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.
Appendix B Review of a Preliminary Report on a Cross-Secffonal Survey of VDT Users at the Baltimore Sun R. Van Harnson This paper contains three major sections: the first section is the summary from the NIOSH study (Smith et al., 1982), the second summarizes and comments on the approach and methods of the study, and the third considers conclusions to be drawn from the study. SUMMARY OF THE NIOSH STUDY The National Institute of Occupational Safety and Health, was asked by a representative of employees of the News- paper Guild, AFL-CIO, representing a large segment of employees at the Baltimore Sunpapers, to undertake an evaluation of the effects of video display terminals (VDTs) "on the environment and health of employees who use them." Included in the request was the statement that there had occurred "several cases of cataracts among VDT users, a high rate of complaints about eye problems such as irritation and blurred vision and headaches, back and neck aches . . .(sic)" Accordingly, we undertook a cross sectional survey, to define the type of eye and body complaints reported by VDT users, and to identify their relation to VDT use; the association between symptoms and the participants' abilities to see clearly (i.e., their refractive abilities relative to the demands for clear vision required by their job); and the prevalence of eye abnormalities, including cataracts and retinal abnormalities, and their relationship to VDT use. R. Van Harrison, who is a study director at the Institute for Social Research at the University of Michigan, was a consultant to the panel. 227
228 We surveyed 379 employees of the Baltimore Sun, 283 of whom were members of the Newspaper Guild. Each participant answered a self-administered questionnaire on personal and job information, symptom complaints, and on a personal assessment of the pressure, pace, autonomy, and security, and satisfaction associated with the job. Each survey participant underwent a complete eye examination. Using a statistical technique known as "factor analy- sis", we found that as participants increasingly reported that they were bothered by the brightness of the VDT screen or characters, by the glare off the screen, by the readability of the characters, or by flicker; they also increasingly reported (1) changes in their visual function, namely, seeing colored fringes around objects, difficulty reading and focusing on characters; (2) pain and stiffness in their neck, shoulders, and back; (3) headaches asso- ciated with work, in particular their usual job; and (4) headaches accompanied by itching, burning, watery eyes, blurry vision, nasal discharge and sweating. As partici- pants tended to report that their VDT use typically involved shifting their eyes between the source document, VDT keyboard and screen; and as they tended to report that they found that they were bothered by the relative height, distance, and tilt of the VDT keyboard and screen; so too they tended to report that their headaches charac- teristically were superficial in location, dull and boring in sensation, beginning on one side of the head, but spreading to involve both sides. As participants reported a greater total number of years of VDT operating experience, they tended to report less that their headaches occurred during periods of stress, worry, and/or tension. As participants reported a greater number of hours per week of VDT operation, they also tended to report less that their head- aches were preceded and accompanied by double and blurry vision. Controlling in the analyses for other characteristics of the participants, which might affect the symptoms being reported, did not change these observed associations in any meaningful way. We did not find any meaningful relationship between adequacy of the participants' refractions, including the wearing of glasses with hi- or multifocal lenses, and the reporting of work~ssociated symptoms. We did not find any signficiant association between VDT use, including hours per week of VDT operation and total years of VDT operating experience; and the prevalence of eye abnor- malities, including cataracts.
229 We note that among VDT users, the average number of years of VDT operating experience was 3.8 years, with a maximum of 9.2 years. If a minimum duration of VDT usage is postulated to be required prior to eye abnormal- ities being detectable, then the group of participants in this survey may well be judged to have had an insufficient amount of VDT usage for us to have found any such pos- tulated associations. Therefore, our survey may well have been inadequate in terms of amount of exposure to resolve such issues as the putative associations of cataracts and VDT usage. This survey has been primarily of value in delineating the relationship between VDT-users' symptoms and vari- ous ergonomic aspects of VDT use. The bothersome visual aspects of the VDT itself, as usually adjusted, explained the plurality of work-associated symptoms, even when other participant and work place characteristics were taken into account. We suggest that future emphasis be placed on research in regard to VDT viewing character- istics, and other aspects of the VDT viewing environ- ment. We feel that these problems are best addressed experimentally. REVIEW OF THE NIOSH STUDY Study Approach and Methods Introduction and Background The study report begins with a review of related literature, identifying from previous studies important variables associated with VDT use and possible health problems. The organization of the analyses and discussion indicates an underlying conceptual model of categories of variables and their expected relationships; however, this model is not explicitly presented in the report. Design The study was performed to identify relationships between VDT use and health problems. It was an exploratory effort, attempting a more rigorous control of confounding factors than has been performed in previous studies. A cros~sectional field study was performed. A population of convenience was identified at the Baltimore Sum-convenient because of the high level of VDT use among the staff and some indication of health problems poten-
230 tially related to VDT use. This population does not necessarily represent any other VDT user population, so findings must be generalized with caution. Indeed, the 49 percent participation rate of the eligible Newspaper Guild members and comparisons between demographic characteristics of participants and some nonparticipants suggest some differences between participants and nonparticipants. For example, the participants were more likely to use a VDT on the job and tended to have more formal education than the nonparticipants. The size of the main analysis sample, Newspaper Guild members who were VDT users (N = 283), and the various subsamples is adequate to detect with confidence somewhat weak relationships, for example, correlations of .25 in the main analysis sample. The design, sample, and sample size are appropriate for exploring possible relationships between VDT use and health problems in a group likely to be at risk. The causal direction reflected in associations cannot be inferred from the data. The sample size is too small to test for increases in very infrequent health problems. These design limitations are clearly indicated in the report. Measures One of the important strengths of the study is the variety of variables included in it: aspects of VDT use; characteristics of the individuals, including demographic characteristics and use of bifocal or multifocal lenses; characteristics of the job, including the work setting (especially lighting) and job characteristics; and health problems, including headaches and other symptoms, and abnormal ophthalmologic findings. Of course, no one study can include all potentially relevant variables. For example, this study does not measure workers' skills in using VDTs. Most data are obtained from self-report questionnaires. Clini- cally specific data, including visual characteristics, refraction, lenses, and other findings, are obtained from an ophthalmologic examination. The self-report data provide information concerning a wide range of variables for relatively little cost. The associated disadvantage is that the data are limited to users' perceptions and users' accuracy in reporting them. For example, people may dif- fer in their reports concerning the flickering of a screen or an illumination contrast. Other studies must determine the extent to which differing reports reflect individual differences in percep- tion and accommodation or other differences in aspects of the work. The measures are certainly adequate to indicate the most likely problems as perceived by the workers.
231 The two instruments, questionnaire and ophthalmologic exam form, appear to be fairly standard, drawn largely from previous instruments. Information on the source of the items would provide a perspective on specific results, particularly if some comparable data were available for specific items. The data collection procedures are fairly standard. The Saturday appointment for the exam may account for some of the nonparticipation rate. The report does not present much data concerning the reliabil- ity of the measures. The factor analyses performed to produce several of the multiple item measures indicate the internal consis- tency of the measures. A more specific examination of the agree- ment among items within the final multiple item measures could have been obtained by calculating the coefficient (alpha) for each measure. No report of interexaminer agreement was presented. Analyses A consistent and well-thought-out analysis plan was followed to describe the range of findings, identify correlations (and therefore potential confounding) between variables, and carry out multivari- ate analyses of the effects of predictor variables on outcomes. The distributions of scores are indicated for all measures except those variables derived through factor analysis and reported as standard scores. Although the standard scores are useful for relative comparisons, they provide little information concerning the range of scores on the original scales (typically a 6-point scale ranging from "never" to "always". For example, the reported data provide no indication of the distribution of scores for time spent with "eyes fixed" on the screen or for frequency of various headache symptoms. The analyses investigating relationships among variables are very well planned. Systematic checking for possible confounding, forward and backward replication of stepwise analyses, and attempts to cros~validate the results in a separate sample evidence a careful and critical examination of the data. Conclusions The authors of the study are careful to point out the major limitations of the study design and to counsel caution in drawing conclusions. ,
232 Substantive Conclusions In this study, simply using a VDT was not automatically associated with health problems. However, particular aspects of \fDT use were associated with types of headaches and other symptoms: poor visual clarity of the screen and improper lighting of the workplace were the most apparent. Their relationships to somatic symptoms were generally independent of the effects of other variables. The importance of VDT screen clarity is underscored by this association of VDT clarity with work-related headaches and with changes in visual functioning. These two relationships were the only findings replicated in a small, independent sample of non-guild VDT users. The authors appropriately recommend that future studies emphasize visual aspects of VDTs and workplace lighting, two factors that directly affect the viewing process. Three other variables related to VDT use had significant independent relationships with one or another of the headache measures. Two of the variables reflect quantitative use of VDTs (years of experience, hours per week of VDT operations), and the third variable ("eyes shifting" mode of use) reflects a qualitative pattern associated with VDT use. The relationships between the quantitative use of VDTs and headaches were negative, indicating more problems among less experienced users. This finding appears to contradict the simple hypothesis that VDT use automatically causes problems, with a greater amount of VDT use related to more problems. One expla- nation for the observed negative relationship would be that less skillful (i.e., less experienced) VDT operators are more likely to have problems operating VDTs, resulting in more problems for them. Future studies will need to examine variables likely to account for this negative relationship. The relationship between the qualitative use of VDTs ("eyes shifting") and headaches was positive. This finding is difficult to interpret because eyes shifting was associated with work that was demanding with little time to do it (job attitude factor 1~. Although this job factor was not related to the particular type of headache measure to which the eyes shifting mode was related, the job factor was related to three other headache and symptom measures. Future studies will have to determine the extent to which the underlying causal effects are due to the pattern of eye movements required with VDT use or the job characteristics of individuals who use the VDT in this mode. An important "nonbinding" was that there was no significant difference between VDT workers and non-VDT workers on the items in the ophthalmologic examination. The authors point out the limitations of the study in that causal processes may take longer than the exposure of the study sample to VDTs and that the
233 sample size is too small to detect infrequent (but important) negative outcomes. The more positive conclusion to be drawn from the study is that an average of 3.8 years of operating VDTs did not produce a sizable increase in negative ophthalmological findings. Studies to detect longer-term effects, very infrequent events, or smaller effects will have to be specially designed for those purposes. Another set of nonfindings concerns wearing bifocal or multi- focal glasses: their use was not associated with any significant increase in headaches or symptoms. Wearing such glasses does not appear to have a sizable impact on headaches or other somatic symptoms. An unfortunate omission in the presentation of results is an indication of the extent to which headaches and somatic symp- toms are a problem among VDT users. It was noted in the previous section that means on these measures (along with all of the measures developed through factor analysis) are reported only as standard scores. While the original scale is somewhat difficult to interpret (a 6-point scale ranging from "never" to "always"), transforming the mean standard scores to this scale would present some information concerning the magnitude of the effect involved. For example, the positive relationship between workstation lighting and pain and stiffness in the axial muscula- ture would be more important to pursue if the pain and stiffness scores for individuals with poor lighting averaged 5 rather than 1.5 on the original 6-point scale. All of the significant relationships with outcomes are with the symptom variables. Although the report identifies the variables associated with these outcomes, it does not indicate the frequency or severity of these outcomes in the study sample. Methodological Conclusions and Limitations The study findings confirm methodological refinements that must be incorporated into subsequent studies of VDT use. Separate and distinct measures of the frequency of operating in an "eyes fixed" mode and an "eyes shifting" mode must be made. Similarly, dim tinct measures must be made between positional problems and visual problems associated with VDT use. These modes and problems have different relationships to demographic factors, job characteristics, and symptoms. The conceptual elaboration and refinement of activities associated with VDTs is shown to be a necessary activity in studies of VDTs. Additional distinctions concerning the types of work and types of user problems are likely to be productive.
234 The results indicated that confounding frequently occurred between VDT use measures, demographic factors, and job char- acteristics and attitudes. However, the effects of the variables were separable. The findings indicate that future studies should include presently identified confounding factors. For example, confounded demographic factors in relationships between mea- sures of VDT use and measures of headache and somatic symp- toms include educational level, years of employment, and age. Future studies should attempt to identify and control the effects of other potential confounding factors, such as a user's skills in using VDTs. A methodological limitation of the findings is that the only significant relationships between VDT use and outcomes were found between self-reports of VDT use and self-reports concern- ing various symptoms. All survey studies may be affected by respondent perceptions and biases that introduce or alter relation- ships. However, in the present study, the findings generally fol- low expected patterns, evidence differential relationships, and are generally interpretable, indicating the construct validity of the measures. The limitations on the generalizability of the preceding con- clusions were noted in the discussion of the study sample. The conclusions cannot automatically be generalized to any other population. However, in the absence of other information, factors that are more important in this population should certainly have a priority for further study in other populations. The relationship between VDT use and health outcomes is complex because of the range of possibilities in VDT work, relevant health outcomes, and confounding personal and job factors. As in any applied area, future research must include a number of methodological approaches. Broad, exploratory studies similar to the present study must be performed in other populations to determine the replicability of relationships and problem magnitude across individuals, jobs, and VDT equipment. More expensive methodologies (e.g., direct observations of VDT use, health diaries) should be used to collect more detailed data on variables of interest. Longitudinal studies can help clarify causal relation- ships. Experimental studies will both clarify causal relationships and evaluate possible improvements in VDT use. The authors of the report specifically recommend that effects of workplace lighting and of visual clarity of the screen now He systematically investigated in experimental studies. The convergence of findings from cross-sectional survey studies will help identify the most important variables to study and to control in laboratory and field experiments to reduce problems associated with VDT use.
