In the last several decades, considerable effort has been devoted to defining and addressing the ethical issues that arise in the context of biomedical experimentation. Central ethical concerns that have been identified in such research include the relationship of risks to benefits and the ability of research subjects to give informed and voluntary prior consent. Assurance of adequate attention to these issues has normally been achieved by review of research protocols by an independent body, such as an Institutional Review Board (IRB). For example, in the United States, institutions engaging in biomedical research and receiving Public Health Service research funds are subject to strict federal governmental guidelines for such research, including review of protocols by an IRB, which considers the risks and benefits involved and the obtaining of informed consent of research subjects. To a large degree, this is a model which has come to be applied to scientific research on human subjects in democratic societies around the world (Brieger et al. 1978).
Although the shortcomings of such an approach have been debated—for example, in a recent Human Research Report, Maloney (1994) says some institutional review boards are not doing well on informed consent—it has many supporters when it is applied to formal research protocols involving human subjects. The deficiencies of the approach appear, however, in situations where formal protocols are lacking or where studies bear a superficial resemblance to human experimentation but do not clearly fall within the confines of academic research at all. The workplace provides one clear example of such a situation. Certainly, there have been formal research protocols involving workers that satisfy the requirements of risk-benefit review and informed consent. However, where the boundaries of formal research blur into less formal observances concerning workers’ health and into the day-to-day conduct of business, ethical concerns over risk-benefit analysis and the assurance of informed consent may be easily put aside.
As one example, consider the Dan River Company “study” of cotton dust exposure to its workers at its Danville, Virginia, plant. When the US Occupational Safety and Health Administration’s (OSHA) cotton dust standard went into effect following US Supreme Court review in 1981, the Dan River Company sought a variance from compliance with the standard from the state of Virginia so that it could conduct a study. The purpose of the study was to address the hypothesis that byssinosis is caused by micro-organisms contaminating the cotton rather than by the cotton dust itself. Thus, 200 workers at the Danville plant were to be exposed to varying levels of the micro-organism while being exposed to cotton dust at levels above the standard. The Dan River Company applied to OSHA for funding for the project (technically considered a variance from the standard, and not human research), but the project was never formally reviewed for ethical concerns because OSHA does not have an IRB. Technical review by an OSHA toxicologist cast serious doubt on the scientific merit of the project, which in and of itself should raise ethical questions, since incurring any risk in a flawed study might be unacceptable. However, even if the study had been technically sound, it is unlikely to have been approved by any IRB since it “violated all the major criteria for protection of subject welfare” (Levine 1984). Plainly, there were risks to the worker-subjects without any benefits for them individually; major financial benefits would have gone to the company, while benefits to society in general seemed vague and doubtful. Thus, the concept of balancing risks and benefits was violated. The workers’ local union was informed of the intended study and did not protest, which could be construed to represent tacit consent. However, even if there was consent, it might not have been entirely voluntary because of the unequal and essentially coercive relationship between the employer and the employees. Since Dan River Company was one of the most important employers in the area, the union’s representative admitted that the lack of protest was motivated by fear of a plant closing and job losses. Thus, the concept of voluntary informed consent was also violated.
Fortunately, in the Dan River case, the proposed study was dropped. However, the questions it raises remain and extend far beyond the bounds of formal research. How can we balance benefits and risks as we learn more about threats to workers’ health? How can we guarantee informed and voluntary consent in this context? To the extent that the ordinary workplace may represent an informal, uncontrolled human experiment, how do these ethical concerns apply? It has been suggested repeatedly that workers may be the “miner’s canary” for the rest of society. On an ordinary day in certain workplaces, they may be exposed to potentially toxic substances. Only when adverse reactions are noted does society initiate a formal investigation of the substance’s toxicity. In this way, workers serve as “experimental subjects” testing chemicals previously untried on humans.
Some commentators have suggested that the economic structure of employment already addresses risk/benefit and consent considerations. As to the balancing of risks and benefits, one could argue that society compensates hazardous work with “hazard pay”—directly increasing the benefits to those who assume the risk. Furthermore, to the extent that the risks are known, right-to-know mechanisms provide the worker with the information necessary for an informed consent. Finally, armed with the knowledge of the benefits to be expected and the risks assumed, the worker may “volunteer” to take the risk or not. However, “volunteer-ness” requires more than information and an ability to articulate the word no. It also requires freedom from coercion or undue influence. Indeed, an IRB would view a study in which the subjects received significant financial compensation—“hazard pay”, as it were—with a sceptical eye. The concern would be that powerful incentives minimize the possibility for truly free consent. As in the Dan River case, and as noted by the US Office of Technology Assessment,
(t)his may be especially problematic in an occupational setting where workers may perceive their job security or potential for promotion to be affected by their willingness to participate in research (Office of Technology Assessment 1983).
If so, cannot the worker simply choose a less hazardous occupation? Indeed, it has been suggested that the hallmark of a democratic society is the right of the individual to choose his or her work. As others have pointed out, however, such free choice may be a convenient fiction since all societies, democratic or otherwise,
have mechanisms of social engineering that accomplish the task of finding workers to take available jobs. Totalitarian societies accomplish this through force; democratic societies through a hegemonic process called freedom of choice (Graebner 1984).
Thus, it seems doubtful that many workplace situations would satisfy the close scrutiny required of an IRB. Since our society has apparently decided that those fostering our biomedical progress as human research subjects deserve a high level of ethical scrutiny and protection, serious consideration should be given before denying this level of protection to those who foster our economic progress: the workers.
It has also been argued that, given the status of the workplace as a potentially uncontrolled human experiment, all involved parties, and workers in particular, should be committed to the systematic study of the problems in the interest of amelioration. Is there a duty to produce new information concerning occupational hazards through formal and informal research? Certainly, without such research, the workers’ right to be informed is hollow. The assertion that workers have an active duty to allow themselves to be exposed is more problematic because of its apparent violation of the ethical tenet that people should not be used as a means in the pursuit of benefits to others. For example, except in very low risk cases, an IRB may not consider benefits to others when it evaluates risk to subjects. However, a moral obligation for workers’ participation in research has been derived from the demands of reciprocity, i.e., the benefits that may accrue to all affected workers. Thus, it has been suggested that “it will be necessary to create a research environment within which workers—out of a sense of the reciprocal obligations they have—will voluntarily act upon the moral obligation to collaborate in work, the goal of which is to reduce the toll of morbidity and mortality” (Murray and Bayer 1984).
Whether or not one accepts the notion that workers should want to participate, the creation of such an appropriate research environment in the occupational health setting requires careful attention to the other possible concerns of the worker-subjects. One major concern has been the potential misuse of data to the detriment of the workers individually, perhaps through discrimination in employability or insurability. Thus, due respect for the autonomy, equity and privacy considerations of worker-subjects mandates the utmost concern for the confidentiality of research data. A second concern involves the extent to which the worker-subjects are informed of research results. Under normal experimental situations, results would be available routinely to subjects. However, many occupational studies are epidemiological, e.g., retrospective cohort studies, which traditionally have required no informed consent or notification of results. Yet, if the potential for effective interventions exists, the notification of workers at high risk of disease from past occupational exposures could be important for prevention. If no such potential exists, should workers still be notified of findings? Should they be notified if there are no known clinical implications? The necessity for and logistics of notification and follow-up remain important, unresolved questions in occupational health research (Fayerweather, Higginson and Beauchamp 1991).
Given the complexity of all of these ethical considerations, the role of the occupational health professional in workplace research assumes great importance. The occupational physician enters the workplace with all of the obligations of any health care professional, as state by the International Commission on Occupational Health and reprinted in this chapter:
Occupational health professionals must serve the health and social well-being of the workers, individually and collectively. The obligations of occupational health professionals include protecting the life and the health of workers, respecting human dignity and promoting the highest ethical principles in occupational health policies and programmes.
In addition, the participation of the occupational physician in research has been viewed as a moral obligation. For example, the American College of Occupational and Environmental Medicine’s Code of Ethical Conduct specifically states that “(p)hysicians should participate in ethical research efforts as appropriate” (1994). However, as with other health professionals, the workplace physician functions as a “double agent”, with the potentially conflicting responsibilities that stem from caring for the workers while being employed by the corporation. This type of “double agent” problem is not unfamiliar to the occupational health professional, whose practice often involves divided loyalties, duties and responsibilities to workers, employers and other parties. However, the occupational health professional must be particularly sensitive to these potential conflicts because, as discussed above, there is no formal independent review mechanism or IRB to protect the subjects of workplace exposures. Thus, in large part it will fall to the occupational health professional to ensure that the ethical concerns of risk-benefit balancing and voluntary informed consent, among others, are given appropriate attention.