From the onset, we wish to make it clear that we are not experts in ethics, nor do we represent ourselves as experts. Just like the rest of you, we are scientists, doing scientific things, searching for truth. In that arena, we are faced with the same issues as you—the difference between right and wrong, good and bad, and objectivity and subjectivity. As researchers, we grapple with difficult questions concerning methods and outcome. And those of us who become administrators agonize over the same questions, especially in regard to policy decisions in developing adequate occupational standards to protect workers.
In preparing this paper, we reviewed a number of books and documents in search of simple answers to complex problems. We looked not only at papers written by occupational safety and health professionals, but also reviewed some of the classic textbooks on ethics.
On the professional side, we read a number of articles and codes of ethics from various research groups. They all have components relevant to occupational health research. Yet the focus of each is quite different, reflecting the kind of research done by each author. Some include numerous pages of what to do and what not to do. Others are more general in content.
On the textbook side, ethical theories abound, from before Socrates up until today. There is no shortage of articles about ethics, codes of conduct, and written discussions of ethical standards. In the United States at least, most of the medical colleges have medical ethicists on staff, and almost every university with a sizeable department of philosophy has an ethicist on the faculty. It is a discipline to which people devote their lifetimes, which confirms the complexity of the issue.
Before we begin this discussion, it is important that we attempt to make it clear what we are talking about. What is meant by the term ethics? In the English language, the terms ethics and morals are used interchangeably. Since we are preparing this paper for a diverse group, we did what we think to be an interesting poll of some Centers for Disease Control and Prevention (CDC) professionals for whom English is their second language. A woman whose first languages are Slavic, German and Russian replied that there are similar words in all of her first languages. She said that in the Slavic language, neither ethics nor morals stands alone as they do in English. For example, she said that you wouldn’t say that someone is without morals, you would say that they exhibit behaviour that is not moral. She said that in the Slavic language you wouldn’t say that someone is without ethics, you would say instead that the person is without ethical principles. A Chinese national said that there are separate Chinese words for both morals and ethics, but they are used interchangeably. Spanish-, French- and German-speaking people said there are words for both in their respective languages and that the words are used interchangeably.
In the textbooks on ethical theory that we reviewed, however, the ethicists made a distinction between ethics and morals which we choose to accept for the sake of clarity. Melden (1955) and Mothershead (1955) both suggest that the word ethics is used when referring to a set of principles or standards for conduct, and that the word morals is used when referring to the conduct of a person or group, i.e., their behaviour. This usage is consistent with the replies of the CDC professionals.
Professor Melden says in his book, “We are all familiar with such rules of conduct. Each society, religion, professional group, or distinguishable community has its principles, its standards of conduct. As persons who are concerned with being responsible in our conduct, we rely ordinarily upon a body of principles for guidance in conduct.” Examples of these principles are all around us. In the Judeo-Christian community, there are at minimum the Ten Commandments. In every society, we have laws at the local, national and international levels which describe and dictate both unacceptable and acceptable behaviour. There are also the scientific method, the International Code of Ethics for Occupational Health Professionals and Guidance on Ethics for Occupational Physicians, to name a few examples. The list could go on and on. The point here is that we are exposed to a number of standards of conduct, or ethics, as we use the term. It is quite fitting that we begin the work of setting some standards for ourselves.
Why do health professionals need standards for our work? As Professor Melden states, we are persons who are concerned with being responsible. To make good science demands the highest responsibility on our part, which leads to the promotion of safety and health. On the other hand, no matter how good the intentions of the researcher may be, compromised science can lead to death, disease, disability and dismemberment, rather than the protection of workers. The bottom line is that workers suffer when science is compromised.
Why does compromised science happen? From our perspective, there are a number of reasons.
Sometimes science is compromised because we don’t know any better. Take for example three workplace tragedies: asbestos, benzene and silica. In the early days, the dangers of these substances were unknown. As technology improved, as the science of epidemiology developed and as medicine became more sophisticated, the obvious became evident. In each of these histories, the problems existed, but scientists did not possess or in some cases apply the tools available to uncover them.
Sometimes science is compromised because it is bad science. We are certain that all of you have seen bad science or have read about it in scientific journals. It is bad because it isn’t science at all. It is opinion expressed in such a way that it appears scientific and therefore factual. This situation is one that can easily be addressed through a rigorous peer review process.
Sometimes science is compromised because the researcher is rushed, due to unrealistic time constraints, lack of funds or influences other than purely scientific analysis. An obvious example of this is a toxicological cancer study in which the test animals’ lives were terminated after less than a third of their normal life span, thus eliminating a sufficient latency period for them to develop cancer as a result of their exposures. Thoroughness was compromised and conclusions reached with only part of the picture considered.
And perhaps worst of all, sometimes science is compromised in pursuit of profit or academic advancement. Likewise, we have all seen evidence of this in the newspapers and professional journals. In some of these instances, the gain to the researcher was academic standing and not financial at all. In others, financial gain, either immediate or future, influenced the outcome. In the first case referred to above, researchers with financial interests in asbestos did not report their own positive findings until many years later, when many thousands of workers had already suffered and died of diseases associated with uncontrolled asbestos exposure (Lemen and Bingham 1994). In some instances, we have seen that those who pay for the research may ultimately influence the outcome.
These are but a few of the cases where a code of ethics could come into play, although any code, no matter how wonderful, will not stop the unscrupulous.
Occupational health is a complex and difficult discipline in which to prevent unethical conduct. Even when we discover methods for prevention of occupational diseases and injuries, the solution to the problem is often viewed as cutting into profits, or the problem is hidden to avoid the expense of the remedy. The profit motive and the complexity of the issues we address can lead to both abuse and shortcuts in the system. What are some of the major difficulties?
Often, occupationally caused maladies have incredibly long incubation periods, giving rise to confounding variables. By comparison, in many infectious diseases results seem quick and simple. An example is a well-managed vaccine campaign for measles in an outbreak situation. In this case, there is a short incubation period, an almost 100% infection rate of those susceptible, a vaccine that is 95 to 98% effective and a total eradication of an epidemic, all accomplished in a few days. That situation is quite different from asbestosis or carpal tunnel syndrome, where some people are affected, but others are not, and most often months or years elapse before disability occurs.
Occupational health concerns are multidisciplinary. When a chemist works with other chemists, they all speak the same language, each has but one interest and the work can be shared. Occupational health, on the other hand, is multidisciplinary, often involving chemists, physicists, industrial hygienists, epidemiologists, engineers, microbiologists, physicians, behaviourists, statisticians and others. In the epidemiological-triad (host, agent, environment), the host is unpredictable, the agents are numerous and the environment is complex. The cooperation of several disciplines is mandatory. A variety of professionals, with totally different backgrounds and skills, is brought together to address a problem. The only commonality between them is the protection of the worker. This aspect makes peer review even more difficult because each speciality brings its own nomenclature, equipment and methods to apply to the problem.
Because of long incubation periods in many occupational diseases and conditions, coupled with the mobility of the workforce, occupational health professionals are often forced to fill in some blanks since many of those workers exposed or at risk cannot be located. This condition leads to a reliance on modelling, statistical calculations, and sometimes compromise in the conclusions. The opportunity for error is great, because we are not able to fill in all of the cells.
Sometimes it is difficult to relate a malady to the work environment or, even worse, to identify the cause. In infectious diseases, the epidemiological triad is often less complex. In the 1990s, CDC staff investigated an outbreak of illness on a cruise ship. The host was well defined and easily located, the agent was easily identified, the mode of transmission was obvious, and the remedial action was evident. In occupationally related disease and injury, the host is defined, but often difficult to find. There are a number of agents in the work environment, often causing synergism, plus other workplace factors which are not directly involved in the health problem but which play an important role in the solution. These other workplace factors include such things as the interests and concerns of the labour force, the management and involved government agencies.
So now to the business at hand—coming up with a code of ethics, a set of principles or standards of conduct, used to guide our conduct, our behaviour, in this complex environment.
As Professor Melden (1955) so clearly writes, “Further, we cannot depend wholly upon such principles for guidance, simply because it is impossible to lay down a set of rules complete enough to anticipate all possible occasions for moral decision.” He goes on to say that “A set of moral principles covering all possible moral eventualities is just as impossible as a set of laws so complete that no further legislation is necessary”. Similarly, Kenneth W. Goodman (1994b), states that “While it is important to realize that science and ethics are closely, even inextricably, linked, there is no reason to suppose that a formal code of ethics will provide closure for all or most disagreements about the nature of data, selection of data, data management, and so forth.” To quote Professor Melden once again, “To be useful, moral principles must be general; but being general, their utility is inescapably limited”.
With the above caveats in mind, we propose to you that the following statements be part of a code of ethics for occupational health.
- That, at a minimum, peer review be required and include tripartite review with worker, industry and government representation, as well as review by academia. This process is difficult because it takes time—time to identify learned reviewers from all three areas, time to bring them together for discussion, and often considerable time to address each of their concerns. For the US National Institute for Occupational Safety and Health, at least, this process is required for all publications. We do not pretend to have all the answers, nor do we alone have all the facts. Much can be learned from labour and industry about workplace situations and problem resolution. Tripartite review is the only way we know of to minimize the effects of special interest groups.
- That even perceived compromises are avoided. Sometimes good science has no credibility because of perceived compromise. Examples of compromises include the funding source for the study, the interest groups selected to review the study, and known bias of reviewers. There are judgement calls on the part of the researcher, and even though the judgement and subsequent decision may be sound, there can be a perceived compromise in the study.
- That study protocols receive review by peers before the study is undertaken. The best-intentioned researcher can build a bias into a protocol. This will become obvious only upon careful protocol review.
- That the scientific method is followed from the start: (a) form a hypothesis, (b) do a literature search, (c) collect data, (d) collate data, (e) test the hypothesis and (f) disseminate results.
- That when using science to develop an occupational health or safety standard, all parties involved in the decision declare their affiliations, their financial interests, their potential conflicts with the industry or substance being regulated, and that all of these facts are clearly delineated in the final documentation of the standard. For any standard or recommended standard, perception is of utmost importance. If it is viewed that the standard was based on a biased interpretation, then the standard will lack credibility. Standards based solely on the interpretation of the science by individuals associated with the industry under consideration would suffer from such an interpretation or, worse yet, could fall short of adequately protecting the workers at risk. Building in check factors such as those described above during the development of the new standard will assure that this will not occur.
We have attempted to discuss a complex and sensitive issue. There are no easy solutions. What we are attempting is right and just, however, because its goal, to protect the worker in the workplace, is right and just. We cannot do this alone, we cannot do it in a vacuum, because the problems we address are not in a vacuum. We need each other, and others, to ferret out our natural instincts for personal gain and glory and to uncover our built-in biases. Such an effort will enable us to contribute to the knowledge and enhance the well-being of humanity.