Discussion

Chapter 16: Discussion

Limitations

Although we were able to involve different types of hospitals from five different areas across the country in this study, only sixteen hospitals were included in our sample. We have no way of knowing whether our findings would have been different if we had interviewed individuals at other hospitals. Similarly, we interviewed only medical oncology, radiation oncology, and cardiology patients who were not hospitalized but were receiving their care at outpatient clinics. Most of them were white, and many were more than sixty years of age. It is quite possible that other types of patients would have answered some questions differently from our patients and that healthy research subjects might have had different attitudes, beliefs, or motivations for participation than patients likely to have serious illnesses did. In the In-Depth Interview component of the study, only people who believed they were or had been research participants were included. The responses of people who had chosen not to participate in research, presumably, would be quite different. It should not be assumed, therefore, that our findings necessarily apply to the entire research enterprise.

An important research question in this project was the degree to which present-day patients know whether or not they are research participants. To answer this question, we interviewed patients and asked them whether they believed they were, or had been, participants in research, and then, with their permission, we checked their records for evidence of research participation. Although this approach provides an estimate of the degree to which present-day patients know or remember if they are research subjects, this estimate is likely to be very rough for two sets of reasons. First, interviewing patients in the way we did may not be the most accurate way of gauging their own understanding of participation in research. This is because they were often approached in a busy clinic setting by an interviewer they did not know. It is also likely that these patients were under stress at the time of the interview, either because of their upcoming appointment with their doctors or because of the very illnesses that brought them to the clinic. In addition, because of necessity the Brief Survey was designed to take only five to ten minutes to complete; we asked patients only about current or former research participation with single questions, rather than a series of questions designed to more completely capture those patients who had experience with research. Moreover, following our review of the medical and research records we did not go back to patients and ask them questions about research once we had an understanding of their medical history and documented research experience.

Second, despite significant attempts to gather information from multiple sources, the method of abstracting medical and research records we used may not have been comprehensive enough to locate all relevant evidence of research participation (e.g., records of research may not be retained at the same institution in which the Brief Survey was conducted, or research participation may have been in the distant past and records may no longer be available). A related problem is that some patients may have been enrolled in studies that purposely do not keep records of participation (e.g., studies where confidentiality is paramount). Finally, while trained abstractors examined records for all patients, health professionals on the Advisory Committee staff only reviewed records where patients' responses differed from the results of the initial records review conducted by the trained abstractors. Health professionals had only a one- to two-day window to perform this confirmatory search of documentation at each institution and thus were not always able to review relevant records because they were unavailable on short notice. Because of these reasons, we do not know the degree to which our estimates are accurate regarding the proportion of patients whose responses about research participation differed from what we found in records.

Implications

A striking finding from this study is the frequency with which people with cancer and heart disease appear to come in contact with biomedical research in the course of their medical care. Notably, nearly 40 percent of the patients we talked with either believed they were or had been subjects in research, had records that showed that they were, or had reported that they had been offered the opportunity to be in research but had declined. Moreover, most patients thought that medical research was a good thing. They had favorable attitudes toward medical research generally, they believed that research did not involve unreasonable risks, and they believed that medical research usually or always advances science. Patients who are or had been participants in research had even more positive attitudes about research than those who had not.

There was evidence in this study that many patients feel free to refuse when physicians and researchers ask them to become research subjects. Nearly 200 patients told us that they had been offered an opportunity to participate, but had declined. Moreover, 40 percent of these patients had chosen at some other time to participate in research, indicating that at least some patients are discriminating in terms of the circumstances under which they are willing to participate in research. There also was little evidence that patients felt coerced or manipulated by health care providers or scientific investigators to participate in research. When we asked patients who were subjects if they had felt pressured by others into becoming research participants, these patient-subjects overwhelmingly said no. Not only did they give the impression that the initial decision to enter a research project was theirs, but many also informed us that they had been told frequently by the investigators that they could drop out of the study at any point, and the patients believed that this was so.

Although the vast majority of both African-American and white patients held favorable beliefs about research, such beliefs were held less often by patients who are African-American. Specifically, as compared with white patients, African-American patients were more likely to believe that people are pressured into research and more likely to believe that research poses unreasonable risks. These findings together suggest that for a small number of patients, distrust as a result of the troubled historical experience of African-Americans in research, as exemplified by the Tuskegee syphilis study, may persist.

We learned a great deal from this project about why patients choose to be in research. The overwhelming majority of the patients we interviewed who were participants in research were subjects in studies investigating medical treatments. Almost all of these patients said that they had enrolled in research because they thought it offered them their best chance of personal medical benefit. Moreover, for many of them, their doctors had recommended it. Often these patients had very serious illnesses and had tried many treatments unsuccessfully; the opportunity to be in research offered them hope that improvement might still be possible. Many of these patients specifically said that they had "no choice" but to participate. They had tried everything else to improve their condition, and nothing else had worked. These patients felt constrained to participate because of their medical situation, not by their providers or the research investigators.

Not surprisingly, then, when asked to describe the research project they were in, most of the patient-subjects we talked with described the project as part of their therapy. Although, when asked, these patients appeared to clearly understand which interventions were associated specifically with the research, they also conceived of the research as their medical treatment. And despite the recognition by most of these patients that the goal of the enterprise of medical research generally was to advance science, when asked about their own specific project, they often believed that the project would benefit them.

It is likely that in some, and perhaps in many, of these cases, it was indeed in the patient's medical best interest to be enrolled in a research project. As demonstrated by the recent push for access to investigational drugs on the part of people with HIV infection and other serious illnesses where there may seem to be no truly efficacious standard therapies, many patients believe that their best chance of extending life is to take treatments that are still experimental. In some cases, patient-subjects were participating in treatment studies involving agents available only through research because their illnesses may have had no known efficacious treatments. From the perspective that holds extending life to be the primary concern, it would be in the patients' best interests to be in the research.

It is a separate issue whether participation in research is in a patient's overall best interests. Investigational interventions for devastating, life-threatening illnesses may be a patient's best chance--however small--of extending life. However, this chance may be at the expense of the person's ability to function and enjoy life for the time affected by participation in the research. Furthermore, the history of experimentation demonstrates that such therapies might also shorten life rather than extend it. Unfortunately, we did not pursue whether these sorts of trade-offs were clearly understood by the patient-subjects we interviewed. In chapter 15, we report some data from the RPRP that bear on this question.

That patients viewed their participation as being in their best interests is consistent with patients' profound trust in their physicians, on whom they depend as their lifelines, and who they could not imagine offering something not in their best interests. We heard from several patients the belief that their doctors are the experts and that they know best what would be helpful. If a doctor recommended or even offered research, patients were certainly more inclined to decide to participate. The trust that patients placed in their physicians often was generalized to the medical and research community as a whole. Patient-subjects frequently expressed the belief that an intervention would not even be offered if it did not carry some promise of benefit; many certainly assumed that the intervention would not be offered if it posed significant risks.

It was largely because of this trust that most patient-subjects considered the consent process somewhat incidental to their decision to participate in research. When asked, almost all patients reported that they had been provided with information, their questions had been answered, and they had been satisfied with the consent process. Nevertheless, doctors' recommendations and patients' own beliefs that the research was their best chance or even their only hope made the research an obvious decision for many patients, and the consent process and consent form were viewed as somewhat of a formality.

This framing of research as therapy is consistent with the very language used to describe research projects. We learned that patients attach very different meanings to the different terms associated with medical research. Experiments are considered by patients to involve unproven treatments of greater risk, often invoking the image of human beings as "guinea pigs," while terms such as clinical investigation or study convey less uncertainty to patients and a greater chance of personal benefit.

The design of this study does not allow us to assess whether patients' expectations of benefit from their therapeutic trials were appropriate for the particular studies in which they were enrolled, or whether their expectations were exaggerated or unrealistic. Moreover, if patients' expectations were exaggerated in some way, we have no evidence to discern whether patients overestimated the expected benefit themselves or whether it was investigators who suggested that the research held more promise than was warranted. It is understandable that patients with poor prognoses may read hope into even the slimmest possibility of benefit. It also is understandable that some physicians, uncomfortable with having little to offer their seriously ill patients, might at such times inadvertently impart more hope than the clinical facts, strictly speaking, warrant.

Hope is a delicate and precious commodity for those with life-threatening illnesses. For clinicians, the balance between support of that hope and honesty is often difficult. At the same time, however, there is a world of moral difference between a physician emphasizing--even inappropriately--slim chances, in order to bolster waning hope, and a physician emphasizing slim chances in order to meet a recruitment goal for a clinical investigation. Feeding hope at the expense of candor is one thing; exploiting the desperation of those whose lives hang in the balance is another. Here again, our data are silent. We cannot know, insofar as physicians contributed to unrealistic expectations among these patient-subjects, how often this was the result of well-meaning reassurances or self-interested misrepresentations.

It seems very much related that we found that a small proportion of patients believed they were subjects in research when it appeared they were not, and other patients believed they were not research subjects when records suggested that they were. These confusions about whether a patient was in research occurred almost exclusively when patients were in (or thought they were in) research investigating potential therapeutic interventions. However, we found that these patients covered the full range in terms of education, income, sex, and race; they came from all three medical specialties studied and all types of hospitals.

At least three-quarters of the patients who apparently were mistaken when they reported they were not research subjects had actually signed consent forms authorizing participation in research. In addition to the limitations of our methods described earlier in the chapter, we can only speculate as to why the discrepancy exists between patients' perceptions and their records. Some patients may not have understood our question and may in fact have known they were research subjects all along. Other patients may not have understood what they were doing when they signed the consent form, perhaps believing that it was a consent for treatment. Still other patients may have had an adequate understanding that they were consenting to participate in research at the time they signed the form and then later forgot. This last explanation is not as troubling as the second, in that it suggests the possibility that in at least some cases valid consents were initially obtained, but it does raise questions about the meaningfulness of these patients' rights to withdraw from research. Such questions are obviously more meaningful in ongoing projects that involve continuing exposure to potential risk, in contrast to those studies where research participation is less burdensome, such as studies involving routine follow-up or only a minor change in a regular therapeutic regimen.

It is often the case in clinical research that the participation of ill people in research and the medical treatment they receive for their illnesses are identical. When this occurs, it is not surprising that some patients conflate their being in research with therapy to the point that they no longer understand or remember that they actually are in a research project. Ironically, it may be especially when patient-subjects feel well cared for that they are most likely to feel like a patient only, and not like a research subject. At the same time, many patient-subjects told us of being reminded by research staff that they could leave the project at any time for any reason. It seems doubtful that the patients we interviewed whose self-report of participation was not consistent with research records had such an experience.

Although most of the patients we interviewed listed a chance at medical benefit as a reason for participating in research, many patients also said that they had participated in research to help others. Some patients described the willingness to participate in research as a civic duty; others wanted to help members of their own families at risk for the same conditions, and still others saw being in research as a means of making a shortened life expectancy more meaningful. Participants in survey research and similar research projects were especially likely to say that they had joined in part because there was no reason not to do so, but also because they hoped they could help others or advance science by doing so. Several patients in therapeutic research who appreciated that there was only a slim chance that the research would provide them with personal benefit, offered that, as a result of their participation, they hoped at least that someone down the road would be better off, if not themselves. This willingness of patients to be altruistic should be tapped explicitly when recruiting participants for research, since it might help to underscore for patients that the primary objective of research is to create generalizable scientific knowledge rather than simply to offer them a chance for some medical benefit. In the end, it is only the benefit of furthering knowledge that can be honestly guaranteed to a potential research subject.