ETHICAL ISSUES IN CLINICAL

EMERGENCY MEDICINE 0733-8627/99 $8.00 + .OO

RESEARCH ETHICS IN
EMERGENCY MEDICINE
Susan S. Fish, PharmD, MPH

A reliable safeguard against unethical research is ”the presence of an

intelligent, informed, conscientious, compassionate, responsible investi-
gator.’14 Most cases of unethical research are the result of ”thought-
lessness and carelessness, not a willful disregard of the patient’s right^."^
This article’s intent is to inform investigators and others of the rules for
conducting ethical research, and by example, show how overlooking
these rules can be harmful not only to the patients involved and to
future patients but also to the entire research community.
Clinical research is necessary to improve the care provided to pa-
tients for illnesses and injuries. This research must be conducted care-
fully, however, because the experiments use humans as subjects, and
their participation in the experiment must be voluntary. This may seem
obvious, but a review of clinical research over the last 50 years shows
that this concept of voluntary participation in research has not always
been foremost in the minds of the researchers. Although the majority
of studies and experiments are performed ethically, the breeches are
noteworthy. Each time a new problem is sensationalized in the press, it
damages the public’s perception of research; thus we must examine the
details of the study, learn from any mistakes made, and determine how
to prevent their recurrence. A review of the more egregious errors and
their lessons is therefore in order here.

From the Department of Emergency Medicine, Boston Medical Center; and the Depart-
ments of Emergency Medicine and Public Health, Boston University Schools of Medi-
cine and Public Health, Boston, Massachusetts

EMERGENCY MEDICINE CLINICS OF NORTH AMERICA

VOLUME 17 * NUMBER 2 MAY 1999 461

462 FISH

THE HISTORY OF UNETHICAL RESEARCH

During World War 11, physicians in Nazi Germany willingly con-
ducted experiments on prisoners in concentration camps, and this "re-
search was conducted without the consent of the s~bjects.~ After World
War 11, a set of standards, called the Nuremberg Code was developed,
with which to conduct hearings of these researchers at the Nuremberg
Trials. This set of standards was the first in modern times to discuss the
ethical conduct of clinical research. Subsequently, at a conference in
Helsinki, Finland, the World Medical Association developed and
adopted a code of ethics for conduct of clinical research in 1964 known
as the Helsinki Declaration. This Declaration has been updated regularly,
in 1975, 1983, and 1989, and is currently under revision again.
During this time of developing various international codes for ethi-
cal conduct of clinical research, however, the US Public Health Service
was continuing to conduct its observational study of the natural history
of syphilis in African-American men, a study known historically as
the Tuskegee study because it took place in Tuskegee, Alabama.lo This
observational study followed 399 untreated men with syphilis, beginning
in 1932 and ending in 1972, long after the use of penicillin had become
standard of care for the treatment of this infection.
Many other human studies of questionable ethical standing took
place and were published in the medical literature. In 1966, Henry
Beecher published a review of 22 studies that were conducted after
World War I1 and after the well-publicized Nuremberg Trials in the
New England Journal of Medicine; for him, these studies were ethically
problematic? In Beecher's review of published literature, he uncovered
50 studies that he considered unethical, but he was given space in the
journal to discuss only 22 of these. Some of the more infamous include
a study at Willowbrook State Hospital (1956-1965) where mentally re-
tarded children were intentionally injected with hepatitis virus to study
the period of infectivity of the disease. Another horrific example is a
study conducted at Jewish Chronic Disease Hospital (1963), where el-
derly debilitated patients were injected with live cancer cells to study
immune response. Studies that have more recently made headlines are
the radiation experiments conducted during the Cold War period of the
1960s and early 1970s. One study included feeding radioactive cereal to
children who were institutionalized at the Fernald State School; these
children had volunteered to be in the Science Club, and neither they nor
their parents were told of the radiation experiments.'

THE ETHICAL PRINCIPLES

In 1974, the National Commission for the Protection of Human
Subjects of Biomedical and Behavioral Research was established by the
US Congress. Five years later, the Commission issued the Belmont Re-
port, named for the location of the ~0nference.l~ Along with setting
guidelines for establishing institutional review boards (IRBs) to review
 RESEARCH ETHICS IN EMERGENCY MEDICINE 463

human research, this report delineated three principles that serve as the
underpinnings for ethical conduct of human research. These principles
are
1. Respect for persons
2. Beneficence/nonmaleficence
3. Distributive justice
The autonomy of the individual is described as respect for persons,
which means that the autonomous person decides for himself or herself
whether to participate in a research study. For someone to decide
whether to participate in a study, the person must understand the
purpose and procedures of the study, as well as the risks and benefits
to participation; this is intended to be accomplished through the informed
consent process. The subject’s autonomous decision to agree to partici-
pate in the research is then documented with an informed consent form.
The discussion of the principle of respect for persons allows that some
people cannot exercise autonomy and that these “vulnerable popula-
tions” require additional protections. Examples of vulnerable popula-
tions include children and mentally disabled people. In these cases,
parents or guardians (i.e., legally authorized representatives) frequently
provide consent for research, but additional scrutiny in reviewing these
studies is also suggested in the Belmont Report.
The principle of beneficence/nonmaleficence, roughly translated as
”do good; do no harm,” is applied through a benefit-risk assessment. The
benefits of research can accrue to the individual subjects, to future
patients with the condition under study, or to society as a whole. The
risks, conversely, usually apply only to the subjects of the study. Thus,
the balance between benefits and risks is not always easy to assess. The
investigator must define these benefits and risks and discuss them with
the potential subjects during the consent process. In addition, it is the
investigator’s responsibility to design the research study to maximize
the benefits while minimizing the risks.
The purpose of distributive justice in research is to ensure that those
who are most likely to benefit from the results or outcomes of the study
are those who are asked to participate-to bear the burdens of the
research. This means that research cannot be done on prisoners or
institutionalized people just because they are a convenient group of
potential subjects, if the results of the study are intended to apply to the
broader population. Distributive justice has also been interpreted to
mean that if the results of a study are intended to apply to women as
well as men, then the study participants should include both women
and men. In addition, distributive justice applies to race, ethnicity, socio-
economic status, and any other category that is appropriate in a particu-
lar study.

REGULATIONS
The ethical principles delineated in the Belmont Report are the basis
for the federal regulations governing human research. Both the Food
464 FISH

and Drug Administration (FDA) and the Department of Health and

Human Services (DHHS) regulate human research in the United States.
Many other countries have similar agencies and similar regulations. The
FDA has oversight for studies involving drugs, devices, and biologic
agents (21 CFR 50), whereas the DHHS has oversight for studies that
are federally funded (45 CFR 46). Both sets of regulations require IRB
review and approval of human research in areas that they regulate, but
many IRBs are set up such that these federal regulations are applied to
all human research at the institution, regardless of the type of study or
the funding source. The Department of Defense has a third set of
regulations for agencies that it oversees.
The regulations that guide IRB review and approval are complex.
The best advice is for each investigator to get to know the chair or
administrator of his or her own IRB. These people can explain the
”rules,” both written and unwritten, and can be very helpful in negotiat-
ing the path through the local committee. In addition, an investigator
should consider becoming a member of the IRB; emergency medicine
researchers can make very valuable contributions to the research com-
munity and to their own field by participating in the committee’s delib-
erations.
One area that is useful to discuss here is the informed consent
process and form. As stated previously, the form is intended to docu-
ment the informed consent process, a dialogue between the investigator
and the potential subject. In the consent process, the investigator ex-
plains the study to the potential subject while encouraging questions, so
that the subject has enough information to make an informed decision
concerning whether to participate in the study. The regulations require
that certain topics to be covered during the informed consent process
(Table 1). The challenge is to present the information in a way that is

Table 1. REQUIRED ELEMENTS OF INFORMED CONSENT

That the study involves research
An explanation of the purpose of the research
The expected duration of the subject’s participation
A description of the procedures of the study and which ones are experimental
A description of the reasonably forseeable risks or discomforts
A description of the reasonably forseeable benefits to subject or others
A description of alternative procedures or treatments
A description of the extent to which research records will be kept confidential
Whether there is compensation for participation in the research
Whether there is compensation or medical treatment for research related
injuries, and whom to contact in the event of a research related injury
Who can answer questions about the research
Who can answer questions about the research subjects’ rights
That participation is voluntary, and the subject can withdraw at any time
without Denaltv

Data summarized from 21 CFR 50.25.

 RESEARCH ETHICS IN EMERGENCY MEDICINE 465

understood completely and clearly by the subject, a skill that must be

developed with experience and guidance.

UNIQUE SITUATIONS IN EMERGENCY MEDICINE

RESEARCH

Patients who come to the ED usually perceive that they have an

emergency medical condition, and in the best of situations are not
necessarily thinking clearly. There are many conditions in which patients
are unable to process cognitively owing to the pain of their illness or
injury, medications they may have taken or received, head injury, or
other distracting illness or injury. In the most extreme case, they are
totally unable to participate in dialogue owing to loss of consciousness.
In these situations, it is difficult or even impossible to obtain informed
consent for research. The doctrine of ”presumed or implied consent” for
treatment (the premise that if a patient presents to the ED for care and
is unable to consent to care because the acute medical condition, the
clinician presumes the patient’s consent to lifesaving treatment) does not
apply to the research setting. Research is elective and not required to
save a life. If informed consent is not possible, however, how is this
research to be conducted? Is all research without informed consent
unethical?
Although debate continues on this subject, most researchers and
ethicists have agreed that some research on emergency conditions must
be p ~ s s i b l eMany
.~ feel that doing research without consent in patients
with an emergency condition differs from the variety of research insults
to patients described above. In the Nazi experiments and the Public
Health Service Syphilis Study and others, the investigator deprived the
subjects of their autonomy; in studies of emergency conditions, it is the
condition that has deprived the subject of autonomy prior to consider-
ation of enrollment into a study (Biros M: Personal communication,
1997). To address the issue on a regulatory basis, the FDA and DHHS
issued new regulations that became effective in late 1996 that govern
the conduct of ”emergency research without informed consent [21 CFR
50.24 and 45 CFR 46.lOl(i)].Studies that seek to apply these regulations
must meet criteria listed in Table 2. These regulations define a new
vulnerable population: patients with a sudden, acute, life-threatening con-
dition, for which the “available treatments are unproven or unsatisfac-
tory, and the collection of valid scientific evidence . . . is necessary to
determine the safety and effectiveness of particular interventions.” There
are many requirements of the regulations that must be met. Among
these are that additional protections for this vulnerable population are
required because it is not possible to obtain informed consent from these
patients. These additional protections include at least consultations with
the communities in which the research will take place and from which
the subjects will come; disclosure to the public of the study, its proce-
dures, its risks and its benefits prior to beginning the study; disclosure
466 FISH

Table 2. EXCEPTION FROM INFORMED CONSENT FOR STUDIES CONDUCTED IN

EMERGENCY SETTINGS
The IRB must find and document that
1. The subject is in a life-threatening condition.
2. Available treatments are unproven or unsatisfactory.
3. This research is necessary to determine safety and efficacy of the proposed
intervention.
4. Obtaining informed consent is not feasible because
a. Subject is unable owing to medical condition.
b. Short time window in which to administer the intervention, making it infeasible
to contact subject’s legally authorized representative.
c. Not feasible to prospectively identify who might be eligible to participate in the
research and thus obtain consent for future participation.
5. Participation might be beneficial to the subject because
a. Life-threatening situation requires intervention.
b. Appropriate animal and preclinical studies support possibility of benefit.
c. Risks of research are reasonable compared to risks of life-threatening medical
condition, and risks and benefits of standard therapy.
6. Research could not practicably be conducted without waiving informed consent
requirement.
7. The investigator will attempt to find the subject’s legally authorized representative as
soon as possible and ask for consent if feasible within the time window of the study.
8. Additional protections of the rights and welfare of subjects will include at least
a. Community consultation with communities where the research will be conducted
and from which the subjects will be drawn.
b. Public disclosure of the research prior to beginning the study
c. Public disclosure of the results after completion of the study
d. Oversight of the research by an independent data monitoring committee
e. Providing information to subject’s family as soon as possible, and asking if they
object to subject’s participation
9. A research protocol that involves a drug, device, or biologic requires a separate
investigational new drug application (IND) or investigational device exemption
(IDE), which must be obtained from the Food and Drug Administration (FDA)
10. If an IRB determines that the study does not meet these criteria because of ”relevant
ethical concerns,” it must report its findings to the sponsor, who must in turn report
the findings to the FDA and to other investigators and IRBs participating or
considering participating in the same or similar studies.

Data summarized from 21 CFR 50.24 and 45 CFR 46.101(i)

of study results to the public and to the scientific community after study
completion; establishment of an independent data-monitoring committee
to oversee the study; and notification of the patient’s family as soon as
possible of the patient’s inclusion in the study. All studies that apply
these regulations must be approved not only by the IRB but also by the
FDA or DHHS. These new rules are intentionally difficult to implement;
waiving the requirement for informed consent should not be easy; how-
ever, it is now possible to perform research on serious, acute conditions
that would not otherwise be possible to study, such as cardiac arrest and
severe head trauma. These federal regulations do not apply in states
with laws preventing research without consent.
The areas of research that remain problematic are those in which
the patient is not fully capable of decision making but is not comatose.
These situations include research on intoxicated patients, moderately
 RESEARCH ETHICS IN EMERGENCY MEDICINE 467

injured patients, patients with cardiac ischemia or acute asthma, and

other conditions in which it is unlikely that the patient can make a truly
informed, uncoerced decision about participation. Such studies require
careful design and implementation to ensure that participation of sub-
jects is voluntary.

SCIENTIFIC MISCONDUCT

Although scientific misconduct is thought to occur rarely, each epi-

sode has the potential to harm patients as well as to jeopardize the
public’s trust in the medical and research communities. Researchers,
therefore, have a responsibility to guard against this misconduct in their
work as well as the work of others. Preventing misconduct, however,
requires understanding what it is and how and why it might occur.
Scientific misconduct is defined by the National Commission on Research
Integrity as ”significant misbehavior that improperly appropriates the
intellectual property or contributions of others, that intentionally im-
pedes the progress of research, or that risks corrupting the scientific
record or compromising the integrity of the scientific practices.”13Re-
searchers are human and are capable of making mistakes, both honest
errors and careless errors. For research findings to be believable, re-
searchers must guard against both types of errors by developing systems
of checks and balances in implementing the methods of a study. Errors
will happen, however, and unintentional errors do not indicate miscon-
duct. Misconduct is essentially an intentional error: falsification, fabrica-
tion, plagiarism, or other deception; even if the intent is thought to be
in the best interest of the patient.
Examples of scientific misconduct illustrate the variety of situations
in which this can occur. Although these examples are not taken from
emergency medicine research, their lessons are nonetheless applicable.
Misconduct in the form of failure to comply with any governing regula-
tions could be due to ignorance of the regulations or could be intentional.
As discussed above, all human research must be reviewed by an IRB,
and interventional research must have the voluntary consent of the
subject or patient. Misconduct in this area includes conducting research
that has not been approved by the appropriate IRB or forging a signature
on a consent form.2O Misconduct can also occur when enrollment criteria
have been altered, as in the case of Roger Poisson, MD, an investigator in
a breast cancer lumpectomy versus mastectomy trial, who intentionally
altered the dates of some subjects’ mammograms so that they would
qualify for enrollment.2This discovery of altered data led to multiple
audits of research records that revealed 115 instances of falsification or
fabrication of data from 99 subjects.15Although these data falsifications
and fabrications did not appear to put the women in the study at risk,
the validity of the data from the entire multicenter trial was questioned,
and the careers of not only Poisson but others, including Bernard Fisher,
MD, the director of the National Surgical Adjuvant Breast and Bowel
468 FISH

Project (NSABP), were permanently altered. In addition, other NSABP

studies funded by the National Cancer Institute were temporarily halted
during the investigation. Other types of methodology falsifications in-
clude altering the randomization scheme and unblinding the treatment
without the permission of the principal investigator. (Unblinding can be
done for patient safety whenever necessary, but usually there is a proto-
col established for this purpose.) Although the impact of these falsifica-
tions may not be intuitively obvious, any breach of the randomization
or blinding may put the validity of the results in question. If the results
of the study are not useful to the scientific and medical communities,
then patients have been put at some risk without any benefit to future
medical care, which is ethically unacceptable.
Data analysis and interpretation is open to misconduct as well. Data
fabrication is making up data rather than performing the experiment.
Some studies have been entirely fabricated.I6The potential for harm to
patients is obvious if clinicians base treatment decisions on fictitious
results. This fabrication can also mislead other researchers, who rely on
these data as support for their own research, and who can then develop
a line of investigation that is totally unfounded. Also possible is data
faIsification; this is another obvious case of lying. This could include
changing a blood pressure number rather than using the actual reading,
or taking additional measurements until an "acceptable" reading is
obtained. This can also lead to data deletion, or dropping data that is
considered unacceptable or out of range. Although there is sometimes
reason to drop outlier data, this procedure is clearly spelled out in a
written description of methods and usually discussed in subsequent
publications. Other misconduct using data includes selective inclusion
of data for statistical testing and selective presentation of data. In other
words, rather than include data from all who met enrollment criteria,
only those data that will support some predetermined conclusion un-
dergo statistical analysis. Selective presentation of data usually entails
presenting only those comparisons that are statistically significant, even
if the primary study question is not answered. At the other end of the
spectrum is data dredging: using the entire data set and performing a
variety of statistical tests on portions of the data to find any comparison
that is statistically significant. All of these data manipulations can be
considered misconduct. The way to avoid such situations of misconduct
is to develop clear methods and operating procedures for a study prior
to the enrollment of the first patient. All data acquisition steps must be
detailed and all statistical tests must be spelled out up front. Once the
data are examined, post hoc testing should be used to suggest future
study, not to answer an additional study question in the present study
Because so much misconduct can occur in data manipulation, ethical
conduct of statisticians has become an increasingly important topic.15
Most researchers are honest people, but many believe that academic
pressures cause some to attempt to bend the rules. Academic competi-
tion for promotion and for research funding, as well as the drive to be
the first to discover a cause or a cure, can be seductive. Junior researchers
 RESEARCH ETHICS IN EMERGENCY MEDICINE 469

sometimes lack education and mentoring from senior researchers, who

in the current healthcare environment must perform more clinical ser-
vices as well as supervise many research fellows and students. Others
believe that this is simply speculation because we have little evidence
as to motivation: the pressure to "publish or perish in the academic
world, or "laziness, desire for fame, greed, and an inability to distinguish
right from wrong."18
Although the specific causes may be unknown, prevention of mis-
conduct can be addressed, primarily by education. Many have called for
required education in research ethics for all researchers, and such educa-
tion frequently is required during postdoctoral training. Many clinical
researchers are clinicians without postdoctoral research training, how-
ever (e.g., physicians, nurses, or pharmacists) who may not have re-
ceived this specific education in research ethics, and any introductory
lectures are long forgotten. By necessity, therefore, the responsibility lies
with the institution in which the research is taking place. Articles such as
this one may also help; lectures, professional conferences, and individual
teaching are also required. Senior researchers have a responsibility to
mentor and supervise all junior researchers, research assistants, and
other support personnel. Ethical problems need honest discussion. All
raw data must be retained for long periods, in case there are questions
of validity after publication of the study. Ideally, data should be recorded
in permanently bound data books; entries must never be erased but
crossed out and initialed if a mistake is made. Institutions and profes-
sional organizations must develop guidelines for. ethical conduct of
research, investigation of accusations of misconduct, protection of those
who make the accusations ("whistle blowers"), and discipline of perpe-
trators.

AUTHORSHIP AND PUBLICATION ETHICS

After developing a study protocol and maybe obtaining funding,

obtaining approval of the study from an IRB, obtaining informed consent
from all subjects, and conducting the study in an ethical manner, the
researcher arrives at the point of preparing a manuscript for publication.
Undoubtedly one did not conduct this study by oneself, but there were
many others involved in parts of the process. Are they all authors on
the paper? Does this include the director of the laboratory or the chief
of the department? This can sometimes be the most uncomfortable part
of research. The best way to avoid this discomfort is to delineate prior
to beginning the study what the requirements for authorship will be.
The International Committee of Medical Journal Editors has developed
specific criteria for authorship (Table 3).9 Using these criteria, the chief
of the department and the person who performed all of the data collec-
tion are not necessarily authors on the paper unless they meet the other
criteria as we1l.l' Although not all journals use these criteria, instructions
470 FISH

Table 3. CRITERIA FOR AUTHORSHIP

Each author should have participated sufficiently in the work to take public
responsibility for the content. Authorship credit should be based only on substantial
contributions to
(a) conception and design, or analysis and interpretation of data; and to
(b) drafting the article or revising it critically for important intellectual content; and on
(c) final approval of the version to be published.
Conditions (a), (b), and (c) must all be met.

From International Committee of Medical Journal Editors: Uniform requirements for manuscripts
submitted to biomedical journals. N Engl J Med 324424428,1991

for authors of each journal will delineate authorship requirements if

they differ.
Once written, a manuscript can be submitted to only one journal at
a time. Many journals require that a cover letter or cover sheet state that
the manuscript is not being submitted to or considered for publication
by any other journal at this time. Most journals do not publish papers
that have already been published in another journal, with rare exception.
If a manuscript is submitted to multiple journals simultaneously and is
accepted to more than one of these, all but one must be withdrawn. This
results in unnecessary work for volunteer reviewers, as well as potential
duplication and confusion for all concerned. Thus, one should submit a
manuscript to a single journal. If it is rejected from that journal, it can
then be submitted to another journal. If the initial journal returns the
manuscript with recommendations for changes that the author does not
wish to make, he or she can decline to resubmit to that journal and
instead submit to another journal.
A journal reviews a manuscript and usually issues one of four
decisions: accept, accept with revisions, revise and resubmit for another
review, or reject. Only the first of these is an actual acceptance; only a
true acceptance can be referred to as being “in press.” Thus, when one
is writing a resume or curriculum vitae (CV), one should list as in press
only those manuscripts that have been accepted. Once the article is
published, the full citation should be used. Although this seems obvious,
ignorance of this issue may account for what some have termed misrep-
resentation. In a review of the CVs of 138 applicants to a gastroenterol-
ogy fellowship program who reported research experience, the investiga-
tors found that 34% of the research activities listed could not be
verified.I9 Of the 53 applicants listing publications, 30% misrepresented
their publications. Errors noted included primarily articles in nonexistent
journals, and nonexistent articles in existing journals. Lest one think that
this situation is unique to gastroenterology, the same study was repeated
in emergency medicine by different investigators reviewing 113 appli-
cants to emergency medicine residencies who reported publications.’
The misrepresentation rate among this cohort was 20%. Errors noted
included those seen in the gastroenterology review as well as actual
articles in actual journals but without the applicant listed as an author.
 RESEARCH ETHICS IN EMERGENCY MEDICINE 471

Another error sometimes seen, although not mentioned in these two

reports, is changing the order of authors. Once published, the order of
authors as well as the rest of the citation is set. To indicate on one’s CV
that, although not the first author, one is in the author list, one can type
one’s name in bold.
Plagiarism is the use of someone else’s words or ideas without
attribution. The rules for medical publications are not substantially dif-
ferent from those learned in high school and college. A unique twist on
this situation can occur in that, once an article is published, the copyright
is usually held by the publisher rather than the author. If one is the
author and wishes to quote his or her own article, one must cite it fully,
just as though the words or ideas were from another author.

CONFLICT OF INTEREST

Potential conflicts of interest exist in a large portion of the clinical

research performed in the United States. A potential conflict of interest
exists in any situation in which the investigator is somehow vested in
the outcome of the study. The most obvious sources of these potential
conflicts are in the form of financial arrangements. Financial sponsorship
of a study, although a potential conflict, is a reality. Pharmaceutical and
biotechnology industries must perform research to support applications
for approval to market drugs and devices. Clinicians must perform that
research, and the costs of the research must be covered. Some arrange-
ments allow for the sponsor to cover the actual cost of the research,
whereas others are such that investigators stand to benefit financially
from performing that research. One recent such case has been profiled
in The Wall Street Journal, in which two well-respected pharmacology
investigators, Richard Borison, MD, and Bruce Diamond, PhD, per-
formed many studies for several drug companies.21There were several
fraudulent claims made concerning these investigators, including having
the sponsor send checks to the investigators’ homes rather than to their
university. The result of many months of investigation is that each
investigator has received a prison sentence (5 years for Diamond and 15
years for Borison) and must repay the university $1.1 million to $4.26
million in restitution. Although an extreme circumstance, a similar situa-
tion can occur if an investigator does not know that he or she should
not enter directly into a financial agreement for research with a sponsor
but must let the employing institution (hospital or university) sign that
contract. Although the investigator eventually has access to the money
and the potential conflict of interest still exists, there is more accountabil-
ity and an investigator is unlikely to fall unknowingly into an unfavor-
able situation.
It is believed by some that only industry sponsorship of research
leads to this potential conflict of interest; however, federal funding of
research can also be coercive. If an investigator has a federal grant,
the pressure exists to produce results consistent with the proposal’s
472 FISH

hypothesis; this is thought to be necessary if one wants to obtain another

future grant. In addition, even unfunded research can be a potential
conflict of interest because in the academic environment, the pressure to
publish for the sake of academic promotion is constant. Publication is
thought to be much easier if results are positive; if no difference is found
between a new drug and the standard drug, the study is less likely to
be published. Thus, one could argue that regardless of source of funding
or even if there were funding, a potential conflict of interest may exist
in all research.
Other sources of potential conflicts include the investigator's own-
ing stock in the sponsor of a study or having stock options not yet
exercised. If the investigator is an employee of or consultant to the
sponsor, this can constitute a conflict. Most universities and hospitals
have conflict of interest policies that must be followed. Although these
policies differ among institutions, the general approach is to acknowl-
edge that a potential conflict exists and state the situation fully. An
official of the institution is charged with reviewing these reports and
investigating any that appear problematic.
What are the problems with a potential conflict of interest? First,
conflict of interest can alter objectivity. If an investigator is invested
either financially or otherwise in the outcome of a study, the potential
for introduction of bias into gathering data or interpreting those data is
obvious. To guard against this bias in the gathering of data, the study
should use all possible methods to ensure that those who assess out-
comes are blinded to the intervention, and outcomes used should be as
objective as possible for a given study. The best way to prevent potential
conflicts of interest from having a negative impact on a study is to
disclose such conflicts and discuss them openly.

GRANTS AND CONTRACTS

Grants and contracts should not be signed solely by the investigator,

partly for reasons mentioned earlier. In addition, the officials in the
institution who are responsible for grants and contracts are experienced
in this area, and clinicians usually are not. The story of Betty Dong,
PharmD serves as an ill~stration.'~ Dong is a well-respected pharmacoki-
neticist who was funded by Flint Laboratories to perform a study of the
bioavailability of four different preparations of levothyroxine (including
Synthroid, which was then marketed by Flint). Although both she and
Flint believed that Synthroid would be shown to have superior bioavail-
ability, this was not the outcome of the research. Dong prepared and
submitted the resultant manuscript for publication, but Flint objected.
What had occurred is that, unknown to Dong, the contract that she had
signed prior to beginning the research gave Flint the ownership of the
research data. Thus, if they did not like the results, they could prevent
the publication of the report. The threats of litigation and the subsequent
negotiations finally resulted in publication of the entire story, as well as
 RESEARCH ETHICS IN EMERGENCY MEDICINE 473

the research manuscript, 7 years after it was first accepted for publica-
tion.'j The issue of who owns the data has become a controversial one.
Many think that the ethically correct situation is for the sponsor to have
no control over data and publications. Others feel that sponsors have no
incentive to support research if they have no control of the data. Drum-
mond Rennie, associate editor of JAMA, posited that there are lessons
for all parties involved in research.17First, he suggests that researchers
should never allow research sponsors the power to veto a publication.
Second, universities and research institutions should not allow such
clauses in contracts from sponsors and that these institutions must
support the rights of their faculty. Third, sponsors should realize that
attempts to prevent publication of research data damages their reputa-
tion among academics and regulators alike. Finally, professional societies
and journals have responsibilities to address these issues by developing
standards that make suppression of research data unacceptable.
Another similar situation occurred when Microfibres, a textile man-
ufacturer, hired David Kern, MD of Brown University to investigate
occupational lung disease among its workers in Pawtucket, RI.'* When
Kern attempted to present scientific evidence of a new disorder in nine
workers at the American Thoracic Society meeting, he was told by
Microfibres to cancel his presentation. Kern's contract with Microfibres
also included a clause that the manufacturer owned the results of the
investigation. In the controversy that ensued, Kern was eventually asked
to leave his positions at both Brown University and Memorial Hospital.
Despite protests from the scientific and academic communities, the data
have been withheld, and Kern must leave his university and hospital
positions.
Patients agree to participate in research because there may be some
benefit to the individual and to future patients. Any results that are
withheld from publication have put the research participants at some
inconvenience and some risk without any benefit to future patients.
Withholding results from clinicians can cause ineffective drugs to be
used, or expensive drugs to be used instead of cheaper but equally
effective ones. Employees can continue to be unknowingly exposed to
toxins that are associated with development of disease, when data are
withheld. Withholding data is a form of deception and is ethically
unacceptable. To protect against this situation, an investigator should be
clear about the details of a contract or grant for sponsorship of research
prior to entering into any agreement. The investigator should think
twice before agreeing to give up control of data to a sponsor.

CONCLUSIONS

Unethical research can harm both present and future patients and
jeopardizes trust in the entire scientific community. Regulations and
rules are intended to provide a framework with which to guide the
research and ensure that research is conducted and its results are pre-
474 FISH

sented in an ethical manner. It is the responsibility of the investigator to

be knowledgeable about these regulations and rules and keep the guid-
ing ethical principles foremost in his or her mind during the entire
research process. “An experiment is ethical or not at its inception; it
does not become ethical post hoc-ends do not justify means. There is
no ethical distinction between end and mean^."^

References
1. Advisory Committee on Human Radiation Experiments, Final Report. Washington,
DC,US Government Printing Office, 1995
2. Angell M, Kassirer J P Setting the record straight in the breast cancer trials. N Engl J
Med 3301448-1450,1994
3. Annas GJ, Grodin M A The Nazi Doctors and the Nuremberg Code: Human Rights in
Human Experimentation. New York, Oxford University Press, 1992
4. Beecher H K Ethics and clinical research. N Engl J Med 274313561360, 1966
5. Biros MH, Lewis RJ, Olson CM, et al: Informed consent in emergency research
Consensus statement from the Coalition Conference of Acute Resuscitation and Critical
Care Researchers. JAMA 273:128>1287, 1995
6. Dong BJ, Hauck WW, Gambertoglio JG, et al: Bioequivalence of generic and brand-
name levothyroxine products in the treatment of hypothyroidism. JAMA 2771205-
1213, 1997
7. Evans I: Conduct unbecoming-the MRC’s approach. Br Med J 316:1728-1729, 1998
8. Gurudevan SV, Mower WR Misrepresentation of research publications among emer-
gency medicine residency applicants. Ann Emerg Med 27327-330, 1996
9. International Committee of Medical Journal Editors: Uniform requirements for manu-
scripts submitted to biomedical journals. N Engl J Med 324:424-428, 1991
10. Jones JH: Bad Blood: The Tuskegee Syphilis Experiment. New York, The Free Press,
1993
11. Kassirer JP, Angell M On authorship and acknowledgments. N Engl J Med 3251510-
1512, 1991
12. Knox RA: Brown public health researcher fired. The Boston Globe, July 7, 1997
13. Marwick D: Spring helps research integrity report resurface. JAMA 275:746, 1996
14. National Commission for Protection of Human Subjects of Biomedical and Behavioral
Research The Belmont Report: Ethical Principles and Guidelines for the Protection of
Human Subjects of Research. Washington DC, DHEW Publication Number (0s)78-
0612. Appendix I, DHEW Publication Number (0s) 78-0013; Appendix 11, DHEW
Publication Number (0s)78-0014
15. Piantadosi S Clinical Trials: A Methodologic Perspective. New York, John Wiley &
Sons, 1997
16. Recer P: Leukemia studies are flawed. The Boston Globe, October 30, 1996
17. Rennie D: Thyroid storm. JAMA 2771238-1243, 1997
18. Rennie D: An American perspective on research integrity. Br Med J 316172&1728,1998
19. Sekas G, Hutson WR Misrepresentation for academic accomplishments by applicants
for gastroenterology fellowships. Ann Intern Med 123:3841, 1995
20. Stecklow S, Johannes L Drug makers relied on clinical researchers who now await
trial. The Wall Street Journal, August 15, 1997
21. Wall Street Journal Staff Reporter: Professor pleads guilty to diverting millions from
medical college. The Wall Street Journal, October 15, 1998
Address reprint requests to
Susan S. Fish, PharmD, MPH
Department of Emergency Medicine
Boston Medical Center
818 Harrison Avenue
Boston, MA 02118