and a 50th percentile male lifting a 45 to 50 pound box.

Using the NIOSH lifting equation, the Lifting Index (LI) and Recommended Weight Limit (RWL) were calculated for segregators. The RWL is the object weight that would, most likely, not produce lower back discomfort for most workers. The Lifting Index (LI) is a ratio of lifted object weight to the RWL. A LI of one indicated that the lifted object weight equals the RWL. As the LI increases, the risk of low back injury increases. Only 1 percent of female and 25 percent of the male population could perform a lifting task with a LI of three without risk of injury. The LI is used to compare the relative severity of lifting risk for the purpose of evaluating and redesigning those jobs. For this study, the RWL was calculated by using measurements taken of segregators loading full apple boxes onto pallets. Measurements included the height of a stack, placement distance of the box, frequency of lifts, height of the conveyor, the angle of twist, and length of time spent lifting during the shift.

3.5 Outcome Measures

3.5.1 Injury Reports

The OSHA 200 Log information for calendar years 1994 to 1998 was collected from each participating company. Employers are required by federal law to report, via the OSHA 200 Log, all work-related injuries involving restricted or transferred work, lost time, or medical treatment requiring more than first aid. The number of reported musculoskeletal injuries was determined by summarizing all injuries and excluding all traumatic injuries including slips, trips, falls, or being caught in or struck by objects. Injuries were further sorted by body site injured.

Workers compensation incidence rates were also obtained for each participating company from the Department of Labor and Industries Public Disclosure Unit, as was the industry wide rate for fruit and vegetable packing, risk class number 2104-02.

3.5.2 Subject Interviews

Information about study participant demographics, chronic health symptoms, and perceptions of risk was obtained via structured interviews. The interview format was adapted from a questionnaire developed by the SHARP program and included questions on work history, general health, and self-reported symptoms (Appendix C). In addition, questions about worker perception of work-related risks were added to the questionnaire. Risk perception questions were developed in collaboration with Karen Snyder, Doctoral Candidate in Anthropology at the University of Washington. Trained bilingual research staff members conducted interviews during regular work hours in either English or Spanish. Each interview took approximately 45 minutes.

3.5.2.1 Work History and General Health

The work history section of the questionnaire included questions about the participant’s current warehouse job, past warehouse jobs, the number of years working in a fruit packing house, and the number of hours typically worked each week. Information about general health was also obtained with particular emphasis on conditions known or suspected to be related to WMSD, such as arthritis, joint problems, chronic disorders such as diabetes, smoking history, hobbies, and medication use.

3.5.2.2 Self-Reported Symptoms

This section of the questionnaire related to chronic pain or discomfort, particularly in neck, shoulder, elbow/ forearm, hand/wrist, back, hip, knee, and foot/ankle. Chronic symptoms were defined as those experienced during the preceding 12 months. The symptom location, onset, frequency, duration, severity, medical treatment, and amount of lost work time was recorded for each body part as well as activities that aggravated symptoms. A problem symptom was considered work related if it occurred at least once a week or lasted one week or more, did not start as the result of an acute trauma, occurred in the last year, and was first noticed on the current job.

3.5.2.3 Risk Perception

The risk perception questions were derived from previous research on risk perception of farmers and farm workers (Arcury 1995;Vaughan 1993) and consisted of three open-ended and two Likert-scale questions. The open-ended questions related to specific packing house tasks, actions an individual could take to avoid injury, and things a company could do to prevent injury. The interviewers were encouraged to fully record participant responses to open-ended questions. The Likert-scale questions (5 choices – “not likely” to “completely likely”) related to effectiveness of safety precautions and likelihood of experiencing a a work related injury or illness.

3.5.3 Across Shift Body Discomfort

Acute changes in pain or discomfort were assessed pre-and post-shift with the aid of a diagram of the human figure or body map (Appendix D). Body maps have been used to describe areas of localized pain, discomfort, or fatigue (Borg, 1990; Corlett and Bishop, 1976) and provide an opportunity to characterize symptoms across several muscle groups. The body map used for this study was adapted from one developed by the SHARP program. Prior to starting the work shift, each study participant was asked to record the severity of any discomfort in 12 body sites. Severity was ranked on a scale of 1 to 5, with 5 indicating the most severe discomfort. At the end of the work shift, each subject completed a second body map. For example, a person might rate the neck as a 2 (mild discomfort) at the beginning of the shift and a 3 (moderate discomfort) at the end of the shift.workrelated injury or illness.

The pre- and post-shift discomfort severity for each participant was calculated by subtracting the pre-shift score from the post-shift score. The across shift change for all body sites was calculated by summing the pre- to post-shift change for each of the 12 body sites and dividing by the number of sites scored. Each of the 12 body site specific scores and the all-body site score were averaged for all participants in each of the six job titles. In addition, the body discomfort map was used as one measure to assess chronic discomfort by reviewing the number of subjects reporting high discomfort (a score of 4 or 5) at the start of the shift.

Back to Table of Contents

---