Response by Jennifer Horner Catt

Four principles guide the ethical analysis in this essay. First, ASHA' s Code of Ethics asserts that speech-language pathologists must hold paramount the welfare of persons they serve professionally. Second, the Declaration of Helsinki (World Medical Association, 1989) states that every person who is both a patient and a clinical research subject (even those in a control group) should be assured of the best proven diagnostic and therapeutic method; in short, clinical research should have an intended benefit to the patient. Third, the Council for International Organizations of Medical Sciences (CIOMS) stipulates that to be ethical, a research study must be scientifically sound.

A fourth guiding principle involves the nature of the patient-therapist relationship. Presumably all patients recruited for the proposed aphasia research study are viable treatment candidates because they have aphasia of various types and degrees of severity. They will (or should) have an established relationship with the speech-language pathologist for the purpose of receiving care. Once established, the patient-therapist relationship implies not only the patient' s trust in and reliance on the therapist but also the therapist' s obligation to care for the patient--to act only for the patient' s good. "The good of the patient is the end and purpose of that relationship." (Pellegrino & Thomasma, 1993, p. 43)

Implicit in the patient-therapist relationship is that the therapist will render standard treatment, which stands in contrast to clinical research and nontherapeutic research as follows. Using standard treatment, therapists apply diagnostic or therapeutic techniques designed solely to enhance the well-being of individual patients. In clinical research, therapists apply procedures of potential diagnostic or therapeutic value to patient-subjects in order to seek to produce generalizable knowledge. In nontherapeutic research, researchers seek to produce generalizable knowledge by applying procedures without intending to benefit directly patient-subjects in order to produce generalizable knowledge. (Goldner, 1993, p. 66)

As designed, does this proposed research meet these widely recognized ethical standards? This essay will address a series of questions to elucidate a number of potential ethical problems with the proposed aphasia research study (hereinafter Clinical Trial).

Is the study necessary? No.

The first issue that arises from the Clinical Trial is whether this particular study is necessary. Persuasive evidence exists that treatment for aphasia is more effective than no treatment, and that acute, intensive, and individualized treatment is superior to subacute or chronic, non-intensive, and group treatment (Holland, et al. 1996; Robey, 1998; cf. Pedersen, et al., 1995). An additional large-scale study to reinforce these conclusions potentially will use resources that could be used more wisely to test more focused hypotheses.

Is family education, aphasia treatment? No.

The second related issue is whether the proposed design assures that patients will receive the therapy to which they are entitled. In the family education arm, the speech-language pathologist will educate and "train" the family in three sessions. This arm does not involve the aphasic patient in a bona fide therapeutic relationship and it does not stipulate the type, intensity or duration of family "intervention" with the patient. Lacking these details, this arm is not comparable to the "volunteer" treatments studied by previous researchers (see David, et al., 1982; Meikle, et al., 1979; Hartman & Landau, 1987). In short, family education is not aphasia treatment. As a result, subjects in the family education arm will serve as no-treatment controls for the individual and group arms of the Clinical Trial.

Is family education integral to standard therapy? Yes.

Family education and training is currently part of the standard of care for aphasia. Family education is not a distinct therapeutic regimen for aphasic persons; rather, it is integral to aphasia therapy. Thus, the compartmentalization of family education as a third arm of the Clinical

Trial--and by implication, its exclusion from the individual and group therapy arms--appears to be contrived for the purpose of the Clinical Trial. This contrivance places all arms of the Clinical Trial below the prevailing standard of care for aphasia.

Are there potential risks to patients? Yes.

In their vulnerable condition, newly aphasic individuals are unlikely to appreciate fully the consequences of their communication disorder, and may not understand the ramifications of agreeing to participate in a randomized clinical trial. As noted above, the Clinical Trial appears to exclude family education from two arms, and compartmentalizes family education in a third arm. As described, the Clinical Trial has no provision for crossover or deferred therapy. Thus, the risk of harm to those patients who volunteer for the Clinical Trial is not trivial.

Although some patients might receive therapy of potential value to them, none will receive standard therapy, and others will receive no bona fide therapy at all.

The risks to patients include unabating aphasia (with associated disorders), the use of maladaptive strategies, frustration and isolation. As Annas & Glantz (1986) remind us, "[r]esearch that may seem trivial to use in terms of risk, discomfort, disorientation, or dehumanizing effects may not seem so trivial to this vulnerable and often frightened population." (p. 1157) Patient (or surrogate) consent to participate in the Clinical Trial does not vindicate the use of a research design that places patients at greater than minimal risk for harm without offsetting benefits. (Human Radiation Experiments, 1996) Granting agency or institutional review board (IRB) approval does not vindicate the therapist' s failure to hold paramount the welfare of her patient. Because this proposed Clinical Trial fails to assure that each patient-subject will receive necessary and sufficient aphasia treatment for his or her disorder, it is ethically suspect.

Does clinical equipoise exist? No.

Clinical equipoise refers to a lack of consensus or "genuine uncertainty" within the expert community about the comparative merits of the various interventions. In effect, the choice among treatments A, B, and C is indifferent; it is an "honest null hypothesis." (Freedman, 1987, pp. 141-144) In the clinical setting, speech-language pathologists are obligated ethically to use the best proven or accepted therapy for the good of the patient--there is no hypothesis; the concept of equipoise does not apply. In contrast, when embarking on clinical research, therapists must have an hypothesis and must be in a state of equipoise before offering to randomize patients to one "treatment" or another.

( See Figure .)

The requirement of equipoise creates problems for nontherapeutic conditions such as the family education arm of this Clinical Trial. On the one hand, if researchers propose to use a novel ("first generation") treatment for previously untreatable conditions, it is possible for equipoise to exist. In such a case, the research question is whether the new treatment is better than nothing. (Freedman, 1996, p. 250). The scenario does not apply to the Clinical Trial because we know that aphasia is treatable, and because the individual and group therapy arms are not new (they are proven to be effective). Aphasia therapy research is well past its infancy so equipoise of this first type does not exist.

On the other hand, if researchers plan to assess the efficacy of an innovative ("second-generation") treatment, the new treatment should be tested against the old (standard) treatment rather than against a nontherapeutic arm. (Freedman, 1996, p. 243). The research question is whether the new treatment is either no different or better than the treatment of proven or accepted value. The problem for our Clinical Trial is that by assigning patients to the family education arm, researchers not only withhold a treatment of proven and accepted value, they give patients a nontherapeutic substitute. (Rothman & Michels, 1994 p. 394) Inclusion of the family education arm of the Clinical Trial -- because it essentially offers no therapy to aphasic individuals in need of therapy--clearly vitiates a fundamental principle of our Code of Ethics and accepted norms of clinical research.

Does the Clinical Trial describe the substance of therapy? No.

The Clinical Trial describes the administrative characteristics of intervention, i.e., the apportionment of therapists' time, the individual versus group milieu, the inpatient or outpatient setting, and the minimum (but not the maximum or optimal) number of sessions. Presumably, researchers will analyze the within-group effects by using pre-post behavioral measurements, and between-group effects by comparing the outcomes across the Clinical Trial' s arms.

This Clinical Trial fails--as did previously published group aphasia efficacy studies--to describe the substantive content of therapy. As a result, the Clinical Trial will fail to illuminate the interaction of individual patient attributes and individual treatment techniques. It will obfuscate the therapeutic responsivity of individual patients or aphasia subtypes because it is designed to yield only group statistics.

Is the Clinical Trial an efficient use of resources? No.

As such, what might we learn from these inter-group comparisons? We may learn that treated patients show significantly better outcomes than non-treated patients, that individual therapy has a modest advantage over group therapy, that both individual and group therapy yield a greater effect than no therapy, that those patients treated for longer durations have better outcomes, and so on. We may also learn that initial severity of aphasia is the single most powerful predictor of ultimate aphasia outcome. The problem is--according to Holland, Wertz, Robey and others--the answers to these questions are already known. The inescapable conclusion is that societal resources might be used more efficiently on single-subject designs or small group studies with focused hypotheses (Robey, 1998).

Is the Clinical Trial replicable? No.

This Clinical Trial' s biggest problem is that it will fail to answer our most vital question in aphasia research, namely, what types of treatments work for which types of patients? This Clinical Trial, by design, will mask within-group individual differences. By merely specifying "how" the therapy will be rendered and failing to specify "what" the therapy actually will be, the study may yield data of use to administrators and institutions, but not to patients and clinicians. In the absence of specific information about aphasic subjects and therapeutic techniques, the study will not be replicable--not replicable by the therapists who must treat individual aphasic persons, and not replicable by clinical researchers who may wish to study focused hypotheses.

Summary

This Clinical Trial may be characterized either as a health services study or as an outcome study. It is not a treatment trial because it does not specify the substance of the treatment. It is ethically suspect because the family education arm satisfies neither the standard of care nor the requirement of equipoise. It is not a scientifically sound study because it is not replicable. Furthermore, by not offering all patients the best proven or accepted treatment, it does not hold paramount the welfare of individual aphasic patients who trust their therapists to care professionally for them. In order to remedy these problems, the Clinical Trial should, at a minimum, 1) incorporate family education in the individual and group treatment arms in a manner consistent with the prevailing standard of care, 2) eliminate the family education arm, and 3) both tailor and specify the within-group treatments to assure replicability.

References

1. American Speech-Language Hearing Association, Code of Ethics (last visited Dec. 26, 1998).

2. World Medical Association, Declaration of Helsinki IV, 41st World Medical Assembly (Hong Kong, Sept. 1989), reprinted in Annas, G.J. & Grodin, M.A., The Nazi Doctors and the Nuremberg Code: Human Rights in Human Experimentation (1992), p. 330.

3. Council for International Organizations of Medical Sciences (CIOMS), International Guidelines for Ethical Review of Epidemiological Studies (Geneva 1991) para. 40, 19 Law, Med. Health Care 247 (1991).

4. Pellegrino, E.D. & Thomasma, D.C., The Virtues of Medical Practice. New York: Oxford University Press (1993), pp. 42-44 (describing the internal morality of professional care, 1) the inequality of the relationship, 2) the fiduciary nature of the relationship, 3) the moral nature of medical decisions, 4) the nature of professional knowledge, and 5) the ineradicable moral complicity of the clinician in whatever happens to the patient).

5. Goldner, J. An overview of legal controls on human experimentation and the regulatory implications of taking Professor Katz seriously, St. Louis U. L.J. 1993;38:63, citing Capron, A.M., Different compensation approaches to bad outcomes from standard treatment, innovative treatment, and research. In Siegler, Mark et al. (eds.), Medical Innovation and Bad Outcomes: Legal, Social and Ethical Responses (1987), p. 149.

6. Holland A., et al., Treatment efficacy: Aphasia. J. Speech Hear Res 1996;39(5):S27-S36 ("people who become aphasic following a single, left-hemisphere, thromboembolic stroke and who receive at least 3 hours of treatment each week for at least 5 months, regardless of the time post-onset of stroke, make significantly more improvement than people with aphasia who are not treated" p. S30).

7. Robey, R.R., A meta-analysis of clinical outcomes in the treatment of aphasia. J Speech Lang Hear Res 1998;41:172-187 (in the acute phase, "the average effect for treated recovery...was nearly twice that for untreated recovery" p. 176; furthermore, "the more intense is treatment,

the greater the change" p. 180).

8. Pedersen et al., Aphasia in acute stroke: Incidence, determinants, and recovery, Ann Neurol 1995;38:659-666 (contending that there was no difference between those patients who did and did not receive aphasia therapy, though acknowledging that the study was not designed to investigate the effect of aphasia therapy). See also Wertz R.T., Aphasia in acute stroke Incidence, determinants, and recovery (Letter) Ann Neurol 1996;40:129-130(refuting Pedersen et al.' s claim); Pedersen et al. Reply (Letter) Ann Neurol 1996;40:129-130 (arguing that "no treatment" is "not a good control").

9. Note: Studies using volunteers found no significant difference in outcome between patients treated by speech therapists or volunteers. David et al (1982); Meikle et al. (1979); Hartman & Landau (1987). Because none of these studies used a no-treatment control group, no conclusion is possible regarding aphasia treatment efficacy. "No difference" using such a design merely tells us whether "one treatment is the same as, better than, or worse than the other." Wertz (1996) p. 129.

10. David R., Enderby, P. & Bainton, D. Treatment of acquired aphasia: Speech therapists and volunteer compared. J Neurol Neurosurg Psychiatry 1982;45:957-961.

11. Meikle, M., Wechsler E., Tupper A., et al. Comparative trial of volunteer and professional treatments of dysphasia after stroke. Br Med J 1979;2:87-89.

12. Hartman J., Landau W.M. Comparison of formal language therapy with supportive counseling for aphasia due to acute vascular accident. Arch Neurol 1987;44:646-649.

13. Wertz R.T., Aphasia in acute stroke Incidence, determinants, and recovery (Letter) Ann Neurol 1996;40:129.

14. Annas, G.J. & Glantz, L.H., Rules for research in nursing homes, New Engl J Med 1986;315:1157.

15. Advisory Committee on Human Radiation Experiments, Research ethics and the medical profession: Report of the advisory committee on human radiation experiments, JAMA 1996;276:403.

"Physician-researchers are often torn between the demands of research and the needs of particular patients. Today this tension has taken on special significance with the growth of research at the bedside and the frequency with which the medical care of seriously ill patients is

intertwined with clinical research." (p. 408).

"Institutional review boards should have the responsibility to determine that the sciences is of a quality to warrant the imposition of risk or inconvenience on human subjects and, in the case of research that purports to offer a prospect of medical benefit to subjects, to determine that participating affords patient-subjects at least as good an opportunity of securing the medical benefit as would be available to them without participating in research." p. 409 [emphasis added]

16. Freedman, B. Equipoise and the ethics of clinical research. New Eng J Med 1987;317:141. See also Freedman, B., et al., Placebo orthodoxy in clinical research I: Empirical and methodological myths, J. Law Med & Ethics 1996;24:243; Freedman, B., et al., Placebo orthodoxy in clinical research III: Ethical, legal, and regulatory myths, J. Law Med. & Ethics 1996;24:252.

17. Rothman, K.J. & Michels, K.B., The continuing unethical use of placebo controls, New Eng J Med 1994;331:394.

Case Study

Responses

• Jason Karlawish
• Jeri Logemann