New Study Violates Research Ethics by Assigning Subjects to Smoke, Instructing Them to Smoke, and Failing to Provide Mandated Care

A study published today in the New England Journal of Medicine describes the results of a clinical trial in which adult smokers who expressed no intent to quit at baseline were randomly assigned to smoke their regular brand of cigarettes or reduced nicotine cigarettes at varying levels of nicotine.

(See: Donny EC, et al. Randomized trial of reduced-nicotine standards for cigarettes. New England Journal of Medicine. Published online ahead of print on October 1, 2015. DOI: 10.1056/NEJMsa1502403.)

According to the study: "Participants were randomly assigned to smoke for 6 weeks either their usual brand of cigarettes or one of six types of investigational cigarettes, provided free. The investigational cigarettes had nicotine content ranging from 15.8 mg per gram of tobacco (typical of commercial brands) to 0.4 mg per gram. The primary outcome was the number of cigarettes smoked per day during week 6."

The results of the study, which were accompanied by an editorial, have been widely reported in the media. However, I have not seen any discussion of the ethics of this study.

The Rest of the Story

A basic principle of human subjects research is that of non-maleficence, or doing no harm. Also critical is the principle of harm minimization. You cannot knowingly put human subjects in harm's way when there is an alternative that is safer and will still answer the research question at hand. In particular, subjects being treated in a medical setting must, at a minimum, receive the standard level of care for the disease, condition, or behavior being studied.

This issue arises because most clinical trials involve randomization of subjects to receive an investigational treatment or no treatment. However, when someone has a medical condition, it is obviously not ethical to provide them with no treatment. Therefore, it is well recognized that subjects must, at a minimum, be offered the standard or usual treatment for that disease, condition, or behavior.

For example, suppose you want to investigate the effectiveness of a potential new treatment for severe hypertension (high blood pressure). The treatment involves a new drug. You cannot randomize patients to receive either the new drug or a placebo because it would be unethical to deprive subjects who have severe hypertension of any treatment. At a minimum, you would have to provide the control group with the minimum standard of care, which would almost certainly involve treatment with a standard drug usually used to treat high blood pressure.

In fact, the failure to provide standard care is one of the main problems in the tragic Tuskegee study, which is what in many ways led to the promulgation of ethical standards in research with human subjects. In that study, subjects with syphilis were randomized to receive antibiotic treatment or no treatment. It was unconscionable that subjects were left untreated when there was a known treatment available.

Believe it or not, this same failure occurred in the present study, although obviously not as blatantly or with as tragic the consequences. Nevertheless, the violation of basic research ethics was similar.

In the present study, smokers were assigned to either smoke their usual brand of cigarettes or to smoke a low-nicotine cigarette (at varying levels of nicotine). They were instructed to smoke the assigned cigarettes for six weeks. Although the subjects were free to quit smoking, they were not encouraged to quit unless they informed study staff that they had tried or intended to make a quit attempt.

According to the study protocol: " Participants will be instructed to use the research cigarettes for 6 weeks, at which point they are to discontinue product use." In other words, study subjects were instructed not to quit smoking for at least six weeks!

Moreover, the study protocol did not call for encouraging research participants to quit smoking, nor did it call for providing all subjects the usual standard treatment for smokers, which consists - at very least - of advising patients to quit. In a non-research setting, it would probably represent medical malpractice for a physician to instruct a patient to continue smoking for six weeks. At very least, there is an obligation - under standard medical principles - to advise the patient to quit smoking. Even less acceptable would be to make no recommendation regarding smoking. But to actively instruct the patient to continue smoking would be unethical.

The problem is compounded by the fact that study subjects were provided with a free 14-day supply of cigarettes. Not only does this make it more difficult for them to quit smoking, but it encourages them to continue smoking and sends the implicit message that if they don't continue to smoke, they are violating study protocol. It appears that the subjects were not explicitly told that quitting smoking is encouraged and will not violate the protocol or endanger their receiving compensation in any way.

Furthermore, offering smokers treatment with a smoking cessation medication and/or referral to a smoking cessation clinic or program is not only usual care, but it is actually mandated under the hospital accreditation standards.

The ethical breach becomes even more serious when one examines the informed consent form used in the study. The form states as follows:

"You will smoke your assigned cigarette for the remainder of the study. We will give you more than enough cigarettes to replace the amount that you usually smoke each week. We will ask you to try to smoke only the cigarettes we give you. We will work with you to achieve this goal as it is very important for the study."

Here, subjects are being told that they will smoke their assigned cigarette for the entire six weeks of the study. Moreover, they are informed that their continuously smoking their assigned cigarette brand for six weeks is a goal of the study and is very important. The implication, of course, is that if a subject quits smoking, he or she is not following the protocol and is compromising the study.

Although subjects are later informed that they may quit if they want, nowhere are they encouraged to quit smoking.

Finally, nowhere in the consent form are subjects told that one of the risks of the study is that by continuing to smoke for six weeks rather than quitting, they are subjecting themselves to the risk of health damage from continued smoking. The form does not acknowledge that one of the risks of participation in the study is that it could impede quitting, at least for six weeks.

To me, what puts the nail in the coffin of the ethical breach in this study is that there is an alternative approach that would have yielded the same information, but without violating ethical principles. An alternative approach would have been to encourage all subjects to quit smoking, refer them to smoking cessation clinics or programs, and offer to provide them with nicotine replacement therapy or other treatment. Because the smoking cessation rate with "FDA-approved" or "hospital-mandated" treatment is so low, such a procedure would not have interfered significantly with the ability of the study to answer the research question at hand. If anything, it would have added to the importance of the research by examining the critical question of whether low-nicotine cigarettes can actually enhance smoking cessation.

The rest of the story is that this study violated basic principles of ethical research conduct by failing to provide standard of care treatment for smokers being seen and treated in a medical setting. Even worse, smokers were in fact assigned to smoke, instructed to smoke, and led to believe that the protocol called for them to continue to smoke for the entire study period (which it did). In addition, subjects were not adequately informed of the study risks.