Contents of this paper:

Introduction
Handwriting as an    example
Skill and timing
Other mechanisrns of    strain and injury
Lifting weights
Lifting within limited constraints
Lifting: perception of    task demands
Choice of lifting strategy
Optimal solutions to    lifting outcomes
Coping with chaos in    lifting
Acquisition and    exercise of skilled    activity
Conclusions
References

Introduction

Upper limb pain in office workers and back pain from lifting are the two commonest types of musculoskeletal strain at work today. Existing medical and ergonomics approaches have failed to solve these problems. We propose that such pains and injuries sometimes arise from unskilful methods of work, not considered in traditional models of ergonomics or medicine. Lack of skill in such situations may cause the use of excess force, poor timing, or poor preparation for movement, and can take several forms:

1. Unnecessary force applied to the task, for example pressing too hard with a pen; or hitting too hard with a hammer or onto a keyboard.
2. Excessively forceful grip, for example holding a handle more tightly than is needed.
3. Co-contraction, or excessive tension in two groups of muscles which oppose one another. A simple example is a tense but empty hand. (Sometimes "co-contraction" is applied to the normal simultaneous contraction of synergist muscles, but this is not the meaning used here.)
4. Abrupt application of force instead of smooth increase during a movement, for example hitting with a hammer in a jerky way or lifting without taking advantage of the momentum of the body, or an unexpected load from a slip or fall.
5. Poor choice of body posture, for example position of the body in relation to loads, failure to adjust seating correctly, and working with wrists severely abducted or dorsiflexed.
6. Poor choice of equipment or material, so that more force is required for performance of the task.

Often it is easy to solve these problems by simple demonstration and explanation.

Handwriting as an example

Pains in the hand and arm have been associated with handwriting for hundreds of years, and many theories of causes and methods of cure have been proposed (Sheehy and Marsden, 1982). These include "focal dystonia" (spastic grip and jerky movements), relearning to write using shoulder movements, and psychoanalysis. Published descriptions and photographs (Sarkari et al., 1976) suggest that widely different phenomena have been studied, different from the so-called RSI epidemic seen in Australia during the 1980's.

These studies considered 133 subjects with symptoms attributed to repetitive work, so-called RSI (repetitive strain injury) or OOS (occupational overuse syndrome). Ten of these attributed hand pains to their work. They confirmed earlier observations (Patkin, 1984) that "trying too hard" was the major factor responsible. The excess effort had separate components of grip, pressing, and co-contraction, and could be demonstrated by palpation of muscle, indentation of writing paper resting on cloth, and electromyography using simple equipment.

Associated factors included use of ball-point pens with a poor or unpredictable flow of ink, multiple copies of documents, poor teaching of children at school, and tense habits during other activities. These observations are complemented by diagrams in old textbooks of handwriting which show much more relaxed hand-grips. Pressing too hard with traditional types of pen is unnecessary for legible writing, and causes damage to the point. Eight of the ten patients in the series achieved quick and complete recovery by attention to these factors.

In computer terminal operators an extra factor is traction on the brachial plexus caused by holding the neck forward instead of upright. This causes tingling or discomfort in the forearm due to irritation of the nerves supplying skin in the same area as underlying tense muscles. Brachial Plexus Tension Tests (BPTT) (Elvey, 1979) are similar to tests for sciatic nerve irritation, though their validity has been disputed. A poked forward position of the head may he due to habit, poor legibility of the screen, or reading-glasses with the usual focal length of 30 em instead of twice that distance to suit office work. Here, skill is a good strategy for choice of posture rather than muscle tension.

Skill and timing

During a task, force can be applied too quickly or too slowly. When pushing a child's swing, for example, it takes little force to gradually increase the amplitude of the swing and its combined kinetic and potential energy, provided that force is not applied too quickly, or out of phase with the movement of the swing.

Case example: A clerk working poorly was transferred to chipping scale inside a blast furnace undergoing overhaul. He was clumsy using a chipping hammer as well as angry at his change of job.

Within a few weeks he developed forearm pain. On examining his style of hammering, lie had short jerky strokes, accelerating and stopping the movements of the hammer too quickly.

A similar example is chopping wood, where the tense novice is advised to "let the axe do the work". Using a sledge hammer to drive heavy nails to secure railway lines, the skilled worker will use the recoil or bounce of the hammer to reduce the amount of muscle work needed to raise the hammer again for the next stroke.

Good timing can be defined as "the rate of application of force which maximizes tile rate at which energy is transferred from one part of a system to another", and is another way of describing resonance, or coupling.
Poor timing transfers energy to the structures of tile body rather than the external object. Strains and tears of muscles and tendons occur when their tensile strength is exceeded. This depends on tile size of tile loud, its direction, the mechanical advantage or disadvantage of tile applied force, and tile rate at which it is applied, since biological tissues have a relatively slow rate of stretch and creep apart from tile contractility of muscle. Such strain may cause mild ache or serious tearing injury.

Other mechanisrns of strain and injury

Two cases of lateral epicondylitis occurred in cleaners. They used a heavy flour-polishing machine and developed their injury pulling (lie polisher up a step while keeping the elbow bent. Their supervisor however kept the elbows straight, using her body weight to counterbalance the weight of the machine. This allowed tile extensor muscles of the forearm to stay relaxed. By contrast, flexion of tile elbows was accompanied by strong contraction of the forearm extensor muscles just below their attachment to tile lateral epicondyle of each elbow.

Cleaners using floor polishers unskilfully have a second risk factor. The rotating head of tile polisher is kept slightly tilted so that it moves first to one side and then to the other. If the tilt is altered too quickly at the end of each transverse movement, the operator has to "fight" the polisher and grip it more fiercely, also straining the origin of the forearm extensor muscles as in tile common circumstances of "tennis elbow".

Lifting weights

Injury may arise not only from errors of technique but also from poor preparation, planning, practice, or experience. We propose that there are important differences between tile skills of competitive weightlifting and the process of lifting loads in the workplace. These include physical conditioning and mental training, and the control of variables.

Expert weightlifters use similar patterns of movement, as shown by observation, and analysis of videotapes and ENIG recordings. However they also show individual characteristics, comparable to differences in handwritten signatures, speech, or choice of words. It becomes possible to consider a particular motion "signature" revealed in kinetic energy, or a "grammar" of movement. Common factors in the patterns of movement of weightlifters include:

1. Preparatory movements (power movements in sport are often preceded by movements in tile opposite direction, or backswing).
2. The sequence in which different groups of muscles become active.
3. Reproducible intra- and inter-weightlifter force-time patterns but with sufficient discrimination to identify an individual lifter. (These are shown with force platforms.)

In contrast to weightlifters, Gormley, Sedgwick and Smith (1989) found that warehouse workers and "ordinary" people used techniques and strategies in lifting that varied to meet changing situational demands and as the task proceeded. Subjects did not operate in the invariant manner of a highly skilled weightlifter or machine. Also they did not perform lifting actions in the single plane of motion often assumed in biomeclranical models and normative standards. These observations support the notion that lifting is a complex set of behaviours regulated according to the changing circumstances of the environment, loads, performer status and the actual performance of the task as well as by the original intent underlying the action. Lifting requires tile operator in the process to code, order, and time the actions according to tile syntax and lexical requirements of the lift. The grammar of lifting is acquired through experience and practice in a similar manner to language acquisition. Attempts to teach lifting as though it is merely a mechanical/biomechanical problem have not been successful. Yet such an approach to competitive weightlifting has been found to be highly successful. Why?

Lifting within limited constraints

The highly constrained conditions of competitive lifts performed with particular styles or techniques allows the development of invariance in the strategies and actions used. Competitive weightlifters learn the codes to produce the actions required to overcome the kinetic and topological constraints through progressive practice and repetition. When a sufficient level of mastery has been achieved the 'lifting action sequence' can be performed at will with only minor modification being needed as the loading demands increase. The action sequence is hierarchically organized and executed with the kinematic and kinetic features becoming increasingly invariant the higher the efficiency of tile lifter. Enoka (1988) showed in a comparison of skilled and less skilled Olympic weightlifters that success of the more skilful performers was due to their ability to both generate a sufficient magnitude of joint power and to organise the phases of power production and absorption into an appropriate temporal sequence rather than through a mere quantitative scaling of power production.

Lifting: perception of task demands

Lifting as it occurs in the garden, on the building site, in tile hospital ward and the office is not as constrained as in tile sport of weightlifting. The person must respond. to the particular circumstances according to their perceptions of and experience in meeting the task demands. Often the situation can be quite nova+ In the lifting required of warehouse workers, the operator has to deal with a complex dynamic set of relationships. The lift actions selected involve assessment of the environmental and task demands, action planning and estimation of risks attached to particular decisions made udder temporal, spatial and kinetic limitations associated with success/failure probability predictions.

Choice of lifting strategy

As reported by Gormley, Sedgwick and Smith (1989), the actions selected as appropriate by warehouse workers are often less than optimal in terms of mechanical effectiveness, safety and control. Yet this study found that the task goal was achieved in the minimum time, with loads moved over minimum distances at low levels of perceived effort. The strategy often chosen was to minimize effort rather than maximize safety and control. The importance of experience and knowledge in effectively using such a strategy is supported by the results of Patterson et al. (1987). The study investigated the effects of load knowledge on the stresses at tile lower back during lifting. Loads were lifted under conditions of verbal or visual knowledge of load magnitude or no load knowledge. Results showed experienced lifters had lower stress levels at L4/L5 and utilized two technique strategies that were dependent upon
the load knowledge conditions, whereas the non-lifters used the same strategy for all lifts. Maximum moment values were significantly higher for the inexperienced lifters
under all conditions. Experienced lifters were able to distribute tile load more effectively between body segments.

Optimal solutions to lifting outcomes

Lifting, except under the most constrained circumstances is a process in which the solution outcome call be achieved in a variety of equally successful ways. Optimal solutions vary from task to task and from individual to individual as well as between and within performances of a particular lifting task. Individuals should be recognised as leaving divergent intra- and inter-individual abilities, knowledge and capacity to undertake tile perceptual, cognitive and physical challenges inherent in tile lifting process. Teaching of so-called 'safe' or 'correct' lifting methods fails to recognize the indeterminacy of the problems confronting those who have to produce the actions.

Coping with chaos in lifting
.
Space does not allow the role of co-contraction in skilled actions to be considered in detail here, but strain and injury are obviously more likely when appropriate synergist muscle groups fail to be recruited at a rate to match the demands of tile task. Examples of this are hamstring injury ill sprinters, and errors in estimating the weight of an object.

Epidemiological evidence (Magora, 1973; Andersson, 1981) suggests that workers exposed to sudden unexpected loads are particularly vulnerable to low-back problems. Troup et al. (1981) suggested that slipping accidents were often due to sudden jerking or twisting actions. Ergonomic adjustment of the environment and the loads goes only part of the way towards reducing the risk of such events occurring. Intra- and inter-person variability cannot be avoided, and slakes the prediction of such events notoriously difficult and unreliable. Since variability is part of tile lifting process it follows that education and training programmes should focus oil teaching tile individual to cope with variation and that adaptability of response rattler than single modes of correct technique is the desirable outcome. Such approaches will still riot snake it possible to predict with high probability estimates when accidents will occur and who the victims may be.

Acquisition and exercise of skilled activity

The three-stage model of skill acquisition developed by Glencross (1977) has been adapted to hand movements in surgery and micro-surgery (Patkin, 1988), with three stages described as coding, modelling, and timing (the last corresponding to the hierarchical level of skill of Glencross). Errors during skilled activities can be considered as slips or mistakes, according to Reason's (1987) Generic Error-Modelling System. This approach call be combined with Rasmussen's (198(1) distinction between skill-based, rule-based and knowledge-based levels of performance to describe error types in a three by five matrix. Tuning allows smoother and more accurate actions, placed in tile rule-based performance category.

Conclusions

Data on pain and skill related to activities as diverse as handwriting and lifting weights provide good evidence that some strains and injuries result from poor skills. Given their widespread nature and high cost in tire workplace, it is important to apply existing knowledge to reduce these problems by teaching improved motor skills and to identify the sources of process-based errors in skilled performance at work.

References

Andersson, G.B. (1981) Epidemiologic aspects of low back pain in industry. Spine, 6,53-60.
Elvey, R.L. (1979) Brachial plexus tension tests for the pathoanatornical origin of arm pain. In Aspects of Manipulative Therapy, Lincoln Institute of Health Sciences, Melbourne, Australia, 105-110.
Enoka, R.M. (1988) Load and skill-related changes in segmental contributions to a weightlifting movement. Medicine and Science in Sports and Exercise, 20(2), 178-187. '
Glencross, D.J. (1977) The control of skilled movements. Psychological Bulletin, 84, 14-29.
Gormley, J.T., Sedgwick, A,W. and Smith, D.S. (1989) Effective, safe and controlled lifting. A mechanistic or process approach to training'! Paper, Annual Conference of the Australian Council of Rehabilitation Medicine, Sydney, March 1989. Magora, A. (1973) Investigation of the relation between low back pain and occupation. 4. Physical requirements: bending, rotation, reaching and sudden maximal effort. Scandinavian Journal of Rehabilitation Medicine, 5, 191-196. Patkin, M. (1984) "Trying too hard". An aspect of overuse injury. Proceedings, 21st Annual Conference, Ergonomics Society of Australia and New Zealand, 289-301. Patkin. M. (1988) Hand and arm pain in office workers. Modern Medicine of Australia, 31(10), 66-76.
Patterson, P., Congleton, J., Koppa, R. and Iluchingson, R.D., (1987) The effects of load knowledge on stresses at the lower back during lifting. Ergonomics, 30(3), ,539-549.
Rasmussen, J. (1980) What can he learned from human error reports? In J. Duncan, M. Gruneberg and D. Wallis (editors), Changes in Working Life, Wiley, London. Reason, J. (1987) Generic error modelling system (GEMS): A cognitive framework for locating common human error forms. In J. Rasmussen, K. Duncan and J. Leplat (editors), New Technology and Human Error, John Wiley, London.
Sarkari, N.B.S., Mahendrn, R.K., Singh, S.S. and Rishi, R.P. (1976) An epidemiological and neuropsycliiatric study of writers cramp. Journal of the Association of Physicians of India, 24, 587-591.
Sheehy, M.P. and Marsden C.D. (1982) Writers' Cramp - a focal dystonia. Brain, 105, 461-490.
Troup, J.D., Martin, J.W. and Lloyd, E.F. (1981) Back pain in industry: A prospective survey. Spine, 6, 61-69.

-o0o-