Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
1 Summaty of Findings INTRODUCTION Background Video display terminals (VDTs)~ are used in a broad range of occupations (e.g., clerical work, printing, computer work, air traffic control), and their use in offices is growing rapidly. The number of VOT operators in the United States was estimated to be approximately 7 million in 1980, with ~10 million VDTs in use (Center for Disease Control, 1980~. Workers and union representatives around the world have expressed concern that harmful effects may result from working with VDTs (see, e.g., Bergman, 1980; New York Committee for Occupational Safety and Health, 1980; Working Women, National Association of Office Workers, 1980; DeMatteo et al., 1981; Canadian Labour Congress, 1982~. Much of this concern has involved visual functions, human factors, 2 radiation, and tVDTs are devices for visually displaying (with symbols, graphics, or both) information that is stored and processed electronically. Keyboards are commonly used to control the processing and display of information. Most VDTs now commercially available use cathode-ray tubes (CRTs) and are similar to television receivers in their display characteristics; however, VDTs that use solid-state display devices instead of CRTs are increasingly coming into use. 2Human factors are characteristics of peoplefor example, size, shape, ability to see and hear, strength, and mental capacities-- that should be considered in the design of equipment and socio- technical systems. The effects of design variables on human performance are studied in the field of human factors in order to develop and apply principles to improve the effectiveness, 5
6 psychosocial aspects of VDT-related work. A number of surveys of VDT operators have reported a high incidence of complaints of visual and musculoskeletal discomfort. In several studies in which various visual functions have been measured, they have been reported to be temporarily altered following work at VDTs. Many workers and labor representatives have expressed concern that VDTs may emit harmful levels of radiation that may cause cataracts and other adverse health effects (Working Women, National Association of Office Workers, 1980; DeMatteo et al., 1981; Canadian Labour Congress, 1982~. There have been anecdotal reports that clusters of VDT operators have had spontaneous abortions and miscarriages and have given birth to children with birth defects (microwave News, 1981~. Reports of skin rashes among VDT operators have recently appeared (W. C. Olsen, 1981; Nilsen, 1982~. Some types of clerical jobs in which VDTs are used have been characterized by some labor represen- tatives as being more stressful than clerical jobs performed using traditional technologies (Working Women, National Association of Office Workers, 1980; Canadian Labour Congress, 1982~. Focus of the Study This report primarily concerns issues involving vision and the visual system. However, because factors that affect operator comfort and performance cannot be elucidated by analyzing only the optical characteristics of VDTs, relevant human factors and psychosocial issues are also considered. Because much of the concern about the possibility of radiation hazards has been based on misinformation, we analyze the results of surveys in which the levels of radiation have been measured and compare those levels with ambient levels of radiation emitted by human-made and natural sources and with current standards for occupational exposure. We did not reopen the question of what is an acceptable level of radiation exposure, a question that has been extensively studied and was beyond our mandate. We discuss whether there is evidence that ocular diseases or abnormalities, including cata- racts, are associated with VDT-related work (see Chapters 3 and 7~. We discuss only briefly the possibility of disorders that do not involve vision (i.e., effects on pregnancy and skin rashes; see efficiency, safety, and comfort of people who use machines. In Europe this field is referred to as ergonomics and somewhat greater emphasis is given to biomechanics and physiological aspects of work than in the United States. The two terms are used interchangeably in this report.
7 Chapter 3) because there are few published data and because we lack the appropriate expertise. Organization of the Report This chapter is intended to provide a nontechnical review of issues and a summary of the panel's findings. The following chapters provide a more extensive analysis of technical issues and litera- ture. Chapter 2 analyzes issues concerning the methodologies encountered in field studies in which VDT workers were asked to respond to questions about visual and other complaints. Short- comings in the methodologies of these studies make it difficult to draw firm conclusions about the factors underlying visual com- plaints and symptoms of workers. The first section of Chapter 3 reviews studies of radiation emissions from VDTs and compares the level of emissions with current occupational standards and background radiation from natural and human-made sources. The second section reviews concerns about cataracts and discusses epidemiological issues. Chapter 4 evaluates what is known about the relationship between specific characteristics of display devices and observers' visual performance, subjective responses, and physiological responses. Chapter 5 analyzes the problems VDT workers sometimes experience with improper workstation lighting and reflections, and Chapter 6 examines the ways in which the comfort and performance of workers are affected by constraints on posture and motion imposed by the physical layout of workstations. Chapter 7 explores what is known about the causes of ocular discomfort and difficulties sometimes reported with vision and discusses the limited efforts that have been made to compare visual tasks in jobs that involve VDTs and jobs that do not. Chapter 8 discusses the influence of job design and organiza- tional factors on the well-being of VDT workers. Chapter 9 presents principles of good design and practice that could alleviate problems encountered in VDT work and discusses the feasibility of standards for VDT design. Last, Chapter 10 discusses research needs. The Literature Base The literature related to visual effects of VDTs is growing rapidly. The number of articles published per year went from ' 1 in 1972 to 43 in 1980 (Matula, 1981~. This literature, however, has done little to answer the questions that have been raised. Only a dozen or so formal studies of visual complaints or changes in visual function among VDT workers have been published, and most
8 fail to meet major criteria for acceptable scientific research. The remaining literature on visual effects consists mainly of nontechnical reviews of the concerns of- workers, technical discussions of standards for VDT and workstation design, and handbooks for workstation design. In contrast to the literature on visual effects of VDTs, there is a substantial technical literature on the quality of visual displays, the effective design of lighting, and human factors. We are not aware of any formal studies of job design in VDT-related work. The Nature of VDT Work This study considers only VDTs used for the display of alpha- numeric information; it does not consider graphic displays (e.g., air traffic control scopes, radar scopes). Nevertheless, the kinds of jobs in which alphanumeric VDTs are used must be at least in the hundreds and include both clerical and professional occu- pations. These jobs, of course, differ greatly on many dimensions: in the function of the VDT within the job as a whole, in the amount of time the worker spends on tasks in which the VDT is directly involved, in the visual tasks required, etc. It seems likely that the nature and incidence of visual and other problems would vary greatly among diverse VDT jobs; thus, generalizations should be made with caution. Unfortunately, there has been no formal analysis of task characteristics in various VDT jobs, and there is no ready clas- sification scheme for such jobs. To illustrate the diversity among jobs in which VDTs are used, we can characterize some jobs by a predominant mode of interaction with the VDT (see Table 1.1~. In data entry work, information that is usually noncontextual (numbers, letters, or symbols) is keyed into the computer, often in a repetitive manner according to a set format. In many cases the data have no intrinsic meaning, especially when specialized sym- bols are used. The work pace in data entry is often quite high-- 8,000-12,000 keystrokes/in is not unusual (Grandjean, 1980--and VDT operators may be expected to meet production quotas. Operators may read from printed or handwritten material or use auditory sources. In many cases the task does not require that the operator often look at the video screen. Operators in jobs that primarily involve data entry work usually have little or no control over the structure of their work. Data acquisition involves calling up information from the computer and reading it from the screen; it is thus more screen- intensive (attention is directed primarily to the screen) than data entry work. Telephone information operators often work pre- dominantly in this mode.
9 o ~ A, ~ icy ~ ~ at; a, v, o To v In ._ o - ._ - ~n ._ o o In J m . O O ~0 o - U' I: ,0 ~ U. ._ - Ct cq 5 - a~ 2 ~~ US PA ~ as o V, o of an Cal V U. Ct .O Cal Cal ~ a, Ct ~ C)'~ U. ~ ~ a, ~ =: ." ._ >- ~ m° ~ 2 oo ._ o t;! Ch C) ~ C) V, ~ ~ ~ ~ Cal U. o C) U. U. ~ r ~ ' C,) o V: ._ := a' o ~ ~ ~ =, t: C: ~ C >, C - C~ C.) ~ C~ 0 C ~ V~ V~ (Q V CQ a' - ._ - - ~ ~3 2 2 ~ ° ._ .° C~ ~ C~ _ . _ ~ ~ ~ o C~ C<, _ ~ ~ _ _ ~_ _ $.: _ _ ._ _ _ ~ 3 3 ._ o o =: ~ ~ ' O.~) U: ._ C~ V oo o o 3 ~ v
10 Interactive communication (sometimes called conversational) work involves both data entry and data acquisition. The data may be more complicated than those involved in data entry jobs, and the task is likely to be more screen-intensive. To some degree an operator sustains a dialogue with the computer and has some opportunity for decision making. Airline reservation clerks are an example of workers who seem to work predominantly in this mode. Word processing involves text entry, text recall, searching text for errors, keying in corrections, and organizing format. The term is often used to refer to secretarial tasks in document preparation, but there are similar operations in such job tasks as layout, formatting, proofreading, and editing. Some of the task elements are source-document-intensive, some are screen-intensive, and word processing jobs usually involve different combinations of these elements at different times. There is wide variation among these jobs in the degree of control an operator may have over the structure and pace of work. Programming, computer-assisted design (CAD), and computer- assisted manufacturing (CAM) involve some aspect of program- ming computers using VDTs. Many professional jobs for example, data analysis, computer programming, scientific research--include such use of VDTs. In these jobs the VDT may be only one of several tools used, and the amount of time a worker spends at a terminal often varies greatly from day to day. A worker's control over the job task is considerable. Obviously many jobs have elements of more than one of these categories, and some jobs may not fit into any of them. The comfort, satisfaction, and performance of VDT workers are affected by interacting factors that range from optical to psychosocial (see Figure 1.1~. Unfortunately, the existing literature on the effects of VDTs has done little to distinguish the relative contributions of these factors. And VDT jobs have not been systematically analyzed and compared with non-VDT jobs. Furthermore, many jobs have been substantially altered by the introduction of VDTs; and it is difficult to determine from existing data whether reported visual problems and other concerns of workers result from the new technology itself or the way in which it is being introduced. FIELD STUDIES OF VDT WORKERS AND WORKSTATIONS Studies of Radiation Emission from VDTs Video display terminals are designed to emit visible radiation (light), but in the process of producing visible light small amounts of several other types of electromagnetic radiation are also
11 Job Design Working Environment Physical Characteristics Physical Characteristics Furniture and equipment design Lighting (illumination, glare, temperature, reflectance) Humidity Noise Human Operator Physiological status (age, pathology, visual functions) Performance capabilities Needs and values Job Design Task demands (acuity, speed, vigilance, workload) Rewards Control of job structure and pace Interpersonal interaction 1 · r Human | Thor VDT Screen Keyboard VDT Screen Regeneration Luminance (symbols/ background ) Contrast Image quality Phosphor characteristics Colors Presence of filter Symbol characteristics (font, size) Angle of screen with regard to operator Location Screen separate or attached to keyboard FIGURE 1.1 Some interacting factors in VAT jobs. Job performance Job satisfaction Strains (physiological and psychological ) VDT Keyboard Layout Type and spacing of keys Force/travel/size of keys Angle and height of keyboard Color (e.g., to separate function) Grouping of keys generated, particularly X radiation and radio frequency radiation in the 15-125 kHz frequency range. In response to concerns expressed by VDT operators and labor representatives that radiation emitted by VDTs might be harmful, field surveys and laboratory studies of radiation emissions from VDTs have been conducted over the last several years by govern- ment agencies in the United States and Europe and by private organizations and independent groups (see references in Chapter 3~. Taken collectively, these studies have examined a wide variety of models and hundreds of terminals. Measurements of emissions from older and newer VDT models have not differed significantly. Measurements have been made both under normal operating conditions and under conditions designed to maximize
12 potential emissions (by using maximum contrast and screen brightness, filling the screen completely with characters, using high line voltage, misadjusting service and user controls, causing component failures, etc.~. These studies have concluded that the levels of all types of electromagnetic radiation emitted are below existing occupational and environmental health and safety stan- dard limits of exposure. The levels of radiation measured in these studies have generally been orders of magnitude below occupa- tional exposure standards.3 In one study (Bureau of Radiological Health, 1981) VDT sets were tested under conditions designed to maximize the emission of X radiation by combining artificially induced worst~ase com- ponent failures and misadjustment of user and service controls. Under those conditions, 8 of the 125 sets tested exceeded the 0.5 mR/h standard for television receivers, although under normal conditions no X-radiation emissions were detected from any of the sets. Those eight sets represented three models that were subsequently recalled by the manufacturers to be redesigned for compliance with standards or were excluded from the U.S. market. It is useful to compare the levels of radiation emitted by VDTs to ambient levels of radiation emitted by natural and human-made sources. A person is exposed to greater radiation levels in all parts of the electromagnetic spectrum from ambient sources than from a VDT. The level of ultraviolet (UV) radiation emitted by VDTs has been found to be far lower than that emitted by ordi- nary fluorescent lights and thousands of times lower than outdoor (sunlight) UV levels. Emissions of visible and infrared radiation from VDTs are less than 1 percent of outdoor levels. Radio fre- quency radiation is emitted from VDTs at levels comparable to ambient levels generated by radio transmitters in metropolitan areas. The level of X radiation emitted by VDTs is far less than the ambient background level of ionizing radiation from natural sources (i.e., cosmic radiation, terrestrial radiation, and internal radionuclides) to which the general population is exposed. Standards for occupational exposure to radiation are based on existing knowledge of both acute and long-term biological effects, and they take into consideration the cumulative exposure of workers to various human-made and natural sources of radiation. There is an enormous literature on the biological effects of radiation (reviews are cited in Chapter 3~; we found no evidence to suggest that levels of radiation emitted from VDTs might produce harmful effects. As noted above, we did not attempt in 3In some cases the instruments used to measure some forms of radiation were not sensitive enough to measure emissions substantially below the standard.
13 this study to evaluate the appropriateness of these standards, an issue with both technical and policy aspects. Neither did we attempt to evaluate issues concerning thresholds for biological effects of low levels of various forms of radiation to which humans are commonly exposed. Cataracts Concern about the possibility of radiation hazards from VDTs has been raised in part by anecdotal reports of cataracts (opacities of the lens of the eye) occurring among some VDT workers. Exposure to high levels of ionizing or microwave radiation is known to cause cataracts, and there is some evidence that chronic exposure to high levels of ultraviolet radiation may also cause cataracts. Both laboratory studies of animals and surveys of humans indicate, however, that the levels of radiation required to produce cataracts are thousands to millions of times higher than the levels emitted by VDTs (see Chapter 3~. Some VDT workers will, of course, develop cataracts, since cataracts occur throughout the general population. The causes of most cataracts are not known. Small, inconsequential opacities of the lens are common; as many as 25 percent of normal people may have such congenital or developmental opacities that do not affect vision. Opacities that substantially interfere with vision are much less common but increase in prevalence with age. Some mild opacities may be precursors of senile cataract. There have been no well-designed studies suggesting an associ- ation of VDT work with cataracts or other ocular abnormalities (see Chapter 3~. We cannot adequately assess anecdotal claims of cataracts resulting from VDT work because data sufficient to document the claims have not been published. We found no scien- tifically valid evidence to support the assertion that cataracts with characteristics of those caused by radiation exposure result from VDT work. The ten anecdotal reported cases of cataracts among VDT workers do not suggest an unusual pattern attributable to VDT work: six of the cases appear to be common, minor opacities not interfering with vision, and each of the remaining four cases had known, preexisting pathology or exposure to cataractogenic agents. Two pilot epidemiological studies that include analyses for cataract were underway at the time this report was written. The National Institute for Occupational Safety and Health (NIOSH) has recently completed a study of VDT workers at the Baltimore Sun (Smith et al., 1982~. Preliminary results of this study indicate that the small size and self-selected nature of the study population preclude any assessment of a relationship between VDT use and
14 the development of cataracts (see Chapter 3 and Appendix B). The Mt. Sinai School of Medicine is conducting a larger study for the Newspaper Guild; however, we were unable to obtain sufficient detail about the study design to evaluate it adequately.. The sample size required for useful epidemiological studies of VDT workers depends on the nature of the cataract in question. If careful ocular examinations revealed a specific, unusual form of cataract in a substantial proportion of workers, samples of the size used in the Baltimore Sun or Mt. Sinai studies might be adequate. But larger sample sizes would be required to detect a small increase in prevalence of common cataracts among VDT workers. Exposure to very high levels of some forms of radiation (millions of times higher than levels emitted by VDTs) produces cataracts of characteristic appearance. Similar cataracts (specifically, posterior capsular and cortical cataracts) may, however, occur idiopathically. The weight of available evidence indicates that an association between VDT work and the development of cataracts is highly improbable. Thus, unless contrary evidence is produced by pilot studies now under way, we believe that large-scale epidemio- logical studies of cataracts among VDT workers are not now justified. Field Surveys Based on Self-Reports of VDT Operators Several studies have been published in which surveys of VDT operator complaints were reported. The findings and conclusions of these surveys and of several experimental studies have been widely cited, especially in nontechnical articles, as evidence that VDT work causes visual problems. Our review of field surveys (see Chapter 2) indicates that existing studies have not established whether VDT work per se produces more visual complaints than comparable non-VDT work. Neither have those studies established the causal factors underlying the complaints of workers regarding visual difficulties. The methods used in field studies have been heterogeneous (see Table 2.1 in Chapter 2~. The surveys have generally used both health questionnaires, with questions on ocular, musculoskeletal, and other physical complaints, and psychological questionnaires with items on psychological states, job satisfaction, and job characteristics. Some surveys have included measurements of Results from the Mt. Sinai study were not available when our report was completed.
15 visual status, such as acuity and phoria, and some have recorded information on workplace conditions, such as lighting levels and postural constraints of workstations. The prevalence of visual and musculoskeletal complaints reported by VDT operators has varied greatly among surveys, probably because of the diverse methods used and differences in the populations studied. In some studies, more than one-half of the VDT workers complained of some degree of visual discom- fort. When comparison groups have been used, the percentages of non-VDT workers reporting the same symptoms have generally`` been lower. In most surveys that have used comparison groups, however, the VDT and non-VDT groups have not been matched to control for differences (e.g., in demographic characteristics, workstation design, job design) other than the use of VDTs. Appropriate multivariate statistical procedures have seldom been used in VDT studies. Several of the studies have other flaws in method (such as low response rates or potential bias in selection of respondents) that severely limit the possibility of interpreting apparent differences. Thus it is not possible to determine from existing studies to what extent complaints reported by VDT operators have resulted from the VDT itself as opposed to such factors as workstation or job design. Video display image characteristics, workstation features such as ambient lighting, and the design of VDT jobs may all affect the visual comfort of workers as well as their performance, job satisfaction, and levels of job-related stress. Extensive, well- designed research would be required to determine the relative contributions of these interacting factors. A careful analysis of job and workstation characteristics (see Chapters 7 and 8) should precede any attempt to design field surveys; in this way appro- priate controls can be selected and appropriate questionnaire items can be designed. We suggest that if future surveys are conducted, they should be designed to compare explicitly the relative influences on worker complaints of interacting variables such as job and employee characteristics, workstation design, and display image characteristics. EQUIPMENT AND WORKSTATION DESIGN VDT Design and Display Quality Although well~esigned video displays are available, many displays in commercial use employ components similar to those in home television receivers, which can be inexpensively manufactured and purchased. These displays are not specifically designed for pro- longed work by operators performing close visual inspection of
16 static alphanumeric characters, often under stressful conditions, in poorly designed working environments. Poor display quality hampers visual performance and probably contributes to the annoyance and discomfort sometimes reported by workers. Much is known about what is required for good~quality displays (see Chapter 4~. Visual performance is affected by a number of display parameters, such as character size, structure, and style, and by image contrast and stability. Television-type VDTs pro- vide less than optimal image quality on several of these param- eters. For example, television-type displays generally use medium-short persistence phosphors and a high refresh rate to prevent blurring of moving images. However, these character- istics can cause noticeable flickering of stationary images in VDT applications. Some operators find this flickering annoying, and it may, therefore, lead to reduced performance. Tradeoffs are required in the choice of some display param- eters. For example, positive contrast (light characters on a dark background) can help reduce the flicker sensitivity of the eye, but negative contrast (dark characters on a light background) helps reduce the effects of veiling reflections on the display screen and may help reduce problems with quickly adapting to the different luminance levels of the VDT screen and surrounding objects (see discussion in the next section and in Chapter 5~.5 Techniques for measuring display parameters and evaluating image quality are available and can be used, for example, to give a reproducible measure of the sharpness and clarity of displayed images. Unfortunately, manufacturers have used such techniques only on a limited basis, and they currently use diverse and gener- ally not very useful ways of describing the display characteristics of their products. Because measures of display quality are neither standardized nor offered in manufacturers' specifications, it is extremely difficult for either buyers or manufacturers to compare the quality of different products. It is sometimes difficult for buyers to make informed choices because they are often unaware that data on display quality are available. sin this report we follow the U.S. convention of calling light characters on dark background positive contrast and calling dark characters on light background negative contrast. (This is different from the European convention of calling light characters on dark background negative presentation and calling dark characters on light background positive presentation.)
17 Lighting and Reflections The results of several field surveys indicate that many VDT operators have reported annoyance with general workplace lighting, glare, and images reflected by the VDT screen (see Chapter 5), and some of the same operators reported ocular discomfort or visual impairment (blurring or flickering of vision and double images) that they attributed to VDT work. However, there has been only fragmentary effort to relate specific work- station lighting conditions quantitatively to the comfort and performance of VDT operators. It is not clear from existing data whether VDT workers on the whole have more problems with lighting than non-VDT workers in comparable job situations. Lighting problems in VDT workplaces are in many ways similar to those encountered in non-VDT workplaces, but some special problems are presented by the VDT (e.g., reflections from the VDT screen). Surveys indicate that lighting conditions in VDT workplaces often do not conform to good illuminating engineering practice. Many problems related to lighting in VDT workplaces have been caused by the introduction of VDTs into offices in which the lighting was originally designed for traditional desk-top work. The design of most VDTs creates new geometrical relationships between working surfaces and light sources and, unless appro- priate modifications in workplace lighting are made, operators may experience problems with glare, images reflected by the VDT screen, and reductions in visibility of the display image. For example, the lighting in most offices is designed on the assump- tion that workers will perform tasks requiring their line of sight to be depressed 20°-40° from the horizontal. VDT operators, whose line of sight must be at or near horizontal to view the screen, are likely to experience discomfort due to glare caused by this elevation, which brings their point of fixation closer to ceiling luminaires that can act as glare sources. Because a VDT itself is a light source, operators may encounter difficulty in successively viewing a VDT screen and other direct or indirect light sources having luminances much different from that of the screen. For example, if an operator looks toward a window or luminaire and then looks back toward the screen, several difficulties may occur: discomfort may be caused by the large differences in luminance between the screen and the window or luminaire, and the visibility of the display image may be reduced for several seconds as the visual system adapts from the high luminance of the window or luminaire to the much lower luminance of the VDT screen. This transient adapta- tion effect may be particularly important when a positive- contrast display (light characters on a dark background) is used.
18 Losses in visibility due to transient adaptation may also occur when the operator successively views the VDT screen and an indirect light source, such as the source document. The loss in visibility in this case would be greatest when a positive-contrast display is used with a negative-contrast (dark characters on light background) source document, such as a typewritten page. The same type of discomfort and visibility loss may occur when secondary task lighting (e.g., a desk lamp) is used to illuminate the source document. Reduced visibility of display images can also be caused by scattering of light within the eye, which reduces contrast at the retina, or by specular reflections that produce a veil of light (called reflected glare) over the display image. Scattering of light within the eye tends to occur more frequently with increasing age because of age-related changes in the optical media. Light reflected from the mirrorlike front surface of the VDT screen forms apparent images of nearby or distant objects, such as keyboards, desk tops, or walls. Reflected images of windows or luminaires can produce a veil of light over portions of the screen, reducing contrast and visibility of the display characters. Because reflected images appear to the eye to be located behind the screen, rather than at its surface, the accommodative and con- vergence systems may fluctuate between reflected images and the display image, resulting in an intermittent or constant blur of the display characters. No well-designed studies have attempted to relate measures of DOT operator performance to subjective reports of problems with glare and reflections, but studies of lighting problems in non-VDT tasks suggest that they may lead to performance decrements. Studies of the effects of reflected glare on performance in non- VDT tasks have shown that even when the visual discomfort produced by reflected glare is slight, reflected images of the glare source may lead to decreases in performance because they are distracting or annoying. If the reflected image is to the side of and much brighter than the display image, it may elicit a phototropic fixation response in which an operator's eyes move away from the display image toward the reflected image. Photo- tropic fixation responses can cause a loss in visibility of the display image through transient adaptation effects. Reflected images have also been shown to cause problems with binocular rivalry and binocular fusion in non-VDT tasks. Differences in task characteristics in VDT jobs might also affect the incidence of complaints related to workplace lighting. For example, operators who spend a large proportion of time viewing the screen (e.g., data acquisition operators) may experi- ence more difficulty with reflected images and screen glare than do data entry operators whose job is typically less screen-inten-
19 sive. Data entry operators, however, may experience more difficulty with luminance differences among the source document, screen, and background. No well-designed studies have examined incidence of complaints as a function of differences in task characteristics. Filters can be placed over VDT screens to reduce glare and reflections to some degree. Because their effectiveness is limited, filters should be considered only as a supplement, never as a replacement, for control of light and reflecting sources through proper lighting design. Several types of filters are available, with different levels of effectiveness, at prices ranging from a few dollars to more than $100. Filters are often used without adequate understanding of the trade-offs involved in their use. Some filters, for example, only slightly reduce glare while substantially reducing character image quality or luminance; the net effect of such filters may be to reduce rather than enhance worker comfort and performance. Human Factors Several field surveys have reported that many VDT operators experience job-related muscular discomfort (see Chapter 6~. Most surveys have been based on subjective reports, but some studies have also included medical observations, measurements of work- station dimensions, or both. Approximately one-half of the surveys have compared the incidence of muscular discomfort in VDT operators with that in workers in non-VDT jobs. Some studies have also compared the incidence of discomfort in specific parts of the body in VDT and non-VDT workers. The results of studies have been conflicting; some have found that VDT operators report more discomfort overall or more in specific parts of the body than do non-VDT workers, and some have found the reverse. Conflicting results appear to be due, in part, to deficiencies in the designs of most studies. The results of one study (Smith et al., 1980; National Institute for Occupational Safety and Health, 1981 suggest that operators in VDT jobs that are characterized by high pressure to perform and low control over the task report more visual and muscular discomfort than do operators whose jobs allow greater autonomy and flexibility. Because of the design of the study, however, its results must be interpreted cautiously (see the discussion of this study in Table 2.1 in Chapter 2 and in Chapter 6~. Although we cannot draw firm conclusions about the com- parative types and incidences of jo~related muscular discomfort in VDT and non-VDT workers, the results of studies indicate that many VDT operators do experience significant discomfort. It is
20 likely that this discomfort is largely caused by inappropriate workstation design. Video display terminals are often designed and introduced into offices without the application of relevant human factors design principles. In many instances poorly designed VDTs are simply installed at desks formerly used for traditional office work or placed on whatever furniture happens to be available. Operators are often required to work in cramped spaces that leave them little room to place document holders or manuscripts in positions that allow comfortable working postures. Operators in such situations are likely to experience visual dis- comfort, muscular discomfort, and fatigue. The physical design and arrangement of workstation com- ponents have implications for both visual and postural task requirements. Working at a VDT places a combination of interacting demands on the human visual and musculoskeletal systems that differs from that in traditional office work. The design of many VDTs creates a number of relatively inflexible fixation points for various parts of the body. For example, the keyboard and the screen in many poorly designed VDTs are permanently attached and the angle of the screen is fixed: this configuration allows the operator to assume only a limited number of working postures. If the operator places the VDT at a com- fortable viewing distance that allows the display to be easily read, the keyboard may then be at a distance that requires the operator to hold his or her hands, wrists, and arms in uncomfortable positions. This may be particularly difficult for operators with presbyopia (the reduction of visual accommodative power with age, causing the near point of focus to recede). The optical correction for near work routinely provided presbyopic people is likely to be inappropriate for the distances at which VDT screens are usually viewed. Multifocal lenses may not be designed to allow a person to view the screen through the segment for near work without tilting the head at an uncomfortable angle. (The problem is analogous to that of a presbyope with bifocals trying to read labels on grocery store shelves that are above eye level.) Unless multifocal lenses are designed specifically for VDT work, there may be no strategy that a worker can adopt to obtain clear vision without postural discomfort. Of course, this statement applies to many non-VDT jobs and situations. Static muscle load and consequent postural stress, discomfort, and fatigue created by relatively fixed postures can often be relieved simply by moving the body around. VDT workstations should be designed so that operators can easily change work postures. VDTs that have detachable keyboards, screens that can be tilted to a comfortable viewing angle, and movable document holders allow operators to change postures and aid in preventing postural stress and discomfort. Appropriate supports, such as
21 armrests and wristrests, can also help reduce static muscle load and discomfort. Although there has been little research on the effects of using adjustable chairs and work tables, their use seems desirable, and there is a great deal of anthropometric data that can be used to design adjustable furniture for different populations. Several manufacturers have recently begun offering adjustable furniture for use with VDTs. Although some of this furniture is well designed, the claims of some manufacturers that their equipment is designed using principles of ergonomics or human factors do not stand up to scientific scrutiny. Older workers, because of visual changes such as presbyopia and increased glare susceptibility' may be especially vulnerable to problems of poor VDT workstation design. This issue is of special concern in part because the workforce is expected to become distributed toward older ages in the coming decades. Of course, good workstation design facilitates the comfort and performance of all workers. THE CONCEPT AND STUDY OF "VISUAL FATIGUE" Surveys of VDT operators have reported that complaints of ocular discomfort and difficulties with vision are fairly common. The complaints have included irritantlike effects (itching, dry, gritty, stinging, or watery eyes), sensations of pain or fatigue involving the eyes, and blurring or other difficulties with vision. Although these surveys suggest that complaints of ocular discomfort are more frequent among VDT workers than among non-VDT workers, it cannot be determined from these studies whether the com- plaints are related to the VDT itself or to other aspects of the job situation, including workstation, lighting, and job design (see Chapters 2 and 7~. The ocular symptoms reported by VDT workers appear to us to be similar to those reported to clinicians by many people of all ages and many occupations. There has been little effort, how- ever, to interview workers in depth to obtain a detailed charac- terization of reported symptoms or comparisons of symptoms among VDT and non-VDT workers. It would be useful to know which aspects of visual tasks might contribute to the experience of ocular discomfort sometimes reported by VDT workers and non-VDT workers and how this experience might be affected by the nonvisual features of a work situation. Unfortunately, neither existing VDT studies nor the general scientific literature provides answers to these questions, and the physiological and psychological bases of ocular discomfort cannot be specified.
22 Ocular complaints of workers have often been discussed in terms of "eyestrain" and "visual fatigue." These terms, however, are vaguely defined and do not correspond to known physiological or clinical conditions.6 Confusion may result because different discussions of visual fatigue may refer to quite different phenom- ena: symptoms of ocular discomfort; changes in oculomotor functions, such as accommodation and vergence;7 or changes in performance of visual tasks, such as reading or visual search. Several investigators have sought physiological correlates of visual fatigue in VDT workers (see Chapter 7~. The hypothesis of most of these studies apparently was that fatigue of oculomotor muscles might underlie sensations of ocular discomfort. The reported effects of several hours of VDT work include transient changes in the near points of accommodation and convergence, resting point of accommodation, so-called accuracy of accom- modative response, time required to shift eye fixation and focus between near and far targets, and visual acuity. All of the studies suffer from flaws in method that make their results difficult to interpret (see Chapter 7 and Appendix A). In particular, some studies have not included non-VDT control groups, and others have used non-VDT so called control groups that differed from the VDT group not only in the use of VDTs but also in many other respects. It is not possible to determine from these studies whether the reported oculomotor changes were specifically related to VDT visual tasks. A few studies have used appropriate controls but suffer from other problems of method or interpretation and can at best be considered preliminary investigations. The oculomotor changes reported to follow VDT work are consistent with a larger body of research in which such changes are commonly found following periods of performing various near-visual tasks. The relationship of these changes to the subjective experience of ocular discomfort is poorly understood. Many studies have reported recession of the near points of accommodation and convergence following prolonged near-visual work such as reading under difficult conditions or carrying out inspection tasks. Shifts of accommodation toward the resting 6We suggest that these terms be avoided whenever possible in scientific studies in favor of such terms as ocular discomfort, changes in performance, and change in oculomotor functions that specifically describe the phenomena discussed. When complaints of "visual fatigue" or "eyestrain" are presented in clinical practice or surveys of workers, it is important to attempt to determine more precisely what phenomenon is being described. 7See Chapter 7 for discussion of technical terms having to do with oculomotor functions.
23 point have been found in subjects reading either hard copy or microfiche. Several aspects of eye movements have been shown to change during performance of various near-visual tasks. Some of these changes in eye movements have been shown to be reversed when subjects are aroused or highly motivated, which suggests that either the changes arise in the central nervous system or there is compensation for fatigue of oculomotor muscles. No evidence has been presented to suggest that these temporary oculomotor changes are harmful, although conceivably they might have some effect on performance of VDT and other visual tasks. None of the studies of VDT workers has provided valid evidence of ocular diseases or abnormalities that can be attributed to VDT work. Several surveys of VDT operators have included tests of such visual functions as acuity, astigmatism, stereopsis, phoria, and color vision. Unfortunately, there has been only limited effort to determine whether the status of these visual functions has any correlation with ocular complaints of VDT workers. There are a number of clinical conditions (especially those involving small uncorrected refractive errors and oculomotor imbalances) that can cause visual difficulties with prolonged near work or critical detail work. It is possible that such conditions might underlie some of the complaints of some VDT workers and non-VDT workers. If so, careful clinical examination of those workers would be important, and appropriate corrective lenses might relieve their symptoms. The lack of objective measures of ocular discomfort has made it difficult to determine possible causal factors. It is easier to determine what factors influence the visual performance of VDT workers, because performance is more readily measured than discomfort (see Chapter 4~. It is often assumed that conditions facilitating effective visual performance are less likely to pro- duce ocular discomfort; however, the complex relationships among visual performance, comfort, and psychological variables have not been thoroughly explored. It is important to ask whether there are factors in VDT visual tasks that are inherently different from those in comparable non-VDT visual tasks and if so, whether those factors might affect worker comfort and performance. We cannot completely answer this question because there has been only limited analysis of visual and cognitive functions in VDT work tasks. Some unique features of VDTs and VDT work are apparent. For example, the reflection of images by the VDT screen (discussed above), can cause dif- ficulties when no preventive measures are taken. Some other features that are not inherent in VDTs or in VDT work are often arranged in a potentially problematic way (e.g., the positioning of VDT screens at angles or distances that are incompatible with
24 conventional designs of bifocal spectacles, noted above). Many features of VDT visual tasks are, of course, similar to those of visual tasks in non-VDT jobs. It seems likely that with proper design of VDT display characteristics, workplace lighting, workstations, and jobs, VDT work would not cause any unique visual problems (see Chapter 9~. JOB DESIGN AND PSYCHOSOCIAL STRESS There has been little formal study of psychosocial aspects of work involving video display terminals (see Chapter 8~. A few studies have attempted to identify psychosocial stressors in VDT work and to relate them to self-reports of employee well-being and in some cases to physiological changes such as increased blood pressure. Although some studies have found that VDT operators report high levels of job-related stress, no psychosocial stressors or health-related outcomes have been shown to be unique to work involving VDTs. However, the research literature is inconclusive because of problems in the designs of most studies. The data that exist suggest that where negative health effects do appear, they may stem from factors in the job itself (including but not limited to the VDT component) or from organizational relationships involving the employees. Jobs in which VDTs are used are not purely "VDT jobs," even when the VDT dominates everything else about the job. The total job is defined less by the particular equipment used than by what the equipment is used for: what the worker is expected to pro- duce, the methods and procedures to be followed, the skills and abilities required, and the interactions with other people on the job. It is possible to design jobs carefully so that the work experience is satisfying and productive, but in practice most jobs develop with little real planning, and whatever planning occurs is generally more concerned with the equipment than the person who uses it. At a public symposium organized by the panel, labor represen- tatives described some VDT jobs that clearly are badly designed; workers in these jobs must perform highly repetitive tasks at a fast pace with no opportunity to vary the structure or pace of the work or even to adjust uncomfortable equipment. Of course, not all VDT jobs are badly designed, but unfortunately, we have no data on what percentage of them are. Existing literature suggests that most complaints are reported by workers in jobs in which a single task (such as data entry) dominates the work day, pay is relatively low, and the workers' responsibility is limited to main- taining output and avoiding errors. Such jobs might be seen as highly undesirable in that they stifle human initiative, creativity,
25 and sense of achievement. The question that demands answering is whether VDT workers who complain of problems are responding to the equipment, the basic nature of the job, or their perceptions of the job and its opportunitites and limitations. Jobs in which VDTs are used- - s well as jobs in which VDTs are not used vary greatly across factors that may act as sources of psychosocial stress. These sources include the design of the job itself, the social and psychological environment of the workplace, and the broader organizational system of which the worker is a part. One useful approach to analyzing the psychosocial stressors in a work situation is to examine the degree of fit between the characteristics of workers and the characteristics of work situa- tions (see Chapter 8~. The assumption underlying this approach is that a lack of fit between the characteristics of worker and work environment leads to strain and poor performance. Two kinds of person-environment fit, which may not always represent two mutually exclusive classifications, can be examined: one is the fit between a worker's needs (or preferences, desires, values, etc.) and the related supplies for these needs in his or her job environ- ment; the other is the fit between a worker's abilities and the demands of his or her jObe This person-environment fit approach has not yet been applied to work involving VDTs, but it has been applied to the study of stressors in other work situations, and it can provide a useful conceptual framework in which to study possible psychosocial stressors in VDT work. For example, VDT jobs, like other jobs, vary greatly on such dimensions as the amount of control given an employee and the employee's opportunity to participate in decisions that affect the way in which his or her work is carried out. VDT jobs vary in the extent to which an employee can control the introduction into the workplace of VDT equipment, the amount of incoming work and associated deadlines, the variety of the work content, the amount and scheduling of time spent at the VDT, and the extent of inter- actions with other people. Although research on VDT use does not permit firm conclusions as to how variations in control influence employee well-being, research on other types of work suggests that lack of control has measurable, undesirable effects on employee well-being. Because individuals vary in their need for control, the person-environment fit approach would predict that lack of (or too much) control can act as a psychosocial stressor for workers in jobs in which there is a misfit between the worker and the job on this dimension. VDT jobs, like other jobs, also vary in a number of other pos- sible sources of stress, including complexity of the work, quanti- tative workload, predictability of events (such as computer system breakdowns or processing delays), threat of job loss, and social support. This list is not exhaustive, of course, but indicates
26 the range of dimensions on which VDT jobs end other jobs very end suggests potential stressors that might usefully be studied within the framework of person~nvironment fit theory. Systematic studies of the relative contribution of each psychosocial stressor as a component of the context in which the VDT is used would provide a much greater increase in knowledge about the psycho- genic health effects of VDT use than studies that have simply compared complaints of job-related stress in a "VDT group" and a "non-VDT group" without attempting to match workers and jobs in other respects. These kinds of evaluations would allow the examination of the contributions of the worker and of the VDT and its work environment to employee well-being, and could help generate options for improving the fit between a person and a job. Most, if not all, psychosocial stressors that may be associated with VDT work under conditions in which the work is not organ- ized with the well being of the worker in mind do not seem to be inherent to VDT technology and software. VDTs, like any other work technology, can be used properly or improperly, and VDT work can be organized so that it reduces stress and increases productivity or increases stress and reduces productivity. In many, but not all, job settings, VDTs have been introduced without consultation with the workers who are to use them. The manner in which this new technology is introduced may influence workers' perceptions of the VDT. The private use of VDTs (e.g., home computers and word processors) similar to those used occu- pationally is growing rapidly. It might be interesting to compare user attitudes in private and industrial settings; there have as yet been no formal studies making such a comparison. DESIGN, PRACTICE, AND STANDARDS Principles of Good Design and Practice VDTs have often been designed and introduced into workplaces with little attention to well-established principles of, and existing data about, good design and practice. There is a large base of knowledge about image quality, workplace design, lighting and reflections, and industrial and organizational psychology that has often been disregarded or inappropriately applied. It is likely that problems with and concerns about the use of VDTs would be greatly alleviated by the appropriate application of this know 1- edge to the design of VDT equipment and VDT jobs (see Chapter 9~. We strongly urge designers of VDT jobs to draw upon well- established principles for organizing work in ways that are
27 conducive to the well-being of workers (see Chapter 9~. Our analysis suggests two principles that should be given special con- sideration: (1) stress can best be reduced by optimizing the fit between a worker and his or her working environment, rather than standardizing environments regardless of individual needs and abilities; and (2) participation in decision making and some degree of individual control over the nature and pace of work allows workers to achieve maximum person-environment fit. Among the implications of these principles is that flexibility is preferable to fixed rest breaks. It should be noted, however, that rigidly de- signed jobs with high quotas for productivity, in which the output of workers is monitored moment by moment, allow little or no flexibility; in these cases, fixed rest breaks may provide the only opportunity to move around or to rest tired eyes. Public Education To a substantial degree, alleviation of problems associated with the use of VDTs depends on educating users, both those who decide to have others use such equipment and those who operate it. Users of VDTs and of related equipment need to be aware that many ergonomic problems can be overcome immediately by applying what is already known about display quality and work- place design; it is not necessary to wait for further research (see Chapters 6 and 9)e Well-designed, high~uality displays and related workplace equipment and furniture are commercially available. Manufacturers should standardize definitions of equipment characteristics and techniques used to measure them and make this information available so that users will be in a position to make informed choices. Users should also learn as much as possible about the characteristics of well-designed equipment so that they can effectively evaluate and compare products. Users can also apply existing knowledge in reducing glare and in adjusting angles of view, visual distances between display and operator, display contrast and luminance, and so on. Both managers and workers can support the use of good employee- management practices in the introduction and use of VDTs. Education about these matters involves more than just making such information available. Consumers need to learn how to dis- tinguish between reports that are based on scientific research and those that argue from uncontrolled collections of cases. The scientific community could help with this goal by taking a more active stance in contributing information to media that are widely available to the general public (while continuing to document findings through scientific publication). The media could help by reporting both the claims made by various parties regarding the
28 health effects of VDTs and the basis for those claims; such report- ing may require better knowledge of scientific methods on the part of reporters and writers. Unfortunately, sensationalized accounts often draw more attention, and sometimes more credence, than do careful analyses of evidence. Standards and Guidelines for VDT Uses Standards have been proposed or enacted, primarily in Europe, for VDT design characteristics and use. Various standards have been specified in labor agreements, directives or guidelines from gov- ernment agencies, or legislation. Some standards cover work- station design features, image display characteristics, and lighting and reflection conditions; some also specify provision of rest breaks, operator training in VDT use, and eye examinations for VDT operators. Most, though not all, standards specify numerical values for such parameters as display character size, luminance levels, key force, and viewing distance. Substantial differences and conflicts in specifications are found among these standards, which appear to be based on varying assumptions (see Chapter 9~. Some specifications do not have a clear empirical basis, and some do not reflect existing knowledge about visual performance. Judicious use of guidelines and standards can be helpful, but we feel that it would be premature to establish mandatory standards. Careful research and discussion are needed to resolve conflicts in the specifications given in existing and proposed standards. There do not appear to have been any careful follow-up studies to evaluate the efficacy of standards now in use. Because VDT technology is rapidly evolving, standards that are premature may impede improvements or become irrelevant. Standardized, appropriate methods for measuring and evaluating image quality are needed in order to develop appropriate guidelines for displays. Without such standard techniques, compliance with mandatory Tin this discussion we use standards to refer to specifications of values for design parameters to which strict adherence is expected. These include legally binding specifications, such as the German Safety Standards; specifications written into contracts, such as the U.S. Military Standard 1472C; and specifications voluntarily adopted by industry, such as those promulgated by the American National Standards Institute. In contrast we use guidelines to refer to specifications that are suggested with the understanding that implementation be flexible, depending on circumstances and needs.
29 standards (including those that specify equipment maintenance procedures and schedules) would be difficult. In addition to careful, systematic research, we suggest that there should be a continuing dialogue among scientists, manufac turers, and VDT users so that useful guidelines can evolve. It seems likely that different sets of guidelines will be appropriate for different kinds of VDT applications and operations. RESEARCH NEEDS - A number of questions raised in our analysis of the research literature on effects of VDT work remain unanswered. Many of these questions could be answered by appropriately designed research. Chapter 2 discusses criteria that should be considered in the proper design of research. We suggest several avenues of research that might be taken, but we urge that competing prior- ities in the field of occupational health be carefully considered (Chapter 10~. Objective measures that can be used to relate visual discom- fort to patterns of visual activity, VDT characteristics, and visual performance are needed. Such research would be relevant to an understanding of performance of a range of near-work tasks in addition to VDT work. The implications of positive- versus nega- tive-contrast displays should be investigated in greater depth. Research on the efficacy of screen filters is needed to evaluate the claims made by manufacturers. Research comparing cathode- ray tube displays to geometrically stable displays (e.g., gas or electroluminescent panels) would also be useful. We suggest that research on factors such as workload, social support, and task complexity that affect all job - -including but not limited to VDT work -be given priority over research with a narrow focus on what is stressful or not stressful about VDT work per se.
