Evidence-Based Practice Forum

Ethical Considerations Related to

Evidence-Based Practice

Charles Christiansen,

E
thical issues are part of every moral one. Professional caregivers are
health care encounter. Moral prin- Jennie Q. Lou responsible for practicing in a manner
ciples, such as truth, fairness, that keeps the patient’s interest foremost
doing the right thing, avoiding harm, Charles Christiansen, EdD, OTR, FAOTA, is by achieving the greatest good and avoid-
and respecting autonomy, lie at the heart George T. Bryan Distinguished Professor ing harm in the process. In their pledges
of these ethical concerns. Each patient and Dean, School of Allied Health to practice ethically, professionals already
care encounter includes issues related to Sciences, University of Texas Medical commit to making decisions that are
the established practices of the discipline, Branch, Galveston, Texas. right for a given patient at a given time.
what the patient prefers, concern for Rogers (1983) stated it thusly, “The ulti-
quality of life, and contextual features Jennie Q. Lou, MD, MSc, OTR/L, is Associate mate question we, as clinicians, are chal-
(Jonsen, Siegler, & Winslade, 1998; Professor of Occupational & Public lenged to answer is: What, among the
Kornblau & Starling, 2000). The contex- Health, College of Allied Health, Health many things that could be done for this
tual features include the backdrop or per- Professions Division, Nova Southeastern patient, ought to be done? This is an eth-
sonal story that places the experience University, 3200 S. University Drive, Ft. ical question” (p. 602).
within a real-life context for the patient. Lauderdale, Florida 33328; Given the widespread and growing
Driven by increased accountability [email protected]. interest in evidence-based practice, it
and the widespread and convenient might appear to an outside observer as
availability of information spawned by This article was accepted for publication though the health care community had
the Information Age, the health care February 8, 2001. just realized that the careful and objec-
community is turning increased atten- tive examination of evidence provides a
tion toward evaluating established prac- Associate Editor sounder basis for making decisions than
tices. This movement, which focuses on Linda Tickle-Degnen tradition, conjecture, or authority.
searching and appraising available evi- However, although a professional’s oblig-
dence on the advantages and disadvan- ation to stay fully informed has existed
tages of various intervention options, is ous levels of existing evidence underlying for centuries, only recently has a practi-
described by occupational therapy and a given intervention approach and care- cal means for doing so (e.g., the personal
other health care professions as evidence- fully appraising that evidence as it computer, the World Wide Web) created
based practice. applies to a specific patient encounter. the environment of expectation and
Typically, the method involves carefully accountability necessary to drive the evi-
Evidence-Based Practice: formulating a clinical question, finding dence-based practice movement to its
Origins and Processes evidence that bears on the question, and current level of influence. As with so
Evidence-based practice is rooted in evaluating the evidence without bias and many endeavors pursued with enthusi-
medicine but has quickly been embraced applying it as appropriate to a given asm, the fervor that accompanies evi-
by the entire health care community. As patient (Tickle-Degnen, 1999). dence-based practice may cause us to
defined by leading authorities (Sackett, Implementing the process of evi- overlook its limitations. We may neglect
Rosenberg, Gray, Haynes, & Richard- dence-based practice has implications for to apply to evidence-based practice the
son, 1996), evidence-based practice is resource use, for professional credibility, fundamental principles of objective
“the conscientious, explicit, and judi- and for improving outcomes. These analysis that are so central to the process
cious use of current best evidence in worthwhile benefits are often emphasized itself. In our zeal to be objective and
making decisions about the care of indi- without sufficient attention to the under- informed, we may forget that clinical
vidual patients” (p. 71). As Holm (2000) lying ethical principles involved. One can decision making, at its core, is an ethical
noted, the process of evidence-based argue that the primary reason for imple- matter, and we may lose sight of the eth-
practice involves being aware of the vari- menting evidence-based practice is a ical dilemmas hidden beneath our efforts

The American Journal of Occupational Therapy 345

Downloaded from http://ajot.aota.org on 04/19/2020 Terms of use: http://AOTA.org/terms
 to produce the most effective medical, of interest to occupational therapists are process of evidence-based practice itself
rehabilitation, and health outcomes. difficult to define or measure, thus mak- does not provide a means for judging the
ing some studies that address important political wisdom of allocating resources,
Evidence-Based Practice and issues more difficult. For example, pain, yet decisions must be made. It can be
Ethical Issues in Occupational which can often result in occupation lim- argued, therefore, that it is even more
Therapy and Rehabilitation itations and diminish social participation, important that consumers be included in
In the sections to follow, we identify and is highly subjective and difficult to mea- the process when competing political
discuss some ethical issues called forth by sure. Similarly, a single, well-validated, interests are decided.
the process of evidence-based practice. quality of life measure has not yet
The focus is on occupational therapy and emerged with sufficient recognition to Evaluating Evidence
rehabilitation, but the issues may be serve as a gold standard. These definitional According to the process of appraising
equally valid for other domains of prac- and measurement issues in rehabilitation available information, evidence should be
tice. Our identification of these issues is must be addressed if the International ranked on the basis of the manner in
not intended to diminish the importance Classification of Functioning, Disability which it was obtained (Holm, 2000).
of the evidence-based practice but, rather, and Health (World Health Organization, Because some methods that have less rig-
to point out that ethical issues are embed- 2000) is to be of maximal use. This tax- orous controls (e.g., expert opinion, anec-
ded throughout health care practices. To onomy emphasizes that disability is the dotal experience) are discouraged in favor
the extent that we are aware of them, we product of environmental circumstances of studies that meet or approach the gold
are better prepared to avoid the dilemmas and functional impairment, which indi- standard (prospective randomized con-
and complications that result when they vidually or in combination can result in trolled trials), practitioners should not
arise. reduced quality of life through limitation lose sight of the values brought by each
of activity and reduced social participa- approach (Culpepper & Gilbert, 1999).
The Status of Evidence tion. For example, within the traditional
As in medicine, much of traditional reha- More research is needed to provide model described by evidence-based prac-
bilitation practice unfortunately lacks evidence to guide interventions that tice, a study using Functional Indepen-
sufficient research for the careful, well- reduce disability for specific conditions, dence Measure (FIM™1) scores would be
documented analyses expected in making populations, and settings. Meanwhile, rated more highly than one based on the
confident, evidence-based decisions. practitioners must base their interven- judgment of a patient or family members
More often than not, available evidence tions on clinical tradition, anecdotal (Granger, 1998). In this instance, issues of
in many health care arenas consists most- results, and expert opinion. In working autonomy come into play. Should dis-
ly of expert opinions. Despite important with patients to plan treatment goals and charge decisions be made on the basis of
advances in occupational therapy research interventions, therapists must acknowl- an FIM score or patient satisfaction with
over the past quarter century, only a few edge the basis upon which a given inter- outcomes achieved? In other words,
areas of intervention have amassed suffi- vention approach is made. instances may exist where the value of
cient data to enable research-based confi- autonomy (the patient’s right to choose)
dence in treatment interventions (Holm, Consumer Autonomy is more important than adherence to an
2000; Law & Baum, 1998; Tickle- Patient and family participation in plan- objective research measure.
Degnen, 1999). ning intervention (autonomy) is an
Moreover, some occupational thera- important dimension of rehabilitation Conflict of Interest
py practitioners underappreciate the practice (Tickle-Degnen, 2000), yet this Although studies designed with high lev-
value of evidence-based practice involvement often does not extend to els of control for competing hypotheses
(Dubouloz, Egan, Vallerand, & von decision making about research agendas. are emphasized as necessary for achieving
Zweck, 1999). The reasons for this lack Although the National Institutes of valid conclusions, the ethical threats
of appreciation remain unclear. It may Health have made significant steps involved in conducting such research are
represent a discomfort with the uncer- toward involving consumers in processes sometimes understated. Again, scientists
tainty associated with interpreting proba- designed to protect human subjects, this rather than patients mostly determine
bilities for relationships and effects of participation (e.g., on Institutional decisions about who participates in stud-
interventions. Many practitioners expect Review Boards) comes after studies have ies and whether intervention continues,
research to provide explanations of the already been designed. Ideally, consumers thus creating a conflict of interest
underlying mechanisms of change. In the should participate more in strategic deci- (Gordon, Sugarman, & Kass, 1998).
new world of evidence-based practice, sions related to the allocation of research Conflicts of interest may be repre-
however, practitioners must become funds for conditions, populations, and sented in other ways associated with evi-
more comfortable with the ambiguity settings. dence-based practice. When scientists are
associated with probability tables. This participation should also induced to conduct clinical trials that
The Institute of Medicine acknowl- include decision making under difficult have associated financial benefits, ques-
edged the dearth of evidence to inform political circumstances, as, for example,
rehabilitation practice in a report entitled when competing or contradictory claims 1
FIM™ is a trademark of the Uniform Data
Enabling America (Brandt & Pope, by different groups result in different System for Medical Rehabilitation, a division
1997). In practice, some of the outcomes interpretations of available evidence. The of UB Foundation Activities, Inc.

346 May/June 2001, Volume 55, Number 3

Downloaded from http://ajot.aota.org on 04/19/2020 Terms of use: http://AOTA.org/terms
 tions arise about the validity of conclu- because of cognitive disabilities. These for the right to terminate his treatment
sions. Although such enticements are populations might face additional risk against medical advice. Because his condi-
most often associated with studies of because of their inability to fully evaluate tion renders him unable to continue pur-
pharmaceutical agents, they can also the risks of research participation. The suing his passion for sculpture, he reasons
occur when clinical investigators in reha- Human Research Ethics Group specifi- that life no longer provides sufficient
bilitation have fiduciary relationships cally recommended improvements in meaning to warrant its continuation. In
with manufacturers of medical and reha- review board requirements to provide a litigation against his caregivers, he sues for
bilitation devices. Researchers should clear and complete statement of the the right to determine his level of care
take care to avoid situations where finan- expected benefits and risks associated and prevails. This drama vividly illustrates
cial inducements or other gains may cre- with enrollment in a study. the current dilemma between interven-
ate the potential for perceived or actual The point is that as the pressure to tions aimed at prolonging life and those
conflict of interest. To be ethically sound, produce evidence to justify existing prac- aimed at providing a life with greater
an investigator should have no actual or tices mounts, researchers may be more quality. The same biases are evident in
potential sources of bias associated with likely to infringe on ethical boundaries in decisions affecting research support. Far
personal or facility gain. The Food and their zeal to collect it. If risks are under- more attention (and resources) are devot-
Drug Administration has issued strict stated and potential benefits are exagger- ed to studying approaches aimed at saving
rules and regulations governing disclosure ated, the cumulative effect may be to bias or continuing lives in proportion to stud-
of financial conflict of interest by scien- significantly and imprudently the deci- ies that address issues related to adapta-
tists involved in clinical trials involving sion of a potential research participant in tion following devastating injury and its
medical devices or drugs (Department of the direction of participation. Here, the effect on quality of life.
Health and Human Services, 1998). ethical principle of greatest societal good
Researchers should also ensure that (which is the expected consequence of Using Evidence
disciplinary loyalty does not introduce conducting the research even in the face Once research evidence has been generat-
potential bias into trials involving inter- of potential risk to participants) is in ed, the relevance of its application to spe-
ventions associated with a given profes- conflict with the principle of autonomy. cific situations must be determined. In
sion. What is of concern to the patient is rehabilitation, the extent to which out-
not the standing of a particular profes- Biases in Research Selection comes are satisfactory highly depends on
sional group but, rather, the efficacy of a Another nettlesome problem in evidence contextual factors of a social and psycho-
particular intervention. generation concerns biases in the selec- logical nature. Outcomes are thus harder
tion and support for research to be to measure and compare, making the
Informed Consent undertaken. Earlier, we addressed the application of group data to individual
Ethical oversights also extend to the mat- need for consumer participation in strate- patients difficult because conditions are
ter of informed consent for those invited gic decision making. Here we discuss never exactly the same for any two
to participate in clinical research. The how the traditions of disciplines and cul- patients.
idea of informed consent is again tures may create biases in the selection Difficult-to-apply data, however, are
grounded in the principle of autonomy. and support of research. better than none at all. The paucity of
This principle respects the right of a per- In particular, we maintain that studies in rehabilitation increases the like-
son to weigh the pros and cons of a deci- research in rehabilitation methods, social lihood that judgments and decisions will
sion and to make a choice on the basis of reintegration, and palliative care may not be based on conjecture or tradition. As
his or her consideration of alternatives. occur as frequently as other types of more the rule than the exception, insuffi-
To make a fully informed choice, poten- research because of prevailing models of cient attention may be given to practices
tial research participants must be advised care in western medicine. These models that have not been validated by research.
of the expected benefits and risks of the tend to emphasize life extension and sur- Additionally, the nature of some occupa-
research trials in which they are asked to vival rather than life quality and accep- tional therapy interventions is such that
participate. tance of death. The net result is that the the perceived risk to the patient is viewed
The Human Research Ethics Group, potential benefits of approaches that as minimal. Here, a bias toward action
administered by the Center for Bioethics emphasize life quality rather than life seems justified by the assumption that
at the University of Pennsylvania Health extension may be underappreciated and, the intervention can do no harm, which
System, extensively reviewed the status of thus, used less frequently, even though we now know from our expanding
existing human subjects protections and they may represent the “right thing to knowledge base that this is not always
federal reports with the aim of making do” under the circumstances. Here the true. Although we know well that poorly
recommendations for improvement and selection of research to provide evidence fabricated splints can cause pressure sores,
reform. As noted in a subsequent report for practice interventions can reflect a we are less confident in concluding that
(Moreno, Caplan, & Wolpe, 1998), fed- bias that leads to practices that may not failed therapeutic encounters affect self-
eral regulations had not been revised in be “right” for a particular patient when image, personal identity, and morale.
nearly 2 decades, and considerable evi- that patient’s quality of life is considered. Unknown consequences to the neuroen-
dence suggested the need for stronger In the play Whose Life Is It Anyway? docrine system may be possible, and to
regulations, especially to protect special by British author Brian Clark, the protag- us, dismissing these out of hand as
populations placed at additional risk onist with high-level quadriplegia files suit inconsequential seems to treat the very

The American Journal of Occupational Therapy 347

Downloaded from http://ajot.aota.org on 04/19/2020 Terms of use: http://AOTA.org/terms
 principle of avoiding harm far too casual- Scholars in occupational therapy Evidence-based practice has clear
ly. Furthermore, although a failed thera- (e.g., Wood, 1998; Yerxa, 1988) have limitations in occupational therapy and
peutic intervention may not result in identified oversimplification as one of the rehabilitation. Currently, these limita-
physical harm, the psychological effect to difficult challenges impeding the devel- tions loom as major obstacles to practice
the patient’s confidence or self-efficacy opment of the field. Perhaps the most behaviors that are better informed and
may be ignored or underestimated. problematic aspect of oversimplification influenced by research. In the United
Moreover, the resource costs of providing and the scientific tendency to objectify States, the implementation of a prospec-
ineffective care also must be considered. phenomena is that they create the illu- tive payment system in rehabilitation will
The opposite also can occur. sion that the unobservable is unimpor- provide increased impetus for research
Because resources are increasingly allocat- tant. In the service of objectivity, we may and for attention to the results of that
ed to proven treatments, the availability unintentionally compromise our values research. As we consider and apply this
of useful options that are yet unsubstanti- of caring and compassion. When super- research to practice, we must do so judi-
ated in clinical trials may be reduced imposed on the reality that many rehabil- ciously, mindful that evidence-based
(Christiansen, 1983; Landry & Mathews, itation patients have sensory or cognitive practice is a gift that comes to health care
1998). This bias may also create a milieu losses or deficits that impede communi- in ethical wrappings. ▲
that unintentionally diminishes interest cation and understanding, the resulting
in behavioral factors related to health or disability affects both the practitioner Acknowledgments
to studies of means to prevent disease or and the patient. If such services become We thank the following experts for providing
injury or promote wellness. Prevention dependent on our ability to show objec- insightful feedback during the preparation of
the manuscript: Barbara Kornblau, JD, OTR,
studies are less often undertaken because tive change, rehabilitation itself may be at FAOTA, Professor of Occupational Therapy &
of the difficulties or expenses associated a disadvantage for receiving the public Public Health, Nova Southeastern University;
with randomization and control in natur- support necessary for any health care William J. Winslade, JD, PhD, James Wade
al environments. Such studies have enterprise to survive and flourish. Rockwell Professor of Philosophy of
engendered less urgency and interest Medicine, The University of Texas Medical
Summary Branch; and Gordon Guyatt, MD, MSc,
because their potential benefits may not Professor of Clinical Epidemiology and
be realized (or known) until many years In this article, we have identified some of Biostatistics and Medicine, Faculty of Health
later. The same bias can be seen in sup- the ethical considerations related to evi- Sciences, McMaster University.
port and interest for studying palliative dence-based practice and surrounding
care and hospice services, where the issues as they bear on occupational thera- References
reports are difficult to obtain and com- py and rehabilitation. We acknowledge Brandt, E. N., Jr., & Pope, A. M. (Eds.).
pare and the situations themselves sym- that practitioners are professionally and (1997). Enabling America: Assessing the role of
bolize a failure of the medical establish- morally obligated to ensure that their rehabilitation science and engineering.
Washington, DC: National Academy Press.
ment to prevent death. decisions are informed and reflect best
Christiansen, C. H. (1983). Research:
Because studies must define and practices. Further, we recognize the value An economic imperative. Occupational
measure variables, the complexity of seri- of encouraging practitioners to assume Therapy Journal of Research, 3, 195–198.
ous illness or injury, which represent responsibility for searching and apprais- Christiansen, C. H. (1999). Defining
important life events by any standard, ing available evidence so that informed lives: Occupation as identity: An essay of
must be simplified for the purposes of options can be shared with patients. competence, coherence, and the creation of
research. Traditional science has a bias Table 1 summarizes the ethical considera- meaning, 1999 Eleanor Clarke Slagle lecture.
American Journal of Occupational Therapy, 53,
toward studying the observable and the tions in evidence-based practice. 547–558.
measurable. As a result, matters difficult Ethical dilemmas are a natural part Culpepper, L., & Gilbert, T. T. (1999).
to observe or measure, such as feelings, of the health care enterprise. They existed Evidence and ethics. Lancet, 353, 829–831.
how people understand and interpret before evidence-based practice became an Department of Health and Human
their worlds, and the personal meaning everyday term, and they present them- Services. (1998). Rules and regulations:
they derive from events, are often not selves whether or not evidence-based Financial disclosure by clinical investigators.
considered. Yet, these issues are impor- practice is introduced into a clinical deci- 63 Federal Register 5233–5254.
tant to lives and personal identities sion. From a moral and professional Dubouloz, C.-J., Egan, M., Vallerand,
J., & von Zweck, C. (1999). Occupational
(Christiansen, 1999). More importantly, standpoint, the dangers of not attending therapists’ perceptions of evidence-based prac-
these issues are fundamental to being well to evidence are just as significant as the tice. American Journal of Occupational
(or well-being). ethical issues attending to its application. Therapy, 53, 445–453.

Table 1
Ethical Considerations in Evidence-Based Practice

Selecting and Using Intervention for a Given Patient Participating in or Advocating for Research
• Base decision on best evidence for given patient, condition, and setting • Monitor ethical practices in research in which one participates
• Involve patients in the choice of intervention • Assure complete and appropriate informed consent
• Appreciate the importance of research evidence over tradition or expert • Advocate for research balancing survival with quality of life
opinion • Respect the right of participant autonomy
• Recognize the psychological and resource costs of ineffective intervention • Involve consumers in strategic decisions about research directions
• Avoid the use of unethical studies

348 May/June 2001, Volume 55, Number 3

Downloaded from http://ajot.aota.org on 04/19/2020 Terms of use: http://AOTA.org/terms
 Gordon, V. M., Sugarman, J., & Kass, Landry, D. W., & Mathews, M. (1998). Tickle-Degnen, L. (1999). Evidence-
N. E. (1998). Toward a more comprehensive Economic evaluation of occupational therapy: Based Practice Forum—Organizing, evaluat-
approach to protecting human subjects: The Where are we at? Canadian Journal of ing, and using evidence in occupational
interface of data safety monitoring boards and Occupational Therapy, 65, 160–167. therapy practice. American Journal of
institutional review boards in randomized Law, M., & Baum, C. (1998). Evidence- Occupational Therapy, 53, 537–539.
clinical trials. Hastings Center Report, based occupational therapy. Canadian Journal Tickle-Degnen, L. (2000). Evidence-
20(1),1–5. of Occupational Therapy, 65, 131–135. Based Practice Forum—Communicating with
Granger, C. V. (1998). The emerging Moreno, J., Caplan, A. L., & Wolpe, P. clients, family members, and colleagues about
science of functional assessment: Our tool for R. (1998). Updating protections for human research evidence. American Journal of
outcomes analysis. Archives of Physical subjects involved in research: Project on Occupational Therapy, 54, 341–343.
Medicine and Rehabilitation, 79, 235–240. Informed Consent, Human Research Ethics Wood, W. (1998). Nationally
Holm, M. B. (2000). Our mandate for Group . Journal of the American Medical Speaking—It is jump time for occupational
the new millennium: Evidence-based practice, Association, 280, 1951–1958. therapy. American Journal of Occupational
2000 Eleanor Clarke Slagle lecture. American Rogers, J. C. (1983). Clinical reasoning: Therapy, 52, 403–411.
Journal of Occupational Therapy, 54, 575–585. The ethics, science, and art, 1983 Eleanor World Health Organization. (2000).
Jonsen, A. R., Siegler, M., & Winslade, Clarke Slagle lecture. American Journal of International classification of functioning, dis-
W. J. (1998). Clinical ethics: A practical Occupational the Therapy, 37, 601–616. ability and health. Geneva, Switzerland:
approach to ethical decisions in clinical medicine Sackett, D. L., Rosenberg, W. M. C., Author.
(4th ed.). New York: McGraw-Hill. Gray, J. A. M., Haynes, R. B., & Richardson, Yerxa, E. J. (1988). Oversimplification:
Kornblau, B. L., & Starling, S. P. W. S. (1996). Evidence based medicine: What The hobgoblin of theory and research in
(2000). Ethics in rehabilitation: A clinical per- it is and what it isn’t. British Medical Journal, occupational therapy. Canadian Journal of
spective. Thorofare, NJ: Slack. 312, 71–72. Occupational Therapy, 55, 5–6.

The American Journal of Occupational Therapy 349

Downloaded from http://ajot.aota.org on 04/19/2020 Terms of use: http://AOTA.org/terms