Human Subjects in Trauma Research

The primary cause of mortality in children and young adults is trauma. Injuries from motor vehicle accidents, bicycle accidents, home injuries, and violent crime take a large toll of our young, otherwise healthy population. Improvements in the availability of prehospital care transport facilities as well as upgraded emergency and surgical care have allowed an increasing number of trauma victims to survive serious injury.

However, much research still needs to be performed. The primary cause of death in trauma patients is hypovolemia (low amounts of circulating blood). An emergency medicine researcher has proposed a project, based on previous animal experiments, to use a new device which he designed to allow the installation of large amounts of intravenous fluids to be delivered to a trauma patient extremely rapidly.


He has requested to do this research only on those trauma patients who arrive in the emergency department in extremely critical condition. This would not allow informed consent to be obtained, both because the patient would either be in shock or be unconscious and because therapy, if successful, would need to be begun immediately. The IRB has been asked to approve this research.

Commentary

Through research and the application of research findings to patients, medicine has developed into a scientifically based discipline. Today, the field of critical care, and especially critical trauma care, is one of the most vital and important areas of medical research.

And while most of this research can be accomplished under the standard ethical and legal guidelines covering the use of human subjects, one unique area cannot. That is the research which must begin during the acute, crisis stage of medical care following critical injury or sudden illness, where the patient is acutely incompetent due to the medical catastrophe which has occurred.

NONTHERAPEUTIC RESEARCH: INVASIVE AND INCIDENTAL


Two distinct types of research can occur in this situation: nontherapeutic and therapeutic. Nontherapeutic research involves investigations which, by design, can be of no benefit to the individual patient.

This type of research can further be classified as either invasive or incidental. Invasive, nontherapeutic research might involve passing a catheter into the patient's heart, when otherwise unnecessary, to monitor physiologic responses, without feedback for immediate patient care. Or it could mean taking biopsies (samples of tissue) solely for the purpose of later study. On its face, this type of behavior is unethical in the treatment of a critically ill patient who is unable to give consent.


Since it is the physician's duty "to remain the protector of life and health of that person on whom biomedical research is being carried out, clearly this research should not be sanctioned. The question of performing invasive nontherapeutic research on a patient who is pronounced brain dead is beyond the scope of this discussion, except to note that this patient would be a potential supplier of a scarce resource (organs) for other living patients and that this resource could be damaged by such research.

The second type of nontherapeutic research is incidental research. This research usually consists of taking patient specimens or making measurements, without doing any harmful or any extra invasive procedure on the patient. This could include the obtaining of a urine sample from the collection bag for subsequent biochemical assays or the drawing of an additional tube of blood at the time diagnostic blood specimens are being drawn, for later study. As long as no delay in treatment is caused by incidental research, there is no reason this should not be permitted.

THERAPEUTIC RESEARCH


The second type of research is therapeutic, in which there is direct intervention into the cause of the patient's illness. How to proceed in this case may not be as clear. Diagnostic or therapeutic intervention under research protocols in emergency medicine and the emergent aspects of critical care is full of uncertainty.


How will this research affect the patient? What is the effect of current therapy? Is the research necessary? Would a prospective patient voluntarily consent, if he could, to the study? This list of questions and problems only begins to broach the dilemmas presented to potential therapeutic research in emergency medical critical care. How, then, should we proceed? What are our options?

OPTION 1: NO RESEARCH


Not doing research is actually the current situation in many aspects of critical care. Researchers have had a very difficult time in obtaining IRB approval for performing research on human subjects in situations where consent could not be obtained. In addition, because many aspects of emergency and critical care medicine which need research fall under a rare event research methodology making the research itself difficult, investigators have tended to find easier avenues of exploration.

By not doing research, however, we face the risk of continuing to use accepted and in some cases potentially harmful diagnostic and therapeutic measures on our critically ill and injured patients. One example is that of using calcium salts in cardiac arrest situations. Although calcium salts have been a standard first line drug for many years, they have had very little scientific study done on them over the years.


And considerable animal and retrospective human work was needed before any IRB was willing to do a genuine controlled, blinded study of calcium in cardiac arrest. In fact, we now know that it may be calcium blockers rather than calcium salts which are actually helpful to patients in cardiac arrests.

OPTION 2: USE OF ALTERNATIVE METHODOLOGIES


Two types of alternative research methodologies are potentially available. The first is the computer model. This would be the ideal and, certainly, the easiest system with which to experiment. It is easily manipulated, without any need for consent, available at virtually any time period, and without the mess of blood or any other bodily fluids.

The computer model would be the researchers dream-come-true. Unfortunately, we do not know enough about either human physiology or human disease states even to suggest the simplest workable computer model for the human body. In fact, we do not even know enough about computer modeling for engineers to model some human-designed systems.

The other alternative methodology, and the one most commonly used for critical care research, is the animal model. Most research in critical care at this time is done on a variety of animal models. However, it is difficult in many cases to find a suitable animal model for specific disease states. And, even once one is designed, it will be difficult, if not impossible, to accurately determine how closely the model itself, the benefits of therapy or accuracy of diagnosis on this model, and the resultant outcomes correlate to the human system.


One example is research into cardiopulmonary resus­citation. Much of this work has been done in dog models. However, because of the difference in even the basic anatomy of the dog's chest and internal thoracic structure, it is not clear which of the therapies which are successful on the dog in cardiac arrest can be successfully applied to humans. Finally, while not getting into the controversy surrounding animal experimentation, we must recognize that animals don't benefit at all from research into emergency or critical care medicine.

OPTION 3: UNRESTRICTED HUMAN RESEARCH


A third option is to have research on human subjects unrestricted by requirements for consent or reviews other than for research validity and medical necessity. This can be defended on the grounds that it is generally accepted that society approves of, desires, and benefits from medical experimentation and research.

Of course, some individuals in society do not approve of or desire intervention from allopathic or osteopathic medical practitioners, and so may neither desire nor approve of medical experimentation. Ex­amples of this group might include herbalists, practitioners of "traditional" medicine, or Jehovah's Witnesses. It should be noted, however, that while these individuals may not expect to benefit from research in emergency and critical care medicine, they are beneficiaries of medical advances, including those in the areas of public health and disease prevention and control.

At any rate, at least, those presenting for care to emer­gency departments and the prehospital care system are presumably those who would agree (or their relatives or friends who sent them in for treatment would agree) that advances in medical care would benefit them.


This option, though, ignores the fact that the well-being of the individual must be the primary consideration, both ethically and medically. It is anathema to presume to follow the Nazi example in human research of ignoring the individual for the supposed benefit of society. The rights of the individual are too important to be overridden. Therefore, unrestricted human research is unacceptable.

OPTION 4: PATIENT-CENTERED APPROACH TO RESEARCH ON HUMAN SUBJECTS


The fourth option is to recognize again that individuals presenting as patients to the emergency medical system are those who would benefit from advances in emergency and critical care medicine. This knowledge can then be utilized in a manner to pro­spectively review research proposals in this area with the view toward the patient being the primary beneficiary.

The standard composition of IRBs is that of noninvolved, knowl­edgeable individuals. And although they might not always follow either prescribed or ideal behavior, in general they do an adequate job. They are competent to judge the four parameters necessary to approve, prospectively, human subject research in the area of emergency and critical care medicine by applying patient self-interest and the doctrine of implied consent.

First is the need to determine whether all applicable prior nonhuman studies have been carried out and that further studies could not be adequately performed, using an alternative methodology. This requires a scientific sophistication among the members of the IRB as well as their willingness to review scientific literature and request expert testimony from other, noninvolved scientific specialists.


The second and related issue is the determination that the results so far obtained indicate a potential benefit to humans. This, of course, presumes that the research is therapeutic in nature. However, as previously noted, our knowledge of even basic physiology surrounding emergency critical care is sorely lacking. Therefore, there may really be no benefit to the individual from purely physiologic studies. This concern, however, can be addressed in a manner similar to that which is described for the therapeutic interventions, by looking at the risk (and occasionally benefit) to the patient. This second factor, that of potential benefits to humans, is a risky area.

Although simplistically treated in most codes governing human subject research, it must be recognized that there is a great uncertainty in all biomedical research. Many advances have come unexpectedly from fortuitous observa­tions. Therefore, this second factor may have to be interpreted somewhat liberally.

Finally, the IRB can make a decision to proceed, based upon both the type of study involved and the patient's self-interest. There are basically two somewhat different sit­uations which have to be addressed separately. Approval in both cases would be based on the probability of informed consent by the competent patient. And while both involve either diagnostic or therapeutic intervention, there are different designs which make it necessary for them to be treated differently.

In the first situation, where both accepted and research methodologies for diagnosis or therapy are performed concurrently, these maneuvers and methodologies can be classified into the categories of "beneficial," "neutral," or "harmful," as they affect the patient. This would be based on both initial nonhuman studies and literature reviews.


Once this is done, it is plain that the reasonable person would want the research done in any of the cases where it was shown to be potentially beneficial to him. It could also be reasonably assumed that there would be no objections if the research was considered to be neutral (that is, neither beneficial nor harmful). If the research is definitely harmful and the accepted diagnostic or therapeutic maneuver is considered to be either beneficial or neutral for the patient, it's unlikely that a reasonable person would accept such a situation, and this research should not be performed. Two questions arise under this format.

The first involves the situation in which both the accepted treatment and the research are considered potentially harmful to the patient. An example might be standard treatment for a gunshot wound through both cerebral hemispheres.

Because of the nature of the injury, standard therapy would be "harmful," in that the chance of the patient surviving (or, in some cases, even of receiving care) is virtually nil. On the other hand, the proposed diagnostic research may be that of a radical craniectomy (removal of the entire top part of the skull from the brain) which has also been shown to have an extremely high morbidity and mortality.

In this case, although experience has shown that neither the accepted treatment nor the research protocol has proven beneficial to patients, there is a slim chance that this combination of circumstances could benefit the patient. This is a question that cannot be answered clearly (and, thankfully, is a rare circumstance) but would have to be examined on a case-by-case basis.


The second problem arises with the research that is thought to be beneficial to the patient while preliminary (animal) studies suggest that the accepted therapy or diagnostic maneuver is harmful. The situation then could be that either the research protocol would completely counteract the expected harm that could be done by the accepted methods or they would neutralize each other.

In the latter case (for example, administering calcium and anticalcium medications during cardiac arrests), should the "accepted care" be allowed to continue? This particular problem has terminated at an early, and scientifically inconclusive, stage a number of important experiments. On the other hand, if there is an overwhelming amount of evidence to support the beneficial effects of the new technique, this must be factored into acceptance of both the research and the protocol design.

Some of the same factors also occur in the second research design, that of the controlled trial. Here, the accepted diagnostic or treatment maneuver is compared against the research methodology. At least in acute situations, only one of the two would be used.


Probable patient acceptance in this case would be somewhat altered and slightly more restrictive. In the case where both accepted and research methodologies are potentially harmful, the same situation applies here as applies with simultaneous research. In a case where the accepted methodology is beneficial to the patient and the research methodology has neutral effects, it is most probable that the patient would not accept a potential denial of known beneficial effects.

Similarly, in cases where the research methodology is thought to have beneficial effects and the accepted therapy neutral or harmful effects, for example, the case of calcium salts in CPR described above, there is a question as to how the IRB should treat continuation of accepted methodology in the research protocol. However, once again, without sci­entific validation of the benefit of the research methodologies in humans as compared to methodologies accepted currently, the new methodologies will probably not be widely accepted and not widely benefit subsequent patient populations.

APPLICATION TO THIS CASE


In the case under discussion, either the controlled trial scenario or the simulta­neous research scenario could apply. In an example of a controlled trial with standard intravenous therapy, standard equipment with lower flow rates could be used on half of the patients presenting with critical trauma, and the new high-flow research equipment could be used on the rest of the patients.


The format as described in Figure could be applied to this situation as long as equity was preserved; that is, no particular subpopulation would assume an unfair burden of being research subjects. This, however, should not be a problem with a scientifically valid study design. In this case, previous studies had suggested that the research methodology could potentially be beneficial to the patient.

The benefit of both "accepted" therapy and the research protocol in these critically injured patients is dependent upon the nature of their injury and the time that has elapsed since the injury. Since the level of benefit to the patient of both therapies appear to benefit the patients at least equally, the IRB can approve this researcher's project. There is one caveat, however; the individual would receive standard, "accepted" therapy in the prehospital care system until arrival in the emergency department.

This brings us to the other scenario which would be that of simultaneous or sequential trials. If all patients, once they arrived in the emergency department, received the new high-volume therapy after receiving standard intravenous fluid therapy in the prehospital care system, the model described in Figure would apply. Once again, that the high-volume fluid therapy was potentially beneficial to the patient would obviate the need to look any farther before approving this research.


One other duty, though, is the responsibility of the IRB, that is, the monitoring of the research, once in progress, for complications and adverse effects on the patient. As with any research, if these are found, the project would have to be reevaluated by use of the same formulations, before it could be allowed to continue.

CONCLUSIONS


Three criteria, therefore, must underlie any IRB selection of research protocols to be used with human subjects in emergency medicine critical care. First, all alternative methodologies of testing hypotheses should have been employed prior to initiation of tests with human subjects.

Second, the data must show that patients could potentially benefit from the new research methodology. And third, underlying the entire exercise of determining appropriate human subject protocols is that the probable individual patient acceptance, based on the idea of individual benefit, must be paramount.

No comments: