Joan D Krizack.

Documentation planning for the U.S. health care system online

. (page 11 of 26)
Online LibraryJoan D KrizackDocumentation planning for the U.S. health care system → online text (page 11 of 26)
Font size
QR-code for this ebook

The formats employed for recording and analyzing data likewise vary
enormously. No single technological advance has necessarily rendered
obsolete paper laboratory notebooks or clinical case files. New software
products, however, are constantly being marketed to capture and manip-
ulate raw data. For every research specialty, there is an ever expanding
array of calculating, computing, monitoring, and testing equipment, all
with digitalized data output. Computer graphics programs, many inte-
grated with word-processing software, routinely record and store the
graphic and tabular results of experiments or trials online.

Scientific laboratory apparatus, along with equipment for the clinical
examinations that investigators may share with hospital colleagues, is
usually expensive. Once acquired, it may be complicated or dangerous to
operate. Interaction between investigators and manufacturers and suppli-
ers can be complex. In fields where the research is at the cutting edge of
technology, scientists themselves sometimes become inventors. This phe-
nomenon was perhaps more pronounced in decades past, before the
advent of microchip-regulated electronics. Even today, however, as con-
sumers in the relatively tiny market for expensive machinery, major
laboratories exercise the power to demand customized equipment. Proper
security arrangements for hazardous and controlled substances and mea-
sures for their safe and legal disposal require sizable investments at many
kinds of laboratories. Biomedical research institutions are also significant
consumers of high-priced multipurpose items, such as computer hard-
ware. Biomedical information transfer and biostatistics have become
crucial services of large medical centers, demanding correspondingly large
outlays for mainframe data processing equipment and network lines, not
to mention the specialists to maintain and operate them.^^


Laboratory animal management is a significant area of specialization
within biomedical research. A wide variety of animal species, from pri-
mates to E. coli, are used in experimentation. The archetypal research


subject may be the guinea pig, although more frequent and extensive uses
are made of rats and mice. Several mutant strains of rodent genera (e.g.,
the nude mouse) have been specially bred and marketed for laboratory
purposes. The experimental use of higher animals, such as cats, dogs, and
monkeys, is also widespread in nearly all disciplines. This is the most
costly aspect of operations, because it requires extensive space for cages,
systems for sanitary feed handling and cleaning of the pens, and the
services of veterinarians.

Contemporary procedures and regulations governing animal experi-
mentation have been profoundly affected by criticism or opposition from
groups ranging frorh antivivisectionists to animal liberation partisans.
Humane societies have worked to stiffen ordinances regulating the sale of
impounded dogs and cats to laboratories and have demanded legislative
or police investigation of certain research institutions. The U.S. Depart-
ment of Agriculture, which is charged with enforcing animal welfare
legislation, has repeatedly revised regulations concerning facilities, proce-
dures, and oversight. Committees for the governance of animal experi-
mentation are now mandatory. Biomedical research institutions across
the country have in recent years expanded and upgraded their animal
care facilities.


Procedures for formal clinical trials are essential to the work of a substan-
tial proportion of biomedical research institutions. Most often these are
employed as means for studying the effectiveness of new drugs, but many
other objectives are possible, including investigations of various elements
significant in diet, fitness, genetic makeup, and social behavior. Depend-
ing on the nature of the study, an experimental laboratory may or may
not be directly involved. Some kinds of clinical trials or aspects of large
projects are entrusted to commercial testing companies. Federal regula-
tions, however, require hiring a statistical and clerical staff trained to
conduct surveys of sizable subject populations. Friedman and colleagues
and Spicker have provided comprehensive definitions of clinical trials.^'*
They indicate that every well-designed study requires a protocol incorpo-
rating written agreements between investigators, subjects, and a scientific
group selected to monitor response variables. Each trial should be con-
structed as a means of grappling with a primary question for which there
are reasonable expectations of verifiable conclusions. The study popula-
tion is a subset of a general population defined by specific eligibility
criteria, out of which the group of subjects actually studied is selected.
Nearly all investigations of this sort require a control group with which the


group receiving the new intervention can be compared. Proper calcula-
tion of the size of the respective groups is essential to ensure the statistical
means of recognizing significant differences in the data.

Effective designs for clinical trials incorporate standard means for
avoiding elements of bias. Some subjects participating in drug tests, for
example, may be asked to receive the experimental substance, rather than
the placebo. To allow such a choice to be made by the subject, however,
could produce seriously skewed data. Minimal standards for a scientific
trial, therefore, hold that it be "single blind." The researchers, for their
part, usually need to demonstrate that they have not favored one group of
subjects over another in administering the experimental substance. In a
double-blind trial, neither subjects nor investigators know which inter-
vention is administered. In a triple-blind study, not even the group
monitoring responses is aware of each intervention assignment.

The agreements with subjects that are essential to conducting a clini-
cal trial are supposed to follow the principle of informed consent. Re-
search applications of informed consent developed implicitly over centu-
ries within general medical ethical and legal precepts. They received,
however, specific articulation at the Nuremberg trials, in the court's
judgment against the Nazi concentration camp investigators. United States
V. Karl Brandt. That decision remains the benchmark for mandating efforts
toward free and enlightened decisions on the part of clinical research
subjects. ^5 Gaining informed consent from subjects of biomedical research
is consequently one of the most elaborate and costly steps of a screening
process. Warnings have regularly been voiced over the years that many
investigators neglect their responsibilities in this matter to one degree or
other. ^^ The federal government has responded by issuing increasingly
lengthy regulations governing these interactions. Chief among them is the
mandate to establish institutional review boards or human subjects com-
mittees. In theory, institutional review boards are charged with scrutiniz-
ing research proposals for the protection of human subjects at any institu-
tion funded by the Department of Health and Human Services. ^^

In certain areas of clinical research, formal clinical trials and formal
applications of informed consent do not apply. Many new surgical inter-
ventions, most pathology research, and various kinds of experimental
psychiatric treatments are in this category. Some forms of clinical investi-
gations, furthermore, are entirely retrospective, involving the study of
inactive medical records. Much epidemiological research, for example,
requires permission from hospital management for secondary analysis of
data not originally created for general knowledge. Whatever the design
and scope of the problem under investigation, some form of rigorous
control is required to reach valid scientific conclusions.



A prerequisite for conducting clinical trials in most instances is that a
biomedical research institution offer clinical care services. Although some
subjects participate in research investigations for reasons not related to
their own health, substantially more are attracted after having first sought
treatment as patients. This is one of several facets of the connections
between hospitals and research units that were alluded to earlier in this
chapter. Depending on the treatment specialty, a research institution itself
may function as a hospital or may offer only outpatient services. All
requirements for hospital or clinic licensing and other regulations that are
discussed in Chapter 2 apply to biomedical research institutions that treat
patients. Patient care services of independent biomedical research institu-
tions are subject to the same centripetal forces of the U.S. health care
system that have linked together originally autonomous hospitals as
medical centers. Institutions may affiliate or merge services voluntarily to
control costs or acquire new facilities. There also have been instances in
which internal reorganization has been forced upon research units by
third-party payers to simplify billing procedures.'^


Biomedical research institutions also engage in educational activities, both
informal and formal. Informal educational activity is necessitated by the
fact that scientific investigations are highly specialized, employing sophis-
ticated concepts and the latest equipment. The junior staff need to be
educated regarding the use of equipment and handling of hazardous
materials. Research units affiliated with academic medical centers and
teaching hospitals may also play significant roles in formal educational
programs. A unit could operate, for example, as a specially funded section
of a basic science department at a medical school, with all the senior staff
holding academic appointments. Students might be involved in projects in
fulfillment of elective course work or perhaps as summer employees.
Postdoctoral fellows render significant contributions to research pro-
grams, with their employment often underwritten by federal grants. (For
further information on the role of research programs in health profession-
als' education, see Chapter 5.)


Most biomedical research, particularly in the nonprofit sector, is intended
to produce publishable findings. The quality of the publication as a rule is



FIGURE 4-2 Psychologist Joanna Grant Nicholas studies communication skills
of hearing-impaired children at Central Institute for the Deaf, St. Louis, 1993.
Source: Central Institute for the Deaf, St. Louis, Marcus Kosa, photographer

important to investigators, and a respected refereed scientific journal is
the medium of choice in most cases. The writing and editorial revision that
take place before a manuscript'^ is submitted for publication are often a
painstaking process. Large research institutions may employ professional
writers and editors to facilitate this process. As organizational activities,
such services are likely to come under administrative purview, but unlike
the activities considered previously in this chapter, they are primarily
focused on the end products of research rather than on beginnings.

Scientific journals in the main are published either by professional
associations or by commercial firms. A significant number, however, are
published directly by research institutions. An even more common varia-
tion is for a professional association to appoint an editor-in-chief and an


editorial board, who then may draw on the resources of a research unit.
Neurology, for example, the chief official organ of the American Academy
of Neurology, has since its inception been edited by a series of distin-
guished investigators in the field, who oversee editorial operations.

An essential ingredient in editing a scientific journal is peer review of
the most promising submissions. Editors normally send each manuscript
to at least two experts in the field for them to judge the quality and
potential of the work. Rarely do reviewers have the opportunity to test
fully the methods and findings described, so their reactions are never
foolproof, but overall the process provides an effective means of quality
control. The identities of reviewers are generally kept confidential, and
their critiques may be moderated by the editors before they render
decisions about publication. The exchanges involved can be an important
part of the work of a research institution.'^^ (For more information on
publishing in the medical field, see Chapter 7.)

Avenues of scientific communication less formal than refereed jour-
nals may also be important to research units. Bulletins and newsletters are
commonly devoted to such purposes as interstaff news (a particular
consideration if the unit operates in decentralized facilities or employs
substantial numbers of visiting or temporary staff) and fund raising.
Conferences, seminars, and symposia are frequently chosen means for
disseminating information or airing common problems among colleagues.
Many research institutions employ public relations personnel to tout
achievements to media and directly to the general public. All these and
other efforts require a significant investment of the unit's resources.

Biomedical research institutions may produce marketable inventions
worth in the aggregate millions of dollars annually as products or by-
products of their investigations. Certain major medical centers generate
enough patentable findings to warrant hiring staffs of patent attorneys to
manage the situation. This is beyond the means, however, of most smaller
nonprofit institutions. Independent research units do have the option of
contracting with patent management organizations, the largest of which
are also equipped to handle product marketing.'^^


The work of archivists in scientific fields other than biomedicine to
document what is termed "discipline history" has significant applications
here. The concept of discipline history originally developed out of concern
for preserving landmark records in the physical sciences and engineering.
The primary goal has been to ensure adequate documentation of subject


areas through joint efforts of archivists in several institutions. One of the
most effective demonstrations of archival cooperation of this nature has
been coordinated by the Center for the History of Physics of the American
Institute of Physics (AIP) in fields such as high-energy physics, space
science, and geophysics.'^^

If we ignore for a moment the obvious differences between physics
and biomedicine and compare specific programs in these two disciplines,
we find many similarities. Like biomedical units, institutions devoted to
research in physics have many different missions and profiles, ranging
from federal governmental agencies to commercial ventures. As in bio-
medicine, the contributions of nonprofit units are very strong. These
institutions include both state-supported and private bodies, some univer-
sity-affiliated, others independent. A combination of federal grant pro-
grams and private philanthropy enables them to conduct similar investi-
gatory programs. To round out the list of similarities, much of the research
performed in physics laboratories has a direct impact on biomedicine.
Investigations in isotope analysis, applications of laser techniques, spect-
roscopy, and ultrasonics are but a few of these areas. Many examples of
close collegial interaction between physicists and biomedical researchers
could be cited at any large academic medical center.


It is not appropriate here to discuss how receptive all research institutions
in physics may be to the model established by the AIP other than to
observe that it appears to work most effectively in the world of megapro-
jects supported primarily by governmental agencies and consortia. There
are areas of biomedical research that lend themselves equally well to
discipline history projects, and for similar reasons. Among the most likely
current subjects is the Human Genome Project, a worldwide research
effort that has the goal of mapping the entire structure of human DNA and
determining the location of the estimated 100,000 human genes. Funded
through the NIH and the Department of Energy, a substantial portion of
the research is being conducted in the laboratories of these two agencies.
In the grandest tradition of extramural programs, the NIH portion of the
appropriation is shared with (at this writing) seven major centers located
at universities throughout the United States. Other related program assis-
tance is available to smaller research teams, with additional money pro-
vided for training grants, technology development, and international

The objectives of the Human Genome Project in a sense are inherently


archival. This is particularly clear from one part of the NIH project, which
is to establish and operate a National Center for Biotechnology Informa-
tion (NCBI) within the National Library of Medicine. NCBI has the
particular mission of creating automated systems for knowledge about
molecular biology, biochemistry, and genetics and of pursuing research in
biological information handling. NCBI is currently conducting investiga-
tions with genome research centers and libraries throughout the country
about the feasibility of transmitting mapping data online. The Human
Genome Project also offers opportunities for archival development
through a small portion (3 percent) of the budget allocated to address
ethical, legal, and social considerations. Arguing that the full implications
of the project on society cannot be understood unless appropriate records
on a wide range of issues are collected and retained, at least two private
organizations have begun discipline history studies on genome research.'*^
Currently, the most comprehensive effort in this regard is being
mounted by the Chemical Heritage Foundation, an organization spon-
sored by the American Chemical Society and the Society of Chemical
Engineers (and openly modeled on the AIP). The CHF project, titled
BIMOSI, or Biomolecular Sciences Initiative, is interested not only in
genome research but in all important investigations pertaining to molecu-
lar biology. At this writing, the staff of BIMOSI have begun to advise
researchers and their organizations about what to preserve and where,
and how to conduct oral histories. A somewhat more limited documenta-
tion project is under way at the National Center for Bioethics Literature, of
the Kennedy Institute of Ethics at Georgetown University, focusing on the
ethical, legal, and social implications of human genome research. Histori-
ans and archivists involved in both projects acknowledge the enormity of
their respective undertakings, in particular challenges related to the diver-
sity of interests of individual researchers and their parent institutions.'^^


The opportunity that has developed for a discipline history project on
genome research is unusual among biomedical research fields. The combi-
nation of factors — novelty, urgency, international collaboration, and
above all, the generous public funding of the project — is more characteris-
tic of the great crash programs in nuclear physics than any recent field of
medical investigation. It is worth examining here why there were no
comparable calls for discipline history centers to address earlier national
mandates for biomedical research, such as the "wars" declared on cancer
from the 1930s through the I980s.'*6


The differences are at least fourfold. First, as already noted, there is
the proliferation of biomedical research institutions in the United States.
Consider, for example, a hypothetical history center for cancer oncology.
Counting only the units that truly specialize in problems and issues
relating to malignancies, one would have to deal not with seven major
centers, as with the Human Genome Project, but perhaps seven times
seventy. '^^ For better or worse (and certainly, in the case of cancer, there
have been many arguments to the effect that the proliferation of research
institutions has resulted in much duplication and waste), the extramural
grant system and private philanthropy have never concentrated their
funding eggs in only a few baskets.

Second, there is the general issue of confidentiality of clinical data.
Essentially, this is an area governed by the same constraints involving
patient records discussed in Chapter 2. Research units within hospitals
and academic medical centers are required to safeguard the identities of
study subjects as completely as they do the identities of regular patients.
For this reason, they normally deny outside researchers (or anyone else)
access to clinical data that they have generated or augmented for investi-
gative purposes.'*^

A third difference reflects the enormous contributions of private,
profit-making research institutions to every biomedical field. Competition
alone dictates that information about their proprietary discoveries not be
shared with other organizations or individuals, at least until the data no
longer have market value. Companies that have developed new drugs for
the treatment of cancer (or HIV infections, or any other focus of a health
crisis) find themselves under great pressure to justify decisions about the
costs, marketing, and distribution of their products. They are certain,
therefore, to be armed with policies concerning what they will dissemi-
nate to the public and what they will withhold.

A fourth difference is more a matter of philosophy and custom than
legal substance. This relates to the traditional reliance on refereed journals
as the primary medium for reporting — and preserving the historical re-
cord of — biomedical discoveries in the academic sector. For most biomedi-
cal scientists, journals are the true archives: at best, they convey succinctly
the nature of discoveries, discuss their implications, and provide necessary
directions for replicating the experiments. Unlike findings from projects in
sciences and technologies that are comYnissioned for national security
operations or commercial enterprises, most investigations in medicine and
allied fields are intended to produce publishable results. Despite fierce
competition among scientists for the attention of editors and review
panels of the most respected journals, findings of most reputable health
science projects appear sooner or later in print."*^ It is no accident that the


titles of well over one hundred biomedical serial publications representing
a wide range of investigative fields begin with variations of the term

The narrow implication of this tradition is that special repositories for
original research data are unnecessary. Several indications, however,
point to a greater realization within the biomedical research community
than in the past that measures must be taken to preserve documentation
generated by significant projects. Hedrick in 1985 offered an extensive
summary of these issues. ^o They include wider opportunities for verifica-
tion, refutation, or refinement of original results; the chance for replica-
tions with multiple data sets; encouragement of new questions and multi-
ple perspectives employing the original data; the creation of new data sets
through data file linkages; reductions in the incidence of faked and
inaccurate results; dissemination of knowledge about analytic techniques
and research designs; and the provision of expanded resources for training
of future scientists. Other authors address the possibility that some re-
searchers who have been supported by public funding may be compelled
to preserve and share their data with others. ^^


Archival repositories located throughout the United States hold extensive

1 2 3 4 5 6 7 8 9 11 13 14 15 16 17 18 19 20 21 22 23 24 25 26

Online LibraryJoan D KrizackDocumentation planning for the U.S. health care system → online text (page 11 of 26)